Photonics Elements for Sensing

and Optical Conversions

This book covers a number of a rapidly growing areas of knowledge that may be
termed as diffractive nanophotonics. It also discusses in detail photonic components
that may find uses in sensorics and optical transformations.

Photonics Elements for Sensing and Optical Conversions covers a number of a rapid-
ly growing areas of knowledge that may be termed as diffractive nanophotonics. The
book examines the advances is computational electrodynamics and nanoelectronics
that have made it possible to design and manufacture novel types of photonic com-
ponents and devices boasting unique properties unattainable in the realm of classical
optics. The authors discuss plasmonic sensors, and new types of wavefront sensors
and nanolasers that are widely used in telecommunications, quantum informatics,
and optical transformations. The book also deals with the recent advances in the
plasmonic sensors based on metal-insulator-metal waveguides for biochemical sens-
ing applications. Additionally, nanolasers are examined in detail, with a focus on
contemporary issues; the book also deals with the fundamentals and highly attractive
applications of metamaterials and metasurfaces. The authors provide an insight into
sensors based on Zernike optical decomposition using a multi-order diffractive opti-
cal element, and explore the performance advances that can be achieved with optical
computing.

The book is written for opticians, scientists, and researchers who are interested in an
interesting section of plasmonic sensors, new types of wavefront sensors and nano-
lasers, and optical transformations. The book will be bought by upper graduate and
graduate level students looking to specialize in photonics and optics.

BK-TandF-KAZANSKIY_9781032572949-230642-FM.indd 1 9/26/23 4:50 PM

BK-TandF-KAZANSKIY_9781032572949-230642-FM.indd 2 9/26/23 4:50 PM
 Photonics Elements
for Sensing and Optical
Conversions

Edited by
Nikolay Lvovich Kazanskiy

BK-TandF-KAZANSKIY_9781032572949-230642-FM.indd 3 9/26/23 4:50 PM

 Designed cover image: Nikolay Lvovich Kazanskiy

First edition published 2024

by CRC Press
2385 NW Executive Center Drive, Suite 320, Boca Raton FL 33431

and by CRC Press

4 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN

CRC Press is an imprint of Taylor & Francis Group, LLC

© 2024 selection and editorial matter, Nikolay Lvovich Kazanskiy; individual chapters, the contributors

Reasonable efforts have been made to publish reliable data and information, but the author and pub-
lisher cannot assume responsibility for the validity of all materials or the consequences of their use.
The authors and publishers have attempted to trace the copyright holders of all material reproduced in
this publication and apologize to copyright holders if permission to publish in this form has not been
obtained. If any copyright material has not been acknowledged please write and let us know so we may
rectify in any future reprint.

Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced,
transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or here-
after invented, including photocopying, microfilming, and recording, or in any information storage or
retrieval system, without written permission from the publishers.

For permission to photocopy or use material electronically from this work, access www.copyright.com
or contact the Copyright Clearance Center, Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923, 978-
750-8400. For works that are not available on CCC please contact [email protected]

Trademark Notice: Product or corporate names may be trademarks or registered trademarks and are
used only for identification and explanation without intent to infringe.

Library of Congress Cataloging‑in‑Publication Data

LoC Data here

LoC Data here
LoC Data here]

ISBN: 978-1-032-57294-9 (hbk)

ISBN: 978-1-032-57391-5 (pbk)
ISBN: 978-1-003-43916-5 (ebk)

DOI: 10.1201/9781003439165

Typeset in Times
by KnowledgeWorks Global Ltd.

BK-TandF-KAZANSKIY_9781032572949-230642-FM.indd 4 9/26/23 4:50 PM

 Contents
Preface.......................................................................................................................xi
About the Author.................................................................................................... xiii
List of Abbreviations................................................................................................ xv

Chapter 1 Silicon Photonic Waveguides: Comparison and Utilization

in Sensing Applications.........................................................................1
1.1 Introduction................................................................................ 1
1.2 Optical Waveguide..................................................................... 3
1.2.1 Ridge Waveguide........................................................... 3
1.2.2 Rib Waveguide..............................................................4
1.2.3 Slot Waveguide..............................................................5
1.2.4 Hybrid Plasmonic Waveguide....................................... 6
1.2.5 Suspended Si Waveguides............................................. 7
1.3 Waveguide Coupling Techniques...............................................8
1.3.1 Grating Coupling...........................................................8
1.3.2 Butt Coupling and End-Fire Coupling........................ 10
1.4 Confinement Factor.................................................................. 14
1.5 Losses of Si Waveguides.......................................................... 15
1.5.1 Scattering Loss............................................................ 15
1.5.2 Absorption................................................................... 16
1.5.3 Radiation Loss............................................................. 16
1.6 Manufacturing of Si Waveguides on Silicon on
Insulator Platform..................................................................... 17
1.6.1 Photolithography and Laser Lithography.................... 17
1.6.2 Wafer Preparation........................................................ 19
1.6.3 Spin Coating and Exposure......................................... 19
1.6.4 Photoresist Developing and Hard Bake.......................20
1.6.5 Etching Process...........................................................20
1.7 Advances in Si Photonics Sensors............................................ 21
1.7.1 Wavelength Interrogation Method............................... 22
1.7.2 Intensity Interrogation.................................................24
1.8 Numerical Simulation of Si Ridge Waveguide by
Combining a Thin Metal Film.................................................25
1.8.1 Design of a Metal-Supported Si Ridge
Waveguide...................................................................26
1.8.2 The Connection between the Waveguide
Dimension and Γ.........................................................28
1.8.3 Influence of Metal Layer on the Propagation Loss...... 29
1.9 Concluding Remarks................................................................ 30
References........................................................................................... 30

BK-TandF-KAZANSKIY_9781032572949-230642-FM.indd 5 9/26/23 4:50 PM

 vi Contents

Chapter 2 Photonic Crystal Cavities in Integrated On-Chip Optical

Signal Processing Components........................................................... 33
2.1 Introduction.............................................................................. 33
2.2 Temporal Integration of Optical Signals..................................34
2.3 In-Plane Optical Pumping of Photonic Crystal Cavity............ 38
2.4 Electrical Pumping of Photonic Crystal Cavity.......................44
2.5 Optical Modulator Based on Coupled PC Cavities..................46
2.6 Application of Photonic Crystal Coupled Cavities for
Increase in Sensitivity of Optical Sensor................................. 53
2.7 On a Si-Based Photonic Crystal Cavity for the Near-
Infrared Region: Numerical Simulation and Formation
Technology............................................................................... 59
2.8 Conclusions...............................................................................64
References........................................................................................... 65

Chapter 3 Nanoplasmonic Sensors: Recent Advances......................................... 69

3.1 Introduction.............................................................................. 69
3.2 Plasmonic Waveguides............................................................. 71
3.3 Critical Issues Associated with Plasmonic Sensing
Devices Established on Metal-Insulator-Metal
Waveguide Formations............................................................. 72
3.3.1 Favorite Plasmonic Materials...................................... 72
3.3.2 Fabrication Methods.................................................... 74
3.3.3 Transmission Loss....................................................... 74
3.3.4 Inquisition Method...................................................... 75
3.4 Performance Characteristics of Surface Plasmon
Polariton Resonator.................................................................. 76
3.4.1 Sensitivity.................................................................... 76
3.4.2 The Figure of Merit..................................................... 77
3.4.3 Newly Developed Refractive Index Sensing Devices.... 77
3.4.4 Examples..................................................................... 78
3.5 Surface Plasmon Resonance Sensor Arrangement................... 85
3.5.1 Surface Plasmon Resonance Sensing with
Nanoparticles and Thin Films..................................... 86
3.5.2 Breakthrough in Localized Surface Plasmon
Resonator Sensing.......................................................90
3.6 Future of Plasmonic Sensing Devices...................................... 91
References........................................................................................... 93

Chapter 4 Plasmonic Nanolasers....................................................................... 100

4.1 Introduction............................................................................ 100
4.2 Overview of a Plasmonic Nanolaser....................................... 101
4.2.1 Surface Plasmons and Plasmonic Materials.............. 101

BK-TandF-KAZANSKIY_9781032572949-230642-FM.indd 6 9/26/23 4:50 PM

 Contents vii

4.2.2 Gain Materials Used in Plasmonic Nanolasers......... 103

4.2.3 Theoretical Model and Structure of a Plasmonic
Nanolaser................................................................... 104
4.3 Experimental Demonstrations of the Plasmonic
Nanolasers.............................................................................. 107
4.3.1 Single Nanoparticle-Based Plasmonic
Nanolasers................................................................. 107
4.3.2 Fabry–Perot Cavity-Based Plasmonic
Nanolasers................................................................. 108
4.3.3 Whispering Gallery Mode Cavity-Based
Plasmonic Nanolasers............................................... 112
4.3.4 Particle and Hole Arrays-Based Plasmonic
Nanolasers................................................................. 114
4.3.5 Summary of Laser Characteristics............................ 118
4.4 Applications of Nanolasers..................................................... 118
4.4.1 Plasmonic Nanolasers for Integrated Photonic
Circuits...................................................................... 118
4.4.2 Plasmonic Nanolasers for Sensing............................ 120
4.4.3 Plasmonic Nanolasers for Biological
Applications............................................................... 122
4.5 Summary and Outlook........................................................... 124
References......................................................................................... 124

Chapter 5 Metasurfaces and Several Well-Studied Applications...................... 130

5.1 Introduction............................................................................ 130
5.2 What are Metasurfaces?......................................................... 131
5.3 Negative Refraction................................................................ 132
5.4 Metasurfaces........................................................................... 133
5.4.1 MS Design Process................................................... 134
5.4.2 State of the Art Metasurfaces and Their
Limitations................................................................ 135
5.5 Applications of Metasurfaces................................................. 136
5.5.1 Metasurface Antennas............................................... 137
5.5.2 Metasurface Lenses................................................... 139
5.5.3 Metasurface Perfect Absorbers................................. 142
5.5.4 Metasurfaces for Sensor Applications....................... 144
5.5.5 Metasurface Waveguides........................................... 146
5.5.6 Cloaking.................................................................... 148
5.5.7 All-dielectric Metasurfaces Visible Light
Modulator.................................................................. 150
5.5.8 Frequency Selective Surfaces Based on
Metasurfaces............................................................. 151
5.5.9 Polarizers Based on Metasurfaces............................ 153
References......................................................................................... 155

BK-TandF-KAZANSKIY_9781032572949-230642-FM.indd 7 9/26/23 4:50 PM

 viii Contents

Chapter 6 Optical Fiber Sensors Based on Diffractive and Fiber Periodic

Microstructures................................................................................. 158
6.1 Introduction............................................................................ 158
6.2 Amplitude Optical Fiber Sensors Based on Diffractive
Optical Elements Matched with Transverse Modes............... 159
6.2.1 Design Principles of Fiber-Optical Converters
Based on Mode Filtering........................................... 159
6.2.2 Analysis of Mode Coupling Due to Periodic
Microbendings of a Graded-Index Fiber................... 161
6.2.3 Study of the Output Mode Power for a
Graded-Index Fiber against the Microbending
Magnitude................................................................. 162
6.3 Spectral Optical Fiber Sensors Based on Periodically
Microstructured Fibers........................................................... 168
6.3.1 Approaches to Measuring Physical Quantities......... 168
6.3.2 Optical Fiber Sensors................................................ 171
References......................................................................................... 174

Chapter 7 Wave Front Aberration Sensors Based on Optical Expansion

by the Zernike Basis.......................................................................... 178
7.1 Introduction............................................................................ 178
7.2 Method of Optical Expansion by the Zernike Basis.............. 181
7.2.1 Theoretical and Numerical Investigations................ 181
7.2.2 Multi-Order Diffractive Optical Elements
for Optical Decomposition........................................ 188
7.2.3 Analysis of the Influence of the Level of
Aberration on the Detection Accuracy...................... 191
7.2.4 Experimental Results................................................ 198
7.3 Application of Wave Front Sensors in Ophthalmology.......... 205
7.3.1 Wave Front Sensors in Ophthalmic Research...........205
7.3.2 Medical Ophthalmological Measurement
Devices......................................................................207
7.3.3 Analysis of Wave Front Sensors in
Ophthalmological Studies......................................... 210
7.3.4 Attempts of an Analysis of Eye Aberration.............. 213
7.4 Conclusion.............................................................................. 231
References......................................................................................... 232

Chapter 8 Optical Computing: Key Problems, Achievements,

and Perspectives................................................................................ 239
8.1 Introduction............................................................................ 239
8.2 Types of Computing and Necessity of Optics
for Computing......................................................................... 242

BK-TandF-KAZANSKIY_9781032572949-230642-FM.indd 8 9/26/23 4:50 PM

 Contents ix

8.3 Optical Processing Architecture............................................ 243

8.4 Types of Optical Components Used to Implement
Optical Functions................................................................... 245
8.4.1 Spatial Light Modulator (SLM)................................246
8.4.2 Plasmonic Switches................................................... 249
8.4.3 Theoretical and Numerical Investigations................ 251
8.4.4 Diffractive NN.......................................................... 254
8.4.5 Theoretical and Numerical Investigations................ 256
8.4.6 Resonant Nanophotonic Constructions..................... 261
8.5 Growth Ideas, Constraints, and Misconceptions....................264
8.6 Conclusion..............................................................................266
References......................................................................................... 267

General Conclusion............................................................................................... 279

Appendix A: L
 isting for Calculating the Sensor Matched with the
Zernike Basis Functions................................................................ 283
Appendix B: Listing for Calculating a Distorted Image................................... 289
Index....................................................................................................................... 293

BK-TandF-KAZANSKIY_9781032572949-230642-FM.indd 9 9/26/23 4:50 PM

BK-TandF-KAZANSKIY_9781032572949-230642-FM.indd 10 9/26/23 4:50 PM
 Preface
Topics of this book cover a number of a rapidly growing areas of knowledge that
may be termed as diffractive nanophotonics. Advances made in computational elec-
trodynamics and nanoelectronics have made it possible to design and manufacture
novel types of photonic components and devices boasting unique properties unat-
tainable in the realm of classical optics. Significantly, procedures for designing and
numerical simulation of such elements and devices do not involve the consideration
of quantum effects and can do with solving exact and approximate Maxwell’s equa-
tions. The monograph authors discuss in detail photonic components that may find
uses in sensorics and optical transformations.

ABSTRACT
This book deals with nanophotonic structures and components as well as applica-
tions thereof for sensorics and optical transformations. Special attention is given to
various waveguides and possibilities of their design and practical uses. Plasmonic
sensors and nanolasers are discussed, methods for design and prospective applica-
tions of metasurfaces and other nanostructures intended to perform optical transfor-
mations and computation are given. Capabilities of new types of wavefront sensors
are analyzed.
The monograph is intended for scientific researchers and specialists working in
optics, nanophotonics, information technology, and optical instrumentation. It may
also prove useful for postgraduate students specializing in these research areas.

Nikolay L. Kazanskiy

xi

BK-TandF-KAZANSKIY_9781032572949-230642-FM.indd 11 9/26/23 4:50 PM

BK-TandF-KAZANSKIY_9781032572949-230642-FM.indd 12 9/26/23 4:50 PM
 About the Author
Nikolay Lvovich Kazanskiy graduated with honors (1981) from Kuibyshev Aviation
Institute (presently, Samara National Research University), majoring in Applied
Mathematics. He received his Candidate in Physics & Mathematics (1988) and
Doctor in Physics & Mathematics (1996) degrees from Samara National Research
University. He is the director of Image Processing Systems Institute of the RAS –
Branch of the Federal Scientific-Research Centre “Crystallography and Photonics”
of the Russian Academy of Sciences, and also holds a part-time position of professor
at Technical Cybernetics department of Samara National Research University. He is
a member of OSA, SPIE, and IAPR. He has co-authored 290 scientific papers and
14 monographs, and has 57 inventions and patents. His current research interests
include diffractive optics, computer vision, optical sensors, mathematical modeling,
lighting devices design, and nanophotonics. Website: http://www.ipsiras.ru/staff/
kazanskiy.htm. Email: [email protected].

xiii

BK-TandF-KAZANSKIY_9781032572949-230642-FM.indd 13 9/26/23 4:50 PM

BK-TandF-KAZANSKIY_9781032572949-230642-FM.indd 14 9/26/23 4:50 PM
 List of Abbreviations
ADC analog-to-digital converter
AI artificial intelligence
ANN artificial neural network
AOC analog optical computing
AOI angle of incidence
AR autorefraction
BC boundary condition
BOX buried oxide
BPSK binary phase-shift key
BS beam splitter
BS–PS beam splitter–phase shifter
BWs bandwidths
CCD continuity of care document
CCT central corneal thickness
CFG command voltage generator
CG chalcogenide glass
CGH computer generated holograms
CMOS complementary metal-oxide-semiconductor
CNN convolutional neural network
CPP channel plasmon polariton
CR contrast ratio
CRf cycloplegic refraction
CROW coupled-resonator optical waveguide
CRR circular ring resonator
DFB distributed feedback
DLs dynamic light scattering
DMD digital micromirror device
DMSO dimethyl sulfoxide
DNA deoxyribonucleic acid
DOC digital optical computing
DOE diffractive optical element
DRS dual rotating Scheimpflug
DUV deep UV, deep ultraviolet
EASLM electrically addressed spatial light modulator
EBL electron beam litho
EFR evanescent field ratio
EIT electromagnetically induced transparency
EM electromagnetic
EO electro-optic
EO-SLM electro-optic spatial light modulator
ER extinction ratio
FBG fiber Bragg gratings

xv

BK-TandF-KAZANSKIY_9781032572949-230642-FM.indd 15 9/26/23 4:50 PM

 xvi List of Abbreviations

FDTD finite difference time domain

FEM finite element method
FF fill factor
FFT fast Fourier transform
FIB focused ion beam
FOBG fiber-optic Bragg gratings
FOM Figure of Merit
FP Fabry–Perot
FPGA field-programmable gate array
FR Fano resonance
FSR free spectral range
FSS frequency selective surface
FT Fourier transform
FWHM full width at half maximum
GRIN graded index
GST Ge2Sb2Te5
HCG high-contrast gratings
HMDS hexamethyldisilane
HOA higher-order aberrations
HPWG hybrid plasmonic waveguide
I/P input
IFTA iterative Fourier transform algorithm
IgG immunoglobulin G
IPA isopropyl alcohol
IR infrared
ITO indium tin oxide
JTC joint transform correlator
KC keratoconus
KR Kretschmann
LASE laser particle stimulated emission
LASIK laser-assisted in situ keratomileusis
LCoS liquid crystal on silicon
LCVWP liquid crystal variable wave plate
LESPR lasing-enhanced surface plasmon resonance
LGs logic gates
LOD limit of detection
LP linearly polarized
LSP localized surface plasmon
LSPR localized surface plasmon resonator
LW leaky wave
MA meta-atom
MAC multiply-accumulate
MBBs meta building blocks
MDM mode division multiplexing
MIBK methyl isobutyl ketone
ML machine learning

BK-TandF-KAZANSKIY_9781032572949-230642-FM.indd 16 9/26/23 4:50 PM

 List of Abbreviations xvii

MM metamaterial
MMI multimode interference
MODAN DOE for laser mode generation (or laser mode generator)
MS metasurface
MS-OASLM metasurface-based optically addressed spatial light modulator
MTF modulation transfer function
MTS metasurface
MWv microwave
MZI Mach–Zehnder interferometer
NA numerical aperture
NIR near-infrared
NN neural networks
NPs nanoparticles
NV nonvolatile
NW nanowall
O/P output
OASLM optically addressed spatial light modulator
OC optical computing
OCT optical coherence tomography
ODS optical differentiation sensor
OF optical fiber
OFBG optic fiber Bragg grating
OLG optical logic gates
OLO optical logic operation
ONC optical neural chip
OSI objective scattering index
OTDR optical time-domain reflectometry
PA perfect absorber
PhA photoacoustic
PB Pancharatnam–Berry
PBG photonic bandgap
PC photonic crystal
PCC photonic crystal cavities
PCM phase change material
PDA personal digital assistant
PDMS polydimethylsiloxane
PH pinhole
PIC photonic-integrated circuit
PL photolithography
PML perfectly matched layers
PMMA polymethyl methacrylate
POL polarization
PR photoresist
PS phase shifter
PScF point scattering function
PSF point spread function

BK-TandF-KAZANSKIY_9781032572949-230642-FM.indd 17 9/26/23 4:50 PM

 xviii List of Abbreviations

PWFS pyramid wave front sensors

R&D research and development
RFID radio frequency identification
RI refractive index
RIE reactive ion etching
RIU refractive index unit
RMS root mean square
RR ring resonator
SCB self-collimated beam
SEM scanning electron microscope
SERS surface-enhanced Raman scattering
SERsp surface-enhanced Raman spectroscopy
SLED superluminescent light-emitting diode
SLM spatial light modulator
SMF single mode fiber
SOA semiconductor optical amplifier
SOI silicon on insulator
SP surface plasmon
SPP surface plasmon polariton
SPR surface plasmon resonance
SR square ring
SRR square ring resonator
SSRR split square ring resonator
SW surface waves
SWG subwavelength gratings
SWM surface wave medium
TBP time-bandwidth product
TE transverse electric
TEM transverse electromagnetic mode
TF transfer function
TIA transimpedance amplifier
TIR total internal reflection
TM transverse magnetic
TMAH tetramethylammonium hydroxide
UC upper cladding
UV ultraviolet
VIS visible
WDM wavelength division multiplexing
WFS wave front sensors
WG waveguide
μ-RR micro-ring resonator
Γ field con

BK-TandF-KAZANSKIY_9781032572949-230642-FM.indd 18 9/26/23 4:50 PM

 1 Silicon Photonic
Waveguides
Comparison and Utilization
in Sensing Applications
Muhammad Ali Butt1,2, Svetlana Nikolaevna
Khonina1,3, and Nikolay Lvovich Kazanskiy1,3
1Samara National Research University, Samara, Russia
2 Warsaw University of Technology, Institute of
Microelectronics and Optoelectronics, Warszawa, Poland
3Image Processing Systems Institute – Branch of the

Federal Scientific Research Centre “Crystallography and

Photonics” of Russian Academy of Sciences, Samara, Russia

1.1 INTRODUCTION
Electric wires are still the source to transmit data in roughly all modern devices,
such as the world’s fastest supercomputers, laptops, or phones. The salient features of
copper wire include reliability and low price. For decades, copper wire can provide
massive quantities of transistors to interconnect and maintain circuitry on smaller
chips. However, the imperfections of copper wire, such as limited bandwidth, cur-
rent leakage, and cross talk between the contiguous wires come into the spotlight as
we have swiftly approached the physical limit of chip miniaturization. Additionally,
the power intake, the heat generated, and the space occupied by copper are also
practical concerns that should be addressed. One of the possible solutions to the
aforementioned problems is based on light-based technologies such as optical fiber
cables, which utilize photons to transport the information at a much faster speed and
bandwidth. Nevertheless, the excessive cost of replacing the electric wires in inte-
grated circuits with cutting-edge photonics is a barrier to commercialization.
It is interesting to realize integrated optics in silicon (Si) due to well-developed
technology and economy. It is a well-studied material, and the processing has been
advanced by the electronics industry to a point that is highly adequate for most inte-
grated optical applications. At present, the smallest feature size for most applica-
tions is ~1–2 µm, which in terms of microelectronics is very old technology, while
there is a tendency toward miniaturization. Besides, novel and enhanced process-
ing becomes available as Si microelectronics continue to progress. Photonic circuits
can be fabricated by direct patterning Si to create optical interconnects to broadcast
data-carrying laser signals. The biggest advantage is that it can transmit massive

DOI: 10.1201/9781003439165-1 1

BK-TandF-KAZANSKIY_9781032572949-230642-Chp01.indd 1 9/26/23 1:17 PM

 2 Photonics Elements for Sensing and Optical Conversions

data while consuming low power with no heating up or causing any deterioration in
the signal.
Si is transparent to low energy in the infrared (IR) region of the wavelength spec-
trum due to its electronic bandgap; however, it is opaque to photons in the visible
section of the spectrum [1]. It is an indirect bandgap, which makes Si a bad emitter.
Even though Si photonics is an ideal solution on paper, scientists have struggled for
more than 30 years to overcome its different shortcomings [2].
For easy understanding of the attributes of Si, we have listed some major advan-
tages of Si photonics, along with a few disadvantages associated with both material
and device characteristics, in Figure 1.1. Mainly all the points are easy to under-
stand. However, we will give a little description of the “Pockels effect,” which is
an important phenomenon that allows changing the optical properties of a certain
material by applying an external electric field. Mostly high-speed modulators and
switches are available in lithium niobate, which is used in optical communication
systems. However, these devices are massive; that’s why their application is limited
to complex optical networks such as data centers and telecom networks, though this
effect is inherently zero in single-crystalline Si because of the inversion symmetry
of the crystal structure. Recently, a record-high electro-optic (EO) response in Si
photonic devices was demonstrated by applying nanometer-thick, crystalline layers
of a material exhibiting the Pockels effect [3]. This can be foreseen as the next step
toward an application area mostly beyond communication technologies.

FIGURE 1.1 A few major advantages and disadvantages of Si photonics.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp01.indd 2 9/26/23 1:17 PM

 Silicon Photonic Waveguides 3

1.2 OPTICAL WAVEGUIDE

Integrated photonics contribute to the integration of numerous photonic components,
such as beam splitters, polarizers, couplers, sensors, interferometers, and detectors
on a single platform. The basic theory of light confinement in the optical waveguides
(WGs) is the same as that in optical fibers where a material having a high refractive
index (core) is surrounded by a material with a lower refractive index (cladding) that
can act as a light trap. The light injected into the core undergoes a phenomenon of
total internal reflection between the interfaces; thus, light cannot escape the core.
As a result, light can be transported from point to point in an integrated photonic
chip. In this section, we focus on several configurations of optical WGs built on a Si
on insulator (SOI) platform such as ridge, rib, slot, Si-based hybrid plasmonic WG
(hereafter represented as HPWG), and suspended WGs.

1.2.1 Ridge Waveguide

Ridge WG is one of the primitive components of Si photonics that has been compre-
hensively studied over the years [4]. They are generally used in different applications
such as lasers, modulators, polarizers, switches, and amplifiers. It comprises a Si
core embedded on a silica (SiO2) substrate as presented in Figure 1.2(a), where WSi
and HSi are the width and height of the core, respectively.
The single-mode condition is vital in the realization of functional photonic
devices; therefore the size of the WG core should be determined to obey the single-
mode condition. The principal requirement is the single-mode propagation of the

FIGURE 1.2 Graphic representation of (a) ridge WG, (b) rib WG, (c) slot WG, (d) Si-based
HPWG, and (e) suspended Si WG.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp01.indd 3 9/26/23 1:17 PM

 4 Photonics Elements for Sensing and Optical Conversions

FIGURE 1.3 The effective refractive index of a ridge WG: (a) TE00 mode, (b) TM00 mode.

TE00 and TM00 mode. In a WG, each mode propagates with a phase velocity of
c/neff, where c represents the speed of light in a vacuum and neff is the effective
refractive index of that mode. It implies the optical power of the mode in the WG
core. The mode is guided when neff is higher than the cladding and lesser than the
core; otherwise it will be radiated into the substrate. Moreover, the higher the neff,
the more the mode will be energetically guided. The dependence of the mode neff
on the WSi and HSi is calculated for TE00 and TM00 modes via the finite element
method (FEM), which is executed in a COMSOL Multiphysics software, as dis-
played in Figure 1.3.

1.2.2 Rib Waveguide

The rib WG can be considered as a special case of ridge WG where the Si layer is
not completely etched. The guiding layer principally consists of the slab with a strip
placed onto it. As shown in Figure 1.2(b), WSi is the rib width, HSi is the total height
of the rib, where h is the slab height. Single-mode silica WGs are comparatively easy
to design. On the other hand, SOI WGs with dimensions larger than a few nanome-
ters support multiple modes [5]. These kinds of WGs are typically unattractive in
photonic circuits as their working mechanism can be acutely compromised by the
existence of multiple modes. However, large rib WGs in SOI can be designed to be
monomodal. In 1991, Soref et al. [6] proposed a straightforward equation to obtain a
single-mode condition of such WGs:

WSi h
≤ 0.3 + ,
HSi 1 − h2

where h is the ratio of the slab height to overall rib height, and WSi/HSi is the ratio of
the WG width to total rib height. The dependence of mode effective refractive index
on h is studied in Figure 1.3 at a fixed value of WSi and HSi. The E-field mapping of
quasi-TE and quasi-TM mode is displayed in the inset of Figure 1.4.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp01.indd 4 9/26/23 1:17 PM

 Silicon Photonic Waveguides 5

FIGURE 1.4 The effective refractive index of a rib WG at TE00 and TM00 mode.

1.2.3 Slot Waveguide

Slot WG was anticipated by Vilson Rosa de Almeida and Carlos Angulo Barrios
[7] in 2003. It is proficient in confining light in a low-index material between two
high-index rail WGs by improving the gap (g), width (W), and height (HSi) of the
strip WGs. The design of the slot WG is shown in Figure 1.2(c). The normal E-field
component of quasi-TE undergoes high interruption at the interface between a high
and a low refractive index material, which oversees providing an eminent amplitude
in the low-index slot region that is relative to the ratio of the refractive indices of the
cladding material to that of the Si:

nSi2
Ex ,g = 2
E x ,Si .
nclad

Here, Ex,g and Ex,Si stand for the E-field confined in the slot (which is air) and
residual field in the Si, respectively. The dependence of mode effective refractive
index on g is presented in Figure 1.5(a) at a fixed value of W and HSi. The image of
the slot WG fabricated via focused ion beam (FIB) milled is shown in Figure 1.5(b).
Conversely, this WG scheme is not able to confine the quasi-TM mode as the E-field
remains continuous at the boundary.
As compared to other conventional WG schemes, slot WG can provide high field
confinement in the slot, which is in general not possible to attain. This attribute
makes slot WG an eye-catching replacement for applications that encompass sturdy
light–matter interaction, for instance sensing and nonlinear photonics. By utilizing
EO polymer material in slot WG it is possible to obtain the Pockels effect. As a
result, it gathered high potential in the field of optical switching and high-speed
modulation even at frequencies of 100 GHz. Several devices like Mach–Zehnder
interferometers (MZIs) and ring resonators (RRs) have been realized utilizing slot
WG phase shifters.
Light is typically injected into a slot WG by matching the propagation constants
of the ridge WG and the slot WG. Nonetheless, due to the mode mismatch and

BK-TandF-KAZANSKIY_9781032572949-230642-Chp01.indd 5 9/26/23 1:17 PM

 6 Photonics Elements for Sensing and Optical Conversions

FIGURE 1.5 (a) The effective refractive index of slot WG; (b) scanning electron microscope
(SEM) image of the FIB milled cross section of a fabricated device (the small hole in the
center was due to the incomplete coating of platinum deposited to protect the WGs during
the FIB milling) [8].

scattering loss, effective coupling between the ridge WG and slot WG is a difficult
challenge. As a result, extensive study is being conducted to leverage the use and
development of slot WGs for integrated optical sensors.

1.2.4 Hybrid Plasmonic Waveguide

A new type of plasmonic WG recognized as HPWG has gathered significant inter-
est due to its ability to offer both subwavelength confinement and long propagation
lengths. It appears like the insulator-metal (IM) WG formation, apart from that a
thin layer of low-index material (SiO2) is sandwiched between the metal layer (Au or
Ag) and a higher-index dielectric layer (Si), as displayed in Figure 1.2(d), where HSi,
W, HSiO2, HAu denote the layer thickness of the Si core, the core width, a layer thick-
ness of the SiO2 layer, and a layer thickness of the Au layer, respectively. Theoretical
studies have shown that a low-index layer can hold a low-loss compact mode whose
propagation length robustly relies on its thickness [9, 10].
The mode-coupling theory may be used to explain how the HPWG works. In
general, the Si core confines the dielectric WG mode, but the metal surface confines
surface plasmons (SPs), which are restricted to the metal surface. When these two
structures are kept in close vicinity, the dielectric WG mode supported by Si ridge
couples to the SP mode supported by the metal surface. Due to the mode coupling,
the light is transferred to the region between the metal and the low index medium.
The real part of neff is calculated by maintaining HAu and W at fixed values, whereas
HSiO2 and HSi are varied, as demonstrated in Figure 1.6.
HPWG integrates the characteristics of the dielectric and plasmonic WGs. The
excitation of surface electron fluctuation in the hybrid plasmonic WG focuses the
photons to reside near the metal surface. This helps in the reduction of radiation loss.
The high optical confinement offered by this WG scheme is beyond the diffraction
limit, while the propagation loss is comparatively low as compared to the Si slot WG.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp01.indd 6 9/26/23 1:17 PM

 Silicon Photonic Waveguides 7

FIGURE 1.6 The effective refractive index of HPWG.

Si photonics has turned out to be very interesting due to its manufacturing compat-
ibility with the regular complementary metal oxide semiconductor (CMOS) micro-
electronics technology. For that reason, it is exciting to realize Si-based HPWG with
basic fabrication processes.

1.2.5 Suspended Si Waveguides

The mid-infrared wavelength (2–20 µm) is interesting as it contains several so-
called water windows: wavelengths of light that can propagate through the Earth’s
atmosphere without being absorbed by water molecules. It is vastly applicable in
spectroscopy and gas-sensing applications. SOI is an attractive platform for the
realization of integrated Si photonic devices at near-infrared wavelengths; however,
it is not suitable for the devices operating in the mid-infrared region for the wave-
lengths between 2.75 and 3 µm and beyond 4 µm due to the strong absorption of
the buried oxide (BOX) layer at longer wavelengths. For that reason, other WG plat-
forms such as Si-on-sapphire, Si-on-nitride or germanium-on-Si are being studied
for mid-infrared applications. Recently, a new technique is proposed that involves
the selective removal of the BOX layer in the SOI platform, resulting in Si mem-
brane WGs as displayed in Figure 1.7(a). As a result, the operational wavelength
range of the SOI platform is stretched to cover the full Si transparency window up
to 8 µm. This allows the ease of manufacturing of such devices via standard SOI
fabrication processes.

FIGURE 1.7 (a) Cross section of suspended Si WG, (b) SEM image of S-bend WG [11].

BK-TandF-KAZANSKIY_9781032572949-230642-Chp01.indd 7 9/26/23 1:17 PM

 8 Photonics Elements for Sensing and Optical Conversions

The manufacturing of such WGs involves two etch steps: In the first step, the WG
core and an array of holes along the sides of the core are patterned and etched. In the
second step, the BOX layer is removed using a hydrofluoric acid passing through the
venting holes. The propagation loss of 3 dB/cm at 2.75 µm of such WGs is reported.
In 2016, J. Soler Penades [11] demonstrated several fundamental photonic functional
blocks such as low loss 90° bends, S-bends, 2×2 multimode interference couplers,
and MZIs based on suspended SOI WGs. An extraordinary propagation loss of as
low as 0.82 dB/cm at 3.8 µm is achieved. The SEM image of the S-bend WG is pre-
sented in Figure 1.7(b).

1.3 WAVEGUIDE COUPLING TECHNIQUES

Coupling light to an integrated optical circuit is conceptually straightforward, but,
it is a non-trivial problem. This is attributed to the small size of the optical WGs,
usually a few microns at most in any cross-sectional dimension. There are a variety
of techniques available to perform the coupling task, the most common being butt
coupling, end-fire coupling, prism coupling, and grating coupling. The end-fire and
the butt coupling are very similar, with light just shining at the end of the WG. The
difference normally made between these two approaches is that the butt coupling
entails merely butting the two devices or WGs close to each other, such that the mode
field of the transmitting device lands on the end face of the second device; whereas
the end-fire coupling uses a lens to concentrate the input beam on the end face of the
receiving device. Consequently, light is transmitted into the end of the WG and can
theoretically excite all modes of the WG. Nevertheless, prism coupling and grating
coupling are distinctly different methods, since they pass an input beam at a specific
angle through the surface of the WG. This allows for phase matching within the WG
to a particular propagation constant, thus enabling a particular mode to be excited.
Prism coupling is not particularly useful for semiconductor WG assessment.
Coupling conditions are such that the material from which the prism is produced
should have a higher refractive index than the WG; this severely limits the possibili-
ties, particularly for Si with a high refractive index of 3.5. Nevertheless, there are
materials available that could be used, such as germanium, although other draw-
backs also indicate that the technique is inferior to other techniques. Such drawbacks
are that the prism coupling will affect the WG’s formation, and it is not suitable if a
surface cladding is to be used. It is best suited to planar WGs, and it is not ideal for
material systems that utilize rib WGs such as Si technology. The remaining three
techniques can be beneficial and will be addressed in effect with Si-based technol-
ogy. Figure 1.8 illustrates the concepts of these four coupling methods.

1.3.1 Grating Coupling
Grating couplings provide a way for coupling to individual modes and are useful
for coupling to the WG of a wide range of thicknesses. Due to the need to insert the
input beam at a specific angle, grating couplers are not reliable enough for consumer
applications, but they are a powerful development tool. To couple light into a WG,
as displayed in Figure 1.8(a, b), the components of the phase velocities must be the

BK-TandF-KAZANSKIY_9781032572949-230642-Chp01.indd 8 9/26/23 1:17 PM

 Silicon Photonic Waveguides 9

FIGURE 1.8 Four light coupling techniques to optical WGs: (a) prism coupling, (b) grating
coupling, (c) butt coupling, and (d) end-fire coupling.

same in the propagation direction (z-direction). This is referred to as the state of

phase-match, and, in this case, the propagation constant must be the same in the
z-direction. Consider an incident beam on the WG’s surface at an angle θ, as dis-
played in Figure 1.9.
The beam propagates in the medium nclad with a propagation constant konclad, in
the direction of propagation. The z-directed propagation constant in the nclad medium
will therefore be as follows:

k z = ko nclad sin θ. (1.1)

Therefore, the phase-match condition will be:

β = k z = ko nclad ⋅ sin θ, (1.2)

where β is the WG propagation constant and is greater than konclad. Hence, the condi-
tion of Eq. (1.2) can never be satisfied since sinθ will be less than unity. Therefore,
a prism or grating is needed to couple light into the WG, since both can satisfy the
condition of phase-match if properly designed.
A grating is a periodic structure, and it is usual to fabricate it on the WG surface
if it is to be used as an input or output coupler. The periodic design of the grating
induces a period modulation of the effective index of the WG. For optical mode

FIGURE 1.9 Light beam incident on the surface of the WG.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp01.indd 9 9/26/23 1:17 PM

 10 Photonics Elements for Sensing and Optical Conversions

with propagation constant βWG when the grating is not present, modulation results
in a series of possible propagation constants βþ, given by:

2þπ ,
βþ = βWG + (1.3)
Λ 

where Λ is the grating period and þ = ±1, ±2, ±3, etc. Such modes are analogous to
diffraction grating orders. Obviously, in the WG, the propagation constants referring
to the positive values of þ cannot occur because the propagation constant βþ is still
less than konclad. Therefore, only the negative values of þ will result in a phase-match.
It is normal to render the grating so that only the value þ = –1 results in a phase-
match with a WG mode. Thus, the propagation constant of the WG becomes:

βþ = β WG − 2π /Λ. (1.4)

Therefore, the phase-match condition becomes:

β WG − 2π /Λ = ko nclad ⋅ sin θ. (1.5)

Expressing βWG in terms of the effective index N, Eq. (1.5) becomes:

ko N − 2π /Λ = ko nclad ⋅ sin θ, (1.6)

and on substituting for ko we obtain:

λ
Λ= . (1.7)
( N − nclad ⋅ sin θ )
Since the medium with refractive index nclad is usually air (nclad = 1), therefore
Eq (1.7) becomes:
λ
Λ= . (1.8)
( N − sin θ )
Accordingly, Eq. (1.8) can be used to calculate the grating period for the desired
input angle in the air for mode coupling with βWG propagation constant.
Surprisingly little work has been done on Si grating couplers, mainly due to pro-
duction difficulties. Since the refractive index of Si is high, the necessary grating
period for the input/output coupling is 400 nm. Ang et al. [12] obtained an output
coupling efficiency of approximately 70% for rectangular grating and 84% for non-
symmetrical profile grating [13].

1.3.2 Butt Coupling and End-Fire Coupling

The variables that affect the performance of the butt coupling and the end-fire cou-
pling are identical, so these methods will be discussed together. When the incident
beam is shined at the end-face of the optical WG, the strength with which the light is

BK-TandF-KAZANSKIY_9781032572949-230642-Chp01.indd 10 9/26/23 1:17 PM

 Silicon Photonic Waveguides 11

coupled to the WG depends on the following aspects. There may also be a numerical
aperture discrepancy where the optical WG’s input angles are not well matched to
the range of angles of excitation, but the latter concept is ignored here.

1.3.2.1 The Excitation and Waveguide Modes Match

Matching the excitation fields is generally determined by conducting the overlap
integral between the excitation field and the WG field. The most straightforward
situation is considered when both the fields are single moded. The overlap integral Γ
between the exciting field Eex and the guided field Eg is indicated as:
∞ ∞

Γ=
∫−∞
dy
∫
−∞
Eex Eg d x
1/2 , (1.9)
∞ ∞ ∞ ∞
 

 ∫ ∫
−∞ −∞
E2 dx d y
∫ ∫ −∞ −∞
Eg2 d x d y 


where the denominator is a standardizing value that guarantees the value of Γ lies
between 0 and 1. A straightforward approach to gain a reasonable idea of the over-
lap between excitation and WG fields is to use a Gaussian approximation. Eq. (1.9)
simplifies this considerably, and the result gives a good understanding of the overlap
values if the fields for both the excitation beam and the WG mode are presumed to
have Gaussian shapes, as described in the following equations:

  x 2 y2  
Eg = exp  −  2 + 2   . (1.10)
  wx w y  

This represents a WG field with 1/e widths in the x and y directions of 2wx and
2wy respectively. If we suppose a circularly symmetrical input beam, then Eex is
expressed as:

 x 2 + y2
Eex = exp  −
( )  . (1.11)
 wo2 

Utilizing the mathematical identity for a definite integral:

∞
π
∫ o
exp  −r 2 x 2  d x =
2r
. (1.12)

Eq. (1.9) reduces to express the coupling efficiency:

1/2
2  1 
 
wo  wx w y 
Γ= 1/2 1/2 .
(1.13)
 1 1   1 1 
 w2 + w2   w2 + w2 
 x o   y o 

BK-TandF-KAZANSKIY_9781032572949-230642-Chp01.indd 11 9/26/23 1:17 PM

 12 Photonics Elements for Sensing and Optical Conversions

The power coupling efficiency is given by:

4  1 
2  
wo  wx w y 
Γ2 = . (1.14)
 1 1   1 1 
 w2 + w2   w2 + w2 
 x o   y o 

1.3.2.2 The Degree of Reflection from the Waveguide Facet

The incoming confined light wave is partially transmitted and partially reflected at
the end facet of a dielectric WG, where the guided backwards traveling fraction of the
optical power strongly depends on the geometry of the WG and the refractive index
contrast. The reflection coefficient r for TE and TM polarization is given as:

n1 cos θ1 − n2 cos θ2
rTE = , (1.15)
n1 cos θ1 + n2 cos θ2

n2 cos θ1 − n1 cos θ2
rTM = . (1.16)
n2 cos θ1 + n1 cos θ2

Using Snell’s law, Eq. (1.15) reduces to:

−sin ( θ1 − θ2 )
rTE = . (1.17)
sin ( θ1 + θ2 )

Consequently, the power reflectivity is determined by R = r2. Hence the reflectivity

for TE and TM polarization is given by:

sin 2 ( θ1 − θ2 ) tan 2 ( θ1 − θ2 )
RTE = rTE
2
= and R = r 2
= .
sin 2 ( θ1 + θ2 ) tan 2 ( θ1 + θ2 )
TM TM

Both the reflectivities are the same for normal incidence, θ1 = 0. Additionally,
end-fire coupling allows light to be incident almost normal to the face. In that case:
2
n1 − n2
RTE = RTM = R = . (1.18)
n1 + n2

The reflectivity is about 31% for a Si/air interface, which causes an extra loss of
1.6 dB/facet, which is significant and is minimized using the anti-reflecting coating
in commercial devices. The coating thickness should be equal to λ/4 so that the
waves reflected from the front of the coating and the facet is in anti-phase. In this
way, the reflections are cancelled to some extent, reduced, or eliminated. It may be
shown that for normal incidence, the net reflectivity is given by:
2
n n − nar2
R= 1 2 , (1.19)
n1n2 + nar2

BK-TandF-KAZANSKIY_9781032572949-230642-Chp01.indd 12 9/26/23 1:17 PM

 Silicon Photonic Waveguides 13

where nar is the refractive index of the anti-reflection coating. The net reflectivity can
be zero, if n1n2 = nar2 .

1.3.2.3 The Quality of the Waveguide End-Face

The WG end-face quality is highly dependent on the preparation technique. The
purpose of any surface preparation technique is to create sufficiently smooth facets
to reduce optical scattering. There are three main options used for the end-face
processing of semiconductor WGs: cleaving, polishing, and etching. Cleaving is
achieved by manually placing a small crack at the edge of the sample and then
applying pressure to the sample to crack along the crystallographic plane. This
procedure is difficult to master and is limited to use in the research laboratory, as
it is unsuitable for commercial applications due to the nature of the procedure. The
results may be promising but appear to be volatile in SOI because the addition of
a buried oxide layer implies that the surface and the substrate layers of Si are not
directly connected. The most common method of preparing a WG facet is prob-
ably polishing. Lapping with abrasive materials with sequentially diminishing grit
sizes polishes the sample end-face. A specified recipe is usually followed, which
may result in an excellent surface finish, although the rounding of the end-face is
a common failure. End-faces may also be prepared using chemical or dry etching.
It goes beyond the scope of this chapter to describe how this is achieved, but it is
sufficient to say that it is a technique that can be developed to the highest level for
commercial application.

1.3.2.4 The Spatial Misalignment of the Excitation and Waveguide Fields

When the two fields are slightly misaligned, the overlap between the exciting and
receiving fields can change. The effect of spatial misalignment is calculated by mea-
suring the overlap integral described by Eq. (1.9), when the exciting and WG fields
are displaced by an amount, relative to each other. Let us assume that the WG field is
still represented by Eq. (1.10), but the exciting field has an offset of X, so:

 x 2 + ( y − X )2 
Eex = exp  . (1.20)
 wo2 

The offset is assumed to be along the y-direction, to simplify the calculation.

In the presence of misalignment, the overlap integral may now be written as:
∞ ∞


Γ ′ = exp  − 2
X 2

2 .
∫−∞
dy
∫
−∞
Eex Eg d x
1/2 , (1.21)
 w y + wo  
∞ ∞ ∞ ∞


 ∫ ∫
−∞ −∞
2
E dx d y
∫ ∫ −∞ −∞
2
E dx d y
g


 X2 
Γ ′ = exp  − 2 2 
. (1.22)
 w y + wo 

BK-TandF-KAZANSKIY_9781032572949-230642-Chp01.indd 13 9/26/23 1:17 PM

 14 Photonics Elements for Sensing and Optical Conversions

1.4 CONFINEMENT FACTOR

It is a clear fact that not all the power propagating in a WG mode stays inside the
core of the WG. It is practical to calculate the fraction of power that stays inside the
core, which enables comparisons between WG structure and modes. The field con-
finement factor (Γ) is used as a Figure of Merit (FOM) to characterize the optical
WGs to describe the interaction of light with the ambient medium, and their ability
to confine an optical mode within a certain region. Γ is defined as the ratio of the
time-averaged energy flow through the area of interest (Ainterest) to the time-averaged
energy flow through the total area, Atotal.

Γ=
∫∫ Ainterest
{ 
*

 } ,
Re  E × H  . e d x d y

Re {  E × H  . e} d x d y
(1.23)
∫∫
*

Atotal  

where E and H * are the electric and magnetic field vectors, respectively, and e is the
unit vector in the propagation direction of light. Depending on the application, the area
of interest can differ. For example, in the case of ridge and rib WG, Ainterest is Si core,
whereas, in the case of slot WG, Ainterest is a gap (air). Finally, in an HPWG, Ainterest is
a thin layer of SiO2 sandwiched between Si and Au, respectively, whereas Atotal is the
integral of energy flow in the core and cladding. The Γ of ridge, rib, slot, and HPWG
is calculated depending on geometric variables of the WGs as displayed in Figure 1.10.

FIGURE 1.10 Γ of (a) ridge WG at TE polarization, (b) ridge WG at TM polarization, (c) rib
WG at TE and TM polarization, (d) slot WG, (e) HPWG.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp01.indd 14 9/26/23 1:17 PM

 Silicon Photonic Waveguides 15

FIGURE 1.11 E-field mapping of mode in (a) ridge WG at TE00 mode, (b) TM00 mode, (c) rib
WG at TE00 mode, (d) TM00 mode, (e) slot WG, and (f) hybrid plasmonic WG.

The maximum Γ of 0.91, 0.78, 0.65, and 0.59 is obtained for ridge, rib, slot, and
HPWG, respectively. This shows that ridge and rib WGs offer a better Γ than slot
and HPWG. However, in the slot and HPWG, ~40% of the mode power is residing
in the cladding; this fact can be visualized as an opportunity to utilize these WGs in
sensor applications.
The E-field mapping of propagating mode in the WG is displayed in Figure 1.11.
The geometric variables of the WGs are labeled in the figure.

1.5 LOSSES OF SI WAVEGUIDES

Optical WGs are vulnerable to losses that occur from three sources: scatter-
ing, absorption, and radiation. The relative contribution of each of these effects is
dependent upon the WG design and the quality of the material in which the WG is
manufactured. Integrated photonics based on a heterogeneous Si platform is very
attractive as it combines the best attributes of Si and III-V platforms in the realiza-
tion of numerous high-performing active and passive optical devices on a single chip
[14, 15].

1.5.1 Scattering Loss

There are two types of scattering losses associated with optical WGs: volume scat-
tering and interface scattering. The former deals with the imperfections in the
bulk WG material, such as voids, inclusions, or crystalline defects, and the latter
is connected to the roughness at the interface between the core and the cladding
layer. Generally, volume scattering is irrelevant in a well-developed WG technology
because the material has been upgraded to an adequate point before the fabrication
process of the WGs. On the other hand, interface scattering may not be neglected,
even for relatively smooth interfaces. The SOI high index contrast platform per-
mits small WG dimensions and tight bends to realize compact photonic devices
with a small footprint. But regrettably, the high index contrast also leads to high

BK-TandF-KAZANSKIY_9781032572949-230642-Chp01.indd 15 9/26/23 1:17 PM

 16 Photonics Elements for Sensing and Optical Conversions

propagation loss primarily subjugated by scattering, which limits the performance of

the devices requiring on-chip delays larger than a few centimeters. The Si WGs are
extremely sensitive to fabrication imperfections such as sidewall roughness, which
induces high loss and phase error. Additionally, these WGs are robustly birefringent,
have considerable temperature dependence, and the chip to fiber coupling loss is
significantly high.
In Si optical WGs, the cross-sectional dimensions play an important role in deter-
mining the propagation loss of the mode. The major sources of propagation loss in
an optical WG are related to the absorption within the materials and scattering loss
from the sidewalls, etc. [16]. As demonstrated in [17–19], typical propagation loss of
single-mode ridge WGs is in the range between 0.5 dB/cm and 2 dB/cm. It is proba-
ble to accomplish low propagation loss utilizing advanced immersion 193 nm lithog-
raphy (litho), but presently it might not be as commonly available in Si photonics.
The manufacturing of slot WGs is a challenging task, as sidewall roughness can
induce high scattering loss. The lowest demonstrated losses of the conventional [20]
and segmented slot WGs [21], both fabricated with E-beam litho, are 10 dB/cm
and 4 dB/cm, respectively. Recently, A. Spott et al. [20] demonstrated the photo-
lithographically fabricated low-loss asymmetric Si slot WGs having losses as low
as 2 dB/cm. Moreover, light injection into such a nanoscale structure is a hefty job.

1.5.2 Absorption
In semiconductor WGs, absorption loss appears due to two probable sources: inter-
band absorption and free carrier absorption. Interband absorption comes into play
when photons having energy superior to the bandgap are used to stimulate the elec-
trons from the valence band to the conduction band. In Si, the band edge wave-
length is roughly 1.1 µm. Below this wavelength, Si is a perfect absorber. That is
why it is one of the most widespread materials used for photodetectors in visible
and near-infrared wavelengths. For the waveguiding purpose, the wise selection of
wavelength range is mandatory. For instance, the attenuation loss of 2.83 dB/cm and
0.004 dB/cm is demonstrated at the wavelengths of 1.15 µm and 1.52 µm, respec-
tively. This shows that interband absorption can be neglected by selecting a suitable
wavelength of operation.
Free carrier absorption is a phenomenon that deals with the transfer of photon
energy to an electron or hole in the conduction or valence band of a semiconductor,
respectively. Due to the absorption of photon energy, excited electrons and holes
transition to a state in the same or another conduction and valence band, respectively.
It is considered a parasitic process, as it hurts the performance of the devices such as
Si solar cells and thermophotovoltaic cells. Unsurprisingly, free carrier absorption
is much stronger in heavily doped semiconductors. In the case of pure Si, this effect
appears at approximately 1 µm.

1.5.3 Radiation Loss

The leakage of optical mode from the guiding layer (core) into the surrounding
region of the WG comes under the category of radiation loss. If the WG geometry is

BK-TandF-KAZANSKIY_9781032572949-230642-Chp01.indd 16 9/26/23 1:17 PM

 Silicon Photonic Waveguides 17

well optimized, this loss can be minimized. Radiation loss from a straight channel
WG should ideally be ignored. However, when an EM wave (electromagnetic wave)
propagates along with a WG with a uniform bending radius, some portion of its
energy is lost. A part of this lost energy is redistributed among the guided modes of
the WG; the residue is swiftly radiated away as tunneling leaky modes.
For WG structures as depicted in Figure 1.1, the probability of mode leakage
exists, if the thickness of the lower cladding layer is finite. In SOI-based WGs, the
BOX layer must be adequately thick, roughly 400 nm for operation in the wavelength
range of 1.3–1.6 µm to avoid optical modes from penetrating the oxide layer and
coupling to the Si substrate. As the dimension of the WG core reduces, resulting in a
less-confined mode, the requirement for a thick BOX layer increases. However, the
required BOX thickness can differ from mode to mode, as each mode penetrates
the cladding to a different depth.

1.6 MANUFACTURING OF SI WAVEGUIDES ON SILICON

ON INSULATOR PLATFORM
In the early 1960s, the semiconductor manufacturing process was originated in
Texas, and CMOS was patented by Frank Wanlass in 1963. Usually, Si WGs are
manufactured from SOI wafers, which consist of a buried oxide (BOX) layer
between the Si wafer and a thin Si layer, using standard CMOS processes. This
includes the patterning of Si via optical litho and etching techniques to form WG
structures.
There are many attractive attributes connected to the manufacturing of optical
devices on an SOI platform, such as low cost of the material, mature and well-
developed processing technique, development and manufacturing in the microelec-
tronics industry, and the prospect of easy integration with electrical components in
the same substrate [22, 23]. The BOX layer has a refractive index of 1.46, consider-
ably lower than that for the crystalline Si layer of 3.5. Therefore, this type of plat-
form setup is a conventional WG structure. In this section, we will briefly explain
the steps involved in the manufacturing of Si WGs based on the SOI platform.

1.6.1 Photolithography and Laser Lithography

Photolithography (PL) is a method for transferring patterns from a photomask to a
thin layer of photosensitive material called a photoresist (PR) that covers the sam-
ple’s surface. These patterns are only transitory features on the PR that must be
transmitted to the device’s underlying layers (Si) to gain the WG characteristics.
When dealing with PL, the photomask is an essential element of transferring pat-
terns onto PR coated samples. To provide high-quality, application-specific PL, a
variety of recipes and process modifications have been devised, but the final device
structure is dependent on the accuracy of all the fabrication processes involved [24].
PL should be done in a clean processing room because dust particles in the air might
stick to the sample or mask, causing flaws in the circuit characteristics. When a dust
particle clings to the photomask, it behaves like an opaque pattern and is transmitted

BK-TandF-KAZANSKIY_9781032572949-230642-Chp01.indd 17 9/26/23 1:17 PM

 18 Photonics Elements for Sensing and Optical Conversions

FIGURE 1.12 Dust particles contaminating the photomask patterns.

to the PR, resulting in a fault on the circuit features, as displayed in Figure 1.12. A
class 100 cleanroom is required for the manufacturing of the most integrated optical
structures, which means that the dust count must be four orders of magnitude lower
than that of regular room air. For PL, however, a class 1 or class 10 cleanroom is nec-
essary, as the dust count is even lower. UV (ultraviolet) litho, extreme UV litho, X-ray
litho, electron beam litho, and nanoimprint litho are only a few of the lithographic
methods available. The resolution limit is the key distinction between these methods.
The mask is in direct touch with the wafer in contact litho, which is also iden-
tified as contact printing. The PR is subjected to a collimated beam of UV light
for a set amount of time. This results in a resolution of ~1 micron, which is quite
high. However, there is one significant downside to this method: when the mask is
brought close to the wafer, a dust particle may become embedded in it. This dust
particle has the potential to irreversibly harm the mask. The second kind of litho
is proximity litho, often known as proximity printing, which includes a tiny space
between the photomask and the wafer during exposure, typically 10–50 microns.
Due to the optical diffraction at the feature edges, this approach is utilized to limit
the risks of photomask damage, but it reduces the resolution to a 2- to 5-micron
range. Figure 1.13 depicts a contact and proximity litho scheme.

FIGURE 1.13 (a) Contact litho, (b) proximity litho.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp01.indd 18 9/26/23 1:17 PM

 Silicon Photonic Waveguides 19

Laser litho is a maskless method that uses a focused laser beam to write com-
plicated patterns with tiny components. It offers significantly improved quality and
throughput. On nearly any material, this method can write arbitrary microstruc-
tures with minimum feature sizes as small as 100 nm. It is frequently utilized in the
manufacture and research and development (R&D) of masks and direct writing of
microstructures.

1.6.2 Wafer Preparation

Before PR spin coating, we should make sure that the wafer is free from dust particles
and has been desorbed of any moisture. Removing any traces of moisture is particu-
larly vital because wafer cleaning is performed with a wet process ending in a deion-
ized water wash and dry. A completely dry surface can be obtained by baking the
wafer at a temperature above 432 K for several minutes before the spin coating of the
PR. To improve the PR adhesion on the Si wafer, hexamethyldisilane (HMDS) is used.

1.6.3 Spin Coating and Exposure

To produce micrometer and submicrometer structures, PR is primarily employed in
microelectronics and microsystems technologies. A thin layer of PR is coated before
litho to transfer the specified patterns to the Si layer. A PR is a radiation-sensitive
chemical that may be used to create a patterned coating on the sample. A photosensi-
tive chemical, a base resin, and an organic solvent are the three components. There
are two sorts of PRs: negative and positive.
The exposed portions of a negative PR become less soluble in the developer,
resulting in the reverse patterns on the PR as shown on the mask. The underly-
ing phenomenon behind this type of feature is that following exposure, the negative
PR initiates a chemical reaction that induces polymer crosslinking. The increased
molecular weight of the crosslinked polymer causes it to become insoluble in the
developer. Consequently, as illustrated in Figure 1.14, the unexposed portions are
eliminated during the development process.
The UV-exposed areas of the positive PR, on the other hand, become soluble in
the developer and are removed. The photosensitive compound is insoluble in the
developer solution before exposure. The photosensitive compound, on the other
hand, may absorb radiations in the exposed patterns during exposure. As illustrated
in Figure 1.14, this helps to modify its chemical structure, which becomes soluble in
the developing solution and is eliminated throughout the development process.
After cleaning, the wafer is straightaway coated with PR. A few drops of resist,
approximately 1–10 mL, are applied to the center of the wafer, which is placed in a
spin coater held via a vacuum seal on a polymer chuck. The wafer is spun at a usual
speed of ~2,000–4,000 rpm for thin resists and between 800 and 2,000 rpm for
thick resists. Once the spin coating is done, a soft bake is used to get rid of most of
the solvents in the resist. Additionally, it can improve resist uniformity and bonding
to the wafer. Soft baking is carried out at 373 K for a couple of minutes. Once the
sample is ready, it is transferred to the mask-aligner where it is exposed with a UV
light passing through a permanent patterned metallic mask.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp01.indd 19 9/26/23 1:17 PM

 20 Photonics Elements for Sensing and Optical Conversions

FIGURE 1.14 Demonstration of negative and positive PRs. The UV light-exposed region of
the PR is displayed in a dark gray color, whereas the unexposed part is presented in a medium
gray color.

1.6.4 Photoresist Developing and Hard Bake

The PR patterns are formed on the wafer during the developing stage. Depending
on the type of PR used, the solution dissolves the exposed part of the resist (or unex-
posed part of resist), leaving behind the resist pattern. The final hard bake process
evaporates the residual solvents in the resist and promotes the resist adhesion to
the wafer surface. Usually, it is carried out at a temperature of 363–413 K for up to
several minutes.

1.6.5 Etching Process
Etching involves the selected and controlled removal of material from the Si wafer
using a chemically reactive or physical process. Two common methods are used to
etch Si: wet and dry etching. Each approach has its limitations, but dry etching is
preferred for the sub microns device features. Crystalline Si can be wet-etched by
using a strong aqueous alkaline media such as NaOH, KOH, or tetramethylammo-
nium hydroxide (TMAH) solutions via:

Si + 2OH + 2 H 2O → Si (OH )4 + H 2 → SiO2 (OH )2 + 2 H 2 .

2
(1.24)

Reactive ion etching (RIE) is a dry etching technique that uses a mix of chemical
and physical interactions between the etching gas and the substrate [25]. The physi-
cal mechanism involves bombarding the material with high-energy ions that break it
down, while the chemical mechanism involves induced interactions on the material

BK-TandF-KAZANSKIY_9781032572949-230642-Chp01.indd 20 9/26/23 1:17 PM

 Silicon Photonic Waveguides 21

FIGURE 1.15 Manufacturing steps of SOI WGs: (a) ridge WG, (b) rib WG.

surface with plasma species to produce volatile species. By adjusting the chemi-
cal and physical variables of the etching process, the etch rate and sidewall angle
may be adjusted. Because RIE-produced integrated optical devices have minimal
propagation losses and smooth sidewalls, the sidewall slope is an important element
in determining device performance. Some pollutants produced by the RIE method
are deemed inappropriate for microelectronic processing. For RIE of Si WGs, the
plasma is usually derived from CF4 gas, which is a stable gas but dissociates into
CF3 and F atoms in a plasma with the single fluorine atom being used as the active
etch element for both Si and Si dioxide materials. If the plasma is based on CF4, then
the etch rate is slow due to the swift recombination of CF3 and F. By adding O2 gas
in the chamber, the Si etch rate can be considerably improved due to the reaction of
oxygen with CF3. It helps in repressing F recombination and consequently increas-
ing the free F concentration. Fluorine-based chemical reactions are considered to
be a foundation of many different Si etching processes. The whole manufacturing
process of Si WGs is presented in Figure 1.15. Ridge WG is formed when the Si layer
is completely etched till the BOX layer (Figure 1.15a), whereas rib WG is produced
when the Si layer is partially etched (Figure 1.15b).

1.7 ADVANCES IN SI PHOTONICS SENSORS

There are several platforms available for the realization of photonic sensors, however
Si technology is one of the powerful and promising tools. Most of the research is
conducted on Si photonic devices for the telecommunication field and little attention
is given toward the sensing applications. Not long ago, Si photonics attracted a lot of
attention toward photonic sensing with applications in environmental, chemical, and
biomedical sensing. An important advantage of integrated photonic sensors is their
miniaturized size, mass producibility, and low cost. In Si technology, it is possible
to integrate several building blocks on the same chip, such as light generation, WGs,
light encoding, grating, detection, packaging the devices, and smart electronics to
control all the building blocks, as presented in Figure 1.16.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp01.indd 21 9/26/23 1:17 PM

 22 Photonics Elements for Sensing and Optical Conversions

FIGURE 1.16 Integration of different building blocks on a Si platform.

In principle, the light-guiding for sensing applications can be based on WGs as

presented in Section 1.2 of this chapter. However, for the realization of a complete
photonic lab-on-a-chip microsystem device, several units ought to be integrated on
the same platform, i.e. the micro/nanodevices, the flow cells and the flow delivery
system, for interferometric sensors, a phase modulation system to convert the peri-
odic output signal in direct phase measurements, integration of the light sources
and the photodetectors, and CMOS processing electronics. Here we are presenting
SOI-based sensors relying on two types of interrogation methods: wavelength inter-
rogation and intensity interrogation.

1.7.1 Wavelength Interrogation Method

For label-free detection of analytes, RRs and MZIs are highly sensitive sensors
based on the interaction between the guiding light and the analytes through its eva-
nescent field. The light–matter interaction can be enhanced by employing the slot
WGs, as the analytes can propagate in the core and can directly interact with the
strongly confined light in the low-index region. The sensor signal of these devices
results from a refractive index change of the ambient medium. For RR sensors,
the signal is a shift of the well-separated and sharp resonance peaks, while for
the MZI, it is a shift of the continuous sinusoidal transmission spectrum. SOI slot
WG-based RRs have been demonstrated to achieve more than three times larger
RR sensitivities compared to conventional ridge WG-based RRs. The history of
MZI goes back to 1891 with the pioneering work of Zehnder [26] and Mach [27]
on free-space optics interferometers, which can be used for sensing macroscopic
samples. Integrated photonics versions of MZIs have also been realized in photonic
crystals, using self-collimation of light, where the ratio of the two MZI outputs is
the sensing signal.
Refractive index sensors have several applications in the field of biochemical
domain and have been intensively studied in recent years, such as solution concen-
tration and pH value, which can be measured on the basis of refractive index change
[28, 29]. A semiconductor nanowire sensor for biosensing offers a sensitivity as high
as 235 nm/RIU (refractive index unit) [30]. It has been established that the highest

BK-TandF-KAZANSKIY_9781032572949-230642-Chp01.indd 22 9/26/23 1:17 PM

 Silicon Photonic Waveguides 23

experimental sensitivity for quasi-TM RRs was 135 nm/RIU [31]. However, by opti-
mizing the WG thickness, the bulk sensitivity of 270 nm/RIU was demonstrated
[32]. This value is within 90% of the maximum achievable sensitivity for a resonator
sensor using a ridge WG.
Here, the sensing capabilities of ridge, slot, and a Si-based HPWG that relies on
wavelength interrogation method are presented. For this purpose, we simulated RR
designs based on the aforementioned WG structures, as presented in Figure 1.17
(first row). The geometric variables of the three designs are kept constant to perform
a reasonable comparison for their spectral behavior. The device height (HSi) and the
gap (g) between the bus WG and the ring are fixed at 220 nm and 10 nm, respectively.
The ring radius is R = 1500 nm (from the outer edge), which is similar for all three
designs. The width of WGs used in RRs is varied between 300 and 350 nm. For
HPWG RR, the Si circular ridge encircles a gold disk in the center. The slot width
(s) of 40 nm is used for RRs based on slot and HPWG. And the material filled in
the slot is air (n = 1.0). Figure 1.17 (second and third rows) presents the E-field map-
ping and transmission spectrum of a ridge, slot, and HPWG RR at their respective
λres. The measurement of changes in λres is the most common interrogation method
used in RRs. The optical resonances are obtained by loading the ambient medium
with a material of n = 1.03 for all three designs. The transmission spectrum and
E-field mapping are simulated using the finite element method. From the analysis
of slot and HPWGs presented in Section 1.2, we can expect that RRs based on these

FIGURE 1.17 RR design based on (a) ridge WG, (b) slot WG, (c) HPWG. E-field mapping
and transmission spectrum of respective RR designs are presented in the second and third
row, respectively.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp01.indd 23 9/26/23 1:17 PM

 24 Photonics Elements for Sensing and Optical Conversions

FIGURE 1.18 (a) The Figure of Merit, (b) sensitivity comparison.

WGs can enhance the evanescent field as compared to a ridge WG RR having the
same design variables. High sensitivity is always desirable in these sensors, which
strongly depends on light polarization, optical loss, and the light–matter interaction.
Consequently, elevated sensitivity can be expected [33–35].
In this case, sensitivity is calculated as S = Δλ/Δn, where Δλ represents the change
of the resonance wavelength in nm and Δn is the difference of the refractive index
in the medium. The FOM is another variable that should also be thought of while
constructing the RR sensor. FOM is expressed as S/FWHM. In Figure 1.18, the
FOM and S of RRs are plotted for the width of the WG for Δn = 0.03. The HPWG
RR offers maximum S as compared to the other two designs at W = 300 nm (where
W = Wo and Wa for slot and ridge/HPWG, respectively), which deteriorate fast as W
increases. The Smax drops from 333.33 nm/RIU to 170 nm/RIU when W changes
from 300 to 350 nm. This fact signifies that the hybrid plasmonic ring WG RR has
less fabrication tolerance as compared to the other two ring schemes. In terms of
FWHM, the slot RR has the lowest value, 2.66 nm at W = 350 nm, whereas the ridge
WG RR has broad, FWHM = 6.65 nm at W = 300 nm. Even though the HPWG
RR has the highest S, its broad FWHM makes its FOM lower than the slot WG RR
design. As we have mentioned earlier, FOM is the ratio between S and FWHM;
therefore, a design with a narrow FWHM can have a larger FOM, as presented in
Figure 1.18. The FOMmax is approximately 75.2, 58.5, and 40.4 for a slot, hybrid, and
ridge WG RR, respectively.

1.7.2 Intensity Interrogation
The ambient refractive index can alter the transmission power of the propagating
mode in the WG. Here, we have proposed a transmission decay (dB) of all the pro-
posed WGs at 1,550 nm when the ambient refractive index is varied between 1 and
1.35, as presented in Figure 1.19. The propagation length of the WGs is fixed at 3 µm.
The power decay in the transmission for all the WGs under consideration is pre-
sented in Table 1.1, where HPWG offers the highest power decay of 2.45 dB when
the ambient refractive index is changed from 1 to 1.35.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp01.indd 24 9/26/23 1:17 PM

 Silicon Photonic Waveguides 25

FIGURE 1.19 Transmission versus ambient refractive index in (a) ridge WG, (b) rib WG,
(c) slot WG, (d) HPWG.

TABLE 1.1
Comparison of Transmission (dB) and S of Rib, Ridge, Slot, and HPWG
Rib WG Ridge WG Slot WG HPWG
TE00 TM00 TE00 TM00 TE TM hybrid
Type of WG mode mode mode mode mode mode
Transmission (dB) 0.127 0.016 0.234 0.216 0.75 2.45
Sensitivity (dB/RIU) 0.36 0.045 0.67 0.617 2.14 7

The sensitivity is calculated using ΔT(dB)/Δn, where ΔT and Δn are the change
in transmission power in dB and the change in ambient refractive index, respec-
tively. The maximum sensitivity of rib, ridge, slot, and hybrid WG is obtained at
0.36 dB/RIU, 0.67 dB/RIU, 2.14 dB/RIU, and 7 dB/RIU, respectively. This undoubt-
edly displays the domination of slot and Si-based-HPWG for sensing applications
due to their high light–matter interaction.

1.8 NUMERICAL SIMULATION OF SI RIDGE WAVEGUIDE

BY COMBINING A THIN METAL FILM
In Section 1.2, we discussed a few highly recognized Si-based optical WGs that can
be used as optical interconnects as well as in sensor applications. We present a new

BK-TandF-KAZANSKIY_9781032572949-230642-Chp01.indd 25 9/26/23 1:17 PM

 26 Photonics Elements for Sensing and Optical Conversions

metal-supported Si ridge WG structure that may be utilized in gas-sensing applica-

tions. When compared to slot WG, this method provides a high evanescent field ratio
(EFR), as well as simplicity of manufacturing and reduced complexity. The ability
of slot WG to restrict light outside of the WG core material has piqued interest in
chemical and biological sensing applications. Though, in comparison to ridge WG,
these structures have a higher susceptibility to sidewall roughness-generated scatter-
ing loss, which raises questions about their efficacy. A thin metal layer is introduced
as an under cladding in the suggested WG design to restrict light confinement in the
substrate area, while the energy is raised in the upper cladding (hereafter represented
as UC) region, enhancing the light–matter interaction. Using the 3D-FEM, Γ, power
distribution, and propagation loss of the suggested structure and the conventional Si
ridge WG are calculated. In comparison to the conventional ridge WG, numerical
studies show that the suggested WG formation can achieve significant light confine-
ment in the UC (upper cladding) area.

1.8.1 Design of a Metal-Supported Si Ridge Waveguide

Figure 1.20 depicts the schematics of a conventional Si ridge WG (Subsection 1.2.1)
and a metal-supported Si ridge WG. When compared to silver (Ag), gold (Au) is
utilized in this design because of its biocompatibility and resistance to oxidation. A
thin SiO2 layer is inserted between Au and Si ridge WG in a metal-supported WG
arrangement. Furthermore, this formation is compatible with current CMOS manu-
facturing methods and may be built in the following steps. On a Si wafer, an Au layer
is deposited, followed by a thin coating of SiO2. Finally, the Si layer is placed on the
surface and etched to create the ridge WG. The layer thickness of Au and SiO2 is rep-
resented by HAu and HSiO2, respectively. In our example, HAu = 100 nm was constant,
while HSiO2 was varied to improve the evanescent field in the UC area. HSi and WSi
are the height and width of the ridge WG, respectively.
Figure 1.21 reveals the E-field mapping of a TE mode in the conventional Si
ridge WG and metal-supported Si ridge WG at λ = 1550 nm, where HSi = 220 nm,
WSi = 400 nm, HSiO2 = 50 nm, and HAu = 100 nm. The conventional ridge WG has a
symmetric evanescent field, which is why a significant portion of light also dwells in
the substrate, as shown in Figure 1.21(a), but the metal-supported ridge WG lifts the
evanescent field toward the top cladding region, as seen in Figure 1.21(b). This prop-
erty makes it easier to improve light–matter interaction, which is useful in gas and
biochemical applications.

FIGURE 1.20 Schematic of Si ridge WG: (a) conventional design. (b) metal-supported
design.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp01.indd 26 9/26/23 1:17 PM

 Silicon Photonic Waveguides 27

FIGURE 1.21 E-field mapping: (a) conventional WG design, (b) metal-supported WG

design [40].

The sensitivity of the device is proportional to Γ, and a significant amount of light

confinement in the UC improves the sensitivity. When the optical confinement is
substantial, the concentration variation of a cover medium, for example an aqueous
solution, has a significant impact on the effective refractive index. As a result, EFR
is a crucial characteristic in the development of gas sensors established on evanes-
cent field absorption. Sensors having a high EFR can intermingle with the absorbing
medium more, improving the S of the sensor [36–39].
Figure 1.22 shows the power distribution in the conventional and metal-supported
ridge core, cladding, and substrate for TE mode at 1,550 nm. The height and width of

FIGURE 1.22 The power distribution of the conventional (a) and metal-supported Si ridge
WG (b–d) at 1,550 nm, where WSi = 400 nm and HSi = 220 nm [40].

BK-TandF-KAZANSKIY_9781032572949-230642-Chp01.indd 27 9/26/23 1:17 PM

 28 Photonics Elements for Sensing and Optical Conversions

the ridge WG are set at 200 nm and 400 nm, respectively, for both the WG configura-
tions. Figure 1.22(a) shows that the maximum power in the UC for the conventional
WG is 34.27%, whereas 18.31% power leaks to the substrate. The power distribution
in a metal-supported WG is determined by the thickness of the SiO2 layer inserted
between the Au and Si ridge WG. For instance, when HSiO2 = 50 nm, the maximum
power of 44% is obtained in the UC and only 9.77% power goes to the substrate, as
shown in Figure 1.22(b). With an expansion in the SiO2 cladding layer, the power in
the UC drops (Figure 1.22c, d).
As a result, using a metal layer lowers the light leakage to the substrate. The light
confinement in the suggested structure’s UC region is greater than that of the con-
ventional arrangement. The metal’s high reflection allows for strong confinement to
be achieved. By changing the WG’s size, as discussed in the following section, one
may build the WG for several combinations of power distribution.

1.8.2 The Connection between the Waveguide Dimension and Γ

To address the impact of light confinement in the cladding area, the WG dimensions
and distance between the WG core and the Au layer are critical factors. As a result,
the impact of both these variables is explored. The Γ in WG cladding, core, and
substrate is shown versus HSiO2 in Figure 1.23(a–c) for thickness range of 50 nm to
100 nm, respectively.
At HSiO2 = 50 nm, the cladding area achieves the highest confinement. Because
these WGs are intended for TE polarization, the Γ is heavily influenced by the WG
width (WSi). To investigate the power distribution in the WG, we set HSi at 220 nm

FIGURE 1.23 Influence of SiO2 layer on Γ of (a) cladding region, (b) core, and (c) sub-
strate [40].

BK-TandF-KAZANSKIY_9781032572949-230642-Chp01.indd 28 9/26/23 1:17 PM

 Silicon Photonic Waveguides 29

and changed WSi to between 350 nm and 400 nm. For two reasons, the optical Γ in
the cladding area decreases when HSiO2 and WSi increase: (1) due to the large WG
size, the TE mode is more restricted in the core region, and (2) the influence of metal
reflection diminishes, as illustrated in Figure 1.23(a). Furthermore, it is worth noting
that metal does not affect the light Γ in the core (Figure 1.23b). When WSi is small,
it even amplifies a few percentage points of light in the core. In metal-supported
WGs, the majority of the light from the substrate moved toward UC, as illustrated in
Figure 1.23(c). As a result, we can claim that for conventional and metal-supported
WG, 10% power is improved in the UC at any given dimensions.

1.8.3 Influence of Metal Layer on the Propagation Loss

As the metal is incorporated into the suggested arrangement, it is necessary to evalu-
ate the losses caused by the metal’s optical absorption. When estimating the propa-
gation loss, the WG dimensions, SiO2 layer thickness, and Au thickness must all
be considered. The following equation is used to determine the propagation loss:
10 × log(Pout/Pin)/propagation length in microns, where Pin and Pout are the input
and output power of the WG, respectively. The propagation loss of the conventional
and metal-supported ridge WG are computed for various values of HAu and HSiO2, as
shown in Figure 1.24.
As previously stated, HAu = 100 nm is constant, but HSiO2 ranged from 50–100 nm
with a step size of 25 nm. The propagation loss tends to decrease when HSiO2 grows
from 50 nm to 100 nm, as seen in Figure 1.24. This is because the metal layer is well
separated from the WG core, and as a result, the metal’s absorption is reduced. It
can, however, impact the WG’s EFR. As a result, there is always a trade-off between
propagation loss and EFR. We also discovered that HAu had the smallest impact on
losses. As a result, a thin layer of Au, such as 25 nm, may be employed in the sensor
design from an inexpensive standpoint.

FIGURE 1.24 The propagation loss of conventional and metal-supported ridge WG [40].

BK-TandF-KAZANSKIY_9781032572949-230642-Chp01.indd 29 9/26/23 1:17 PM

 30 Photonics Elements for Sensing and Optical Conversions

1.9 CONCLUDING REMARKS

This chapter has discussed several attributes of Si photonics that can be fabricated
by direct patterning of Si to create optical interconnects to broadcast data-carrying
laser signals. The biggest advantage is that it can transmit gigantic data while con-
suming low power with no heating up or causing any deterioration in the signal. The
modal characteristics are studied via the finite element method (FEM), which give
a general guideline of different types of WGs. The effective refractive index, Γ, and
losses are thoroughly studied. Based on their analysis, one can choose an appropriate
WG arrangement for on-chip communication and sensing applications. The manu-
facturing of optical elements is also discussed in detail, which provides a general
idea of the technology used for this platform. There are many attractive attributes
connected to the manufacturing of optical devices on an SOI platform, such as low
cost of the material; mature and well-developed processing technique, development,
and manufacturing in the microelectronics industry; and the prospect for easy inte-
gration with electrical components in the same substrate. The sensing performance
of the aforementioned WGs is presented, which shows the supremacy of slot and
Si-based HPWG over conventional WG formations. At the end of the chapter, a novel
WG scheme is proposed that utilizes a metal layer buried under an oxide layer that
separates metal from the WG core. This scheme suppresses the light confinement in
the substrate region, whereas the mode is elevated in the UC region, which enhances
the light–matter interaction. This kind of WG scheme can be used as a highly sensi-
tive evanescent field gas absorption sensor.

REFERENCES
1. Zhang, Y. et al. 2018. High transparent mid-infrared silicon “window” decorated with
amorphous photonic structures fabricated by facile phase separation. Opt. Express.
26(14):18734–48. https://doi.org/10.1364/OE.26.018734.
2. Reed, G. T., and Knights, A. P. 2008. Silicon photonics: The state of the art. Wiley.
ISBN: 978-0-470-02579-6.
3. Abel, S. et al. 2019. Large Pockels effect in micro-and nanostructured barium titanate
integrated on silicon. Nat. Mater. 18:42–7. https://doi.org/10.1038/s41563-018-0208-0.
4. Dong, P. et al. 2010. Low loss shallow-ridge silicon waveguides. Opt. Express.
18(14):14474–9. https://doi.org/10.1364/OE.18.014474.
5. Butt, M. A., Kozlova, E. S., and Khonina, S. N. 2017. Conditions of a single-mode
rib channel waveguide based on dielectric TiO2/SiO2. Computer Optics. 41(4):494–8.
https://doi.org/10.18287/2412-6179-2016-40-4-494-498.
6. Soref, R. A., Schmidtchen, J., and Petermann, K. 1991. Large single-mode rib wave-
guides in GeSi-Si and Si-on-SiO2. IEEE J. Quantum Electron. 27(8):1971–4. https://
doi.org/10.1109/3.83406.
7. Almeida, V. R., Xu, Q., Barrios, C. A., and Lipson, M. 2004. Guiding and confin-
ing light in void nanostructure. Opt. Lett. 29(11):1209–11. https://doi.org/10.1364/
OL.29.001209.
8. Wang, X. et al. 2013. Silicon photonic slot waveguide Bragg gratings and resonators.
Opt. Express. 21(16):19029–39. https://doi.org/10.1364/OE.21.019029.
9. He, X. et al. 2018. Ultralow loss graphene-based hybrid plasmonic waveguide with deep-
subwavelength confinement. Opt. Express. 26(8):10109–18. https://doi.org/10.1364/
OE.26.010109.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp01.indd 30 9/26/23 1:17 PM

 Silicon Photonic Waveguides 31

10. Zenin, V. A. et al. 2017. Hybrid plasmonic waveguides formed by metal coating of dielec-
tric ridges. Opt. Express. 25(11):12295–302. https://doi.org/10.1364/OE.25.012295.
11. Penades, J. S. et al. 2016. Suspended silicon mid-infrared waveguide devices with sub-
wavelength grating metamaterial cladding. Opt. Express. 24(20):22908–16. https://doi.
org/10.1364/OE.24.022908.
12. Ang, T. W., Reed, G. T., and Vonsovici, A. et al. 1999. Grating couplers using silicon on
insulator. Proc. SPIE. 3620:79–86.
13. Ang, T. W., Reed, G. T., and Vonsovici, A. et al. 1999. Blazed-grating coupler in uni-
bond SOI. Proc. SPIE. 3896:360–8. https://doi.org/10.1117/12.370333.
14. Pollock, C. R., and Lipson, M. 2003. Integrated photonics. New York: Springer
Science+Business Media. ISBN: 978-1-4419-5398-8.
15. Liu, J.-M. 2005. Photonic devices. Cambridge: Cambridge University Press. ISBN:
978-0-521-55195-3.
16. Rickman, A. G., Reed, G. T., and Namavar, F. 1994. Silicon-on-insulator optical rib
waveguide loss and mode characteristics. J. Lightw. Technol. 12(10):1771–6. https://doi.
org/10.1109/50.337489.
17. Selvaraja, S. K. et al. 2014. Highly uniform and low-loss passive silicon photonics
devices using a 300 mm CMOS platform. Optical Fiber Communication Conference:
Th2A.33. https://doi.org/10.1364/OFC.2014.Th2A.33.
18. Kobayashi, N., Sato, K., Namiwaka, M., Yamamoto, K., Watanabe, S., Kita, T.,
Yamada, H., and Yamazaki, H. 2015. Silicon photonic hybrid ring-filter external cav-
ity wavelength tunable lasers. J. Lightw. Technol. 33(6):1241–6. https://doi.org/10.1109/
JLT.2014.2385106.
19. Gao, F., Wang, Y., Cao, G., Jia, X., and Zhang, F. 2005. Improvement of sidewall sur-
face roughness in silicon-on-insulator rib waveguides. Appl. Phys. B. 81:691–4. https://
doi.org/10.1007/s00340-005-1951-x.
20. Baehr-Jones, T., Hochberg, M., Walker, C., and Scherer, A. 2005. High-q optical reso-
nators in silicon-on-insulator based slot waveguides. Appl. Phys. Lett. 86(8):081101.
https://doi.org/10.1063/1.1871360.
21. Wang, G., Baehr-Jones, T., Hochberg, M., and Scherer, A. 2007. Design and fabrica-
tion of segmented, slotted waveguides for electro-optic modulation. Appl. Phys. Lett.
91(14):143109. https://doi.org/10.1063/1.2793618.
22. Spott, A. et al. 2011. Photolithographically fabricated low-loss asymmetric silicon slot
waveguides. Opt. Express. 19(11):10950–8. https://doi.org/10.1364/OE.19.010950.
23. Zhang, L. 2014. Silicon process and manufacturing technology evolution: An overview
of advancements in chip making. IEEE Consum. Electron. Mag. 3(3):44–8. https://doi.
org/10.1109/MCE.2014.2317896.
24. Tiginyanu, I., Ursaki, V., and Popa, V. 2011. Nanocoatings and ultra-thin films:
Technologies and applications. Sawston, Cambridge: Woodhead Publishing Limited.
ISBN: 978-1-84569-812-6.
25. Jansen, H. et al. 1996. A survey on the reactive ion etching of silicon in microtechnol-
ogy. J. Micromech. Microeng. 6:14–28. https://doi.org/10.1088/0960-1317/6/1/002.
26. Zehnder, L. 1891. Ein neuer Interferenzrefraktur. Z. Instumentenkd. 11:275–85.
27. Mach, L. 1892. Uber einen Interferenzfractur. Z. Instrumentenkd. 12:89–93.
28. Vaiano, P. et al. 2016. Lab on fiber technology for biological sensing applications. Laser
Photonics Rev. 10(6):922–61. https://doi.org/10.1002/lpor.201600111.
29. Guo, T. 2017. Fiber grating-assisted surface plasmon resonance for biochemical and
electrochemical sensing. J. Lightw. Technol. 35(16):3323–33. https://doi.org/10.1109/
JLT.2016.2590879.
30. Wang, Y. L. et al. 2017. Refractive index sensor based on leaky resonant scattering of
single semiconductor nanowire. ACS Photonics. 4(3):688–94. https://doi.org/10.1021/
acsphotonics.7b00064.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp01.indd 31 9/26/23 1:17 PM

 32 Photonics Elements for Sensing and Optical Conversions

31. Xu, D. X. et al. 2010. Real-time cancellation of temperature induced resonance shifts
in SOI wire waveguide ring resonator label-free biosensor arrays. Opt. Express.
18(22):22867–79. https://doi.org/10.1364/OE.18.022867.
32. Fard, S. T., Schmidt, S., Shi, W., and Wu, W. X. 2017. Sensitivity of silicon-on-insulator
(SOI) strip waveguide resonator sensor. Biomed. Opt. Express. 8(2):500–11. https://doi.
org/10.1364/BOE.8.000500.
33. Butt, M. A., Khonina, S. N., and Kazanskiy, N. L. 2019. A Serially cascaded micro-ring
resonator for simultaneous detection of multiple analytes. Laser Phys. 29(4):046208.
https://doi.org/10.1088/1555-6611/ab0371.
34. Butt, M. A., Khonina, S. N., and Kazanskiy, N. L. 2018. Hybrid plasmonic waveguide-
assisted metal-insulator-metal ring resonator for refractive index sensing. J. Mod. Opt.
65(9):1135–40. https://doi.org/10.1080/09500340.2018.1427290.
35. Butt, M. A., Khonina, S. N., and Kazanskiy, N. L. 2018. Highly sensitive refrac-
tive index sensor based on hybrid plasmonic waveguide microring resonator. Waves
Random Complex Media. 30(2):292–9. https://doi.org/10.1080/17455030.2018.1506191.
36. Butt, M. A., Khonina, S. N., and Kazanskiy, N. L. 2018. Silicon on silicon dioxide slot
waveguide evanescent field gas absorption sensor. J. Mod. Opt. 65(2):174–8. https://doi.
org/10.1080/09500340.2017.1382596.
37. Butt, M. A., Khonina, S. N., and Kazanskiy, N. L. 2018. Modelling of Rib channel
waveguides based on silicon-on-sapphire at 4.67 µm wavelength for evanescent field
gas absorption sensor. Optik. 168:692–7. https://doi.org/10.1016/j.ijleo.2018.04.134.
38. Butt, M. A., Degtyarev, S. A., Khonina, S. N., and Kazanskiy, N. L. 2017. An evanescent
field absorption gas sensor at mid-IR 3.39 µm wavelength. J. Mod. Opt. 64(18):1892–7.
https://doi.org/10.1080/09500340.2017.1325947.
39. Butt, M. A., Khonina, S. N., and Kazanskiy, N. L. 2019. Enhancement of evanes-
cent field ratio in a silicon strip waveguide by incorporating a thin film. Laser Phys.
29(7):076202. https://doi.org/10.1088/1555-6611/ab1414.
40. Butt, M. A., Khonina, S. N., and Kazanskiy, N. L. 2019. Sensitivity enhancement of
silicon strip waveguide ring resonator by incorporating a thin metal film. IEEE Sens. J.
20(3):1355–62. https://doi.org/10.1109/JSEN.2019.2944391.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp01.indd 32 9/26/23 1:17 PM

 2 Photonic Crystal
Cavities in Integrated
On-Chip Optical Signal
Processing Components
Pavel Grigoryevich Serafimovich1 and
Nikolay Lvovich Kazanskiy1,2
1I mage Processing Systems Institute – Branch of the
Federal Scientific Research Centre “Crystallography and
Photonics” of Russian Academy of Sciences, Samara, Russia
2 Samara National Research University, Samara, Russia

2.1 INTRODUCTION
Optical signal processing components, such as optical modulators, sensors, all-
optical integrators, and differentiators implemented on a chip, are important for
developing computer technology. All-optical, fully integrated, on-chip computing
components will increase the speed of information processing by several orders of
magnitude. Moreover, such components enable the processing of not only real but
also complex values. In this regard, it is important to implement the basic computing
operations optically. In recent years, all-optical integrators and differentiators based
on Bragg gratings [1] and ring resonators [2] have been proposed. Such elements
can be used in both digital and analog signal processing. Among the digital signal-
processing applications are the use of optical integrators and differentiators as pulse
counters and ultrafast memory elements [3]. Analog signal-processing applications
include all-optical solution of differential equations of various orders [4].
Integrators and differentiators based on Bragg gratings are a few millimeters
in size. Integrators and differentiators based on ring resonators are more compact.
Their size is on the order of tens of micrometers on the chip plane. Photonic crystal
structures are currently the subject of intensive research [5–7]. In this chapter, we
numerically describe and study the most compact optical integrators and differentia-
tors based on photonic crystal (hereafter represented as PC) cavities [8].
The model of a two-component nanocavity is also described. In this model, the
minimum details of the structure are found only in the periodic component of the
resonator. The advantages of such a structure include a promising way to construct
an electrically pumped photonic cavity, the ease of introducing nonlinear optical
materials in the nanocavity, the possibility of formation of the desired energy dis-
tribution in the far zone, and the possibility to develop dynamic systems based on
nanocavities.

DOI: 10.1201/9781003439165-2 33

BK-TandF-KAZANSKIY_9781032572949-230642-Chp02.indd 33 9/26/23 4:51 PM

 34 Photonics Elements for Sensing and Optical Conversions

The optical modulator is one the most important optical signal processing
components. In recent years, these devices have been significantly improved [9].
Nevertheless, some problems remain. In particular, the size of the device and the
structural complexity must be reduced to provide functional flexibility and enable
effective electrical modulation. Photonic crystal structures are widely used as
platforms for implementing various nanophotonic elements [10]. Several optical
modulators based on a single PC cavity structure have been proposed, including
high-frequency modulating PC lasers [11], thermal tuning PC modulators [12], and
optical [13, 14] and electrical [15] pumping PC modulators. An electro-optic modu-
lator constructed with two-dimensional (2D) PC slab cavity has been described in
[16]. A two-component structure with several 2D PC slab cavities has been used in
[17] to increase the frequency bandwidth of the device by employing wavelength
division multiplexing. Electro-optic modulators built with a PC nanobeam cavity
have been demonstrated in [18, 19]. The switching energy is an important character-
istic of any electro-optic modulator. The value of this energy depends, in particular,
on the resistance and the capacitance of the device. Reducing the footprint of the
modulator decreases the resistance and capacitance values and therefore diminishes
the total energy consumption. The footprint of the PC nanobeam cavity is, as a rule,
several times smaller than the one of the 2D PC slab cavity. Thus, a PC nanobeam
cavity has the potential to exhibit reduced switching energy compared to its 2D PC
counterpart [19].
Today, much attention attracts the developing of the refractive index sensors
based on optical microcavities [20–22]. Such sensors are used, for example, in bio-
logical [23] and chemical investigations [24], when measuring the temperature and
mechanical stresses in acoustics. The values of the Q-factor of the cavities at the
level of 106 and higher are reached. The application of the active cavities with light
and electron pumping is the way to increase the values of the Q-factor. The high
values of the Q-factors of the optical cavities ensure the high sensitivity of the sen-
sors on their basis.
Moreover, the special features of the resonant sensors’ functioning offer the pos-
sibility to obtain the dynamic response in real time, as well as the wide range of the
examined samples. The samples, for example, can be different liquid or gaseous
chemical compounds or biomolecules. The optical resonant sensors demonstrate
good compatibility with the requirements of the microfluidics. All these make them
suitable for using as a part of integrated solutions in the forms of biochips or “lab-
on-a-chip.” Another advantage of the optical resonant sensors is the eliminating of
the need to use the fluorescent tags.

2.2 TEMPORAL INTEGRATION OF OPTICAL SIGNALS

Figure 2.1 shows a scheme of the coupled-resonator optical waveguide (CROW).
The variable ai, i = [1, N] is the complex amplitude of the resonant mode in the i-th
resonator; κi–1 and κi, i = [1, N], are the left and right coupling coefficients of the
i-th resonator, respectively; ri, i = [1, N] is the energy loss of the i-th resonator to
the exterior space; and pin, prf, and ptr are the amplitude of the input, reflected, and
transmitted fields, respectively.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp02.indd 34 9/26/23 4:51 PM

 Photonic Crystal Cavities 35

FIGURE 2.1 Scheme of coupled-resonator optical waveguide.

Consider an array in which the resonators have identical resonant frequencies.

Then, according to the temporal coupled-mode theory, the transmission function of
the system can be written as:

2(−i)
N −1
ptr κ 0κ N κ1κ 2 κ N −1
TN ( s ) ≡ =− , (2.1)
pin det (M)

where M is the corresponding tridiagonal matrix [9],

ptr = − i 2κ N aN = −2 κ 0κ N  M −1  N ,1 pin ,

and det (M) is the determinant of the matrix M.

For N = 1, Eq. (2.1) is reduced to the form

2κ 0
T1 ( s ) = − . (2.2)
s + 2κ 0

Hereafter, for simplicity, we neglect the losses to the exterior space.

Let us consider how accurately Eq. (2.2) approximates the integrator of the first
order. The polarized electric field with envelope Pin(t) can be written as:

E ( x , t ) = Pin ( t − x /vg ) exp ( im0 x − iω 0t ) =

∞
(2.3)
=
∫ R (ω − ω ) exp ( im (ω ) x − iωt ) dω,
−∞
0

where R(ω) is the envelope spectrum signal, m(ω) is the wave number [m0 = m(ω)],
and vg is the group velocity.
A linear system described by the complex transfer function (TF) H(ω) converts
the envelope of the input pulse [Eq. (2.3)] to

Ptr ( t ) =
∫ R (ω ) H (ω ) exp ( iωt ) dω = P (t ) ∗ h (t ) ,
−∞
in (2.4)

where the symbol “*” denotes the convolution operation, and h(t) is the spectrum of
the TF H(ω).

BK-TandF-KAZANSKIY_9781032572949-230642-Chp02.indd 35 9/26/23 4:51 PM

 36 Photonics Elements for Sensing and Optical Conversions

The impulse response of a linear system with the TF T1(s) is:

h1 ( t ) = −κ 0 exp ( −κ 0t ) u ( t ) , (2.5)

where u(t) is the Heaviside step function.

Substituting Eq. (2.5) into Eq. (2.4), we obtain an expression for the envelope of
the output pulse:
t

∫
Ptr ( t ) = −κ 0 Pin ( T ) exp ( − iκ 0 ( t − T )) dT .
−∞
(2.6)

The right side of this equation expresses the integral of the input pulse envelope with
exponential weight.
Figure 2.2 shows the result of integration of the envelope of an optical pulse with
a duration of 100 ps by resonators with Q-factors of 3 × 104 and 5 × 104. The Q-factor
is related to κ0 by the ratio Q = ω0/(4κ0). Figure 2.2 shows that the higher the Q-factor
of the resonator is, the more slowly the integrated signal envelope decays. For reso-
nators with Q-factors of 3 × 104 and 5 × 104 we estimate an integration time window
(defined as the decay time required to reach 80% of the maximum intensity) of 12.5 ps
and 19.5 ps, respectively.
For N = 2, Eq. (2.1) can be written as

i2κ 0κ1
T2 ( s ) = . (2.7)
( s + κ 0 )2 + κ12
Let us calculate the parameters of the PC nanobeam cavity that integrates the
optical signal. Compared with the resonators in the 2D PC layer [25], PC nanobeam
cavities [26] have a smaller area and are naturally integrated into the waveguide
geometry of the chip.

FIGURE 2.2 Result of first-order integration of optical pulse with duration of 1 ps by reso-
nators with Q-factors of 3 × 104 and 5 × 104.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp02.indd 36 9/26/23 4:51 PM

 Photonic Crystal Cavities 37

FIGURE 2.3 Schemes of (a) PC nanobeam cavity and (b) array of two such cavities.

Figure 2.3(a) illustrates a possible embodiment of a resonator based on a PC

nanobeam. The decreasing radius of holes in the tapering region forms a defect in
which the resonant mode is excited. An array of two PC resonators is shown in
Figure 2.3(b), where ntap is the number of holes in the tapering region. The coupling
value between resonators in the array is determined by nreg, the number of holes with
the maximum radius between defects.
The resonance cavity characteristics were computed using the parallel three-
dimensional (3D) finite-difference time-domain method. The waveguide in our sim-
ulations has a width of 490 nm and a height of 220 nm. It is composed of silicon (Si)
and deposited on silica substrate. Air-filled holes in the regular part of the waveguide
have a radius of 100 nm and are spaced 330 nm apart. The lattice parameters and
the radii of the holes near the defect (ntap = 12) are (in nm): a1 = 40, b1 = 255, a2 = 55,
b2 = 350, a3 = 65, b3 = 365, a4 = 75, b4 = 375, a5 = 85, b5 = 385, a6 = 95, b6 = 395. These
parameters demonstrate the existence of an energy bandgap for transverse electric
polarization in the waveguide. The left and right borders of the bandgap are 1.46 µm
and 1.67 µm, respectively. The size of the bandgap is about 210 nm. The resonant
mode wavelength (1.57 µm) is placed in the center of this region. The free spectral
range (FSR) of the cavity can be estimated to 100 nm (~12.5 THz). The Q-factor of
the resonator is 3.6 × 104, and nreg = 6. The integration time window for the first-
order integrator based on such cavity is about 14 ps. Figure 2.7 shows the frequency
response of the integrator.
The time-bandwidth product (TBP) of the integrator can be defined as the mul-
tiplication of the integration time window (in ns units) and the FSR of the cavity (in
GHz units) [27]. Thus, the TBP of the calculated integrator is 0.014 × 12,500 = 175.
That value is comparable to the TBP of the passive integrators based on Bragg grat-
ing [11] and ring resonator [2]. The suggested integrator based on PC cavity has a
large potential of increasing of the achieved TBP value. At least two methods can be
proposed. Both of them are related to increasing of Q-factor of the resonator while
the size of the bandgap remains the same. The first one is based on the fine-tuning of
the tapering region and increasing of nreg. By this way, the Q-factor value achieved
in this work can be enhanced by several orders of magnitude [28]. Another way is to
use an active cavity [29]. The development of the active photonic crystal cavities with
the optical and electrical pumping is discussed in Sections 2.2 and 2.3, respectively.
Figure 2.4(a) shows the result of integration of the first derivative of a Gaussian
pulse with a duration of 150 fs. The Q-factor of the resonator is 3.6 × 104, and nreg = 6.
The result of integration of the second derivative of a pulse with the same duration by

BK-TandF-KAZANSKIY_9781032572949-230642-Chp02.indd 37 9/26/23 4:51 PM

 38 Photonics Elements for Sensing and Optical Conversions

FIGURE 2.4 Results of integration of corresponding derivatives of Gaussian pulse with

duration of 150 fs by (a) one PC resonator and (b) an array of two PC resonators.

second-order integrator is shown in Figure 2.4(b). The second-order integrator con-

sists of two resonators, as shown in Figure 2.3(b). The coupling coefficients between
the resonators in each integrator are given by Eq. (2.8).
PC cavities can also be used as optical differentiators. This application is
described in [30].

2.3 IN-PLANE OPTICAL PUMPING OF PHOTONIC

CRYSTAL CAVITY
The cavity’s Q-factor can be defined as:
E0
Q ≡ 2π , (2.8)
E1
where E0 is the quantity of energy in the cavity and E1 is the quantity of energy lost
per oscillation. The cavity’s mode volume Vm is defined as [31]:
3
∫ dVε E 2  nmax 
Vm ≡ 2   , (2.9)
ε max max  E   λ 

BK-TandF-KAZANSKIY_9781032572949-230642-Chp02.indd 38 9/26/23 4:51 PM

 Photonic Crystal Cavities 39

where the integration is performed over the cavity volume; εmax and nmax, respec-
tively, denote the permittivity and the refractive index at the point of the maximal
value of the square of the electric field modulus |E|2; and λ is the wavelength of light.
In an optical cavity, light is confined in the form of one or several resonance modes.
Note that the nonlinear optical transformations require that the mode spatial overlap
should be maximal. For two orthogonally polarized resonance modes, the coefficient
of spatial overlap of orthogonal resonance modes can be derived from the formula:

ε NL dV
∫ ∑ i , j ,i ≠ j
ETE,i ETM, j
γ≡ NL
, (2.10)
∫ ∫
2 2
dVε ETE dVε ETM

where ∫ NL denotes the integration over the volume of a nonlinear material in the
resonance cavity. ETE,i, ETM,j are the electric field components of the orthogonal reso-
nance modes.
Two orthogonally polarized modes can be excited in such a nanocavity, with the
modes overlap coefficient γ found in the range 0.76–0.78 [32]. The disadvantages of
the nanocavity include a large waveguide thickness (about three operating wave-
lengths) and a complicated process of independently tuning the frequencies of two
resonance modes.
In such a structure, there are two separate light input channels for two orthogo-
nally polarized resonance modes, thereby simplifying the implementation of optical
switches [16]. Note that the waveguide thickness is not larger than a quarter of the
operating wavelength. Besides, the structure under discussion provides for the flexible
tuning of frequencies corresponding to two orthogonal resonance modes. However,
the overlap coefficient γ, defined according to Eq. (2.3), is not larger than 0.07.
Two general approaches can be used to construct nonlinear optical elements. For
structures studied in [32, 33] the PC waveguide itself should be fabricated from an
optically nonlinear material to employ nonlinear effects. Under another approach,
nonlinear materials are introduced into the cavity [34, 35] and its shape is optimized.
In this chapter, we use the latter approach and propose a resonance cavity that has
the coefficient γ [33].
In the ridge PC waveguides, the total internal reflection (TIR) impedes the propa-
gation of light in the transverse directions. As far as the longitudinal direction is
concerned, it is the PC that ensures the propagation of light in the nanocavity [36].
To design an optical nanocavity with high Q-factor, we employ a structure com-
posed of three parts (Figure 2.5a). First, it includes PC mirrors containing iden-
tical equidistant holes in the waveguides. The operating wavelength is assumed
to be 1.55 µm. The crossing waveguides each have width w = 530 nm and height
h = 227 nm. The waveguide holes of radius 95 nm are air-filled and located 365 nm
apart. Such geometric parameters provide the formation of a bandgap for the TE and
TM waves in the horizontal and vertical waveguides, respectively. The region under
simulation contains five holes in each of four cavity arms. The cavity is made from
Si (n = 3.48) placed in the ambient air.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp02.indd 39 9/26/23 4:51 PM

 40 Photonics Elements for Sensing and Optical Conversions

  
FIGURE 2.5 Schematic structure of the optical nanocavity under study (a) and the distribu-
tion of the E-field modulus for one of the orthogonal modes (b).

TABLE 2.1
Positions and Radii of the Holes in the Transition Region for
a Basic Resonance Cavity (Nm)
a1 b1 a2 b2 a3 b3 a4 b4 a5 b5
50 285 55 400 65 400 75 400 85 390

The second system’s component is a transition region between the PC waveguide

and the resonance cavity region, which serves to reduce the energy loss in the cav-
ity. The holes of designed radii are arranged in it in such a manner as to provide the
maximal Q-factor. The transition region contains five air-filled holes. The holes’
radii and spacing between them are given in Table 2.1. In Table 2.1, the hole nearest
to the cavity of radius a1 is offset from the cavity by the distance b1 (Figure 2.5a).
The third system’s component is the proper resonance cavity. An optical reso-
nance cavity like that shown in Figures 2.5(a) and 2.6(a) [33]. What we have changed
in the cavity under study is the waveguide’s geometric parameters and the holes’
shape (which were assumed to be rectangular in [33]).
The resonance cavity characteristics were calculated using the parallel FDTD
(finite difference time domain) method [37]. Absorbing layers were put at the bound-
aries of the 3D region under calculation. The computational mesh spacing was cho-
sen in such a way as to obtain a converging solution.
Figure 2.5(b) shows the distribution of the E-field modulus for one of the orthogo-
nal modes. The cavity’s Q-factor is ~9,000.
Figure 2.6(a) depicts the configuration of the basic resonance cavity. A detailed
pattern of the E-field modulus of the resonance mode is depicted in Figure 2.6(b).
With the geometry of the horizontal and vertical waveguides being identical, it is
only possible to generate the degenerate resonance modes in the form of TE and TM
waves. However, as reported [33], the use of waveguides with different geometry
enables the excitation of nondegenerate resonance modes. For the degenerate reso-
nance mode shown in Figures 2.5(b) and 2.6(b), the value of γ is 0.03, which is equal
to that reported in [33].

BK-TandF-KAZANSKIY_9781032572949-230642-Chp02.indd 40 9/26/23 4:51 PM

 Photonic Crystal Cavities 41

  

FIGURE 2.6 (a) A detailed scheme of the basic resonance cavity and (b) the pattern of the
E-field modulus for the resonance mode.

The resonance cavity regions that contain the nonlinear material are character-
ized by a smaller refractive index than that of the waveguide. With this approach,
the electromagnetic waves can efficiently be concentrated in low refractive-index
regions of the resonance cavity. This results in the enhanced interaction of light with
the filling material of the said regions. In addition, the waveguide internal energy
loss, e.g. due to two-photon light absorption in Si, is also reduced.
Next, the resonator shown in Figure 2.7(a) is investigated. There are four holes
of radius 100 nm in the resonance cavity’s central region. The diagonal distance
between the holes is 245 nm. The holes’ positions and radii have been calculated
from the condition of the maximal Q-factor. Table 2.2 gives the positions and radii
of the transition region holes.

  

FIGURE 2.7 (a) Schematic of a resonance cavity with circular holes and (b) the pattern of
the resonance mode E-field modulus.

TABLE 2.2
Positions and Radii of the Transition Region Holes for the
Cavity in Figure 2.3(a) (Nm)
a1 b1 a2 b2 a3 b3 a4 b4 a5 b5
50 245 55 395 65 395 75 395 85 390

BK-TandF-KAZANSKIY_9781032572949-230642-Chp02.indd 41 9/26/23 4:51 PM

 42 Photonics Elements for Sensing and Optical Conversions

  

FIGURE 2.8 (a) Schematic of the slit resonance cavity and (b) the pattern of the E-field
modulus of a resonance mode.

Shown in Figure 2.7(b) is the pattern of the E-field modulus for one of the orthogo-
nal modes. The mode energy is seen to be concentrated in the cavity’s central region,
which is where the holes are located. The Q-factor is ~5,000. The value of γ is 0.05.
Let us consider the resonance cavity with a different shape of holes in the form of
slits, as shown in Figure 2.8(a). Using a slit resonance cavity [38], the E-field within
2
the slit can be enhanced by the value of nwg /nsl2 , where nwg is the waveguide refractive
index and nsl is the slit-filling material refractive index.
To make use of the enhanced E-field in the slit resonance cavity, four rectangular
holes are made in it, as shown in Figure 2.8(a). For each rectangular slit, the length is
given by sw and the width is sh. The diagonal distance between the slits is g = 245 nm.
Table 2.2 gives the positions and radii of the slits of the resonance cavity.
Shown in Figure 2.8(b) is the pattern of the E-field modulus for one of the orthog-
onal modes for the rectangular slits of size 160 nm × 115 nm, g = 210 nm. The slits’
size and positions have been calculated from the condition of the maximal Q-factor.
The cavity Q-factor is ~8,000. The value of γ is 0.15.
Changes in the slit width affect the cavity’s Q-factor, the mode overlap coefficient,
and the mode volume. Shown in Figure 2.9(a) are the distributions of the E-field
modulus |E| along a line that passes through the cavity’s center and is rotated by 45°
relative to the waveguide axis for several slit widths. The peak of the field amplitude
is achieved when the slit width is small. Figure 2.9(b) shows in which way the cavity’s
Q-factor and the mode volume depend on the slit width.

FIGURE 2.9 The distribution of the E-field modulus for (a) a resonance mode on the cav-
ity’s central diagonal for several slit widths and (b) the cavity’s Q-factor and mode volume
for several slit widths.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp02.indd 42 9/26/23 4:51 PM

 Photonic Crystal Cavities 43

  

FIGURE 2.10 The distribution of the E-field modulus for a resonance mode along the cen-
tral diagonal of the cavity plane for the slits of width 120 nm and height of (a) 240 nm and
(b) 180 nm.

Figure 2.10 shows that it is possible to reduce the transverse losses in the cavity.
Up to this point in the work, the slit height in the cavity has been assumed to be
fixed and equal to the waveguide width. With decreasing slits’ height, they are trans-
formed into intra-waveguide hollows. Although these structures are more difficult
to fabricate, such hollows allow the cavity’s Q-factor to be increased and the mode
volume to be made smaller.
Figure 2.10(a) shows the distribution of the E-field modulus |E| in a plane that
passes through the cavity’s center and is rotated by 45° relative to the waveguide axis
for the slits that go as far as the entire waveguide length of 225 nm. Figure 2.10(b)
shows the distribution of |E| for the slits of height 170 nm.
For the through-length slits in Figure 2.10(a), the coefficient γ is 0.15, the Q-factor
is 7,800, and the mode volume is 0.32. For the closed slits in Figure 2.10(b), these
values are, respectively, 0.14, 7,900, and 0.29.
Figure 2.11 illustrates the simulation results when the slits are filled with some
material, e.g. chalcogenide glass (CG) [36]. The chalcogenide glasses have a refrac-
tive index ranging from 2.3 to 2.8 at wavelength 1.5 µm. Note that the CGs have the
optical nonlinearity factor by three orders of magnitude higher than for Si, the low-
level two-photon absorption, and fast response time (<100 fs).

  

FIGURE 2.11 (a) The distribution of the E-field modulus for a resonance mode on the cen-
tral diagonal of the cavity plane. The slit material refractive index is 2.5, slit width is 120 nm,
slit height is 240 nm. (b) The plots for the cavity’s Q-factor and mode volume at different
values of the slit-filling material refractive index.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp02.indd 43 9/26/23 4:51 PM

 44 Photonics Elements for Sensing and Optical Conversions

Figure 2.11(a) depicts a cross section analogous to those shown in Figure 2.6 for
the through-cavity holes filled with the material of refractive index n = 2.5. When
compared with Figure 2.6(a), the simulated mode volume is larger by a factor of two.
Q-factor is also increased slightly. Figure 2.11(b) shows the plots for the Q-factor and
mode volume against the slit-filling material refractive index.

2.4 ELECTRICAL PUMPING OF PHOTONIC CRYSTAL CAVITY

Most of the existing technologies in use to create high-Q PC nanocavities suggest
fine-tuning of the resonance chamber geometry by changing the parameters of the
PC. Such parameters may be, for example, the radius of the hole in the PC period
and/or the hole periodic spacing. To simplify the solutions for the problems cited, the
authors theoretically investigated a two-component PC cavity. The first component
of such a cavity is a periodic structure based on a PC nanobeam. Compared with the
2D structure based on a PC slab, the area of PC nanobeam is smaller and is naturally
integrated into the waveguide geometry of connections on a chip. The second com-
ponent is a fragment of a complementary material that occupies an area of several
lattice constant of the PC. The shape and size of the fragment were determined from
the given parameters of PC cavity. While combining the two components, a defect
forms in the resulting nanostructure. The resonant mode of the corresponding fre-
quency can be excited in this defect.
The proposed approach to creating two-component PC cavities is illustrated
through the structure shown in Figure 2.12. The first component of the resonator was
a PC nanobeam. The nanobeam was made of Si and was placed on a silica substrate.
The holes in the nanobeam were of the same radius; they were equidistant from
each other and filled with air. PC nanobeam parameters are given in the caption
to Figure 2.12. With these parameters, PC band gap was created for TE-dominant
polarization radiation in the wavelength range of 1.4–1.7 μm. The second component
of the nanocavity was an elliptical piece of Si, placed on a silica substrate.

FIGURE 2.12 The geometry of the resonator calculated by (a) top view and (b) side view.
PC nanobeam (n = 3.46) lies on the substrate (n = 1.45). PC nanobeam width is d = 0.5 μm,
thickness t w = 0.26 μm. Circular holes have a radius of R = 75 nm and are filled with air;
distance between holes a = 0.34 μm. The elliptical shape (ellipse parameters A and B) (n =
3.46) lies on the substrate (n = 1.45). Thickness of ellipse te = 100 nm.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp02.indd 44 9/26/23 4:51 PM

 Photonic Crystal Cavities 45

Therefore, both the cavity’s components have a structure of silicon on insulator

(SOI) wafers. The surface roughness of Si wafers can be as low as several hundred
picometers at 1–300 μm length scales. This makes it possible to tightly combine two
Si surfaces of wafers, as shown in Figure 2.12.
In subsequent calculations, the thicknesses of the PC nanobeam and elliptical
defect were assumed to be 260 nm and 100 nm, respectively. These thicknesses pro-
duced an optimal FF (fill factor) change in the cavity. Increasing the thickness of the
nanobeam necessitates increase in the thickness of the elliptical defect.
The resonance cavity characteristics were computed using the parallel FDTD
method. In particular, the cavity (Q = 3.05 × 104) was calculated with the parameters
of the ellipse A = 6.8 μm (20 holes below the ellipse) and B = 0.5. To achieve a high-Q
nanocavity, five additional holes were placed in the PC nanobeam at both ends of
the ellipse. Thus, the total length of the cavity was (20 + 5 × 2) × 0.34 = 10.2 μm.
Figure 2.13(a) shows the distribution of Hz in the vertical plane passing through the
axis of the nanobeam. There is some vertical asymmetry of the resonance mode
due to flow of energy in the elliptical defect. Figure 2.13(b) shows the distribution in
the horizontal plane, just above the elliptical Si fragment (in silica). Hz values along
the intersection line of these two planes are represented by the dotted line in the
graph of Figure 2.13(c), and the values directly below the PC nanobeam (in silica) by
the dashed line on the same chart. The solid line represents the function cos(πx/a)
exp(–σx2) with σ = 0.23, a = 0.34 microns. Good agreement between the distribu-
tions of Hz and an analytic function demonstrates that the shape of the resonant
mode’s envelope is Gaussian. Assuming a linear dependence of γ on x, the relation
γ(x) = a/π∫σdx ≈ x/40 can be obtained. In [39], quadratic tapering of PC nanobeam
width was used to form the defect. In the paper [28], for a nanocavity with a length
of 60 periods of the PC, the relation γ(x) ≈ x/120 was implemented. Thus, it can be

FIGURE 2.13 (a) The distribution of Hz in the vertical plane passing through the axis of the
waveguide, (b) the distribution of Hz in the horizontal plane just above the elliptical fragment
(in quartz), (c) the dotted line – Hz values along the line of intersection of the planes (a) and
(b) the dashed line – Hz values just below PC nanobeam (in quartz), the solid line – function
cos(πx/a)exp(–σx2) for σ = 0.23, a = 0.34 μm.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp02.indd 45 9/26/23 4:51 PM

 46 Photonics Elements for Sensing and Optical Conversions

FIGURE 2.14 An example of geometry (not to scale) for P-type (left) and N-type (right)
doping regions.

concluded that the two techniques used in creating a defect are almost equivalent.
The nanocavity with an elliptical defect is three times shorter than the one with vari-
able nanobeam width. Accordingly, the rate of change γ in the nanocavity with an
elliptical defect is three times faster [40].
The two-component nanocavity proposed in this chapter has in our opinion two
main advantages when compared to existing solutions. First, the proposed structure
allows for the development of an integrated on-chip light source with vertical elec-
trical pumping. Integrated on-chip light-emitting diodes with a laterally doped p-i-n
structure, based on the nanobeam PC cavity, were demonstrated recently. Electron
beam lithography steps can be used to implant N- and P-type dopants to the first
and second components of the structure, respectively. Figure 2.14 shows an example
of geometry for P-type and N-type doping regions. Such geometry makes it pos-
sible to focus current flow to the active region of the cavity, thereby, in comparison
with lateral electrical pumping, improving efficiency and reducing threshold. The
P-type parts that adjoin the elliptical defect have a small intersection with the reso-
nance mode. Therefore, Q-factor suffered low degradation, especially in case of B >
d (Figure 2.12a).
Using hybrid metal/PC nanocavities is another possible approach to realize verti-
cal electrical pumping. The two-component structure of the cavity assumes addi-
tional flexibility in choice of electrical current pathways. Although Q-factor in this
case can be reduced to several hundreds, this could be enough for development of an
optical amplifier integrated on-chip.
The second advantage is that the creation of nanocavities with nonlinear
properties is simplified. The supplementary component of the structure can be
used to bring a nonlinear or optically active material directly into the nanocav-
ity region.

2.5 OPTICAL MODULATOR BASED ON COUPLED PC CAVITIES

We describe a compact optical modulator based on two PC nanobeam cavities cou-
pled by a waveguide. Compared with previously proposed single-cavity modula-
tors there are two potential advantages of the suggested modulator. First, the shift
range of the modulator frequency is scalable and depends on switching energy level.
Second, it is possible to modulate pumping in the low-intensity region of the resonant
mode, allowing for effectively electrically controlled modulation.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp02.indd 46 9/26/23 4:51 PM

 Photonic Crystal Cavities 47

FIGURE 2.15 Scheme of two resonators coupled through a waveguide.

Figure 2.15 shows the general scheme of two resonators coupled by a waveguide.
The variable κj, j = {1,2} is the coupling coefficient of the j-th resonator of the wave-
guide; rj, j = {1,2} is the energy loss of the j-th resonator to the exterior space; and pin,
prf, and ptr are the amplitudes of the input, reflected, and transmitted fields, respec-
tively. To simplify the notation, we characterize the resonators with the inverse val-
ues of the resonator lifetime parameters.
The system depicted in Figure 2.15 may be interpreted as a Fabry–Perot (FP)
etalon. The transmission function of the mirrors in the etalon is defined by the trans-
mission function of the single resonator. According to temporal coupled-mode the-
ory, the transmission function of the single resonator can be written as:

2κ j
T1 ( s j ) = , s j = i(ω − ω j ) + rj , j = {1,2} , (2.11)
s j + 2κ j

where ωj , j = {1,2} is the resonant frequency of the j-th resonator. The Q-factor
of the j-th resonator Q j is given by 1/Q j = 1/Q jκ + 1/Q jr, where Q jκ = ωj /(4κj) and
Q jr = ωj /(2rj).
The transmission function of an FP etalon, with mirror transmission function T1
and reflection function (1 – T1), can be written as:

ptr T1 ( s1 ) T1 ( s2 ) iθ
T2 ( s1 , s2 ) ≡ = , (2.12)
pin 1 − (1 − T1 ( s1 ))(1 − T1 ( s2 )) i 2θ

where θ is the phase shift in the waveguide connecting the resonators. We neglect
the waveguide dispersion. Equation (2.12) describes the FP etalon with two direct-
coupling resonators. A similar structure but with two side-coupling resonators in
the 2D PC slab was considered in [41] and the effect of electromagnetically induced
transparency (EIT) was observed. This effect is missed in the case of the FP etalon
with direct-coupling resonators.
The dashed line in Figure 2.16 shows the transmission function of a single resona-
tor, T1. The Q-factor of this resonator is Q = 1,200; the resonance wavelength is λ0 =
1.488 μm. The energy loss of the resonator is determined by the value of Qr = 3.2 ×
105. The solid line in Figure 2.16 shows the transmission function of two such reso-
nators coupled by a waveguide with a phase shift of θ = π/2. This phase shift corre-
sponds to the situation in which the length of the connecting waveguide is equal to 0.
Dynamic adjustment of the variable θ can be implemented by changing the
refractive index of the waveguide, e.g. by thermal tuning or free carrier generation.
Free carrier generation provided dynamic control of the PC cavity Q-factor in [12].

BK-TandF-KAZANSKIY_9781032572949-230642-Chp02.indd 47 9/26/23 4:51 PM

 48 Photonics Elements for Sensing and Optical Conversions

FIGURE 2.16 The transmission function of one cavity (dashed line) and two coupled reso-
nators with θ = π/2 (solid line).

Pulsed optical pumping was used to change the phase shift in a Si PC waveguide
over the range of 0 to π. Free carrier generation can provide an ultrafast nonlinear
optical tuning on the order of 10 s of GHz [14].
The optical modulator proposed here does not require such a large range of phase
shifts. Figure 2.17(b) shows the transmission function of the two resonators with
θ1 = θ0 = π/50 (solid line) and θ2 = θ0 + θs = π/50 + π/250 (dashed line). It can be seen

FIGURE 2.17 (a) The transmission functions of two resonators with phase shifts in the
waveguide of π/25 (solid line) and π/25 + π/125 (dashed line). (b) The transmission functions
of two resonators with phase shifts in the waveguide of π/50 (solid line) and π/50 + π/250
(dashed line). (c) The transmission function of two mismatched resonators (resonance wave-
lengths of 1.486 μm and 1.488 μm) with phase shifts in the waveguide of π/50 (solid line)
and π/50 + π/250 (dashed line). (d) The transmission function of two mismatched resonators
(resonance wavelengths of 1.486 μm and 1.488 μm) with phase shifts in the waveguide of
π/25 (solid line) and π/25 + π/125 (dashed line).

BK-TandF-KAZANSKIY_9781032572949-230642-Chp02.indd 48 9/26/23 4:51 PM

 Photonic Crystal Cavities 49

that for values of θ π / 2, the two degenerate modes of the resonators split. The width
of the mode splitting equals 10 nm for θ0 = π/50. In addition, shifting the resonant
mode by one line width requires an additional phase shift of only θs = π/250.
Figure 2.17(a) shows the transmission function of two resonators with θ1 = π/25
(solid line) and θ2 = π/25 + π/125 (dotted line). The width of the mode splitting
equals 5 nm. Thus, the proposed optical modulator can scale the resonant frequency
shift range using the initial pumping energy, which corresponds to the initial phase
shift θ0 [14]. Thermal pumping can be used for fine-tuning θ0. The timescale for this
process is on the order of microseconds. The scaling feature of this optical modula-
tor can be used to increase the frequency range of optical communication systems,
optical sensors, and spectrometers integrated on a chip [42, 43].
In the next step we explore how the resonance mode wavelength shift affects
device function. Such a shift can appear due to the errors in PC resonator manufac-
turing. Figure 2.18 demonstrates the case of θ = π/2. A mismatch between resonant
modes wavelengths leads to a significant change in the transmission function. The
solid line in Figure 2.18 shows the transmission function of two resonators with
degenerate resonant modes at a wavelength λ0 = 1.488 μm, i.e. no mode wavelength
mismatch. If the resonance mode wavelength of one of the resonators is shifted by
1 nm, such that λ0 = 1.487 μm, then the transmission function undergoes significant
changes (dashed line in Figure 2.4). These changes become even greater if the mis-
match of the resonant modes is 2 nm (λ0 = 1.486 μm), as shown by the dotted line
in Figure 2.18.
Figures 2.17(c) and (d) demonstrate the stability of the transmission function to
resonance wavelength shifts of mismatched resonators coupled by θ sufficiently less
than π/2. The parameters of the coupled resonators in Figures 2.17(c) and (d) are the
same as those in Figures 2.3(a) and (b), except for the resonator resonant frequen-
cies. In Figures 2.17(c) and (d), the resonant frequencies are tuned to 1.488 μm and
1.486 μm in the first and second resonators, respectively. Errors in PC resonator
manufacturing usually produce some shift of the resonant frequency. Thus, the pro-
posed optical modulator allows for high tolerance in manufacturing. The differences

FIGURE 2.18 The transmission function of two resonators with resonance wavelengths
1.488 μm and 1.488 μm (solid line), 1.487 μm and 1.488 μm (dashed line), and 1.486 μm and
1.488 μm (dotted line).

BK-TandF-KAZANSKIY_9781032572949-230642-Chp02.indd 49 9/26/23 4:51 PM

 50 Photonics Elements for Sensing and Optical Conversions

FIGURE 2.19 (a) Scheme of PC nanobeam cavity. (b) Scheme of array of two PC nano-
beam cavities. (c) The electromagnetic field distribution of the “basic” resonant mode (λ0 =
1.488 μm). (d) The electromagnetic field distribution of the “shifted” resonant mode (λ0 =
1.492 μm). (e) The FDTD calculation of the transmission function of two resonators coupled
by a waveguide of length d = 490 nm (solid line) and the transmission function of the tuned
modulator; the figure inset shows the tuned parts of the waveguide (gray rectangles).

between the transmission functions in Figures 2.17(a) and (b) and Figures 2.17(c) and
(d) are the smaller for smaller phase shift θ. This result can be explained by using
theory of FP etalon transmission.
Figure 2.19(a) illustrates a possible embodiment of a resonator based on PC nano-
beam cavities. Compared with resonators in a 2D PC slab [8], PC nanobeam cavi-
ties have a smaller area and are naturally integrated into the waveguide geometry
of the chip. In the tapering region, the radii of the holes decrease, forming a defect
in which the resonant mode is excited. An array of two PC resonators is shown in
Figure 2.19(b), where ntap is the number of holes in the tapering region. The coupling
between resonators in the array is determined by nreg, the number of holes in the
regular part of the PC waveguide, and d, the length of the coupling waveguide. Let us
calculate the parameters of the particular PC nanobeam cavities that would realize
our suggested optical modulator.
The 2D FDTD method is used to calculate the characteristics of the resonator.
The waveguide has a width w = 500 nm and an index of refraction n = 2.97. This
value corresponds to the effective refractive index of a Si waveguide with thick-
ness of 270 nm, which lies on a glass substrate and is surrounded by air. Air-filled
holes in the non-tapering part of the waveguide have a radius of 93 nm and are
spaced 350 nm apart. Table 2.3 lists the lattice parameters and the radii of the
holes near the defect (ntap = 4). These parameters demonstrate the existence of an
energy bandgap for transverse electric polarization in the waveguide. For n reg = 4,

BK-TandF-KAZANSKIY_9781032572949-230642-Chp02.indd 50 9/26/23 4:51 PM

 Photonic Crystal Cavities 51

TABLE 2.3
Geometric Parameters of PC Resonator Shown
in Figure 2.19(a)
a1 b1 a2 b2 a3 b3 a4 b4
65 200 80 290 90 310 93 304

the resonance wavelength is 1.488 μm and Q = 1,230. The energy loss of the resona-
tor is Qr = 3.19 × 105.
Several methods can be used to tune the phase shift θ, e.g. optical and electrical
pumping used in [13] and [15], respectively, to generate free carriers. In both cases,
the pumping is performed in the cavity region, where the amplitude of the reso-
nant mode is maximal. Increasing the pumping energy sharply reduces the value of
Q-factor [13, 15]. Thus, shifting the resonant wavelength by greater than one line
width becomes problematic.
Next, we analyze the electromagnetic field distribution in the suggested modu-
lator. Figure 2.19(c) shows the amplitude of the “basic” resonant mode (λ 0 =
1.488 μm). Figure 2.19(d) demonstrates the amplitude of the “shifted” resonant
mode (λ0 = 1.492 μm). The resonators share two regions of low amplitude of both
resonant modes (denoted in Figures 2.19(c) and (d) by gray trapezoids). Increasing
the value of n reg can increase the size of those regions. The overlapping of the
contacts and the doped areas with the field of the resonant mode increases the
optical absorption and reduces the Q-factor. In previously suggested electro-optic
modulators, the doped areas were placed near regions of the maximal amplitude
of resonant modes. Increasing the optical absorption in those regions drastically
deteriorates the Q-factor of the cavity. There are at least two approaches to resolve
this problem. First, the side slabs were used in [18, 19] to separate the PC nano-
beam cavity and the doped areas. The thickness and lateral size of those slabs are
the trade-off between the high Q-factor and the low switching energy. Second, a
two-component parabolic PC nanobeam cavity was suggested in [40] to separate
the region of the maximal amplitude of resonant mode and the doped areas. Both
approaches can be used with the modulator described in this work. The regions of
low resonant modes amplitude represent the part of the device that can be electri-
cally tuned without deteriorating the Q-factor of the resonance modes. Electrodes
providing the electrical pumping may be placed near the pumping area. Figure
2.19(e) shows the transmission function of two resonators coupled by a waveguide of
length d = 490 nm (solid line). The dashed line in the same figure represents the
transmission function of the tuned modulator. The tuned parts of the waveguide
are shown in the inset of Figure 2.19(e) by the gray rectangles. The refractive index
of the tuned parts of the waveguide equals to n = 2.96. The length of each rectangle
equals to four-unit cell size of the PC waveguide lattice (4 × 350 nm). The position
of the tuned “shifted” resonant mode on Figure 2.19(e) corresponds to the one on
Figure 2.17(a) (d = 489 nm). The position of the tuned “basic” resonant mode on
Figure 2.19(e) is shifted to the short wavelengths due to the PC waveguide refrac-
tive index decreasing near the central part of the device.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp02.indd 51 9/26/23 4:51 PM

 52 Photonics Elements for Sensing and Optical Conversions

In our following numerical investigation, we model phase shift θ by adjusting

d, the length of the connecting waveguide, instead of adjusting the PC waveguide
refractive index. Let us find the correspondence between θ and d. The wavelength
of the resonant mode in the waveguide is 1,448/2.97 = 500 nm. Then, the addi-
tional phase shift θs = π/2 corresponds to additional waveguide length d = 500/4 =
125 nm. The initial phase shift θ0 = π/2 corresponds to an initial waveguide length
d = 350 nm. However, the results of FDTD simulation coincide with the analytical
relation Eq. (2.12) with θ = π/2 when d = 370 nm. This discrepancy may be because
the Bragg frequency of the PC waveguide does not coincide with the resonance
mode frequency. In this case, to realize the condition θ0 = π/2, it is necessary to add
some additional waveguide fragment between the resonators [44].
Figure 2.20(a) shows the transmission function of two resonators coupled by
a waveguide of length d = 490 nm (solid line) or d = 489 nm (dashed line). These
parameters correspond approximately to the parameters of the analytic curves in
Figure 2.17(a). The magnitude of the resonance mode splitting in the FDTD model
was also about 10 nm. FDTD modeling confirms that the displacement of the spec-
tral peak width at half-maximum occurs with a phase shift θs = π/2/125 = π/250.
Figure 2.20(b) demonstrates the transmission function of two resonators coupled
by a waveguide of length d = 484 nm (solid line) or d = 480 nm (dashed line). This
matches the parameters of Figure 2.17(b). Resonance mode splitting is about 5 nm.

FIGURE 2.20 (a) The FDTD calculation of the transmission function of two resonators
with connecting waveguide lengths of d = 490 nm (solid line) and d = 489 nm (dashed line).
(b) The FDTD calculation of the transmission function of two resonators with connecting
waveguide lengths of d = 484 nm (solid line) and d = 480 nm (dashed line). (c) The FDTD cal-
culation of the transmission function of two mismatched resonators (resonance wavelengths
at 1.486 μm and 1.488 μm) with connecting waveguide lengths of d = 490 nm (solid line)
and d = 489 nm (dashed line). (d) The FDTD calculation of the transmission function of two
mismatched resonators (resonance wavelengths at 1.486 μm and 1.488 μm) with connecting
waveguide lengths of d = 484 nm (solid line) and d = 480 nm (dashed line).

BK-TandF-KAZANSKIY_9781032572949-230642-Chp02.indd 52 9/26/23 4:51 PM

 Photonic Crystal Cavities 53

Decreasing θ0 decreases the maximum of “shifted” spectral peaks, as seen in

Figures 2.20(a) and (b). This is due to mismatched modes in the PC and connecting
waveguide sections. This situation could be enhanced, e.g. by tapering hole radii
and/or adjusting hole spacing in the central part of the FP etalon [45].
Thus, FDTD modeling confirms the ability of our designed optical modulator
to scale the additional phase shift θs. The value of θs necessary to shift the spec-
tral peak width at half-maximum depends on the initial value of θ0. The theoretical
minimum value of θs is determined by the free spectral zone, which is given by the
size of the band gap for PC resonators. In this chapter, we assume resonators with
free spectral bandwidths of about 300 nm. Therefore, to shift the resonance mode
by the peak width at half-maximum requires providing θ0 of about 10 −5. Practically,
before operating at such small θ0 values, the decreasing of “shifted” spectral peaks
should be diminished.
Finally, we investigated the impact of manufacturing errors on the proposed
design of the optical modulator. We modify the resonant wavelength of one of the
two resonators by reducing the radii of the cavity central holes pair from 65 nm to
63.5 nm; the resonant wavelength redshifts by about 2 nm. Thus, the optical modula-
tor consists of two mismatched resonators with resonant wavelengths of 1.486 μm
and 1.488 μm, respectively. Figure 2.20(c) shows the transmission function of two
mismatched resonators with d = 490 nm (solid line) and d = 489 nm (dashed line).
The transmission function in Figure 2.20(c) is only slightly distorted in compari-
son with Figure 2.20(a). A similar trend can be seen comparing Figure 2.20(b) and
2.20(d) for d = 484 nm (solid line) and d = 480 nm (dashed line).

2.6 APPLICATION OF PHOTONIC CRYSTAL COUPLED CAVITIES

FOR INCREASE IN SENSITIVITY OF OPTICAL SENSOR
There are examples of application of the optical cavities of different kinds in the
refractive index sensors. Among them, the sensors based on the spherical cavities
were suggested in [45]. The disk cavities used as the components of the optical sen-
sors are discussed in paper [46].
Perhaps, the refractive index sensors based on the ring cavities are the most fre-
quently used solution [47]. They are easily integrated on the crystal and the technol-
ogy of their manufacturing is well developed. The optical sensors based on the PC
cavities are discussed in paper [48].
As a rule, in the optical sensors the light comes to the resonant chamber through
the monomode waveguide. The transmission spectrum of the cavity has the form
of the Lorenz curve. The curvature of this curve depends on the value of the cavity
Q-factor, and it defines the sensor sensitivity.
The optical sensors based on the ring cavities are more compact compared with,
for example, the sensors based on the spherical or disk cavities. Their dimensions are
of the order of some tens of micrometers along both directions in the crystal plane.
The solutions suggested in the present publication are demonstrated using the
most compact optical refractive index sensor known today based on the photonic
crystal cavities (PCC) [8, 44]. Usually, their dimensions do not exceed the value of
several wavelengths of the light that has been used.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp02.indd 53 9/26/23 4:51 PM

 54 Photonics Elements for Sensing and Optical Conversions

FIGURE 2.21 Scheme of optical system consisting of two coupled cavities with phase delay.

In Figure 2.21 we present the proposed scheme of the optical system consisting of
two coupled cavities. The phase delay of the connection of the cavities has the value
of θ. In this figure, the variable κj, j = {1,2} defines the connectivity coefficient of
the j-th cavity with the waveguide; rj = {1,2} are the coefficients of the spatial energy
losses of the j-th cavity; and pin, prf, ptr are the amplitudes of the input, reflected, and
transmitted fields, respectively.
According to the time depended coupled mode theory, the transmission function
of one cavity can be written in the form:
2κ j
T1 (s j ) = , s j = i(ω − ω j ) + rj , j = {1,2} , (2.13)
s j + 2κ j

where ωj, j = {1,2} is the resonant frequency of the j-th cavity. The Q-factor of the
j-th cavity is defined by the relation 1/Q j = 1/Q jκ + 1/Q jr, where Q jκ = ωj /(4κj) and
Q jr = ωj /(2rj).
When θ = π/2, the transmission function of the system of two cavities shown in
Figure 2.21 is written as:
i2κ 0κ1
T2 ( s ) = . (2.14)
( s + κ 0 )2 + κ12
In this equation, we neglected the waveguide dispersion. The phase incursion θ = π/2
corresponds to the situation in which the coupling waveguide is absent.
In Figure 2.22, the dashed line corresponds to the transmission function of
one cavity calculated according to Eq. (2.13) and the solid line is the transmission

FIGURE 2.22 Transmission functions for one cavity (dashed line) and two coupled cavities
for θ = π/2 (solid line).

BK-TandF-KAZANSKIY_9781032572949-230642-Chp02.indd 54 9/26/23 4:51 PM

 Photonic Crystal Cavities 55

FIGURE 2.23 (a) Cavity based on ridge photonic-coupled waveguide in Si and (b) geo-
metrical parameters of the resonant chamber.

function of the same coupled cavities defined by Eq. (2.14). We see that when two
cavities are used in place of one cavity, the Q-factor of the system increases. The
Q-factor of the system with one cavity and the resonant wavelength were Q1 = 1.5 × 105
and λ0 = 1.488 μm, respectively. The extrinsic losses of this cavity are defined by the
quantity Q jr = 2.5 × 105.
As the concrete PCC, let us calculate the cavity based on the 2D ridge PC wave-
guide. Such cavities have a small area, and they are integrated naturally into the
waveguide geometry of the junctions on the chip. In Figure 2.23(a) the cavity based
on the PC ridge waveguide is shown. The defect where the resonant mode excites
is formed by reduction of the radius of the apertures in the vicinity of the resonant
chamber.
In Figure 2.23(b), the geometrical parameters of the apertures in the vicinity of
the resonant chamber are shown. For the given wavelength of the light, these geo-
metrical parameters ensure the appearance of the band gap for the TE polarization
mode in the waveguide.
To calculate the characteristics of the cavity we used the 2D FDTD method. The
width of the waveguide was w = 500 nm, the refractive index of its material was
equal to 2.97. This value corresponds to the effective refractive index of the Si wave-
guide whose width is 270 nm that is placed on the glass substrate and surrounded
by the air.
In Figure 2.24(a) we show the 3D cavity based on the PC waveguide without a
substrate with three regular apertures. The system consisting of two such cavities
is shown in Figure 2.24(b). The apertures in the regular part of the waveguide
are filled with the air and they are spaced apart at intervals of a = 350 nm. The
wavelength of the resonant mode is equal to 1.48827 μm. We denote the dis-
tance between two cavities (that is the distance between the neighbor holes in
the central part of the structure) by d. The Q-factor of one of these cavities with
eight regular apertures from each side is equal to 156,413. The extrinsic loss of
the given cavity is defined by the value Q jr = 2.56 × 105. These parameters are
approximately equal to the ones of the cavity whose transmission function is
shown in Figure 2.22.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp02.indd 55 9/26/23 4:51 PM

 56 Photonics Elements for Sensing and Optical Conversions

FIGURE 2.24 (a) Examples of 3D cavity based on the PC waveguide without a substrate
with three regular apertures, (b) system consisting of two coupled 3D cavities.

In Figure 2.25, the readings of the transmission function for the systems consist-
ing of one and two cavities are shown. To calculate the readings, we used the FDTD
method. The parameters of the basic cavity correspond to the parameters shown in
Figure 2.24(b). We see that with good accuracy the curves in Figure 2.25 correspond
to the curves in Figure 2.22.
In Figure 2.26 we show the transmission functions for the systems consisting
of two cavities calculated analytically (solid line) and with the aid of the FDTD
method (the readings are interpolated by the dotted line). The period of the regular
part of the PC waveguide is equal to 350 nm. That is the reason why the absence of

FIGURE 2.25 Readings of transmission functions for one and two cavities interpolated
with the aid of FDTD method by dotted and solid lines, respectively.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp02.indd 56 9/26/23 4:51 PM

 Photonic Crystal Cavities 57

FIGURE 2.26 Transmission functions for systems consisting of two cavities for θ = π/2 cal-
culated analytically (solid line) and with the aid of FDTD method (readings are interpolated
by dotted line).

the coupling waveguide (θ = π/2) corresponds to the value d = 350. However, from
Figure 2.26 we see that almost perfect correspondence between the curves defined
by the analytical expression (2) for θ = π/2 and the result of the FDTD method
simulations is achieved for d = 354 nm. It can be explained by the fact that the Bragg
frequency of the PC waveguide does not match the resonant frequency. In this case,
to accomplish the condition θ = π/2 we need to add a fragment of the waveguide
between the cavities [24].
The Q-factor of the resonant system can be estimated from the slope of the spec-
tral peak. For example, in the refractive index sensors the slope of the spectral peak
corresponds to the rate of the changes of the transmission function due to the changes
of the refractive index of the environment. In Figure 2.27 we show the derivatives of
the transmission functions in the region near the spectral peaks calculated with the
aid of the FDTD simulations for the systems consisting of one (the dotted line) and
two (the solid line) cavities, respectively.

FIGURE 2.27 Derivatives of transmission functions in regions near spectral peaks for
systems consisting of one (dotted line) and two (solid line) cavities, respectively.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp02.indd 57 9/26/23 4:51 PM

 58 Photonics Elements for Sensing and Optical Conversions

From Figure 2.27, we see that for the system of two coupled cavities the maxi-
mal value of the derivative of the transmission function in the point of the spec-
tral peak is approximately two times larger than the respective value in the case
of one cavity. As a result, the Q-factor of the system consisting of two cavities
is better compared to one of the systems with one cavity. However, as it is seen
from Figure 2.25, for the parameters of the cavities in use, this improving of the
Q-factor is obtained when increasing the input energy level. Nevertheless, the obvi-
ous analytical calculations show that the difference between the maximums of the
transmission functions for one and two cavities can be reduced by the way of reduc-
tion of the energy losses. For example, for Q1 = 1.5 × 105 and Q jr = 2.5 × 106, this
difference is about 3%.
With the aid of numerical simulations, let us show how the errors in manufactur-
ing of single and coupled cavities influence the displacement of the resonant mode
frequencies of the cavities and the Q-factors of resonant systems.
In Table 2.4, we present the results of computer simulations of the influence of
the errors in manufacturing for one PC cavity. We varied the radii of the holes in the
PC waveguide. It was assumed that the errors in the values of the radii were random
numbers uniformly distributed inside the interval ±1, ±2, ±3, ±4, or ±5 nm. We used
so simple a scheme when modeling the influence of the errors of the manufacturing
mainly due to calculation time. Several hours of the computing time are necessary to
calculate the characteristics of one cavity with the aid of the FDTD method using the
computational cluster of the peak capacity equal to 1 TFLOP. From Table 2.4 we see
that when the error is equal to ±5 nm the shift of the resonant frequency can reach
5 nm. At this point, when the error increases, the value of the Q-factor of the cavity
can both decrease and increase.
Table 2.5 demonstrates the results of computer simulation of the influence of the
manufacturing errors for the system consisting of two coupled PCC. In this case, the
changes in the transmission function depend mainly on the mismatching of these
two cavities in the resonant wavelengths. In the presence of a mismatching, the reso-
nant modes of the cavities are no longer the degenerated ones, and that results in a
broadening of the spectral peak of the transmission function.

TABLE 2.4
Values of Resonant Wavelengths and
Q-factors for Different Values of Errors
for the System Consisting of One PCC
Error, nm Wavelength, nm Q-factor
±0 1,488.2 156,413
±1 1,488.2 156,435
±2 1,488.3 153,050
±3 1,489.2 183,318
±4 1,490.7 110,346
±5 1,491.2 97,482

BK-TandF-KAZANSKIY_9781032572949-230642-Chp02.indd 58 9/26/23 4:51 PM

 Photonic Crystal Cavities 59

TABLE 2.5
Values of the Resonant Wavelengths and
Q-factors for Different Values of Errors
for System Consisting of Two PCC
Error, nm Wavelength, nm Q-factor
±0 1,488.2 252,739
±1 1,487.1 172,461
±2 1,484.8 110,686
±3 1,483.6 312,560
±4 1,482.8 180,881
±5 1,482.5 297,793

Consequently, the individual tuning of the resonant wavelengths in each resonator

allows us to compensate, at least partially, the nondegeneracy of the resonant modes.
Dynamic tuning of resonant wavelength of each resonator can be done by changing
the refractory index of the waveguide. For this purpose, one can use, for example,
the Kerr effect, the injection of the free carriers, or the thermal effect. Among these
methods, the thermal effect permits the implementation at the minimal frequency.
However, for the resonant system described here the high rate of the change of the
resonant wavelength is not necessary.
Recently, injection of the free carriers is most frequently used to change the
refractory index in PCC. The injection of the free carriers into PCC may be done
using optical or electron pumping. When comparing optical and electron pumping,
only electron pumping provides the possibility of practical realization of the device
proposed in this chapter.

2.7 ON A SI-BASED PHOTONIC CRYSTAL CAVITY

FOR THE NEAR-INFRARED REGION: NUMERICAL
SIMULATION AND FORMATION TECHNOLOGY
Si is the most used material in the complementary metal–oxide–semiconductor
(CMOS) technology of constructing electronic circuits [49]. However, as a rule, as
materials for optoelectronic components, III–V semiconductors, e.g. GaN, InP, or
GaAs, are used [50]. The monolithic integration of Si and III–V semiconductors in
a crystal presents difficulties because of the difference between the crystal-lattice
periods of these materials. The use of Si in various fields, including optoelectronics,
would simplify the problem of the integration of photonic and electronic compo-
nents. Si is rather easily oxidized. Stable Si oxide possesses a refractive index of ~1.5,
which creates the required contrast with Si, whose refractive index is ~3.5 in the tele-
communication wavelength region. An important element of nanophotonic devices is
the optical cavity [51, 52]. The basic parameters of the cavity are the quality factor,
frequency, and mode composition. High-quality-factor cavities with a small mode
volume make it possible to enhance the interaction of light with the medium, to

BK-TandF-KAZANSKIY_9781032572949-230642-Chp02.indd 59 9/26/23 4:51 PM

 60 Photonics Elements for Sensing and Optical Conversions

reduce the size of the photonic element, and to optimize its dispersion characteristics.
Nanocavities are formed based on PCs [53]. High-quality-factor nanocavities make
possible the production, e.g. of optical switches, filters, modulators, and coherent
light sources with a low threshold power [54]. The fabrication of a nanocavity within
a 3D PC is a technologically difficult problem. Therefore, 2D PCs in the form of a
membrane are often used. In such a membrane, light propagates because of total
internal reflection (TIR) [55, 56]. Further simplification of the problem is attained by
using ridge PC waveguides. In such waveguides, the TIR effect prevents the propaga-
tion of light in transverse directions. In the longitudinal direction, the localization
of light in the nanocavity is provided by the PC [26, 57]. In this study, we explore a
cavity based on a ridge PC waveguide. The structure of such a waveguide consists of
three parts. The first part is the PC mirrors formed as identical equidistant holes in
the waveguide; these holes constitute the PC lattice. The second element of the struc-
ture is the transition zone between the PC waveguide and the cavity region as such.
This zone is used to reduce energy losses in the cavity. The transition zone involves
several holes, whose radii and positions are optimized to make possible a reduction in
the losses. The third element of the structure is the cavity proper. A rise in the quality
factor of the ridge PC cavity is attained by reducing losses in both longitudinal and
transverse directions. Losses in the longitudinal direction can be reduced merely by
increasing the number of periods of the PC [40, 58]. To reduce losses in the transverse
directions, some finer optimization of the structure of the nanocavity is required.
In this section, we report the results of numerical simulation of a PC cavity based
on a Si-based ridge waveguide and demonstrate the production technology of cavi-
ties of this type. As shown by the results of technological experiments, the procedure
of ion-beam etching used to produce cavities here can lead to the PC cavity having
cone-shaped holes rather than cylindrical holes. In the study, we simulate the conic-
ity of the hole shape and show the possibility for compensating the effects of the
deviation in the hole shape from cylindrical. We consider the influence of the height
of the ridge waveguide on the parameters of the PC cavity.
The nanocavity was fabricated at the Institute for Physics of Microstructures,
Russian Academy of Sciences, Nizhny Novgorod, Russia. The PC cavity structures
were formed in several stages, using electron-beam and ion-beam lithography. In the
preliminary stage, we formed waveguides by means of electron-beam lithography in
combination with ion etching. Then, by precision etching with a sharply focused ion
beam, we formed the end faces of the waveguides and the PC cavities themselves as
a sequence of holes at the waveguide surface; the holes were different in diameter
and positioned at a certain distance from each other. The cavities were formed on
standard silicon-on-insulator (SOI) substrates (Soitec) with a thickness of the buried
Si-oxide layer of ~1 μm and a thickness of the upper “bearing” Si layer hSi ~220 nm.
To create a pattern, we first deposited a two-layer positive resist (polymethyl meth-
acrylate PMMA 495 and PMMA 950) on the substrate; the total width of the resist
was ~200 nm. After exposure to an electron beam and development in a solution
of methyl isobutyl ketone (MIBK) in isopropyl alcohol (IPA), MIBK:IPA (1:3), a
~20-nm-thick vanadium layer was deposited onto the resist, and the lift-off proce-
dure was carried out in acetone with ultrasonic activation. As a result of the lift-off
lithography process, a vanadium mask, formed as 20 × 2-μm strips, was created.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp02.indd 60 9/26/23 4:51 PM

 Photonic Crystal Cavities 61

This mask was further used upon etching of the Si layer with an ion beam. Primary-
ion etching was conducted through the thickness of the Si layer in an atmosphere of
argon. This was done using a system with a hot cathode (a Kaufman setup, an Ar+ ion
energy of ~1,000 eV). The remains of the vanadium mask were removed by wet etch-
ing in a solution of hydrogen peroxide. As a result of the procedures just described,
Si strips to be used as blanks for the waveguides were formed. The dimensions of the
blanks were somewhat larger than the required dimensions. Final treatment of the
waveguides, specifically, the formation of necessary-sized strips, holes in the strips,
and the end faces with specified parameters was carried out by local etching with a
sharply focused ion beam. Etching was conducted with Ga+ ions with an energy of
30 keV, using a Neon 40 (ZEISS, Germany) two-beam microscope. The ion-beam
current was varied from 1 to 10 pA; the beam diameter was ~7–15 nm. Figure 2.28
shows micrographs of the PC cavities fabricated by the procedure just described.
The micrographs were obtained with the same scanning microscope, immediately
after the final stage of etching.
The images were obtained with secondary electrons, using an electron probe at
different angles of sample inclination. The waveguide length and width were 11.5
and 0.5 μm, respectively; the hole diameter was varied and increased along the
direction from the center of the waveguide to its edges. As shown in Figure 2.28(b),
in the image of the cavity at the angle of sample inclination of 54°, we can see a char-
acteristic contrast ratio indicative of the conical shape of holes of the cavity. This can
be clearly seen in the micrograph of the cavity cross section made with the use of a
sharply focused ion beam (Figure 2.28d). In this case, we can see in detail the shape
of the holes and determine their dimensions.
Simulation of the PC cavities was carried out by the finite-difference time-domain
(FDTD) method. In this study, we used the parallel 3D FDTD method. Calculations for
PC cavities were performed for the telecommunication wavelength region (~1.5 μm).
In accordance with the perfectly matched layers (PML) method, it was assumed

FIGURE 2.28 Micrographs of a waveguide with the photonic crystal cavity at different
magnifications: (a) the whole cavity; (b) the central part of the cavity measured at an angle of
sample inclination of 54°; (c) the central part of the cavity (view from above); and (d) the cross
section of the cavity along waveguide central axis.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp02.indd 61 9/26/23 4:51 PM

 62 Photonics Elements for Sensing and Optical Conversions

FIGURE 2.29 Geometrical parameters of the PC cavity.

TABLE 2.6
The Radii of Holes and the Spacings between Them in
the PC Waveguide (in Nm)
b1 a1 b2 a2 b3 a3 b4 a4 b5
200 65 290 82 310 90 305 85 345

that there is an absorbing layer at the boundaries of the 3D calculation region. The
thickness of this layer was half the resonance wavelength. For the radiation source,
we used a Gaussian pulse. The frequency width of this pulse was about 10% of the
resonance frequency. The resolution of the computational mesh was chosen from
the condition of convergence of the results [58].
Figure 2.29 shows the geometrical parameters of the PC cavity to be simulated.
Here, a1 and a2 are the hole radii and b1 and b2 are the distance from the center and
the spacing between neighboring holes, respectively. Table 2.6 lists these param-
eters. The parameters are close to those of the PC cavity fabricated in the study. In
the PC cavity to be simulated, there are eight holes in the central transition zone and
eight regular holes on each side of the central zone. The PC waveguide width and
height are, correspondingly, w = 500 nm and h = 260 nm. The waveguide consists of
Si with the refractive index 3.45, lies on a Si-oxide layer with the refractive index 1.45,
and is surrounded by air. The holes in the regular part of the waveguide are 185 nm
in diameter, filled with air, and spaced from each other by a distance of a = 350 nm.
The wavelength of the resonance mode in the cavity with such parameters corre-
sponds to λ = 1,457.41 nm (here and below, we give the wavelengths in vacuum). The
quality factor of the cavity with eight regular holes on each side is 69,064.
It can be shown that a change in the PC waveguide height (h) yields a change in
the band gap of the PC and, as a consequence, the frequency region of the cavity. The
results of numerical simulation show that there exists an optimal height h, for which
the band gap is maximal. The quality factor and cavity-mode wavelength obtained
as functions of the waveguide height by theoretical calculations for the PC cavity
just described are listed in Table 2.7. The quality factor of the cavity was calculated
by the method described elsewhere [57]. The results of simulation (Table 2.7) show
that the maximum quality factors of the PC cavity are attained on condition that the
ratio of the height h to the lattice period of the PC (a) is unity. As the PC waveguide
height is increased, the resonance wavelength increases and, at large values of h, is
beyond the band gap of the PC. For example, the height h = 510 nm given in Table 2.7
corresponds to another resonance mode with a quality factor much (about an order of
magnitude) lower than that of the preceding mode.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp02.indd 62 9/26/23 4:51 PM

 Photonic Crystal Cavities 63

TABLE 2.7
Dependences of the Parameters of the PC Cavity
on the Waveguide Height
Height h, nm Wavelength, nm Quality Factor
220 1,423.59 54,482
240 1,439.17 62,153
260 1,457.41 69,064
280 1,476.52 74,530
300 1,493.63 75,365
330 1,516.72 79,201
370 1,541.73 75,742
410 1,562.43 73,472
450 1,579.71 60,487
490 1,594.31 45,135
510 1,451.56 6,768

As shown in Section 2.2, some undesirable effects associated with variations in

the parameters of the PC cavity can be produced during ion etching. Specifically,
during etching, holes in the PC waveguide can acquire a conical shape. It is known
that, during etching with an ion beam, the conicity of a hole is influenced by several
factors: the relation between the diameter of the ion beam and the dimensions of the
region of etching, the scanning mode, the ion energy, etc. [59]. Let us consider the
influence of the hole conicity on the parameters of the PC cavity.
A conical hole is described by the cone height h equal to the waveguide height,
the radii of the lower (r) and upper (R) bases, and the point of intersection of the cone
with a cylinder of specified diameter (RC).
This point is defined by the distance L reckoned from the upper surface of the
waveguide. The cone-shaped hole and the projection of its intersection with an ideal
cylinder in the waveguide are schematically shown in Figure 2.30.

FIGURE 2.30 Projection of the intersection of a cone-shaped hole with an ideal cylinder
in the waveguide.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp02.indd 63 9/26/23 4:51 PM

 64 Photonics Elements for Sensing and Optical Conversions

TABLE 2.8
Calculated Parameters of the PC Cavity as Functions of the
Degree of Hole Conicity
r
L 5d/12 d/3 d/4
h/4 λ = 1,461.35 nm; λ = 1,476.15 nm; λ = 1,490.51 nm;
Vcn/Vcl = 0.896; Vcn/Vcl = 0.806; Vcn/Vcl = 0.731;
Q = 49,816 Q = 28,089 Q = 11,197
h/2 λ = 1,484.52 nm; λ = 1,466.16 nm; λ = 1,483.17 nm;
Vcn/Vcl = 1.009; Vcn/Vcl = 1.037; Vcn/Vcl = 1.083;
Q = 8,311 Q = 18,968 Q = 13,272
3h/4 λ = 1,426.14 nm; λ = 1,450.91 nm; λ = 1,472.59 nm;
Vcn/Vcl = 1.398; Vcn/Vcl = 1.925; Vcn/Vcl = 2.583;
Q = 5,186 Q = 10,020 Q = 12,951

In the numerical experiment, the values of L were chosen as fractions of the

waveguide height [3h/4, h/2, h/4] and the radius of the lower cone base was defined
in terms of its ratio to the diameter of the distorted cylindrical hole r [5d/12, d/3,
d/4]. Thus, the value of h was 260 nm and the value of d corresponded to the diam-
eter of each cylinder distorted. It was assumed that all the holes in the PC cavity
were distorted in a similar manner. The resultant values of the resonance-mode
wavelength (λ), the ratio between the volumes of the conical and cylindrical holes
(Vcn /Vcl), and the quality factor of the cavity in relation to the parameters r and L are
given in Table 2.8.
As follows from Table 2.8, the quality factor and the resonance-mode wavelength
depend on the volume and the degree of hole conicity. In this case, the maximum
quality factor corresponds to a hole with a minimum degree of conicity (L = h/4,
r = 5d/12). Eventually, the hole conicity influences the effective refractive index of
the medium. From the results of numerical simulations, it follows that, by optimiz-
ing the volume and conicity of holes, we can tune the cavity to the desired wave-
length and, at the same time, retain a high-quality factor of the cavity.

2.8 CONCLUSIONS
We describe and numerically investigate an all-optical temporal integrator based
on PC nanobeam cavities. We show that an array of PC cavities enables high-order
temporal integration. This integrator is more compact than any of those previously
suggested. Its dimensions depend linearly on the order of integration. The ways to
increase the time-bandwidth product of the integrator by using an active cavity are
discussed. For in-plane optical pumping, the nanocavities in the cross section of
ridge PC waveguides are described. The structures proposed have been shown to
concentrate the electromagnetic field energy in the cavity’s regions that contain
the material introduced to the resonance cavity. For electrical pumping, the model

BK-TandF-KAZANSKIY_9781032572949-230642-Chp02.indd 64 9/26/23 4:51 PM

 Photonic Crystal Cavities 65

of two-component nanocavity with the possibility of electrical pumping is also

described. In this model, the minimum details of the structure are found only in
the periodic component of the resonator. The advantages of such a structure include
a promising way to construct an electrically pumped photonic cavity, the ease of
introducing nonlinear optical materials in the area of the nanocavity, the possibility
of formation of the desired energy distribution in the far zone, and the possibility to
construct dynamic systems based on nanocavities. The production technology of a
PC cavity formed as a group of holes in a Si strip waveguide by ion-beam etching is
described. The parasitic effect associated with hole conicity, which develops upon
hole formation by the given technology, is studied. Numerical simulation shows that
the hole-conicity-induced decrease in the cavity quality factor can be compensated
with consideration for the hole volume. The influence of the waveguide thickness on
the resonance wavelength and quality factor of the PC cavity is analyzed.

REFERENCES
1. Ngo, N. Q. 2007. Design of an optical temporal integrator based on a phase-shifted
Bragg grating in transmission. Opt. Lett. 32(20):3020–2. https://doi.org/10.1364/
OL.32.003020.
2. Ferrera, M. et al. 2010. On-chip CMOS-compatible all-optical integrator. Nat.
Commun. 1:29. https://doi.org/10.1038/ncomms1028.
3. Ding, Y., Zhang, X., Zhang, X., and Huang, D. 2009. Active microring optical inte-
grator associated with electro absorption modulators for high speed low light power
loadable and erasable optical memory unit. Opt. Express. 17(15):12835–48. https://doi.
org/10.1364/OE.17.012835.
4. Slavík, R. et al. 2008. Photonic temporal integrator for all-optical computing. Opt.
Express. 16(22):18202–14. https://doi.org/10.1364/OE.16.018202.
5. Babiker, S. G., Shuai, Y., Sid-Ahmed, M. O., and Xie, M. 2014. One-dimensional Si/
SiO2 photonic crystals filter for thermophotovoltaic applications. WSEAS Trans. Appl.
Theor. Mech. 9:97–103.
6. Anagnostakis, E. A. 2010. A qualitative comprehension of nanophotonics. Proceedings
of the European Conference of Systems, and European Conference of Circuits
Technology and Devices, and European Conference of Communications, and European
Conference on Computer Science, ed. V. Mladenov, K. Psarris, N. Mastorakis, A.
Caballero, G. Vachtsevanos, 84–100. Wisconsin, US: World Scientific and Engineering
Academy and Society (WSEAS).
7. Singh, S., and Sarin, R. K. 2007. Enhanced performance of microstrip-fed wide slot
antenna using periodic gaps in dielectric substrate. Proc 11th Conf on 11th WSEAS
Int Conf on Communications (ICCOM’07) 11:127–9.
8. Akahane, Y., Asano, T., Song, B.-S., and Noda, S. 2005. Fine-tuned high-Q
photonic-crystal nanocavity. Opt. Express. 13(4):1202–14. https://doi.org/10.1364/
OPEX.13.001202.
9. Reed, G. T., Mashanovich, G., Gardes, F. Y., and Thomson, D. J. 2010. Silicon optical
modulators. Nat. Photonics. 4:518–26. https://doi.org/10.1038/nphoton.2010.179.
10. Takano, H., Song, B.-S., Asano, T., and Noda, S. 2006. Highly efficient multi-channel
drop filter in a two-dimensional hetero photonic crystal. Opt. Express 14(8):3491–6.
https://doi.org/10.1364/OE.14.003491.
11. Englund, D., Altug, H., Ellis, B., and Vučković, J. 2008. Ultrafast photonic crystal
lasers. Laser Photon. Rev. 2(4):264–74. https://doi.org/10.1002/lpor.200710032.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp02.indd 65 9/26/23 4:51 PM

 66 Photonics Elements for Sensing and Optical Conversions

12. Faraon, A., and Vučković, J. 2009. Local temperature control of photonic crystal
devices via micron-scale electrical heaters. Appl. Phys. Lett. 95(4):043102. https://doi.
org/10.1063/1.3189081.
13. Tanaka, Y., Upham, J., Nagashima, T., Sugiya, T., Asano, T., and Noda, S. 2007.
Dynamic control of the q factor in a photonic crystal nanocavity. Nat. Mater. 6:862–5.
https://doi.org/10.1038/nmat1994.
14. Fushman, I., Waks, E., Englund, D., Stoltz, N., Petroff, P., and Vučković, J. 2007.
Ultrafast nonlinear optical tuning of photonic crystal cavities. Appl. Phys. Lett.
90(9):091118. https://doi.org/10.1063/1.2710080.
15. Englund, D., Ellis, B., Edwards, E., Sarmiento, T., Harris, J. S., Miller, D. A. B., and
Vučković, J. 2009. Electrically controlled modulation in a photonic crystal nanocavity.
Opt. Express. 17(18):15409–19. https://doi.org/10.1364/OE.17.015409.
16. Tanabe, T., Nishiguchi, K., Kuramochi, E., and Notomi, M. 2009. Low power and fast
electro-optic silicon modulator with lateral p-i-n embedded photonic crystal nano-
cavity. Opt. Express. 17(25):22505–13. https://doi.org/10.1364/OE.17.022505.
17. Debnath, K., O’Faolain, L., Gardes, F. Y., Steffan, A. G., Reed, G. T., and Krauss,
T. F. 2012. Cascaded modulator architecture for WDM applications. Opt. Express.
20(25):27420–8. https://doi.org/10.1364/OE.20.027420.
18. Schmidt, B., Xu, Q., Shakya, J., Manipatruni, S., and Lipson, M. 2007. Compact electro-
optic modulator on silicon-on-insulator substrates using cavities with ultra-small modal
volumes. Opt. Express. 15(6):3140–8. https://doi.org/10.1364/OE.15.003140.
19. Shakoor, A., Nozaki, K., Kuramochi, E., Nishiguchi, K., Shinya, A., and Notomi, M.
2014. Compact 1D-silicon photonic crystal electro-optic modulator operating with
ultra-low switching voltage and energy. Opt. Express. 22(23):28623–34. https://doi.org/
10.1364/OE.22.028623.
20. Quan, Q., Floyd, D. L., Burgess, I. B., Deotare, P. B., Frank, I. W., Tang, S. K., and
Loncar, M. 2013. Single particle detection in CMOS compatible photonic crystal
nanobeam cavities. Opt. Express. 21(26):32225–33. https://doi.org/10.1364/OE.21.
032225.
21. Kazanskiy, N. L., Serafimovich, P. G., and Khonina, S. N. 2010. Harnessing the guided-
mode resonance to design nanooptical transmission spectral filters. Opt. Mem. Neural
Netw. 19(4):318–24. https://doi.org/10.3103/S1060992X10040090.
22. Serafimovich, P. G., and Kazanskiy, N. L. 2015. Active photonic crystal cavities for opti-
cal signal integration. Opt. Mem. Neural Netw. 24(4):260–71. https://doi.org/10.3103/
S1060992X15040050.
23. Vollmer, F., Braun, D., Libchaber, A., Teraoka, I., and Arnold, S. 2002. Protein detec-
tion by optical shift of a resonant microcavity. Appl. Phys. Lett. 80(21):4057. https://doi.
org/10.1063/1.1482797.
24. Robinson, J. T., Chen, L., and Lipson, M. 2008. On-chip gas detection in silicon
optical microcavities. Opt. Express. 16(6):4296–301. https://doi.org/10.1364/OE.16.
004296.
25. Liu, H. C., and Yariv, A. 2012. Designing coupled-resonator optical waveguides based
on high-q tapered grating-defect resonators. Opt. Express. 20(8):9249–63. https://doi.
org/10.1364/OE.20.009249.
26. Kazanskiy, N. L., and Serafimovich, P. G. 2014. Coupled-resonator optical waveguides
for temporal integration of optical signals. Opt. Express. 22(11):14004–13. https://doi.
org/10.1364/OE.22.014004.
27. Asghari, M. H., Wang, C., Yao, J., and Azaña, J. 2010. High-order passive photonic
temporal integrators. Opt. Lett. 35(8):1191–3. https://doi.org/10.1364/OL.35.001191.
28. Quan, Q., and Loncar, M. 2011. Deterministic design of wavelength scale, ultra-high
q photonic crystal nanobeam cavities. Opt. Express. 19(5):18529–42. https://doi.org/
10.1364/OE.19.018529.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp02.indd 66 9/26/23 4:51 PM

 Photonic Crystal Cavities 67

29. Huang, N., Li, M., Ashrafi, R., Wang, L., Wang, X., Azaña, J., and Zhu, N. 2014.
Active Fabry-Perot cavity for photonic temporal integrator with ultra-long operation
time window. Opt. Express. 22(3):3105–16. https://doi.org/10.1364/OE.22.003105.
30. Kazanskiy, N. L., Serafimovich, P. G., and Khonina, S. N. 2013. Use of photonic crystal
cavities for temporal differentiation of optical signals. Opt. Lett. 38(7):1149–51. https://
doi.org/10.1364/OL.38.001149.
31. Coccioli, R. 1998. Boroditsky, M., Kim, K. W., Rahmat-Samii, Y., and Yablonovitch,
E. Smallest possible electromagnetic mode volume in a dielectric cavity. IEE Proc.,
Optoelectron. 145(6):391–7. https://doi.org/10.1049/ip-opt:19982468.
32. Zhang, Y., McCutcheon, M. W., Burgess, I. B., and Loncar, M. 2009. Ultra-high-q
TE/TM dual-polarized photonic crystal nanocavities. Opt. Lett. 34(17):2694–6. https://
doi.org/10.1364/OL.34.002694.
33. Rivoire, K., Buckley, S., and Vučković, J. 2011. Multiply resonant photonic crystal
nanocavities for nonlinear frequency conversion. Opt. Express. 19(22):22198–207.
https://doi.org/10.1364/OE.19.022198.
34. Schriever, C., Bohley, C., and Schilling, J. 2010. Designing the quality factor of infil-
trate photonic wire slot microcavities. Opt. Express. 18(24):25217–24. https://doi.
org/10.1364/OE.18.025217.
35. Yamamoto, T., Notomi, M., Taniyama, H., Kuramochi, E., Yoshikawa, Y., Torii, Y.,
and Kuga, T. 2008. Design of a high-Q air-slot cavity based on a width-modulated line-
defect in a photonic crystal slab. Opt. Express. 16(18):13809–17. https://doi.org/10.1364/
OE.16.013809.
36. Fan, S., Winn, J. N., Devenyi, A., Chen, J. C., Meade, R. D., and Joannopoulos, J. D.
1995. Guided and defect modes in periodic dielectric waveguides. J. Opt. Soc. Am. B.
12(7):1267–72. https://doi.org/10.1364/JOSAB.12.001267.
37. Taflove, A., and Hagness, S. C. 2005. Computational electrodynamics: The finite-
difference time-domain method. 3rd ed. Norwood, MA: Artech House. ISBN:
978-1-58053-832-9.
38. Almeida, V. R., Xu, Q., Barrios, C. A., and Lipson, M. 2004. Guiding and confin-
ing light in void nanostructure. Opt. Lett. 29(11):1209–11. https://doi.org/10.1364/
OL.29.001209.
39. Ahn, B. H., Kang, J. H., Kim, M. K., Song, J. H., Min, B., Kim, K. S., and Lee, Y.
H. 2010. One-dimensional parabolic beam photonic crystal laser. Opt. Express.
18(6):5654–60. DOI: 10.1364/OE.18.005654.
40. Serafimovich, P. G., Kazanskiy, N. L., and Khonina, S. N. 2013. Two-component cavity
based on a regular photonic crystal nanobeam. Appl. Opt. 52(23):5830–4. https://doi.
org/10.1364/AO.52.005830.
41. Yang, X., Yu, M., Kwong, D. L., and Wong, C. W. 2009. All-optical analog to electro-
magnetically induced transparency in multiple coupled photonic crystal cavities. Phys.
Rev. Lett. 102(17):173902. https://doi.org/10.1103/PhysRevLett.102.173902.
42. Gan, X., Pervez, N., Kymissis, I., Hatami, F., and Englund, D. 2012. A high-resolution
spectrometer based on a compact planar two dimensional photonic crystal cavity array.
Appl. Phys. Lett. 100(23):231104. https://doi.org/10.1063/1.4724177.
43. Deotare, P., Kogos, L., Quan, Q., Ilic, R., and Loncar, M. 2012. On-chip integrated
spectrometer using nanobeam photonic crystal cavities. Conf on Lasers and Electro-
Optics: CM3B.4. https://doi.org/10.1364/CLEO_SI.2012.CM3B.4.
44. Velha, P., Rodier, J. C., Lalanne, P., Hugonin, J. P., Peyrade, D., Picard, E., Charvolin,
T., and Hadji, E. 2006. Ultra-high-reflectivity photonic-bandgap mirrors in a ridge
SOI waveguide. New J. Phys. 8(9):204. https://doi.org/10.1088/1367-2630/8/9/204.
45. Arnold, S., Khoshsima, M., Teraoka, I., Holler, S., and Vollmer, F. 2003. Shift of whis-
pering-gallery modes in micro-spheres by protein adsorption. Opt. Lett. 28(4):272–4.
https://doi.org/10.1364/OL.28.000272.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp02.indd 67 9/26/23 4:51 PM

 68 Photonics Elements for Sensing and Optical Conversions

46. Boyd, R. W., and Heebner, J. E. 2001. Sensitive disk resonator photonic biosensor. Appl.
Opt. 40(31):5742–7. https://doi.org/10.1364/AO.40.005742.
47. White, I. M., Oveys, H., and Fan, X. 2006. Liquid-core optical ring-resonator sensors.
Opt. Lett. 31(9):1319–21. https://doi.org/10.1364/OL.31.001319.
48. Lee, M., and Fauchet, P. M. 2007. Two-dimensional silicon photonic crystal based
biosensing platform for protein detection. Opt. Express 15(8):4530–5. https://doi.org/
10.1364/OE.15.004530.
49. Baba, T. 1997. Photonic crystals and microdisk cavities based on GaInAsP-InP system.
IEEE J. Sel. Top. Quantum Electron. 3(3):808–30. https://doi.org/10.1109/2944.640635.
50. Pottier, P., Seassal, C., Letartre, X., Leclercq, J. L., Viktorovitch, P., Cassagne, D.,
and Jouanin, C. 1999. Triangular and hexagonal high Q-factor 2-D photonic bandgap
cavities on III-V suspended membranes. J. Lightw. Technol. 17(11):2058–62. https://doi.
org/10.1109/50.802995.
51. Little, B. E., Chu, S. T., Haus, H. A., Foresi, J., and Laine, J. P. 1997. Microring
resonator channel dropping filters. J. Lightw. Technol. 15(6):998–1005. https://doi.
org/10.1109/50.588673.
52. Yamamoto, Y., and Slusher, R. E. 1995. Optical processes in microcavities. In Confined
electrons and photons: New physics and applications, ed. E. Burstein, C. Weisbuch,
871–8. New York, US: Springer. https://doi.org/10.1007/978-1-4615-1963-8_46.
53. Labilloy, D., Benisty, H., Weisbuch, C., Krauss, T. F., Bardinal, V., and Oesterle, U.
1997. Demonstration of cavity mode between two-dimensional photonic-crystal mir-
rors. Electron. Lett. 33(23):1978–80. https://doi.org/10.1049/el:19971321.
54. Beaky, M. M., Burk, J. B., Everitt, H. O., Haider, M., and Venakides, S. 1999. Two-
dimensional photonic crystal Fabry Perot resonators with lossy dielectrics. IEEE
Trans. Microw. Theory Tech. 47(11):2085–91. https://doi.org/10.1109/22.798003.
55. Painter, O., Lee, R. K., Scherer, A., Yariv, A., O’Brien, J. D., Dapkus, P. D., and Kim, I.
1999. Two-dimensional photonic band-gap defect mode laser. Science. 284(5421):
1819–21. https://doi.org/10.1126/science.284.5421.1819.
56. Chow, E., Lin, S. Y., Johnson, S. G., Villeneuve, P. R., Joannopoulos, J. D., Wendt, J.
R., Vawter, G. A., Zubrzycki, W., Hou, H., and Alleman, A. 2000. Three-dimensional
control of light in a two-dimensional photonic crystal slab. Nature. 407(6807):983–6.
https://doi.org/10.1038/35039583.
57. Kazanskiy, N. L., Serafimovich, P. G., and Khonina, S. N. 2012. Enhancement of spa-
tial modal overlap for photonic crystal nanocavities. Computer Optics. 36(2):199–204.
58. Kazanskiy, N. L., Serafimovich, P. G., and Khonina, S. N. 2012. Use of photonic
crystal resonators for differentiation of optical impulses in time. Computer Optics.
36(4):474–8.
59. Hopman, W. C. L., Ay, F., Hu, W., Gadgil, V. J., Kuipers, L., Pollnau, M., and de Ridder,
R. M. 2007. Focused ion beam scan routine, dwell time and dose optimizations for
submicrometre period planar photonic crystal components and stamps in silicon.
Nanotechnology. 18(19):195305. https://doi.org/10.1088/0957-4484/18/19/195305.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp02.indd 68 9/26/23 4:51 PM

 3 Recent Advances
Nanoplasmonic Sensors

Muhammad Ali Butt1,2, Nikolay Lvovich Kazanskiy1,3,

and Svetlana Nikolaevna Khonina1,3
1Samara National Research University, Samara, Russia
2 Warsaw University of Technology, Institute of
Microelectronics and Optoelectronics, Warszawa, Poland
3Image Processing Systems Institute – Branch

of the Federal Scientific Research Centre

“Crystallography and Photonics” of Russian
Academy of Sciences, Samara, Russia

3.1 INTRODUCTION
The study of light–matter interaction via surface plasmon polaritons (SPPs) or surface
plasmons (SPs) is known as plasmonics. SPs are mutual vibrations of free electrons
(e−) at a metal surface that are optically stimulated. SPPs are quanta of collective
charge density oscillations that are strictly restricted at the boundary of negative and
positive permittivity materials. SPs may, however, be described in classical electro-
dynamics without restoring to quantum physics. They represent genuine mechanical
oscillations of the e− gas controlled by the incident light’s electromagnetic (EM) field.
SP modes are enclosed in nanosized, i.e. subwavelength metal structures ̶ localized
surface plasmons (LSPs). Light interaction with nanoscopic matter has been seen
for eras, such as in the stained-glass goblets of the Roman Empire or magnificent
medieval cathedral windows. More than a century ago, Michael Faraday and Gustav
Mie theorized about it. Our capacity to create and manipulate matter on a nanoscale
has only recently given birth to what is known as nanoplasmonics.
Photonics, biological imaging, molecular spectroscopy, and sensing, to mention a
few, have all benefited from the advancement of nanoplasmonic technology during
the previous decade. Subwavelength confinement, a high cross section for scattering
and absorption of light and, most importantly for sensing purposes, significantly
increased EM fields in the immediate vicinity of nanostructures are all key char-
acteristics that enables such progress. Through field amplification, nanoplasmonic
resonances are sensitive to small changes in the surrounding refractive index (RI).
This assumption is the foundation for nanoplasmonic sensing, which is essentially
RI sensing with the strong addition that the investigated volumes are nanoscale.
In a variety of nanoplasmonic sensing systems, even single biological or chemical
molecular substances can be identified.
The metal-insulator-metal waveguide configuration is among the most fre-
quently utilized plasmonic-based n-structures for the development of dense optical

DOI: 10.1201/9781003439165-3 69

BK-TandF-KAZANSKIY_9781032572949-230642-Chp03.indd 69 9/26/23 4:55 PM

 70 Photonics Elements for Sensing and Optical Conversions

circuits. Metal-insulator-metal waveguides are plasmonic structures with two metal

claddings around an insulator. The essential aspects of this system are its simple
construction and the ability to restrict light at the subwavelength level. Researchers
have studied the construction of a wide range of devices utilizing metal-insulator-
metal waveguides to build highly integrated optical circuits. Filters [1], couplers
[2], sensing devices [3], demultiplexers [4], switches [5], modulators [6], and split-
ters [7] are among the devices included. Photonic crystals are used to create a
variety of efficient devices, including optical filters, logic gates, sensing devices,
and demultiplexers, which may be built in plasmonic waveguides with less than
one footprint area.
RI-sensing devices offer a wide range of uses in the biological and chemical
sciences, and have been extensively explored recently, including the measure-
ment of unknown concentration of the solution and pH value via RI variations
[8, 9]. As illustrated in Figure 3.1(a), an EM field is created by activating the
sensor element with light, which creates extremely focused SPs at the metal’s
surface. When the material under investigation comes into contact with the sen-
sor, the metal-insulator-metal waveguide’s neff variations cause λres to redshift,
as illustrated in Figure 3.1(b). Variations in the RI in the neighborhood of the
surface are extremely sensitive to SPs. Deviations in one of the properties of
the light connected to the SP, such as the λ res, intensity, or phase, can be used to
calculate the ΔRI.

FIGURE 3.1 (a) Diagram of a sensing method in plasmonic sensing devices. (b) Redshift in
λres because of the variation in RI in the adjacent medium.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp03.indd 70 9/26/23 4:55 PM

 Nanoplasmonic Sensors 71

3.2 PLASMONIC WAVEGUIDES

In terms of speed and data transfer rate, semiconductor-integrated circuits and recent
electronic-integrated devices are swiftly nearing their basic limits. One possible
answer to these issues is to employ light rather than e−s in the fundamental process-
ing components. Nevertheless, due to the prevailing diffraction limit of light when
the size of photonic device is near to or smaller than the λ of light in the material, the
feasibility of creating nanoscale photonic devices may be constrained. Bypassing the
diffraction limit, SPP EM waves linked to charge oscillation at the metal-dielectric
interface can accomplish localization of EM energy in nanoscale areas that are sig-
nificantly smaller than the λ(s) of light in the material. The EM field perpendicular
to the metal-dielectric boundary decays exponentially from the metal surface while
preserving long-range EM energy transmission from the surface. SPPs may be uti-
lized to make photonic components and optical signal processing devices at the sub-
wavelength scale because of this significant feature.
In recent years, plasmonic waveguides such as metal-insulator-metal, insulator-
metal-insulator, metal grooves, metal strips, metal wedge, and hybrid Bragg wave-
guides have been suggested. These waveguides may confine EM waves close to the
boundary, well beyond the diffraction limit of light. Metal-insulator-metal waveguide
formations are one of the most attractive approaches for building nanoscale photonic-
integrated circuits, and they are particularly well suited to domains of optical commu-
nication and sensing. As illustrated in Figure 3.2(a), this waveguide system is made up
of three layers, with a dielectric core of either air or another low RI material inserted
between two metal claddings. In [10], the dispersion relationship for metal-insulator-
metal configurations is analytically explained, as shown in Figure 3.2(b). Gordon [11]
and Dionne et al. [12] provided thorough analytical analyses of metal-insulator-metal
waveguides in 2006, while Bozhevolnyi et al. experimentally showed channel plasmon
subwavelength waveguide elements such as interferometers and ring resonators (RRs)
[13]. An insulator-metal-insulator is one more plasmonic waveguide construction in
which metal is inserted between two insulator claddings that have been employed in
several eye-catching utilizations for active [14] and passive [15] elements. In Japan, an
SPP-based research article for the fast detection of COVID-19 was recently published.

FIGURE 3.2 (a) Graphical illustration of a metal-insulator-metal waveguide and (b) disper-
sion plot of a metal-insulator-metal layered assembly adapted from [10].

BK-TandF-KAZANSKIY_9781032572949-230642-Chp03.indd 71 9/26/23 4:55 PM

 72 Photonics Elements for Sensing and Optical Conversions

Antibiotic-coated Au nanoparticles (NPs) experience a resonance peak shift when

the viruses are captured, resulting in a unique color transition [Kurabo. (2020).
Available online at: https://www.kurabo.co.jp/news/products/ (accessed July 7, 2020).].
Comparable techniques are widely used in pregnancy testing.
Several plasmonic waveguides, such as groove waveguides, multilayer metal-
insulator-metal waveguides, and nanowires, have recently been proven to allow the
long-distance propagation of SPPs. Fundamental and higher-order modes occur in
metal groove waveguides, generally identified as channel plasmon polariton (CPP)
modes [13]. Depending on the modes, the energy might be restricted to various
places in the groove. The taper angle can be used to change the modes and propaga-
tion distance. To properly direct the SPPs, the groove depth must not be substantially
less than the basic mode’s penetration depth. [13] described the setup, manufac-
turing, and characterization of CPP-based subwavelength waveguide components
working at telecom λ(s), such as Y-splitters, Mach–Zehnder interferometers (MZI),
and RRs. [5] showed a passive imbalanced MZI established on multilayer metal-
insulator-metal plasmonic waveguides, paving the path for a small EO modulator.
Owing to their single crystal formations and flat surfaces, chemically produced
nanowires have a low loss. The cylinder’s Bessel function [16] may represent the
basic modes of wires, and the transmission of SPPs on a metal wire can create a
helix. When the wire is put on a substratum, the top and bottom of the wire exhibit
faster and slower modes, respectively.

3.3 CRITICAL ISSUES ASSOCIATED WITH PLASMONIC

SENSING DEVICES ESTABLISHED ON METAL-
INSULATOR-METAL WAVEGUIDE FORMATIONS
In this section, we offer a short explanation of the most extensively utilized plas-
monic materials, fabrication methods, transmission losses, modal solutions, and
inquisition methods.

3.3.1 Favorite Plasmonic Materials

Polarization describes the interaction between a material and an EM wave. Although
polarization can be electric and/or magnetic in nature, for frequencies greater than
several hundred THz, the magnetic polarization of naturally occurring materials
is insignificant. The electric polarization can be defined by the complex electric
permittivity or dielectric function of material, which is signified by ε(ω). Whereas,
in fact, the real part of the dielectric function (ε′) defines the intensity of the polar-
ization induced by an exterior electric field, the imaginary part (ε″) explains the
losses faced in polarizing the material. Therefore, a material with insignificant loss
is related to a low value of ε″.
Primary loss mechanism in the soft UV, visible (VIS), and near-IR frequencies may
be arising from events associated with conduction e− and bound e− (interband effects).
Losses of conduction e− are caused by collisions between e−-e− and e−-phonon and by
scattering because of lattice defects or grain boundaries. As the conduction e− have

BK-TandF-KAZANSKIY_9781032572949-230642-Chp03.indd 72 9/26/23 4:55 PM

 Nanoplasmonic Sensors 73

almost a continuum of vacant states, their contact with an EM field is perfectly esti-
mated by classical theory. Through considering conduction e− as a three-dimensional
(3D) free e− gas, the Drude theory explains this phenomenon. The permittivity of a
material can be expressed according to the generalized Drude principle:

ω 2p
ε ( ω ) = ε ( ω )′ + iε ( ω )′′ = ε int − , (3.1)
i
ω  ω + 
 τ

where

ne 2
ω 2p = (3.2)
ε 0 m*

In Eq. (3.1), εint is an input caused by interband transitions (it is unity for a
perfectly free e−-gas) and τ is the mean relaxation time of conduction e−. The plasma
frequency (ωp) is assumed by Eq. (3.2), where n is the conduction e− density, and m*
is the effective optical mass of conduction e−. Generally, εint counts on λ, which is
usually described by adding the Lorentz oscillators terms, but for some frequency
domains, it can be estimated as constant, as presented in Table 3.1.
Plasmonic utilizations necessitate materials with negative ε′. This condition is
met by materials with a plasma frequency greater than the preferred application
frequency. Metals are frequently chosen because of their high electric conductivity
and large plasma frequencies. The material properties for high-conductivity metals,
as indicated in the references, are summarized in Table 3.1. Among the metallic
elements, Ag has the lowest size and is the best alternative at optical frequencies.
Because of its minimal imaginary part of permittivity and superlative e− conduc-
tion, Ag is regarded the most promising material in the near-infrared (NIR) range.
The disadvantage of Ag is that it is easily oxidized, which limits its utilizations
and allows Au to be an acceptable option in this frequency range. At lower NIR
frequencies, Au is frequently used because it is chemically stable under a variety
of circumstances. In the VIS continuum, however, Au suffers substantial interband
losses at λ(s) below or about 500 nm. Cu, too, suffers from substantial interband loss
over most of the VIS continuum. Al is the most suited material for UV or deep UV
(DUV) utilizations, for example, 193 nm light for DUV optical litho.

TABLE 3.1
Drude Model Parameters for Metals
Metals εint ωp (eV) Γ (eV) ωint (eV) Reference
Silver (Ag) 3.7 9.2 0.02 3.9 [17]
Gold (Au) 6.9 8.9 0.07 2.3 [17]
Copper (Cu) 6.7 8.7 0.07 2.1 [17]
Aluminum (Al) 0.7 12.7 0.13 1.41 [18]

BK-TandF-KAZANSKIY_9781032572949-230642-Chp03.indd 73 9/26/23 4:55 PM

 74 Photonics Elements for Sensing and Optical Conversions

For plasmonic utilizations around the optical frequencies, Ag and Au were pri-
marily the materials of choice. Future plasmonic utilizations, however, need even
lower losses to make full use of their prospective. Cu is an alternative plasmonic
material with almost comparable interband transition and optical damping as Au
in the wavelength range 600 nm to 750 nm. Unfortunately, Cu is also susceptible
to oxidation. To overcome the corrosion problem, graphene has recently been used
on top of Cu or Ag. Because graphene is chemically and mechanically inert, it is
kept away from fluidics. Cu-graphene and Ag-graphene optical fiber surface plas-
mon resonance (SPR) sensors provide long-term stability and reliability. Niobium is
another one-of-a-kind plasmonic material with a high chemical resistance and high
mechanical strength. The linking between niobium film and silica glass is so strong
that no extra bonding layer is required. SPR sensors based on indium tin oxide have
lately drawn a lot of interest because of its low bulk plasma frequency. Furthermore,
it has the same optical damping as Ag and Au.

3.3.2 Fabrication Methods

Metals are more difficult to microfabricate than semiconductors (mainly noble metals).
With the use of sputtering or evaporation methods, plasmonic metals may be deposited
in a vacuum. The deposition rate is generally restricted to <1 nm/s to achieve adequate
uniformity and quality. Because the necessary plasmonic system height is usually
modest, the slow deposition rate is not a big problem. The electroplating [19] technique
can produce faster growth rates, but the surface quality may deteriorate, making it
unsuitable for SP waves. A prototype shift from the photoresist (PR) to noble metals
is difficult due to the restricted number of efficient etching techniques for Cu, Ag, and
Au. In Cl2-based settings, however, Al may be plasma etched.
For pattern metals, two main techniques are widely employed. The first method
involves image reversal and lift-off, which necessitates resist overhang after advance-
ment. MicroChem Corp. [20] sells special PRs for the lift-off process, for example
lift-off resist. Ion milling is a second method for patterning metal. A focused ion
beam may be employed to mill designs directly onto Au or Ag film [21]; neverthe-
less, redeposition from the beam source is common. Although an Ar ion beam [22] is
a viable option to etch metals, it necessitates the use of hard masks and may increase
the manufacturing stages and processing complication. Several techniques, includ-
ing e−-beam and ion-beam litho, may be used to fabricate n-structures for plasmonic
sensing devices, enabling the cost-effective production of n-structures across vast
regions. Furthermore, wet chemical etching or vapor deposition methods may be
used to make metal NPs quickly and easily. The purpose of this chapter, however, is
not to go through these approaches in depth. We recommend some literature [23, 24]
that provides an in-depth discussion of this essential issue as an alternative.

3.3.3 Transmission Loss
Unlike dielectric waveguides, where transmission loss is unimportant, plasmonic
waveguides are very lossy due to the presence of metal in the waveguide system.
Major ohmic losses impact the transmission of directed SPs, limiting the maximum

BK-TandF-KAZANSKIY_9781032572949-230642-Chp03.indd 74 9/26/23 4:55 PM

 Nanoplasmonic Sensors 75

propagation length. Numerous structures have been created utilizing collections of

nanosize dimension characteristics to offset these losses [25, 26]. A system featuring
a thin lossy metal sheet resting on a dielectric substratum and enclosed by a varied
dielectric superstrate was used to attain the longest propagation length of 13.6 mm [26].
In general, transmission loss and mode confinement are trade-offs in plasmonic
waveguides. The bigger the transmission loss, the smaller the mode size. The modes
validated by plasmonic waveguides are usually more complicated in form. As a result,
determining the mode confinement of plasmonic waveguide is difficult. This problem
has been debated, with the conclusion that the definition of the mode region should be
determined by the specific purpose [27]. Energy dissipation occurs in both electronic
and plasmonic circuits. The SP transmission length during which the strength of the
SPP falls to 1/e of its initial value is used to characterize the transmission loss.
Other variables, like as the spacer and insulator medium in the slot, also influence
SPP propagation in metal-insulator-metal waveguides. However, if the waveguide is
tens of kilometers long and anagogic to electronic devices, the losses caused by the
metal absorption are considerable. As a result, relaying the signals to configure SPP-
based devices is critical.

3.3.4 Inquisition Method
Various light properties are used to investigate the transmission spectra of plasmonic
sensor devices. As a result, plasmonic sensing devices may be divided into three
groups based on the inquisition scheme used, for instance intensity, phase, or wave-
length. Lasers or superluminescent light-emitting diodes (SLEDs) with narrowband
filters are mandatory to excite the plasmonic modes. When compared to broadband
sources, these light sources have superior stability and power, resulting in a higher
signal-to-noise ratio. Any variation in the intensity of light linked to a plasmonic
mode is often monitored with a one-dimensional (1D)-PDA (personal digital assis-
tant) or 2D detectors like CCD (continuity of care document) or CMOS (complemen-
tary metal–oxide–semiconductor) cameras.
Phase inquisition is the process of converting phase variations into physically
detectable intensity signals by interacting a data-carrying light with a reference
beam. When common route interferometry methods are employed, the primary
advantage of phase investigation is evident, because the reference and signal beams
are accessed over the same optical channel and hence are impacted by the same
noise components. As the measured signal is constantly referenced, phase inquisi-
tion eliminates background and nonspecific noise from the environment. Different
ways to such evaluations have recently been presented for plasmonic sensing devices
that use the phase inquisition method. Retrieving phase information from intensity
data is one of the most intriguing ideas. The most common method is polarime-
try, which involves extracting phase information from the interference of s- and
p-polarized light using a polarizer. The electric field perpendicular to the plane of
incidence is denoted by s-polarization, whereas the electric field parallel to the plane
of incidence is denoted by p-polarization. Their relative phase difference is deter-
mined from a series of intensity measurements using angular modulation created
by a rotating analyzer or phase modulation formed by a liquid crystal variable wave

BK-TandF-KAZANSKIY_9781032572949-230642-Chp03.indd 75 9/26/23 4:55 PM

 76 Photonics Elements for Sensing and Optical Conversions

plate (LCVWP) and a static output analyzer [28]. Because this is effectively the
equivalent idea as ellipsometry and is a mature approach for the depiction of thin
films, phase inquisition-based plasmonic sensing devices might benefit from com-
mercially available equipment.
The most common inquisition approach employed in RRs based on plasmonic
systems is the measurement of variations in the λres [29]. λ inquisition-based optical
systems often employ a polychromatic light source, for instance a halogen lamp or a
SLED, to span the whole continuum where λres is likely to be seen. In terms of light
continuum, halogen lamp technology is superior. When using the fixed incidence
angle configuration and λ inquisition approach, this light source is preferable. A
CCD-, CMOS-, or PDA-based spectrometer is often used to collect continuum of
light coupled to an SP, and the spectral location of the plasmonic feature is scruti-
nized utilizing suitable feature-tracking algorithms.

3.4 PERFORMANCE CHARACTERISTICS OF SURFACE

PLASMON POLARITON RESONATOR
In this section, the most studied metal-insulator-metal sensing device configurations
for RI-sensing purposes, for example circular ring resonators (CRRs), split square
ring resonators (SSRRs), notched RRs, gear-shaped nanocavities, square ring reso-
nators (SRRs), and hexagon cavities, among others, are reviewed. Scientists pro-
posed these sensing devices employing either a side-coupled cavity or edge-coupled
cavity. Because of the short coupling length, edge-coupled resonators have tinier
FWHM in contrast to side-coupled configurations [30]. Most metal-insulator-metal
plasmonic resonator sensing devices are currently being proposed theoretically and
numerically. However, it is necessary to realize these concepts experimentally and
to make them available in biomedical purposes. Now, we will discuss two major
attributes of plasmonic sensing devices based on how their performance is evaluated.

3.4.1 Sensitivity
The capacity to detect variations in the RI is the most utilized performance param-
eter of plasmonic sensing devices. This is commonly expressed as the bulk RI S,
which is defined as sensitivity S = Δλres /Δn, where λres is the λ at which the SP exci-
tation arises, and Δn is the RI shift. Table 3.3 shows the S of a variety of plasmonic
sensing device designs that have been published. The S varies significantly depend-
ing on the type of the EM mode, resonant λ, excitation geometry, and substratum,
among other aspects. S has been studied in relation to the geometric characteristics
of metal-insulator-metal waveguide RRs encased by a (homogeneous) sample by
numerous researchers [31–50].
Plasmonic metamaterials (MMs) made from Au nanorods (abbreviated as NRs)
are used to represent the dispersion of SPPs localized on a metal sheet [51], therefore
enhancing S. The plasmonic MMs support a guided mode with resonant excitation
conditions equivalent to the SPP mode of a smooth metal sheet when the spacing
between the NRs is smaller than the incident λ. This MM has 3,000 nm/RIU S to RI
variations in the medium between the rods. Circular dichroism can be used instead

BK-TandF-KAZANSKIY_9781032572949-230642-Chp03.indd 76 9/26/23 4:55 PM

 Nanoplasmonic Sensors 77

of extinction-based sensing methods for chiral nanostructures. Their spectral char-

acteristics are sharper than chiral particles because of their POL-dependent (polar-
ization-dependent) spectra. These characteristics offer chiral plasmonic nanohelices
with astonishing RI S of 1,091 nm/RIU and FOM* = 2,800 RIU−1 [52].

3.4.2 The Figure of Merit

The capability of a plasmonic sensing device to detect small variations in the RI is
proportionate to S and inversely related to the spectral width of the resonance feature
that is being tracked. The Figure of Merit (FOM) is expressed as the product of these
characteristics and is represented as FOM = S/FWHM. Even though the FOM is
quite high in certain studies [34, 41, 46, 48], it is computed differently in each study
and is typically referred to as FOM*. For example, the expression ΔR/(RΔn) at a fixed
λ is used in [53] to calculate an extraordinary FOM* = 2.33 × 104, where R signifies
the variation in reflection intensity due to variations in the refractive index (n)
of the ambient medium and R is the reflection rate in the sensing device system.
FOM = 4.05 × 104 is calculated in [46] using the equation ΔT/TΔn, where T indicates
the transmittance in the suggested formations and ΔT/Δn signifies the transmission
variation at a constant λ caused by a RI modification. FOM is frequently utilized to
evaluate diverse plasmonic sensing devices from a sensing standpoint.
Fano resonances (FRs) are a technique for improving FOM by lowering the spec-
tral width of a plasmonic feature. The goal of mode coupling research has been to
achieve a small FWHM. Cetin et al. [54] revealed that Fano resonant asymmetric ring/
disk systems have spectrally sharp features with FWHM as small as 9 nm and a high
S as big as 648 nm/RIU. FR is a sort of resonant scattering phenomenon that happens
when a discrete state and a continuous state are coupled and interfere [55]. In metallic
n-structures, FR has been explored as the ideal characteristic for preventing the dif-
fraction limit of light [56] produced by SPPs. Consequently, several plasmonic forms
have been suggested to produce FR, including rectangular cavities [57], plasmonic
n-clusters [58], n-slits [59], and metal-insulator-metal waveguide structures [60].
Because of their sharp and asymmetric line structure [61], which allows the trans-
mission continuum to be rapidly decreased from peak to trough, plasmonic sensing
devices based on FR are expected to be very sensitive. The transmission continuum’s
FWHM is quite small, which increases the sensing device’s sensing resolution substan-
tially [62]. In 2019, several RI-sensing devices based on FR were disclosed, although
only a handful are cited here [63–71]. In terms of FOM, Fano resonant systems may
be able to outperform Lorentz resonant systems; nevertheless, the insufficient con-
trast of plasmonic spectral characteristics may restrict the system’s capacity to convert
advances in FOM to the capacity to identify smaller variations in the bulk RI.

3.4.3 Newly Developed Refractive Index Sensing Devices

Many metal-insulator-metal resonator models have been suggested and studied for
RI-detecting purposes in recent years [30–75]. Figure 3.3(a–h) depicts a summary of
such arrangements. Some display Lorentzian line forms, whereas others show FRs,
which enhance FOM considerably.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp03.indd 77 9/26/23 4:55 PM

 78 Photonics Elements for Sensing and Optical Conversions

FIGURE 3.3 (a) Semi-RR, (b) side-coupled hexagonal cavity resonator, (c) side-coupled RR,
(d) double-sided coupled RR, (e) gear-shaped nanocavity resonator, (f) notched RR, (g) two
metal-insulator-metal waveguide-coupled circular RR, and (h) side-coupled split circular RR.

Plasmonic sensing devices are smaller and easier to incorporate on-chip than fiber
sensors [72, 73]. The S, on the other hand, is not as high as that of fiber sensors. As
a result, the major focus of research is on improving the S of plasmonic sensing
devices. In [32], the author has proposed an RI-sensing device with an S higher than
most previously described sensing devices.
A metal-insulator-metal waveguide-coupled asymmetric resonator for biosensing
and slow light purposes was presented. The transmission continuum of the device
had electromagnetically induced transparency (EIT)–like features. EIT denotes
a quantum mechanical phenomenon involving quantum destructive interference
between resonant excitation channels at the higher atomic level, which can affect
an atomic system’s optical properties. This sensor’s S and FOM are 806 nm/RIU
and 66, respectively [76]. Because of steep dispersion and strong resonance at the
transparency window, many utilizations such as ultrafast switching and filtering are
based on the EIT effect [77, 78].
Given the progress of highly integrated photonic circuits, the mechanism of multi
FRs in single one subwavelength metal-insulator-metal formations has received the
most attention. As a result, metal-insulator-metal waveguide-based composite con-
figurations, such as groove-cavity composite formations and cascaded groove for-
mations, have been successfully suggested and explored [34, 44, 46, 70]. In the IR
wavelength region, dual Fano transmission peaks with asymmetrical line forms have
been obtained. Chen et al. [79] suggested a novel method for producing double FRs
by combining two equal stub resonators combined to a rectangular cavity offers S
and FOM of 1,100 nm/RIU and 91, respectively. An end-coupled slot resonator with
extraordinary S and FOM was used to create multiple FRs [80].

3.4.4 Examples
Here, we have presented two detailed examples for designing the plasmonic sensor
for RI sensing purpose via the finite element method (FEM). In the first illustration,
the multichannel metallic dual nanowall (NW) square split RR is explained, which

BK-TandF-KAZANSKIY_9781032572949-230642-Chp03.indd 78 9/26/23 4:55 PM

 Nanoplasmonic Sensors 79

can be employed as a filter/sensor, whereas in the second, the plasmonic sensor with
nanodots is presented for enhanced S.

3.4.4.1 Numerical Modeling of a Multichannel Metallic

Dual Nanowall Split Square Ring Resonator
This section numerically analyzes a new metal-insulator-metal waveguide construc-
tion consisting of a multichannel dual metallic NW SSRR side paired with a metal-
insulator-metal bus waveguide. A square ring geometry has the benefit of revealing
additional resonance modes that are attributed to the involvement of bends in a rect-
angular ring construction [81]. The suggested tool is customized by embedding a
dual metallic NW within a standard metal-insulator-metal square ring, which dis-
rupts the metal-insulator-metal RR’s symmetry and disqualifies these resonators
from being classified as traveling-wave resonators. By appropriately placing the
NW, the resonance modes may be extremely stimulated or repressed. The effects of
NW on filtering and sensing properties are quantitatively investigated using FEM.
Other related characteristics are carefully chosen, such as the relative permittivity
(εr) of Ag and the RI of the air. The relative permittivity of Ag abides by the Drude–
Lorentz model as follows:

ω 2p
εm (ω ) = ε∞ − , (3.3)
ω ( ω + iγ )

where ε∞ = 3.7 is the dielectric constant at an infinite frequency, ωp = 9.1 eV is the

bulk plasma frequency of free conduction e−, and γ = 0.018 eV is the e− collision
frequency. ω is the angular frequency of the incident light in a vacuum. Because of
its small imaginary part of εr in the NIR band, Ag is used in the arrangement. As a
result, as compared to Au and Al, its power consumption is minimal. Furthermore,
the Ag layer has great adherence to the substrate and can be easily patterned with
strong etch selectivity in HNO3 and H2O.

3.4.4.1.1 Selective Wavelength Filter Configuration

Metal-insulator-metal waveguide filter configuration in this example comprises four
dual NWs SSRR, two on each side of the bus waveguide. The outer length of rings
is signified by l1, l2, l3, and l4, where l1 > l2 > l3 > l4. The inner length of the ring
is li–d, where i = 1, 2, 3, and 4 and d is the cavity width. The distance between bus
waveguide and rings is represented by g, which is static during the examination pro-
cess. The dual NWs are introduced in the ring; the width of the wall is represented
by S. All the four rings are coupled with the output ports, Pi that transmits the filtered
λ contingent on the resonance conditions. The graphical illustration of the filter con-
figuration is presented in Figure 3.4. The geometric variables of the dual NW SSRR
are specified in Table 3.2.
When a TM-polarized light is introduced into a metal-insulator-metal waveguide,
it excites SPPs that propagate along the metal’s interface. Only the fundamental TM
mode can be sustained in this configuration since the waveguide width is less than

BK-TandF-KAZANSKIY_9781032572949-230642-Chp03.indd 79 9/26/23 4:55 PM

 80 Photonics Elements for Sensing and Optical Conversions

FIGURE 3.4 Graphical illustration of a multichannel λ filter.

TABLE 3.2
Geometric Variables of the Dual NW Square Split Ring Resonator (SSRR)
W_bus (nm) l1 (nm) l2 (nm) l3 (nm) l4 (nm) s (nm) d (nm) g (nm)
30 100 90 80 70 5–25 5–20 20

the incident λ. Figure 3.5(a) shows the E-field mapping at the cross section of the
metal-insulator-metal waveguide. The λres may be calculated by changing the RR’s
effective index (neff ). As shown in Figure 3.5(b), the neff of the resonator is completely
dependent on the size of the metal-insulator-metal waveguide. Using the Maxwell
equations and boundary conditions, the neff in the waveguide may be calculated.
Furthermore, by adding a material with a marginally higher RI into the surrounding
medium, neff may be changed. Consequently, the resonant modes shift to longer λ(s)
as the RI of the medium increases. This capability permits this configuration to be
used in sensor purposes.
As the NWs are positioned in the middle of each ring, integer modes in the VIS
and near-IR λ domain are stimulated. Figure 3.6(a–d) displays the excitation of
modes at λres for the corresponding rings.

3.4.4.1.2 Biological-Sensing Applications

In addition to filter applications, the suggested design may be employed in biosens-
ing utilizations [82–85]. By increasing the RI in the medium, λres can be moved to

BK-TandF-KAZANSKIY_9781032572949-230642-Chp03.indd 80 9/26/23 4:55 PM

 Nanoplasmonic Sensors 81

FIGURE 3.5 (a) E-field mapping at the cross section of a metal-insulator-metal waveguide.
(b) Real part of the neff of a metal-insulator-metal waveguide for bus waveguide width 25 nm,
30 nm, 35 nm, and 40 nm.

FIGURE 3.6 E-field mapping in the SSRR at λres: (a) 838 nm, (b) 753 nm, (c) 675 nm,
(d) 598 nm.

higher λ(s). As a result, the suggested structure can simultaneously serve as a sensor
and a filter. The S, FOM, and Q-factor are utilized to determine the RI sensor’s
performance quantitively. We filled the adjacent medium with RI = 1.3 to 1.5 to esti-
mate the S of the suggested configuration and displayed the transmission continuum
attained from P1, P2, P3, and P4, as revealed in Figure 3.7. A significant redshift
in λres is seen when the RI of the medium increases. This improvement implies that
altering the RI in the cavity might provide another way to adjust the structure’s
filtering properties.
S, FOM, and Q-factor all grow linearly with the size of the SSRR, as shown in
Figure 3.8(a)–(c). Dual NW SSRR 1 has the best S, FOM, and Q-factor of 793.3 nm/
RIU, 52.9, and 82.1, respectively. In Table 3.3, the sensor’s best continuum charac-
teristics are listed.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp03.indd 81 9/26/23 4:55 PM

 82 Photonics Elements for Sensing and Optical Conversions

FIGURE 3.7 The transmission field of dual NW SSRR configurations for different values
of n. The output from (a) P1, (b) P2, (c) P3, (d) P4.

FIGURE 3.8 (a) S versus L, (b) Q-factor, (c) FOM versus L.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp03.indd 82 9/26/23 4:55 PM

 Nanoplasmonic Sensors 83

TABLE 3.3
Spectral Features of a Dual NW SSRR
SSRR 1 SSRR 2 SSRR 3 SSRR 4
S (nm/RIU) 793.3 711.4 626 537.8
Q-factor 82.1 73.8 65.9 57.7
FOM 52.9 47.4 41.7 35.9

3.4.4.2 A Technique for Boosting the Sensitivity of a Conventional

Plasmonic Metal-Insulator-Metal Square Ring Resonator
In this illustration, a square ring resonator (SRR) arrangement combined with
Ag nanodots is presented. The width of the bus waveguide, the side length of the
ring, cavity width, and the gap between bus and ring are signified by w, L, c, and
g, respectively. The ring waveguide is connected to the output port, which transfers
the filtered λ reliant on the resonance conditions. A surface integration of the output
port yields the transmission continuum. As seen in Figure 3.9(a), there are seven
nanodots on either side of the cavity. The nanodots are put in a cavity with a period
of 3r, where r is the nanodots’ radius. The sensor’s structural variables are listed
in Table 3.4. The E-field mapping of light in the conventional ring arrangement at
λres = 712 nm is shown in Figure 3.9(b).
As seen in Figure 3.9(c), an array of nanodots acts as a grating, allowing the
energy in the cavity to build up. The inset illustrates the amplification of SPPs at
nanodot’s borders, which allows for more light contact with the surrounding medium
and therefore increases the sensor’s S.
Filling the surrounding medium with a variety of RIs is used to determine the
resonator’s continuum response. For the λ range of 600 nm to 1,200 nm, the normal-
ized intensity at the sensor’s out is computed for n = 1.0, 1.1, 1.15, 1.2, 1.25, 1.30, and
1.35. The S is computed in Figure 3.10(a) concerning the number of nanodots. Seven
nanodots are initially inserted in one side length of the ring in the cavity, followed by
the addition of the other three sides regularly. The nanodot’s radius is set at 2.5 nm.

FIGURE 3.9 (a) Schematic illustration of a metal-insulator-metal SRR formation with

nanodots embedded in the cavity. (b) E-field mapping in the conventional SRR at λres.
(c) E-field mapping in the nanodots embedded SRR formation at λres.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp03.indd 83 9/26/23 4:55 PM

 84 Photonics Elements for Sensing and Optical Conversions

TABLE 3.4
Geometric Variables of the Sensor
W (nm) L (nm) g (nm) c (nm) r (nm)
30 100 10 15 2, 2.5, 3, 3.5, 4

TABLE 3.5
The S of the Nanodots Embedded SRR
Nanodots on Nanodots on Nanodots on Nanodots on
Standard Ring One Side Two Sides Three Sides Four Sides
Smax (nm/RIU) 670 720 740 732 738

The inclusion of nanodots in the cavity results in a significant increase in S. Table 3.5
shows the resonator’s maximum Smax.
The resonator’s S, Q-factor, FOM, and λres are all affected by the size of the
nanodots. As a result, we looked at the influence of nanodots radius on spectral com-
petence by keeping the number of nanodots at 28 (seven on each side of the ring),
as shown in Figure 3.10(b). The resonator’s Smax, FOM, and Q-factor are shown in
Table 3.6. The FOM and Q-factor of the resonator fall dramatically as the size of
the nanodots increases, resulting in a substantial increase in S. The widening of the
FWHM is linked to a reduction in the FOM and Q-factor. As shown in Figure 3.11(a),
the highest FOM and Q-factor have been obtained for the resonator with nanodots
of size 2 nm, whereas Smax = 907 nm is obtained for the nanodots of radius 4 nm due
to the intensification of SPPs at the metal-insulator-metal bus waveguide bound-
ary. Furthermore, as seen in Figure 3.11(b), λres exhibits a redshift as the size of the
nanodots increases.

FIGURE 3.10 The S of the SRR by placing nanodots of (a) 2.5 nm radius on different sides
of the cavity; (b) various sizes in all four sides of the cavity.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp03.indd 84 9/26/23 4:55 PM

 Nanoplasmonic Sensors 85

TABLE 3.6
Spectral Competence of the Nanodots Embedded Metal-Insulator-Metal
S-R-R
Conventional Nanodots Nanodots Nanodots Nanodots Nanodots
Variables RR r = 2 nm r = 2.5 nm r = 3 nm r = 3.5 nm r = 4 nm
Smax (nm/RIU) 670 720 738 780 827 907
Q-factor 52.4 52.6 45.9 46.6 45.7 41
FOM 49.2 50.4 43.4 44.3 43.7 40.8

FIGURE 3.11 (a) FOM and Q-factor versus the size of nanodots; (b) λres shift versus nanodots.

3.5 SURFACE PLASMON RESONANCE SENSOR ARRANGEMENT

The resonance of SP is the secret to great performance in the design of SPR sensors.
A p-polarized light can best couple the resonance state at the interface of the thin
metal and dielectric medium at a given λ and incidence angle. When the SP’s wave
vector (βSP) value equals the incident light’s propagation constant, this happens. To
have a better understanding of this phenomenon, the SP wave vector in the nano-
metal film surface is hypothesized in the following equation:

 2π εM εD 
βSP = Re  , (3.4)
λ εM + εD 

where λ is the wavelength of the incident light, while εM and εD are the real part value
of the metal and medium dielectric constant, respectively. One more important factor
is the propagation length (Plength) of the SP wave articulated by the following formula:
3
π εiM  ε M ε D  2
Plength = , (3.5)
2λ ε 2M  ε M + ε D 

where the εiM is the imaginary part value of the metal-dielectric constant. A certain λ
necessitates a specific configuration in the SP generation to meet the incident light’s

BK-TandF-KAZANSKIY_9781032572949-230642-Chp03.indd 85 9/26/23 4:55 PM

 86 Photonics Elements for Sensing and Optical Conversions

propagation constant at a precise angle of incidence. The availability of a light source

determines the priority of investigation and modification of the overall sensor setup.

3.5.1 Surface Plasmon Resonance Sensing with

Nanoparticles and Thin Films
SPR-based methods make use of the optical phenomenon of a fixed λ resonance
involving free e− on a metal surface. Metallic NP coloring might be the most persua-
sive proof of this optical phenomena. When Au particles are smaller than a nanometer,
they lose their typical yellow or bulk Au appearance. For example, spherical Au NPs
with a diameter of 13 nm look brilliant red. This optical phenomenon is described by
the fact that light with the suitable wavevector can resonate with the metal’s free e−,
causing a collective oscillation of the metal’s free e− known as SP and instigating the
intensity of the light beam to be attenuated at a specific λ, as shown in Figure 3.12.
The simplicity with which LSPR may be measured explains its rising popularity.
For NPs or n-structured substratum immobilized on a transparent surface, UV-VIS
equipment is often employed, whereas dark-field microscopy may also be used for
NPs immobilized on solid support [86]. The color of the light absorbed is deter-
mined by the dielectric constant or RI of the chemical environment in the vicinity of
the NP, in addition to the type and geometric variable of the metallic nanomaterials.

FIGURE 3.12 Graphical illustration of biosensing experiments with LSPR. (a) Biosensing
scheme on an Au NP, onto which a chemical layer is placed to selectively capture a molecule
in the occurrence of interfering agents. (b) Excitation in extinction spectroscopy by a beam
of light entering resonance at a precise λ with the free e− cloud. (c) The extinction continuum
redshifts with the binding of molecules to the Au NP.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp03.indd 86 9/26/23 4:55 PM

 Nanoplasmonic Sensors 87

The λ of the resonance λ performs a redshift as the RI of the solution increases,

which is conventionally referred to as the bulk S, which is measured in nm/RIU. SPR
has a detecting range of a few tens to hundreds of nanometers, which is restricted by
the depth of penetration SPR probes have into the solution. The λ shift seen in SPR
sensing is frequently determined by the quantity or size of molecules adsorbed on
metallic surface, and this notion lies at the heart of SPR-sensing devices. SPR mate-
rials having a shallow penetration depth are more responsive to the development of a
molecular adsorbate, and, as a result, to analyte detection. High S is a dynamic field
of study in SPR sensing, as it is in other analytical methods, to further increase the
detection of even lower concentrations of analytes.

3.5.1.1 Kretschmann Configuration

The pioneering work of Otto Kretschmann (KR) and Raether in the late 1960s dis-
covered that the SPR phenomenon occurs when the light at a metal/liquid interface
induces SP waves, which are linked with collective excitations of e− in the metal. The
majority of commercially available SPR biosensors still employ KR and Raether’s
technique of focusing p-polarized light through a glass prism and then reflecting it
from a thin Au layer of 50 nm placed on its surface. When the tangential x-compo-
nent of the incident optical wave vector, k x, is equal to kSP, the SP wave vector, the
pumping light energy is reassigned to the SPs. Under total internal reflection circum-
stances, the SPR effect results in the form of a strong resonant dip in the reflected
light intensity at a specified angle (θ). In reality, because of the strict need for optical
and evanescent wave vector matching, an alteration in the RI of the ambient medium
that touches Au (within 200–300 nm) would have a significant impact. Figure 3.13
shows a graphical representation of the KR setup.
Most SPR-sensing devices monitor RI variations because of binding events near the
Au layer that cause SPR to arise. More specifically, the aggregation of biological mate-
rial within that thin layer, which occurs because of interactions with macromolecules

FIGURE 3.13 SPR detection mechanism in the KR setup.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp03.indd 87 9/26/23 4:55 PM

 88 Photonics Elements for Sensing and Optical Conversions

found in the liquid and species formerly immobilized on the sensing device’s surface,
influences the RI consequently results in a change in θ observed in real time. The
detection limit (DL) of existing equipment for RI change is, at best, 10−7 RIU. It results
in a DL of around 0.1 pg mm−2 of biomaterial collecting on the biosensor surface.
Angle scanning, λ scanning, and SPR imaging are three types of equipment used in
SPR sensing investigations. Angle scanning instruments are the highly popular com-
mercial equipment, and they are often attained by optical or mechanical means.
For analyzing samples on a few channels, such instruments offer great S. An SPR
image-sensing device’s function is based on measuring the intensity of a reflected
beam incident at a certain angle and λ. An imaging camera detects a change in inten-
sity due to variations in SPR response. The SPR-sensing device designates areas for
various analytes. SPR image-sensing devices are useful because they provide multi-
plexing capabilities for evaluating numerous analytes. SPR sensing based on the KR
arrangement has evolved into a sophisticated method for examining biomolecular
interactions without the use of labels. The ability of SPR sensing to determine bind-
ing constants and binding partners has been particularly useful in biological sci-
ences. We have tabulated a few of the sensing device configurations based on the KR
configuration in Table 3.7 for the reader’s convenience.

TABLE 3.7
The Summary of SPR-Sensing Device Advancement
No. Scheme Target Sample Performance Ref
1 KR, SPR, λ = 632.8 nm, the µ-array Si coating SAM Decay length ~4 µm 87
sensing membrane
2 KR, dielectric mirror TiO2/SiO2 Glucose solution Res: 1.28 × 10–5 RIU Dynamic 88
sensing formation, λ = 632 nm range: 1.331–1.5 RIU
3 KR, angular int., λ = 632.8 nm Sugar content in LOD: 0.01–0.05% 89
carbonated drink
4 KR, Au/SnO2 sensing film, angular Ammonia gas S 0.055 o/ppm (0.5–250 ppm) 90
int., λ = 633 nm
5 KR, Au/ZnO sensing film, angular DNA of N. LOD: 5 ng/µL 91
int., λ = 633 nm meningitidis
6 KR, intensity mod. at dual λ(s) DNA hybridization LOD: 2 × 10–6 RIU 92
references
7 KR CaF2 prism, TiO2/Au layer on CO2 5X S improvement 93
sensing, angular int., λ = 633 nm
8 KR, angular int., diverging laser Ethanol solution LOD: 5 × 10–6 RIU 94
beam, λ = 637 nm
9 KR with the rotating diffuser, SPRi, IgG, BSA Proof of concept for 95
CCD camera detector, λ = 633 nm, multi-sample detection
high throughput, and disposable
sensing configuration

DNA – deoxyribonucleic acid; IgG – immunoglobulin G; BSA – bovine serum albumin; LOD – limit of detection

BK-TandF-KAZANSKIY_9781032572949-230642-Chp03.indd 88 9/26/23 4:55 PM

 Nanoplasmonic Sensors 89

FIGURE 3.14 The configuration of the SP grating coupler.

3.5.1.2 Grating Coupled Surface Plasmon Resonance

Grating-enhanced SPR excitation has been a remarkable pioneer in SPR arrangement.
Figure 3.14 demonstrates the grating configuration of SP excitation. Wood pub-
lished the first SPR excitation phenomena on the grating arrangement in 1902 [96].
Enhancements were subsequently made until Rayleigh established an analytical
explanation to the Wood’s anomalies explaining the diffraction angles used in grat-
ing methods by means of the following formula [97]:

λ
sin ( θp ) = sin θ + p , (3.6)
Λ
where θ is the incidence angle of p-polarized light, θp is the diffraction angle, λ is
the λ of the incoming light, and Λ is the period of groove. This narrative permits
the control of diffraction angle of any scattered order from the grating period p, the
incidence angle of light (θ), and the λ. The passing-off of the order n occurs when
sin (θp) = ±1. Hence, from Eq. (11), the λ of a continuum producing the passing-off
of a diffracted order are expressed as [98]:

λ
n = − sin ( θm ) ± 1; m = ±1, ±2, ±3, .... (3.7)
Λ
The λ of the passing-off defined earlier is known as the Rayleigh λ. The rela-
tion of incident light vector and grating configuration in resonance condition can be
articulated by [99]:

λ  εM εD 
nDsin θ + m = ± Re   + ∆nef , (3.8)
Λ  εM + εD 
where nD is the RI of the medium and ∆nef = Re{∆β (λ /2π)}, and Δβ is the propa-
gation constant fluctuating in the presence of the grating formation. In the grating
coupler-based SPR-sensing devices, the prism does not need to be directly coupled
to the chip.
The sensing device chip substratum is embossed with a diffractive grating that
separates the incident light beam into several beams that exit at different angles.
These beams can be further collected and separate by prism sets, making it possible
to observe the intensity of the spectra. Since a sensing device chip is not directly

BK-TandF-KAZANSKIY_9781032572949-230642-Chp03.indd 89 9/26/23 4:55 PM

 90 Photonics Elements for Sensing and Optical Conversions

linked to a prism, the systems needed for chip handling can be significantly sim-
plified. This method also eliminates the requirement that the width of the plas-
mon-forming metal layer be precisely regulated, although the need to emboss the
diffraction grating on the chip complicates the production process. Another impor-
tant consideration is that since the light must pass through the sample to enter the
sensing device chip, the optical clarity of both the analyte and the buffer solution.
We are referring to a few recently proposed grating-based SPR-sensing devices that
might interest the readers [100–103].

3.5.2 Breakthrough in Localized Surface Plasmon Resonator Sensing

Developing new plasmonic materials that are highly sensitive to bulk RI directs to
greater variations in the SPR interaction and low DL. Colloidal chemistry has been
advanced to a great extent over the past decade to realize metallic NPs of different
forms and materials, containing spheres, cubes, rods, and pyramid-shaped NPs. The
excitation λ and S are vastly dependent on the geometric variables of the NPs. Even
though widespread investigation has been dedicated to improving the analytical prop-
erties of metallic NPs, yet there is no agreement on the optimal shape and size of the
NP. Nanospheres are simple to fabricate but at the cost of the low S of ~76 nm/RIU
[104]. Au NRs are more sensitive than nanospheres (366 nm/RIU) and are stimulated
in the near-IR, a spectral range that is relatively free of interference, but long-term
strength of Au NRs must be enhanced for sensing utilizations [104]. Certain forma-
tions such as cubes, hollow spheres, and pyramids are ideal substratum for enhanced
Raman scattering, but they have not been commonly used in SPR sensing.
The current advancement in producing precise dimension of NP and n-structured
materials have facilitated the advancement of novel analytical approaches using
localized surface plasmon resonator (LSPR) sensing. Some of these analytical tac-
tics exploit the plasmonic coupling method. Metallic NPs placed in close vicinity
to an Au film or another metallic NP results in large variations in coloration, which
may offer sufficient S for single molecular recognition. Proteins, enzymes, and DNA
interactions can be effectively examined with the configuration of a colorimetric-
sensing device based on the combination of Au NP. Analytes with multiple binding
sites or the interaction of binding partners on different NPs can induce aggregation. In
addition, colorimetric-sensing devices can be built into an immunochromatographic
test-chip, analyzing or a paper-based detection scheme designed to generate low-
priced biosensors. Various uses other than sensing are also envisaged. Au NPs are
being studied for photothermal therapy, using heat dissipation around Au NP while
excited by a laser beam to eradicate cancer cells. This method can result in tumor
staining and marginally invasive therapy with suitable surface chemistry to identify
tumors. The low penetration of light in tissues because of absorption and scattering
makes this treatment inadequate for cancer therapy. Therefore, the manufacturing
of n-structured surfaces has created a multitude of different plasmonic materials
that have been immobilized onto a substratum. This enables the advancement of
SPR-sensing devices, allowing the multiplexing of SPR-sensing devices based on
NPs. For instance, e-beam or photolitho focus ion beam milling and colloidal litho
techniques were used to fabricate n-structured plasmonic sensing devices, such as

BK-TandF-KAZANSKIY_9781032572949-230642-Chp03.indd 90 9/26/23 4:55 PM

 Nanoplasmonic Sensors 91

nanotriangle arrays, films over nanospheres, nanocrescents, nanohole arrays, and

a collection of other exotic structures. Chip-based LSPR-sensing devices are being
built in a test-strip with an optical reader system or in a multiplexed array of sens-
ing devices. Such advancements in nanomaterials have had a significant impact on
sensing devices established on LSPR by offering reproducible means for producing
n-structures on a simple to the manageable substratum, enhancing S and creating
colorimetric assays for biomolecules.

3.6 FUTURE OF PLASMONIC SENSING DEVICES

In just about a decade, n-plasmonic sensing devices made a remarkable evolution
from the theoretical idea to the operational devices. It is apparent that this tech-
nology is proficient of providing high S, small size, readily accessible multiplexing
competences that can be apprehended and applied in a wide range of areas, both
in investigation and even consumer devices. The functional surface chemistry has
progressively become the central point in the advancement of n-plasmonic sensing
devices that target at point-of-care diagnostics. New conceptual configurations have
been revealed to further simplify n-plasmonic sensing devices such as the recent
advances in paper-based substratum for MM devices that can potentially be used for
quantitative examination in biochemical sensing purposes, as shown in Figure 3.15
[105]. Experimental demonstrations are carried out by patterning μm-sized metal-
insulator-metal resonators on paper substratum and measuring the resonance shift

FIGURE 3.15 (a) Graphical illustration of μm-sized MM resonators sprayed on paper

substratum with a predefined micron stencil. (b) Photo of a paper-based THz MM device.
(c) Optical microscopy photo of one part of manufactured paper MM sample [105].

BK-TandF-KAZANSKIY_9781032572949-230642-Chp03.indd 91 9/26/23 4:55 PM

 92 Photonics Elements for Sensing and Optical Conversions

brought by engaging various concentrations of glucose solution on the metal-insulator-

metal surface.
A highly responsive label-free sensing method established on asymmetric FRs in
plasmonic nanoholes with comprehensive repercussions for point-of-care diagnostics
is established. By utilizing exceptional light transmission phenomena through high
Q-factor ∼200 subradiant dark modes, FOM as great as 162 for intrinsic dynamic light
scattering (DLs) exceeding that of the Au standard prism coupled surface-plasmon
sensing devices (KR configuration) have been demonstrated [106]. The examples
of the recent progress are the selective decoration with analyte-capturing chemis-
try of nanostars [107, 108], nanocrescents [109], nanotriangles [110], and others. In
combination with surface chemistry, proper configuration of plasmonic n-structures
can also yield significantly increased and “localized” S [111]. It is expected that
the interplay of numerous n-plasmonic resonances in one structure would generate
n-plasmonic sensing devices with the ability to resolve the tridimensional analyte
structure [112]. It is likewise incredibly clear that applications outside biological and
chemical sensing are increasingly sought by n-plasmonic detectors.
Materials science would greatly profit from the wider application of specific
n-plasmonic sensing methods. Intriguingly, since the most sophisticated and com-
mercially productive plasmonic sensing platforms are those using thin noble metal
films to propagate SPPs, there is a drift to merge both [113]. This is realized, for
instance, in plasmonic nanowires. Extending the sensing device’s probing volume
further, as with SPR technology, while retaining intense enhancement of the EM
field characteristic of the n-plasmonic sensing devices, is certainly beneficial. On
the side of n-plasmonic sensing device’s fundamental characteristics, it is imperative
that consistent predictive models emerge that explicitly link the structure with opti-
cal and sensing properties of n-plasmonic sensing platforms.
On one more functional side, scalable manufacturing methods that can deliver
large arrays of n-plasmonic structures are expected to be further developed. At the
same time, nanofabrication of an industrial scale that aims at multiplexed array for-
mat is essential for further reduce the price per chip in n-plasmonic-based sensing.
All in all, we will most likely see a gradual growth of n-plasmonic sensing devices
for different purposes in the coming years. Nevertheless, there are still fascinating
fundamental research prospects in the configuration principles and function of sens-
ing devices. Methods have been developed that allow low-cost substratum, such as
plastics and polymers, to manufacture these n-structures. Plasmonics is one of the
most exciting research areas in modern science, which will continue to evolve. In
the years to come, faster, smaller, and cheaper sensing device architectures based on
plasmonics will be reported, capitalizing on recent breakthroughs such as nanoflow
through LSPR single-molecule detection and highly sensitive FRs.
Nevertheless, incorporating these innovations into biomedical devices is a chal-
lenge for the industry. Soon, biosensors will be universal as they are needed for
significant transformative approaches that have been expected, such as routine
point-of-care-based diagnosis and personalized medicine. Low-cost biosensors are
also required for rapid diagnosis of infectious diseases in developing countries and
remote locations. Therefore, plasmonic biosensors are well prepared not only to be
at the center of an emerging healthcare revolution but also to propel it. Furthermore,
advancements in perpetual issues in surface chemistry, such as nonspecific binding

BK-TandF-KAZANSKIY_9781032572949-230642-Chp03.indd 92 9/26/23 4:55 PM

 Nanoplasmonic Sensors 93

and consistent molecular immobilization schemes, have helped the growth of the
field. The variety of established sensing devices and the various sensing schemes
suggest that a plasmonic solution can be used for several biochemical or biomedical
problems, ranging from fundamental protein–protein interaction studies and single
protein dynamics to disease diagnosis by detecting DNA fragments in complex bio-
logical samples at femtomolar concentrations.

REFERENCES
1. Chen, L., Lu, P., Tian, M., Liu, D., and Zhang, J. 2013. A subwavelength MIM wave-
guide filter with single-cavity and multi-cavity structures. Optik. 124(18):3701–4.
https://doi.org/10.1016/j.ijleo.2012.11.025.
2. Pu, M., Yao, N., Hu, C., Xin, X., Zhao, Z., Wang, C., and Luo, X. 2010. Directional
coupler and nonlinear Mach-Zehnder interferometer based on the metal-insulator-
metal plasmonic waveguide. Opt. Express. 18(20):21030–7. https://doi.org/10.1364/
OE.18.021030.
3. Duan, G., Lang, P., Wang, L., Yu, L., and Xiao, J. 2016. An optical pressure sensor based
on π-shaped surface plasmon polariton resonator. Mod. Phys. Lett. B. 30(21):1650284.
https://doi.org/10.1142/S0217984916502845.
4. Kou, Y., and Chen, X. 2011. Multimode interference demultiplexers and splitters in
metal-insulator-metal waveguides. Opt. Express. 19(7):6042–7. https://doi.org/10.1364/
OE.19.006042.
5. Kamada, S., Okamoto, T., El-Zohary, S. E., and Haraguchi, M. 2016. Design optimiza-
tion and fabrication of Mach-Zehnder interferometer based on MIM plasmonic wave-
guides. Opt. Express. 24(15):16224–31. https://doi.org/10.1364/OE.24.016224.
6. Babicheva, V. E., and Lavrinenko, A. V. 2013. A Plasmonic modulator based on a
metal-insulator-metal waveguide with a barium titanate core. Photonics Lett. Poland.
5(2):57–9. https://doi.org/10.4302/plp.2013.2.08.
7. Butt, M. A., Khonina, S. N., and Kazanskiy, N. L. 2020. Ultra-short lossless plas-
monic power splitter design based on metal-insulator-metal waveguide. Laser Phys.
30:016201. https://doi.org/10.1088/1555-6611/ab5577.
8. Kuchmizhak, A., Pustovalov, E., Syubaev, S., Vitrik, O., Kulchin, Y., Porfirev,
A., Khonina, S. N., Kudryashov, S. I., Danilov, P., and Ionin, A. 2016. On-fly fem-
tosecond-laser fabrication of self-organized plasmonic nanotextures for chemo- and
biosensing applications. ACS Appl. Mater. Interfaces. 8(37):24946–55. https://doi.org/
10.1021/acsami.6b07740.
9. Kudryashov, S. I., Danilov, P. A., Porfirev, A. P., Saraeva, I. N., Nguyen, T. H. T.,
Rudenko, A. A., Khmelnitskii, R. A., Zayarny, D. A., Ionin, A. A., Kuchmizhak, A. A.,
Khonina, S. N., and Vitrik, O.B. 2019. High-throughput micropatterening of plasmonic
surfaces by multiplexed femtosecond laser pulses for advanced IR-sensing applications.
Appl. Surf. Sci. 484:948–56. https://doi.org/10.1016/j.apsusc.2019.04.048.
10. Economou, E. N. 1969. Surface plasmons in thin films. Phys. Rev. 182(2):539–54.
https://doi.org/10.1103/PhysRev.182.539.
11. Gordon, R. 2006. Light in a subwavelength slit in a metal: Propagation and reflection.
Phys. Rev. B. 73(15):153405. https://doi.org/10.1103/PhysRevB.73.153405.
12. Dionne, J. A., Sweatlock, L. A., Atwater, H. A., and Polman, A. 2006. Plasmon slot
waveguides: Towards chip-scale propagation with subwavelength-scale localization.
Phys. Rev. B. 73(3):035407. https://doi.org/10.1103/PhysRevB.73.035407.
13. Bozhevolnyi, S. I., Volkov, V. S., Devaux, E., Laluet, J.-Y., and Ebbesen, T. W. 2006.
Channel plasmon subwavelength waveguide components including interferometers and
ring resonators. Nature. 440(7083):508–11. https://doi.org/10.1038/nature04594.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp03.indd 93 9/26/23 4:55 PM

 94 Photonics Elements for Sensing and Optical Conversions

14. Nikolajsen, T., Leosson, K., and Bozhevolnyi, S. I. 2004. Surface plasmon polariton
based modulators and switches operating at telecom wavelengths. Appl. Phys. Lett.
85(24):5833–5. https://doi.org/10.1063/1.1835997.
15. Charbonneau, R., Scales, C., and Breukelaar, I. et al. 2006. Passive integrated
optics elements based on long-range surface plasmon polaritons. J. Lightw. Technol.
24(1):477–94. https://doi.org/10.1109/JLT.2005.859856.
16. Ditlbacher, H. et al. 2005. Silver nanowires as surface plasmon resonators. Phys. Rev.
Lett. 95(25):257403. https://doi.org/10.1103/PhysRevLett.95.257403.
17. Johnson, P. B., and Christy, R. W. 1972. Optical constants of the noble metals. Phys.
Rev. B. 6(12):4370–9. https://doi.org/10.1103/PhysRevB.6.4370.
18. Ehrenreich, H., Philipp, H., and Segall, B. 1963. Optical properties of aluminum. Phys.
Rev. 132(5): 1918–28. https://doi.org/10.1103/PhysRev.132.1918.
19. Giurlani, W. et al. 2018. Electroplating for decorative applications: Recent trends in
research and development. Coatings. 8(8):260. https://doi.org/10.3390/coatings8080260.
20. MicroChem Corp. 2012, March. http://www.microchem.com.
21. Kannegulla, A., and Cheng, L.-J. 2016. Metal assisted focused-ion beam nanopatterning.
Nanotechnology. 27(36):36LT01. https://doi.org/10.1088/0957-4484/27/36/36LT01.
22. Hindmarch, A. T., Parkes, D. E., and Rushforth, A. W. 2012. Fabrication of metallic
magnetic nanostructures by argon ion milling using a reversed-polarity planar magne-
tron ion source. Vacuum. 86(10):1600–4. https://doi.org/10.1016/j.vacuum.2012.02.019.
23. Masson, J.-F., Murray-Methot, M.-P., and Live, L. S. 2010. Nanohole arrays in chemical
analysis: Manufacturing methods and applications. Analyst. 135(7):1483–9. https://doi.
org/10.1039/c0an00053a.
24. Cao, J., Sun, T., and Grattan, K. T. V. 2014. Gold nanorod-based localized surface plas-
mon resonance biosensors: A review. Sens. Actuators B Chem. 195:332–51. https://doi.
org/10.1016/j.snb.2014.01.056.
25. Maier, S. A., Barclay, P. E., Johnson, T. J., Friedman, M. D., and Painter, O. 2004.
Low-loss fiber accessible plasmon waveguide for planar energy guiding and sensing.
Appl. Phys. Lett. 84(20):3990. https://doi.org/10.1063/1.1753060.
26. Maier, S. A., Friedman, M. D., Barclay, P. E., and Painter, O. 2005. Experimental dem-
onstration of fibre-accessible metal nanoparticle plasmon waveguides for planar energy
guiding and sensing. Appl. Phys. Lett. 86:071103. https://doi.org/10.1063/1.1862340.
27. Oulton, R. F., Bartal, G., Pile, D. F. P., and Zhang, X. 2008. Confinement and propaga-
tion characteristics of subwavelength plasmonic modes. New J. Phys. 10:105018. https://
doi.org/10.1088/1367-2630/10/10/105018.
28. Otto, L. M., Mohr, D. A., Johnson, T. W., Oh, S. H., and Lindquist, N. C. 2015.
Polarization interferometry for real-time spectroscopic plasmonic sensing. Nanoscale.
7(9):4226–33. https://doi.org/10.1039/C4NR06586G.
29. Homola, J., Sinclair, S., and Gauglitz, G. 1999. Surface plasmon resonance sensors:
Review. Sens. Actuators B. 54(1–2):3–15. https://doi.org/10.1016/S0925-4005(98)00321-9.
30. Butt, M. A., Khonina, S. N., and Kazanskiy, N. L. 2021. Metal-insulator-metal nano
square ring resonator for gas sensing applications. Waves Random Complex Media.
31(1):146–56. https://doi.org/10.1080/17455030.2019.1568609.
31. Zhang, Z., Yang, J., He, X., Zhang, J., Huang, J., Chen, D., and Han, Y. 2018. Plasmonic
refractive index sensor with high figure of merit based on concentric-rings resonator.
Sensors. 18(1):116. https://doi.org/10.3390/s18010116.
32. Wu, T., Liu, Y., Yu, Z., Peng, Y., Shu, C., and Ye, H. 2014. The sensing characteristics
of plasmonic waveguide with a ring resonator. Opt. Express. 22(7):7669–77. https://doi.
org/10.1364/OE.22.007669.
33. Wei, W., Zhang, X., and Ren, X. 2015. Plasmonic circular resonators for refrac-
tive index sensors and filters. Nanoscale Res. Lett. 10:211. https://doi.org/10.1186/
s11671-015-0913-4.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp03.indd 94 9/26/23 4:55 PM

 Nanoplasmonic Sensors 95

34. Chen, Z., and Yu, L. 2014. Multiple Fano resonances based on different waveguide
modes in a symmetry breaking plasmonic system. IEEE Photonics J. 6(6):4802208.
https://doi.org/10.1109/JPHOT.2014.2368779.
35. Gaur, S., Zafar, R., and Somwanshi, D. 2016. Plasmonic refractive index sensor
based on metal insulator metal waveguide. IEEE International Conference on Recent
Advances and Innovations in Engineering (ICRAIE-2016): 1–4. https://doi.org/10.1109/
ICRAIE.2016.7939557.
36. Zhang, Z., Luo, L., Xue, C., Zhang, W., and Yan, S. 2016. Fano Resonance based on
metal-insulator-metal waveguide-coupled double rectangular cavities for plasmonic
nanosensors. Sensors. 16(5):642. https://doi.org/10.3390/s16050642.
37. Binfeng, Y., Hu, G., Zhang, R., and Yiping, C. 2016. Fano resonances in a plasmonic
waveguide system composed of stub coupled with a square cavity resonator. J. Opt.
18(5):055002. https://doi.org/10.1088/2040-8978/18/5/055002.
38. Yan, S., Zhang, M., Zhao, X., Zhang, Y., Wang, J., and Jin, W. 2017. Refractive index
sensor based on a metal-insulator-metal waveguide coupled with a symmetric struc-
ture. Sensors. 17(12):2879. https://doi.org/10.3390/s17122879.
39. Zhao, X., Zhang, Z., and Yan, S. 2017. Tunable Fano resonance in asymmetric MIM
waveguide structure. Sensors. 17(7):1494. https://doi.org/10.3390/s17071494.
40. Butt, M. A., Khonina, S. N., and Kazanskiy, N. L. 2018. Hybrid plasmonic waveguide-
assisted metal-insulator-metal ring resonator for refractive index sensing. J. Mod. Opt.
65(9):1135–40. https://doi.org/10.1080/09500340.2018.1427290.
41. Rakhshani, M. R., Tavousi, A., and Mansouri-Birjandi, M. A. 2018. Design of a plas-
monic sensor based on a square array of nanorods and two slot cavities with a high fig-
ure of merit for glucose concentration monitoring. Appl. Opt. 57(27):7798–804. https://
doi.org/10.1364/AO.57.007798.
42. Wang, L., Zeng, Y.-P., Wang, Z.-Y., Xia, X.-P., and Liang, Q.-Q. 2018. A Refractive
index sensor based on an analogy t shaped metal-insulator-metal waveguide. Optik.
172:1199–204. https://doi.org/10.1016/j.ijleo.2018.07.093.
43. Butt, M. A., Khonina, S. N., and Kazanskiy, N.L. 2018. Plasmonic refractive index
sensor based on M-I-M square ring resonator. Int. Conf. on Computing, Electronic
and Electrical Engineering (ICE Cube): 1–4. https://doi.org/10.1109/ICECUBE.
2018.8610998.
44. Fang, Y., Wen, K., and Li, Z. et al. 2019. Multiple Fano resonances based on end-
coupled semi-ring rectangular resonator. IEEE Photon. J. 11(4):4801308. https://doi.
org/10.1109/JPHOT.2019.2914483.
45. Chen, Y., Xu, Y., and Cao, J. 2019. Fano Resonance sensing characteristics of MIM
waveguide coupled square convex ring resonator with metallic baffle. Results Phys.
14:102420. https://doi.org/10.1016/j.rinp.2019.102420.
46. Yu, S., Zhao, T., Yu, J., and Pan, D. 2019. Tuning multiple Fano resonances for on-
chip sensors in a plasmonic system. Sensors. 19(7):1559. https://doi.org/10.3390/
s19071559.
47. Asgari, S., and Granpayeh, N. 2019. Tunable mid-infrared refractive index sensor com-
posed of asymmetric double graphene layers. IEEE Sens. J. 19(14):5686–91. https://doi.
org/10.1109/JSEN.2019.2906759.
48. Shi, Y., Zhang, G.-M., An, H.-L., Hu, N., and Gu, M.-Q. 2017. Controllable Fano
resonance based on coupled square split ring resonance cavity. Acta Photonica Sinica
46(4):0413002. https://doi.org/10.3788/gzxb20174604.0413002.
49. Zafar, R., and Salim, M. 2015. Enhanced figure of merit in Fano resonance-based
plasmonic. IEEE Sens. J. 15(11):6313–7. https://doi.org/10.1109/JSEN.2015.2455534.
50. Butt, M. A., Khonina, S. N., and Kazanskiy, N. L. 2019. A multichannel metallic dual
nano-wall square split-ring resonator: Design analysis and applications. Laser Phys.
Lett. 16(12):126201. https://doi.org/10.1088/1612-202X/ab5574.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp03.indd 95 9/26/23 4:55 PM

 96 Photonics Elements for Sensing and Optical Conversions

51. Kabashin, A.V. et al. 2009. Plasmonic nanorod metamaterials for biosensing. Nat.
Mater. 8(11):867–71. https://doi.org/10.1038/nmat2546.
52. Jeong, H. H. et al. 2016. Dispersion and shape engineered plasmonic nanosensors. Nat.
Commun. 7:11331. https://doi.org/10.1038/ncomms11331.
53. Danaie, M., and Shahzadi, A. 2019. Design of a high resolution metal-insulator-metal
plasmonic refractive index sensor based on a ring shaped Si resonator. Plasmonics.
14:1453–65. https://doi.org/10.1007/s11468-019-00926-9.
54. Cetin, A. E., and Atlug, H. 2012. Fano resonant ring/disk plasmonic nanocavities on
conducting substrates for advanced biosensing. ACS Nano. 6(11):9989–95. https://doi.
org/10.1021/nn303643w.
55. Limonov, M. F., Rybin, M. V., Poddubny, A. N., and Kivshar, Y. S. 2017. Fano Resonances
in photonics. Nat. Photonics. 11:543–54. https://doi.org/10.1038/nphoton.2017.142.
56. Song, M. et al. 2016. Nanofocusing beyond the near-field diffraction limit via plas-
monic Fano resonance. Nanoscale. 8:1635–41. https://doi.org/10.1039/c5nr06504f.
57. Wang, Q., Ouyang, Z., Sun, Y., Lin, M., and Liu, Q. 2019. Linearly tunable Fano reso-
nance modes in a plasmonic nanostructure with a waveguide loaded with two rectan-
gular cavities coupled by a circular cavity. Nanomaterials. 9(5):678. https://doi.org/
10.3390/nano9050678.
58. Ye, J. et al. 2012. Plasmonic nanoclusters: Near field properties of the Fano resonance
interrogated with SERS. Nano Lett. 12(3):1660–7. https://doi.org/10.1021/nl3000453.
59. Liu, J., Liu, Z., and Hu, H. 2019. Tunable multiple Fano resonance employing
polarization-selective excitation of coupled surface-mode and nanoslit antenna
resonance in plasmonic nanostructures. Sci. Rep. 9(1):2414. https://doi.org/10.1038/
s41598-019-38708-2.
60. Zhan, S., Peng, Y., He, Z., Li, B., Chen, Z., Xu, H., and Li, H. 2016. Tunable nanoplas-
monic sensor based on the asymmetric degree of Fano resonance in MDM waveguide.
Sci. Rep. 6(1):22428. https://doi.org/10.1038/srep22428.
61. Deng, Y., Cao, G., Yang, H., Li, G., Chen, X., and Lu, W. 2017. Tunable and high-
sensitivity sensing based on Fano resonance with coupled plasmonic cavities. Sci. Rep.
7:10639. https://doi.org/10.1038/s41598-017-10626-1.
62. Wen, Y. et al. 2019. High sensitivity and FOM refractive index sensing based on Fano
resonance in all-grating racetrack resonators. Opt. Commun. 446:141–6. https://doi.
org/10.1016/j.optcom.2019.04.068.
63. Wei, G., Tian, J., and Yang, R. 2019. Fano resonance in MDM plasmonic wave-
guides coupled with split ring resonator. Optik. 193(5):162990. https://doi.org/10.1016/
j.ijleo.2019.162990.
64. Chen, Y., Xu, Y., and Cao, J. 2019. Fano resonance sensing characteristics of MIM
waveguide coupled square convex ring resonator with metallic baffle. Results Phys.
14:102420. https://doi.org/10.1016/j.rinp.2019.102420.
65. Chen, F., Zhang, H., Sun, L., Li, J., and Yu, C. 2019. Temperature tunable Fano reso-
nance based on ring resonator side coupled with a MIM waveguide. Opt. Laser Technol.
116:293–9. https://doi.org/10.1016/j.optlastec.2019.03.044.
66. Eisorbagy, M. H. et al. 2019. Performance improvement of refractometric sensors
through hybrid plasmonic Fano resonances. J. Lightw. Technol. 37(13):2905–13. https://
doi.org/10.1109/JLT.2019.2906933.
67. Chen, F., Zhang, H., Sun, L., Li, J., and Yu, C. 2019. Electrically tunable Fano res-
onance based on ring resonator coupled with a stub. Optik. 185:585–91. https://doi.
org/10.1016/j.ijleo.2019.03.161.
68. Liu, H. et al. 2019. Metasurface generated polarization insensitive Fano resonance for
high performance refractive index sensing. Opt. Express. 27(9):13252–62. https://doi.
org/10.1364/OE.27.013252.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp03.indd 96 9/26/23 4:55 PM

 Nanoplasmonic Sensors 97

69. Zhang, Y., and Cui, M. 2019. Refractive index sensor based on the symmetric MIM
waveguide structure. J. Electron. Mater. 48(2):1005–10. https://doi.org/10.1007/s11664-
018-6823-3.
70. Li, Z. et al. 2019. Refractive index sensor based on multiple Fano resonances in
a plasmonic MIM structure. Appl. Opt. 58(18):4878–83. https://doi.org/10.1364/
AO.58.004878.
71. Chen, F., and Li, J. 2019. Refractive index and temperature sensing based on defect
resonator coupled with a MIM waveguide. Mod. Phys. Lett. B. 33(3):1950017. https://
doi.org/10.1142/S0217984919500179.
72. Qi, L., Zhao, C. L., Yuan, J. Y., Ye, M. P., Wang, J., Zhang, Z., and Jin, S. 2014. Highly
reflective long period fibre grating sensor and its application in refractive index sensing.
Sens. Actuators B Chem. 193:185–9. https://doi.org/10.1016/j.snb.2013.11.063.
73. Wu, D. K. C., Kuhlmey, B. T., and Eggleton, B. J. 2009. Ultrasensitive photonic
crystal fibre refractive index sensor. Opt. Lett. 34(3):322–4. https://doi.org/10.1364/
OL.34.000322.
74. Shen, Y., Zhou, J. H., and Liu, T. R. et al. 2013. Plasmonic gold mushroom arrays
with refractive index sensing figures of merit approaching the theoretical limit. Nat.
Commun. 4(4):2381. https://doi.org/10.1038/ncomms3381.
75. Yue, T., Zhang, Z., Wang, R., Hai, Z., Xue, C., Zhang, W., and Yan, S. 2017. Refractive
index sensor based on Fano resonances in metal-insulator-metal waveguides coupled
with resonators. Sensors. 17(4):784. https://doi.org/10.3390/s17040784.
76. Ali, A., Ghafoorifard, H., Abdolhosseini, S., and Habibiyan, H.. 2018. Metal-insulator-
metal waveguide-coupled asymmetric resonators for sensing and slow light applica-
tions. IET Optoelectron. 12(5):220–7. https://doi.org/10.1049/iet-opt.2018.0028.
77. Lu, H., Liu, X., and Mao, D. 2012. Plasmonic analog of electromagnetically induced
transparency in multinanoresonator-coupled waveguide systems. Phys. Rev. A. 85(5):
053803. https://doi.org/10.1103/PhysRevA.85.053803.
78. Boller, K., Imamoglu, A., and Harris, S. E. 1991. Observation of electromagneti-
cally induced transparency. Phys. Rev. Lett. 66(20):2593–6. https://doi.org/10.1103/
PhysRevLett.66.2593.
79. Chen, Y., Chen, L., Wen, K., Hu, Y., and Lin, W. 2019. Double Fano resonances based on
different mechanisms in a MIM plasmonic system. Photonics Nanostruct. 36:100714.
https://doi.org/10.1016/j.photonics.2019.100714.
80. Wen, K., Chen, L., Zhou, J., Lei, L., and Fang, Y. 2018. A plasmonic chip-scale refrac-
tive index sensor design based on multiple Fano resonances. Sensors. 18(10):3181.
https://doi.org/10.3390/s18103181.
81. Zand, I., Abrishamian, M. S., and Berini, P. 2013. Highly tunable nanoscale metal-
insulator-metal split ring core ring resonators (SRCRRs). Opt. Express. 21(1):79–86.
https://doi.org/10.1364/OE.21.000079.
82. Hirsch, L. R., Jackson, J. B., Lee, A., Halas, N. J., and West, J. L. 2003. A whole
blood immunoassay using gold nanoshells. Anal. Chem. 75(10):2377–81. https://doi.
org/10.1021/ac0262210.
83. Guler, U., Suslov, S., Kildishev, A. V., Boltasseva, A., and Shalaev, V. M. 2015. Colloidal
plasmonic titanium nitride nanoparticles: Properties and applications. Nanophotonics.
4(3):269–76. https://doi.org/10.1515/nanoph-2015-0017.
84. Behnam, M. A., and Emami, F. et al. 2018. Novel combination of silver nanoparti-
cles and carbon nanotubes for plasmonic photothermal therapy in a melanoma cancer
model. Adv. Pharm. Bull. 8(1):94–5. https://doi.org/10.15171/apb.2018.006.
85. Zand, I., Mahigir, A., Pakizeh, T., and Abrishamian, M. S. 2012. Selective-mode opti-
cal nanofilters based on plasmonic complementary split-ring resonators. Opt. Express
20(7):7516–25. https://doi.org/10.1364/OE.20.007516.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp03.indd 97 9/26/23 4:55 PM

 98 Photonics Elements for Sensing and Optical Conversions

86. Willets, K. A., and Van Duyne, R. P. 2007. Localized surface plasmon resonance
spectroscopy and sensing. Annu. Rev. Phys. Chem. 58:267–97. https://doi.org/10.1146/
annurev.physchem.58.032806.104607.
87. Grigorenko, A. N., Beloglazov, A. A., and Nikitin, P. I. 2000. Dark-field surface plas-
mon resonance microscopy. Opt. Commun. 174(1–4):151–5. https://doi.org/10.1016/
S0030-4018(99)00676-8.
88. Lin, C.-W., Chen, K.-P., Hsiao, C.-N., Lin, S., and Lee, C.-K. 2006. Design and fabrica-
tion of an alternating dielectric multi-layer device for surface plasmon resonance sen-
sor. Sens. Actuators B. 113(1):169–76. https://doi.org/10.1016/j.snb.2005.02.044.
89. Wan, Y. W. Y., Chuah, H. P., and Mahmood, M. Y. W. 2007. Optical properties and
sugar content determination of commercial carbonated drinks using surface plasmon
resonance. Am. J. Appl. Sci. 4(1):1–4. https://doi.org/10.3844/ajassp.2007.1.4.
90. Paliwal, A., Sharma, A., Tomar, M., and Gupta, V. 2016. Surface plasmon resonance
study on the optical sensing properties of tin oxide (SnO2) films to NH3 gas. J. Appl.
Phys. 119:164502. https://doi.org/10.1063/1.4948332.
91. Kaur, G., Paliwal, A., Tomar, M., and Gupta, V. 2016. Detection of Neisseria menin-
gitidis using surface plasmon resonance based DNA biosensor. Biosens. Bioelectron.
78:106–10. https://doi.org/10.1016/j.bios.2015.11.025.
92. Zybin, A., Grunwald, C., Mirsky, V. M., Wolfbeis, O. S., and Niemax, K.. 2005.
Double-wavelength technique for surface plasmon resonance measurements: Basic
concept and applications for single sensors and two-dimensional sensor arrays. Anal.
Chem. 77(8):2393–9. https://doi.org/10.1021/ac048156v.
93. Herminjard, S., Sirigu, L., Herzig, H. P., Studemann, E., Crottini, A., Pellaux, J.-P.,
Gresch, T., Fischer, M., and Faist, J. 2009. Surface plasmon resonance sensor show-
ing enhanced sensitivity for CO2 detection in the mid-infrared range. Opt. Express
17(1):293–303. https://doi.org/10.1364/OE.17.000293.
94. Karabchevsky, A., Karabchevsky, S., and Abdulhalim, I. 2011. Fast surface plasmon
resonance imaging sensor using radon transform. Sens. Actuators B Chem. 155(1):361–
5. https://doi.org/10.1016/j.snb.2010.12.012.
95. O’Brien, M. J. 2nd, Pérez-Luna, V. H., Brueck, S. R. J., and López, G. P. 2001. A sur-
face plasmon resonance array biosensor based on spectroscopic imaging. Biosens.
Bioelectron. 16(1–2):97–108. https://doi.org/10.1016/s0956-5663(00)00137-8.
96. Wood, R. W. 1902. On a remarkable case of uneven distribution of light in a diffrac-
tion grating spectrum. Proc. Phys. Soc. London. 4:396–402. https://doi.org/10.1088/
1478-7814/18/1/325.
97. Kooyman, R. P. H. 2008. Physics of surface plasmon resonance. In Handbook of surface
plasmon resonance, Ch 2, ed. R. B. M. Schasfoort, and A. J. Tudos, 403. Cambridge,
UK: The Royal Society of Chemistry. https://doi.org/10.1039/9781847558220-00015.
98. Maystre, D. 2012. Theory of wood’s anomalies. In Plasmonics: From basics to
advanced topics, ed. S. Enoch, and N. Bonod, 39–83. Berlin, Heidelberg: Springer-
Verlag. https://doi.org/10.1007/978-3-642-28079-5_2.
99. Homola, J., ed. 2006. Surface plasmon resonance based sensors. Berlin, Heidelberg:
Springer-Verlag. ISBN: 978-3-540-33918-2.
100. Rossi, S., Gazzola, E., Capaldo, P., Borile, G., and Romanato, F. 2018. Grating-coupled
surface plasmon resonance (GC-SPR) optimization for phase-interrogation biosensing
in a microfluidic chamber. Sensors 18(5):1621. https://doi.org/10.3390/s18051621.
101. Dai, Y., Xu, H., Lu, Y., and Wang, P. 2018. Experimental demonstration of high sensi-
tivity for silver rectangular grating-coupled surface plasmon resonance (SPR) sensing.
Opt. Commun. 416:66–70. https://doi.org/10.1016/j.optcom.2018.02.010.
102. Bonod, N., Popov, E., and McPhedran, R. C. 2008. Increased surface plasmon res-
onance sensitivity with the use of double Fourier harmonic gratings. Opt. Express.
16(16):11691–702. https://doi.org/10.1364/OE.16.011691.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp03.indd 98 9/26/23 4:55 PM

 Nanoplasmonic Sensors 99

103. Popov, E. K., Bonod, N., and Enoch, S. 2007. Comparison of plasmon surface waves
on shallow and deep metallic 1D and 2D gratings. Opt. Express. 15(7):4224–37. https://
doi.org/10.1364/OE.15.004224.
104. Sepulveda, B., Angelome, P. C., Lechuga, L. M., and Liz-Marzan, L. M. 2009.
LSPR-based nanobiosensors. Nano Today. 4(3):244–51. https://doi.org/10.1016/j.
nantod.2009.04.001.
105. Tao, H. et al. 2011. Metamaterials on paper as a sensing platform. Adv. Mater. 23:3197–
201. https://doi.org/10.1002/adma.201100163.
106. Yanik, A. A., Cetin, A. E., Huang, M., Artar, A., Mousavi, S. H., Khanikaev, A.,
Connor, J. H., Shvets, G., and Altug, H. 2011. Seeing protein monolayers with naked
eye through plasmonic Fano resonances. Proc. Natl. Acad. Sci. USA. 108(29):11784–9.
https://doi.org/10.1073/pnas.1101910108.
107. Hrelescu, C. et al. 2011. Selective excitation of individual plasmonic hotspots at the tips
of single gold nanostars. Nano Lett. 11(2):402–7. https://doi.org/10.1021/nl103007m.
108. Dondapati, S. K. et al. 2010. Label-free biosensing based on single gold nanostars as
plasmonic transducers. ACS Nano. 4(11):6318–22. https://doi.org/10.1021/nn100760f.
109. Unger, A. et al. 2009. Sensitivity of crescent-shaped metal nanoparticles to attachment
of dielectric colloids. Nano Lett. 9(6):2311–15. https://doi.org/10.1021/nl900505a.
110. Beeram, S. R. et al. 2009. Selective attachment of antibodies to the edges of gold nano-
structures for enhanced localized surface plasmon resonance biosensing. J. Am. Chem.
Soc. 131(33):11689–91. https://doi.org/10.1021/ja904387j.
111. Feuz, L. et al. 2010. Improving the limit of detection of nanoscale sensors by directed
binding to high-sensitivity areas. ACS Nano 4(4):2167–77. https://doi.org/10.1021/
nn901457f.
112. Liu, N. et al. 2011. Three-dimensional plasmon rulers. Science 332(6036):1407–10.
https://doi.org/10.1126/science.1199958.
113. Bolduc, O. R. et al. 2011. Advances in surface plasmon resonance sensing with
nanoparticles and thin films: Nanomaterials, surface chemistry, and hybrid plasmonic
techniques. Anal. Chem. 83(21):8057–62. https://doi.org/10.1021/ac2012976.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp03.indd 99 9/26/23 4:55 PM

 4 Plasmonic Nanolasers
Van Duong Ta1 and Hanh Hong Mai2
1  epartment of Optical Devices, Le Quy Don
D
Technical University, Hanoi, Vietnam
2 Department of Quantum Optics, Faculty of

Physics, VNU University of Science, Vietnam

National University, Hanoi, Vietnam

4.1 INTRODUCTION
Since 1960 when Theodore Maiman demonstrated the first laser device [1], laser
technology has had a great impact on our society. Nowadays, lasers play a crucial
role in optical communications, medical therapy, precise manufacturing, scientific
research, and everyday life applications. For example, laser emission spectra have
been used for long-distance transmission of data over optical fibers [2]. Micrometer-
sized lasers, the so-called microlasers, can be integrated into a single cell for cell-
tracking and intracellular sensing [3]. The sensitivity of a microlaser-based biosensor
can be down to the level of a single nanoparticle and single virus [4].
Since invention of the laser, enormous effort has been spent on laser miniaturiza-
tion to create a more compact size and lower power consumption. Great success was
achieved in shrinking the laser size following the introduction of the semiconduc-
tor as a gain material in 1962 [5]. By 1989, the first vertical-cavity single-quantum
well microlaser was reported with a minimum size of 3 µm (about three times the
wavelength of the emitted light) [6]. In the 1990s, the size of lasers was reduced to
wavelength scales as demonstrated by microdisc lasers [7], microsphere lasers [8],
photonic crystal lasers [9], nanowire lasers [10], and so on. However, further minia-
turization of laser size, beyond the wavelength scales, was a challenging issue; con-
ventional laser cavities confine light by using differences in the refractive index of
dielectrics so the minimum optical mode size and consequently the size of the laser
is determined by the diffraction limit [11].
Recently, the diffraction barrier has been resolved in metallic-based nanostruc-
tures via the surface plasmonic effect [12]. Metallic nanoplasmonics is promising in
terms of faster speed, smaller size, and more efficient electronics (Figure 4.1) [13].
Owing to ultra-high optical confinement and ultrafast relaxation processes, plasmon-
ics can overcome the speed limit of semiconductor electronics (caused by heat gen-
eration) and the size limit of dielectric photonics (determined by the diffraction law).
Owing to the unique properties of plasmonics, the size of lasers that rely on plas-
monic cavities (the so-called plasmonic nanolasers) can be shrunk to less than the
wavelength of the emitted light [14–16]. These novel kinds of light sources have a
size of ten to hundreds of nanometers and the mode dimension is below the diffrac-
tion limit of emission. This breakthrough in laser miniaturization is promising for

100 DOI: 10.1201/9781003439165-4

BK-TandF-KAZANSKIY_9781032572949-230642-Chp04.indd 100 9/26/23 4:57 PM

 Plasmonic Nanolasers 101

FIGURE 4.1 Comparison of operating speed and device sizes rely on typical material prop-
erties including semiconductors, insulators, and metals. The dashed lines represent the physi-
cal limitations of different technologies. (Redrawn from [13].)

high-volume data storage, ultra-high resolution imaging, and sensitive sensing. In

the literature, there have been several review articles covering different aspects of
nanolasers, such as theory and experiment [17–23], semiconductor nanowire-based
plasmonic lasers [24], structural engineering [25], and applications [26].
In this chapter, we review the recent progress on plasmonic nanolasers, and
cover various aspects of nanolasers. We first introduce surface plasmons, typical
gain materials, and the working principle of plasmonic nanolasers. We then describe
experimental demonstrations of these devices classified on the confinement method
used for light amplification, including single particle, Fabry–Perot cavity, whisper-
ing gallery mode cavity, and metallic particle/hole array, followed by a summary of
important parameters of nanolasers. After that, we discuss applications of plasmonic
nanolasers in integrated photonic circuits, sensing, and biology. Finally, a summary
and future prospects of plasmonic nanolasers are given.

4.2 OVERVIEW OF A PLASMONIC NANOLASER

4.2.1 Surface Plasmons and Plasmonic Materials
Surface plasmons (SPs) or plasmonics are waves that exist at the interface between
a conductor (generally a metal) and a dielectric material (Figure 4.2a) [12]. These
waves are formed by the interaction between the external electromagnetic field of
light and the free electrons of the conductor. In principle, when the incident light
comes to the metallic surface, free electrons in the metal respond by oscillating in
resonance with the incident light. This electron oscillation creates electromagnetic
fields outside (as well as inside) the metal. As a result, the incident light wave creates
charge associated electromagnetic waves, and these two waves can interact with each

BK-TandF-KAZANSKIY_9781032572949-230642-Chp04.indd 101 9/26/23 4:57 PM

 102 Photonics Elements for Sensing and Optical Conversions

FIGURE 4.2 (a) Surface plasmon at the interface between a dielectric material and a metal.
(b) The field component (E z) perpendicular to the surface is enhanced near the surface but its
strength reduces exponentially with distance away from it. The decay length of the field in
the dielectric medium and in the metal is characterized by δd and δm, respectively. (c) Typical
values of δd and δm in comparison with light wavelength (λ) and the propagation length of the
SP mode (δSP) for aluminum and silver surfaces. (Redrawn from [12].)

other. The resonant interactions between these two waves constitute the SP and give
it unique properties.
The properties of SP strongly depend on the metallic structure. Recent advanced
technology allows the metal surface to be structured at the nanoscale, which enables
the control of SP properties for a wide range of applications, for example in subwave-
length optics, biophotonics, data storage, and microscopy [12].
SPs have two different properties in comparison with light waves. First, the
momentum of the SP mode is greater than that of a free-space photon of the
same frequency. This mismatch momentum is characterized by the SP disper-
sion relation [27]:

εd εm
kSP = k0 , (4.1)
εd + εm

where, kSP and k0 are the SP and the free-space photon wavevector, respectively,
and εm and εd are the permittivity of the metal and dielectric material, respectively.
The mismatch momentum is needed to consider when using light for generating
SPs. Second, the SP field perpendicular to the surface reduces exponentially with
distance away from it [Figure 4.2(b)]. That means SPs are bound to the surface, and,
therefore, their energy cannot propagate away from the surface.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp04.indd 102 9/26/23 4:57 PM

 Plasmonic Nanolasers 103

SP can propagate but will quickly diminish due to the absorption in the metal.
The propagation length, δSP, can be calculated as follows [12]:

3
c  ε ′ + ε d  2 ( ε ′m )
2
1
δ SP = =  m , (4.2)
′′ ω  ε ′m ε d  ε ′′m
2 kSP

where, kSP ′′ is the imaginary part of the complex SP wavevector kSP = kSP ′ + ikSP
′′ ;
ε ′m and ε ′′m are the real and imaginary parts of the dielectric function of the metal
ε m = ε ′m + iε ′′m . In free-space, the propagation distance of the SP can be found from
the real and imaginary parts of the dielectric function of the metal. The real part
characterizes the electron oscillation to the external electromagnetic field, and the
imaginary part determines optical loss.
Figure 4.2(c) shows the propagation distance of SP on silver and aluminum sur-
faces at two wavelengths. Due to low losses, the propagation distance of SP on a
silver surface is about 10–100 µm in the visible range. At a wavelength of 1.5 µm,
it can reach around 0.5 mm. For aluminum, the propagation distance of SP is much
shorter, only 10 µm at a wavelength of 0.5 µm.
Plasmonic materials (generally metals) play an important role in a plasmonic cav-
ity; therefore, choosing a suitable material is crucial for plasmonic lasing. Adapted
data from [28, 29], Wu et al. plotted the real and imaginary parts of the dielectric
function of four common metals as a function of wavelength [23]. Generally, for
the wavelength ranging from 200 to 1,200 nm, aluminum (Al) has the smallest ε′,
followed by silver (Ag) with a medium ε′, and gold (Au) and copper (Cu) have the
highest ε′. Concerning the imaginary part, Ag has the smallest ε″, thus Ag exhibits
the lowest loss in the visible to near-infrared (NIR) region. This is the advantage of
Ag. However, Ag is easily oxidized in ambient conditions so coating a thin dielectric
layer on the Ag surface is necessary to avoid degradation. In contrast to Ag, Au is
chemically stable. Au also has a relatively small ε″. The disadvantage of Au is its
high cost. Despite that, Au is a widely used metal in plasmonic waveguides and
nanolasers. Cu and Al are more cost-effective compared with Au and Ag. Both Cu
and Al have high ε″ in the visible and NIR regions. Interestingly, Al possesses a rela-
tively small ε″, which is lower than that of Au in the wavelength from 200–400 nm,
thus it may be suitable for use in plasmonic waveguides and nanolasers at this wave-
length range.
Recently, alternative plasmonic materials (besides the four common metals just
mentioned), such as transparent conducting oxides, metal nitrides, perovskite oxides,
and two-dimensional (2D) materials have been investigated [30, 31]. For instance,
indium tin oxide (ITO) exhibits a low loss, which is comparable to Ag, in the wave-
length range from 500–1,000 nm [32]. Titanium nitride (TiN) demonstrates a lower
loss than Au in the violet region [33].

4.2.2 Gain Materials Used in Plasmonic Nanolasers

Active materials with high optical gain are desired in a plasmonic laser because
the optical losses of a plasmonic cavity are high (due to the absorption of metallic

BK-TandF-KAZANSKIY_9781032572949-230642-Chp04.indd 103 9/26/23 4:57 PM

 104 Photonics Elements for Sensing and Optical Conversions

material). In general, inorganic semiconductors and organic dyes are the most widely
used materials.
Inorganic semiconductors provide high gain with wavelength ranging from ultra-
violet (UV) to NIR region. Typical II–IV semiconductors such as ZnO, GaN, and
ZnS (for UV) and CdS, CdSe (for visible) have been attracting great research interest
[34]. For the NIR region, GaSb, GaAs, and InP are important laser materials [34]. In
some cases, the emission wavelength can be tuned. For example, material InxGa1–xN
(0 ≤ x ≤ 1) offers not only high gain but also is wavelength tunable from NIR (InN,
0.65 eV) to near UV (GaN, 3.4 eV) [35]. Recently, semiconductor colloidal quantum
dots have been demonstrated as a promising laser material with a tunable spectral
range [36].
Organic dyes are conventional laser materials that provide emission from near
UV to NIR [37]. In general, organic dyes have a broad gain spectral range, thus they
can be employed for wavelength-tunable or short-pulsed lasers. Another advantage
of dyes is their high solubility in various solvents. Consequently, they can be easily
incorporated into a range of matrix materials, including polymers and liquid crys-
tals, to create compact laser configurations such as microspheres, microdiscs, and
microfibers [38]. The disadvantages of dyes are their toxicity and photobleaching
(due to the oxidation of dye molecules and the thermal effect caused by high-density
excitation).
Recently, perovskite has emerged as a new type of promising gain material, with
flexible bandgap engineering, a large absorption coefficient, and low defect state
density [39, 40]. Generally, the chemical formula of perovskite is AMX3 or A2MX4,
where A is a cation such as Cs+, CH3NH3+, NH2CH=NH2+; M is a divalent metal-
lic cation including Pb2+, Sn2+, Mn2+, Fe2+; and X is halogen anion, Cl−, I−, Br− [41].
Especially, perovskite can be synthesized into reduced-dimensional structures such
as nanoplatelets, nanowires, and quantum dots, which are crucial for the develop-
ment of plasmonic nanolasers.

4.2.3 Theoretical Model and Structure of a Plasmonic Nanolaser

Similar to a conventional laser, a plasmonic nanolaser consists of three main com-
ponents: a plasmonic cavity (instead of an optical cavity in a conventional laser), an
active medium, and a pumping source. Radiative emission is provided from active
material via the pumping process. The emission is then amplified in the plasmonic
cavity. When the optical gain is larger than the total optical losses, the system starts
to lase. The cavity feedback can be provided by either the resonant recirculation or
localized plasmonic mode.
The working principle of a plasmonic laser can be described by a three-level
gain system (Figure 4.3) [42]. The external energy source (electrically or optically)
excites electrons in the active medium from the ground state (N0) to the excited level
(N2). These excited atoms then quickly relax to the lower level (N1) and form exci-
tons. Next, the excitation recombines and transfers its energy (nonradiative) to the
plasmonic cavity mode (dash arrows). The interaction between the cavity mode and
excitations at excited states provides stimulated emission for the generation of the
laser. The laser operation can be described by rate equations, in which the excited

BK-TandF-KAZANSKIY_9781032572949-230642-Chp04.indd 104 9/26/23 4:57 PM

 Plasmonic Nanolasers 105

FIGURE 4.3 Schematic diagram showing energy transition in a typical plasmonic nanolaser.

carrier number N and the photon number s in the cavity mode are time-dependent
functions [17, 23]:

dN N
= R − ΓFP β 0 A ( N − N tr ) s − ΓFP β 0 AN − , (4.3)
dt τ nr

ds
= ΓFP β 0 A ( N − N tr ) s + ΓFP β 0 AN − γ p s, (4.4)
dt

where R is the pumping rate, Г is the confinement factor, FP is the Purcell factor, β0
is the spontaneous emission factor that contributes to the lasing mode without the
Purcell effect, A is the spontaneous emission rate, Ntr is the transparent excited car-
rier density, τ nr is the nonradiative transition lifetime, and γp is the damping rate of
the cavity.
The Purcell factor can be evaluated as follows [43]:
3
3 Q  λ
Fp =   , (4.5)
4π 2 V  n 

where Q is the quality factor of the cavity, V is the cavity mode volume, and n is the
refractive index of the medium.
The lasing threshold can be characterized as [23, 26]:

γp  1  1
Rth = 1 +  1 +  , (4.6)
2Γ  β  ζ

where β is the spontaneous emission factor that contributes to the lasing mode with
the Purcell effect, and ζ = γp/(ΓFpβ0ANtr) is the ratio of cavity loss γp to material loss
ΓFpβ0ANtr .
Plasmonic nanolasers can be realized on various architectures regarding
the mechanism for light amplification. In a single metal nanoparticle structure
[Figure 4.4(a)], the localized cavity mode is confined in all three dimensions (3D),
thus the mode volume and the laser size is very small. However, this structure suffers

BK-TandF-KAZANSKIY_9781032572949-230642-Chp04.indd 105 9/26/23 4:57 PM

 106 Photonics Elements for Sensing and Optical Conversions

FIGURE 4.4 Typical configurations of plasmonic nanolasers according to the feedback

method used for light amplification: (a) localized SP, (b) Fabry–Perot, (c) whispering gallery
mode, and (d) metal particle array.

a high optical loss, making it challenging to realize lasing. For structures with lower
confinement such as in 2D, the cavity mode propagating is along a waveguide where
light is amplified by the Fabry–Perot cavity [Figure 4.4(b)]. In other structures, the
cavity mode can be trapped in a surface where light is enhanced by whispering gal-
lery mode [Figure 4.4(c)]. In this case, the light is confined only in 1D at the direc-
tion perpendicular to the surface plane. Recently, as shown in Figure 4.4(d), a metal
particle array has demonstrated potential for arrays of nanoscale light sources with
unique properties such as directional emission [25].
In general, the electric-field intensity of a cavity mode of plasmonic nanolas-
ers can be simulated using COMSOL Multiphysics [44, 45]. Figure 4.5(a) shows a

FIGURE 4.5 Simulation of electric-field intensity (in grayscale) for different structures: (a)
a gold nanoparticle, (b) a GaAs nanowire separated from a metallic surface by a nanoscale
dielectric gap, (c) whispering gallery mode in a ZnO hexagonal cavity.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp04.indd 106 9/26/23 4:57 PM

 Plasmonic Nanolasers 107

simulation of the electric-field intensity of a gold nanoparticle. Figure 4.5(b) demon-

strates the cavity mode in the gap area between a GaAs nanowire and a silver surface
(using data in [46]). For a surface structure such as a hexagonal ZnO disc, light is
trapped by total internal reflection at the interface of the cavity and the air, and is
amplified by whispering gallery mode (Figure 4.5c).

4.3 EXPERIMENTAL DEMONSTRATIONS OF
THE PLASMONIC NANOLASERS
4.3.1 Single Nanoparticle-Based Plasmonic Nanolasers
A single nanoparticle-based plasmonic nanolaser, or a spaser, is a nanoplasmonic
device generating surface plasmons. The term “spaser” is short for surface plas-
mon amplification by stimulated emission of radiation, which was first proposed by
Stockman and Bergman in 2003 [47]. In a spaser, a metal nanoparticle plays the role
of a resonator with the replacement of its dielectric environment with an amplify-
ing material, and localized plasmons play the role of photons. The main difference
between spasers and lasers is that in a spaser plasmonic modes are nonradiatively
generated, amplified, and localized on the nanoparticles, whereas in a conventional
laser, wave modes are traveled within the cavity. That means conventional laser
mode contains photons while spaser mode contains plasmons. However, similar to
conventional lasers, when the supplied optical gain is larger than total optical losses,
surface plasmons generated in metallic nanostructures are amplified. In comparison
to a conventional laser, a spaser holds two unique properties. First, the localiza-
tion length of spacing modes is only restricted by the inhomogeneity of the metal
and the nonlocality radius (~1 nm). This provides an exclusive possibility to achieve
nanoscopic optical sources [48]. Second, regardless of how small a resonator can
be reached, it is still able to provide optical feedback due to local field confinement.
Thus, in the spaser configuration, it is not necessary to implement high reflective
mirrors like in a conventional laser [20, 42]. As a result, the spaser function is based
on two main factors: (1) the intrinsic electromagnetic properties of the nanoparticle
resonator (e.g. the nanoparticle size, the Q-factor, and other parameters) and (2) the
interaction conditions of the amplifying medium with the nanoparticle plasmonic
mode. The two simplest and most studied spaser configurations are (1) nanoshell
configuration, with the amplifying medium at the center of a metal nanoparticle
[49] and (2) nanoparticle configuration, with the active medium outside of a metal
nanoparticle [48, 50]. In these configurations, the presence of metal near the active
medium makes the possibility of transferring its excitation energy to the plasmonic
mode, thus, plasmonic excitation occurs.
Spasers were first experimentally demonstrated by Noginov with nanoparticle
configuration [15]. In this work, metal nanoparticles were 14-nm diameter gold nano-
spheres synthesized covered with a dielectric silicon shell containing Oregon green
488 dye molecules. The dye molecules have an emission wavelength of 510 nm,
which coincided with the plasmonic resonance wavelength of the as-synthesized
gold nanoparticles. The silicon dioxide layer thickness was 15 nm. The total size
of the 3D spaser was 44 nm in diameter. The spasers were optically pumped by

BK-TandF-KAZANSKIY_9781032572949-230642-Chp04.indd 107 9/26/23 4:57 PM

 108 Photonics Elements for Sensing and Optical Conversions

nanosecond pulse 488 nm laser; the lasing emission was observed at 525 nm with a
Q-factor of 14.8. Since the nanolaser had a small size in comparison with the lasing
wavelengths, the loss was comparably minor and the major loss was due to the gold
core. The ratio between the number of dye molecules and the number of nanopar-
ticles was 2:7 × 103. All the measurements were carried out with an ensemble of
spasers at a concentration of 3 × 1011 cm−3 in aqueous suspension.
An example of a nanoshell spaser configuration is a dielectric core doped with an
optical gain material that was covered by a metallic shell [51–53]. It was proved theo-
retically and experimentally that light can be amplified through a plasmonic struc-
ture containing gain material [15, 54]. The emission properties were numerically
examined by finite element method, and frequency domain nonlinear solver (Wave
optics module, COMSOL Multiphysics). The core was chosen to be either a mesopo-
rous silica sphere or polystyrene that can load dye molecules; the shell was set as Ag
or Au. It was shown that with this configuration an extremely high local field inten-
sity can be achieved with an enhancement factor that reached 108. Additionally, this
configuration could also create an enhancement factor of surface-enhanced Raman
spectroscopy (SERsp) with the order of 1016 –1017 [55]. The proposed nanoshell spaser
configuration has several advantages. First, the surface plasmon resonance emission
can be variable from the visible to the near-infrared regions by adjusting the edge
length and wall thickness of the spasers. Second, the surface plasmon resonance is
strong due to the energy transfer from the gain material, which will compensate the
radiative and nonradiative damping that normally occurs in surface plasmon-based
devices. Thus, the amplitude and energy of the spaser are amplified, and lasing can
be attained [55]. As a result, coupling among surface plasmon polariton, active atom
states and light is facilitated. Third, nanoshell spaser configuration can overcome a
drawback of nanoparticle configuration. Since the semiconductor cell in nanopar-
ticle configuration must contain dye molecules, the dye molecules must be able to
bind covalently with semiconductor material via a characteristic group. Besides, due
to the damping, the concentration of dye molecules must be sufficient for compensat-
ing plasmonic losses [21].

4.3.2 Fabry–Perot Cavity-Based Plasmonic Nanolasers

Technically speaking, in Fabry–Perot (FP)-based plasmonic nanolasers, optical
feedback will be obtained by an FP cavity formed by the two ends of a waveguide
while surface plasmonic waves are amplified by the gain medium.
In 2009, Hill et al. first demonstrated a plasmonic nanolaser-based waveguide
configuration consisting of a stack of metal-dielectric-metal layers. Their waveguide
consisted of a rectangular cross section InP/InGaAs/InP nanopillar surrounded by a
20-nm-thick layer of silicon nitride (SiN). The pillar was then encapsulated within
two Ag walls. The InGaAs semiconductor layer formed the gain materials, while the
SiN layer was used to separate gain materials. The two Ag walls were used to reduce
quenching, as demonstrated in Figure 4.6(a) [56]. Due to the difference in refractive
index between the InGaAs and InP, light can be confined vertically in the wave-
guide, and the two ends of the waveguide formed an FP cavity. The devices [SEM
image shown in Figure 4.6(b)] were fabricated by using several techniques, such

BK-TandF-KAZANSKIY_9781032572949-230642-Chp04.indd 108 9/26/23 4:57 PM

 Plasmonic Nanolasers 109

FIGURE 4.6 Fabry–Perot based plasmonic nanolasers. (a) Schematic diagram showing the
structure of cavity formed by a rectangular semiconductor pillar encapsulated in Ag. (b)
Scanning electron microscope image showing the semiconductor core of one of the devices.
The scale bar is 1 micron. (c) Above threshold emission spectra with pump current 180 µA
at 78 K. Inset: emission spectra for 20 (the lowest intensity), 40 (the middle), and 60 (the
highest intensity) µA, respectively, at 78 K. (d) Lasing mode light output (crosses), integrated
luminescence (circles), as a function of pump current at 78 K. Reprinted with permission
[56]. Copyright © 2009, Optical Society of America. (e) Side view of an Au-film plasmonic
waveguide-based nanolaser; the Au waveguide is sandwiched between two dies contain-
ing the QW stacks. (f) Emission spectra are narrower with increasing pump intensity. Inset:
Spectrum at 118 kW/cm2, far above the threshold. (Reprinted with permission from [58].
Copyright © 2011, Optical Society of America.)

as employing epitaxy, electron beam lithography, dry etching, and various material
deposition techniques. The thickness of the InGaAs semiconductor layer could be
reached down to 90 nm, which is about half of the diffraction limit in width. Such
small thickness can support the high confinement of TM01 mode. The nanolaser
was electrically pumped by injecting electrons via the top of the pillar and inject-
ing holes via the p-InGaAsP layer and the lateral contact area. Under cryogenic

BK-TandF-KAZANSKIY_9781032572949-230642-Chp04.indd 109 9/26/23 4:57 PM

 110 Photonics Elements for Sensing and Optical Conversions

temperatures, the nanolaser emitted light at telecommunication wavelength [Figure

4.6(c)]. The inset in Figure 4.6(c) shows the spectra of the device around and above
the threshold. Below the threshold, a broad spectrum from spontaneous emission
was observed. Just below the threshold, the full width half maximum (FWHM) of
the resonance was 4 nm, indicating a cavity Q-factor on the order of 370 at a cryo-
genic temperature of 78 K and Q-factor ~140 at room temperature. Lasing mode
light output and integrated luminescence as a function of pump current are shown in
Figure 4.6(d). When the laser threshold was reached (~40 µA), a narrow lasing mode
was obtained at 180 µA, and the FWHM was reduced to 0.7 nm. Later, a plasmonic
laser with a similar structure of InP/InGaAsP/InGaAs layer stack was introduced
whichcould function at room temperature with a Q-factor of 300. The laser included
a semiconductor core of about 140 nm [57]. Within the laser, a distributed feedback
(DFB) was implemented. The DFB structure was produced by applying electron
beam lithography and reactive ion etching together with surface oxidation and wet
chemical method. The metallic Bragg gratings provided a broad stopband of ~500
nm, and the grating coupling coefficient was more than 5,000/cm. Due to the grating,
spontaneous emission was strongly suppressed; thus, the lasing threshold was sig-
nificantly reduced in comparison with that of similar length FP waveguide cavities.
Furthermore, the reduced threshold pumping allowed achieving room temperature
lasing of the plasmonic nanolaser.
Another configuration for FP-based plasmonic nanolaser is demonstrated in
Figure 4.6(e) [58]. The nanolaser included an Au waveguide thin film located inside
InGaAs quantum well active medium. The waveguide length and width were 1 mm
and 100 µm, respectively. The InGaAs quantum wells were grown by metal-organic
chemical vapor deposition; Au thin film was formed by thin film deposition. A
50-nm thick InP spacer layer was implemented between the quantum wells and the
Au film to avoid coupling to the lossy mode, and also to minimize the quenching
of photoexcited electron-hole pair modes in the Au waveguide. The two ends of the
Au waveguide formed an FP resonator. The InGaAs active medium was pumped
by 1.06-µm laser pulses and the nanolaser threshold was ~60 kW/cm2, which is in
good agreement with calculations [Figure 4.6(f)]. Lasing was observed at room tem-
perature at the telecommunication wavelength of 1.46 µm. It was also shown that by
improving the Au film quality and bonding, the plasmonic losses can be reduced to
an intrinsic ohmic level, thus a threshold of ~500 A/cm2 could be obtained, which is
slightly higher than modern laser diodes in the laser wavelength range [58].
A new configuration of an FP-based plasmonic nanolaser implementing an active
medium nanowire on top of a metal substrate was first demonstrated by Oulton et al.
in 2009 [16]. In this work, a semiconductor CdS nanowire was placed on top of an
Ag substrate. The CdS dielectric nanowire and the Ag substrate were separated by
a nano thickness MgF2 layer. The CdS nanowires were used as the active medium.
The feedback was provided by wave reflection from the nanowire ends due to the
large difference in refractive index between the semiconductor and the air, creating
an FP resonator along the nanowire. The laser devices were optically pumped by an
fs pulse laser at a wavelength of 405 nm. The lasing output observed at the nanowire
ends was around 489 nm wavelength at below 10 K. In this configuration, the close
proximity of the semiconductor and metal interfaces confined light into an extremely

BK-TandF-KAZANSKIY_9781032572949-230642-Chp04.indd 110 9/26/23 4:57 PM

 Plasmonic Nanolasers 111

small area, which is a hundred times smaller than a diffraction-limited spot. The
transverse size of mode localization was extremely small, about 10 nm. However,
the tight confinement also led to a high cavity loss, which consequently resulted in
a high lasing threshold of ~60 MW cm−2. In comparison with photonic lasing of a
CdS nanowire placed on a MgF2 substrate, which showed a cutoff frequency when
the nanowire diameter reached down to 140 nm, the hybrid plasmonic lasing mode
was not limited in this dimension. In fact, the lasing from the plasmonic mode was
still observable with a nanowire diameter of only 52 nm.
Another approach based on a similar configuration was successfully demon-
strated by Wu et. al [59]. Their plasmonic nanolaser consisted of a bundle of InGaN/
GaN nanorods placed onto a 50-nm-thick colloidal Au triangular plate, separated by
a 5-nm-thick SiO2 gap layer. The nanolaser emitted in the green region with a lower
threshold of 300 kW cm−2. Afterward, several works with similar confinement con-
figurations were successfully developed to reduce the lasing threshold, operating the
lasing at room temperature, broadening the lasing spectrum, or pursuing even tighter
optical confinement of the lasing mode [60–63].
To obtain multiple color plasmonic nanolasers, several approaches have been
applied. One of the approaches is to use a semiconductor gain media with mul-
tiple compounds [34, 35]. Since the operation band of plasmonic nanolasers mostly
depends on the surface plasmon frequency, and on the electronic bandgap of the
gain media, therefore, by controlling the elemental composition of semiconductor
gain media, lasing emission can be tuned. For example, Lu et al. showed remarkable
work in developing a tunable InGaN/GaN core-shell hexagonal nanowire nanolaser
[Figure 4.7(a)–(b)] [35]. The nanolaser exhibited a single-mode, continuous, multi-
color subwavelength plasmon, operating at 7 K. The nanocavity included a core-shell
nanowire of InGaN/GaN with a diameter of ~30–50 nm and a length of 100–250 nm
placed on a 28-nm Ag film. A 5-nm Al2O3 film was implemented between the nanow-
ire and the Ag film. By tuning the percentage of In in the gain medium, varying from
~27% to 53%, the electronic bandgap was altered from ~2.65 to 1.93 eV, leading to a
tunable plasmonic lasing in the range of 474–627 nm [35].
A Fabry–Perot-based plasmonic nanolaser could also be obtained by coupling an
active medium nanowire with a metal nanowire. In 2013, Wu et al. reported a plas-
monic lasing structure based on a longitudinally hybridized cavity [64]. As shown
in Figure 4.7(c)–(d), a curved CdSe nanowire was coupled to a small point at the
side of a curved Ag nanowire with a coupling efficiency of 20%. This configura-
tion produced a longitudinal hybrid cavity, in which the lasing mode circulating was
alternately changed between a photonic mode in the CdSe nanowire and a plasmonic
mode in the Ag nanowire. Furthermore, it also provided the ability to separate spatial
contribution from plasmonic and photonic components at the emission output [24].
By optically pumping the CdSe nanowire at room temperature, lasing emission from
the plasmonic mode supported by the Ag nanowire was observed around 723 nm
wavelength, with an estimated mode area of 0.008λ2 [23]. The lasing emission also
exhibited a strong polarization dependence, indicating the electromagnetic nature
of the surface plasmon polariton waves. Although the nanolaser still has challenges
in minimizing metal loss and in improving the overall gain efficiency, the nanowire
lasers offer significant advantages, such as improved thermal management.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp04.indd 111 9/26/23 4:57 PM

 112 Photonics Elements for Sensing and Optical Conversions

FIGURE 4.7 Fabry–Perot-based plasmonic nanowire lasers. (a) Scheme of a multicolor

InGaN/GaN core-shell nanowire plasmonic laser. Inset on the upper right is the SEM
image, which indicates that the laser emissions are from an individual single nanorod. (b)
All-color, single-mode lasing images observed from single nanorods with an emission
line width ~4 nm. Reprinted with permission [35]. Copyright © 2014, American Chemical
Society. (c) Schematic diagram of a hybrid photon–plasmon nanowire laser composed of a
Ag NW and a CdSe NW coupled into X-shape. (d) Polarization-sensitive lasing spectra from
the Ag nanowire end-facet. The emission polarizations oriented parallel and perpendicular
are demonstrated as light gray and dark gray, respectively. The insets are dark-field micro-
scope images showing the dependence of polarization on lasing outputs. The arrows indicate
the directions of the polarization. (Reprinted with permission from [64]. Copyright © 2013,
American Chemical Society.)

4.3.3 Whispering Gallery Mode Cavity-Based Plasmonic Nanolasers

Besides Fabry–Perot configuration, whispering gallery mode configuration for plas-
monic nanolasers also attracts a lot of attention. In this configuration, the feedback
along wave propagation is obtained by the total internal reflection effect. This plas-
monic nanolaser configuration was first demonstrated in a planar structure shape by
Ma et al. [65]. The nanolasers had a planar nanosquare with a 45-nm CdS semicon-
ductor on top of a Ag substrate, isolated by a 5-nm-thick MgF2 layer. Here, the CdS
semiconductor functioned as an active material. Due to the proximity of the CdS

BK-TandF-KAZANSKIY_9781032572949-230642-Chp04.indd 112 9/26/23 4:57 PM

 Plasmonic Nanolasers 113

nanosquare and Ag surface, the modes of the CdS square were able to hybridize
with the surface plasmon polariton of the metal-dielectric interface. This resulted
in strong optical confinement of λ/20 in the MgF2 gap region with a relatively low
metal loss. The hybrid surface plasmon polariton mode was internally reflected on
the CdS nanosquare’s side borders. Since the device is ultrathin, the total internal
reflection regime of the device is only satisfied by plasmonic waves, not for photon
waves. Thus, only plasmonic modes were formed in this device. The total internal
reflection regime can also mitigate radiation loss. With this construction, the pro-
posed nanolaser had a strong field localization with low losses in the metal, and low
radiation loss. As a result, a high Q-factor approaching 100 and a Purcell factor of
18 were obtained. By controlling the geometry of the laser’s structure, a single-mode
lasing was also achieved at room temperature. Similarly, a whispering gallery mode-
based nanolaser was demonstrated in the form of a nanodisc [66]. The nanolaser
consisted of a 235-nm InP disc with four InAsP quantum wells embedded in the
middle of the disc. A glass substrate was placed on top of the disc for lasing emission
collection. The bottom and the surrounding areas of the disc were covered by Ag to
construct a nanopan [see Figure 4.8(a)–(d)] [66]. The Ag nanopan served as the laser
cavity for plasmonic mode confinement, and the four InP quantum wells act as an
active medium. The nanolasers were optically pumped by a 980-nm pulse laser and
exhibited a lasing mode at 1,308 nm at a cryogenic temperature of 8 K, and the lasing
threshold was ∼120 kW·cm−2 [Figure 4.8(e)] [66].
Also based on a similar configuration, Khajavikhan et al. represented a contin-
uous-wave telecommunications-frequency nanolaser that can function at room tem-
perature [67]. The laser cavity is based on a nanoscale coaxial-shaped cavity design
which is well known to electrical engineers and widely used in transmission lines

FIGURE 4.8 Whispering gallery mode-based plasmonic nanolasers. (a) Schematic diagram
of the plasmonic nanolaser-based nanodisc/nanopan structure including an InP nanodisc with
four InAsP quantum wells in the middle. A transparent glass substrate was placed on top of
the nanodisc, and the bottom and the sidewalls of the disc were covered by Ag creating a
nanopan. (b) Schematic illustration of the removal of the silver nanopan. (c) SEM image of
the InP disc on glass before silver deposition. (d) SEM image of the silver film without the
disc. The arrow shows damage by the separation of the disc from the silver film. Scale bars
in (c) and (d) are 400 nm. (e) Lasing emission at 8 K. (Reprinted with permission from [66].
Copyright © 2010, American Chemical Society.)

BK-TandF-KAZANSKIY_9781032572949-230642-Chp04.indd 113 9/26/23 4:57 PM

 114 Photonics Elements for Sensing and Optical Conversions

in the microwave regime. The cavity had a metallic rod at its center and was sur-
rounded by air, six InGaAsP-based quantum walls, and a thin layer of SiO2. The
whole assembly was then coated with Ag/Al alloy. The metal coverage of the devices
and the air and silica “plugs” served to enforce the mode confinement.
Additionally, the metallic coating functioned as a heat sink that allowed the
nanolaser to work with continuous-wave operation at room temperature. The lasers
were then optically pumped using a commercial 1064 nm laser. The output room-
temperature and continuous lasing modes operating at telecommunication wave-
lengths of 1260 nm and 1590 nm were obtained with practically no threshold. Due
to the small size of the coaxial structures, the laser also exhibited a direct couple of
~99% of spontaneous emission into lasing mode. In fact, by tailoring the nanolaser’s
geometry, such as for the core radius, the ring width, the low index plugs, and the
gain medium’s height, the laser’s modal content can be further modified [65]. It is
noted that the advantage of a whispering gallery mode cavity is low optical loss,
which is significant for obtaining a low lasing threshold. As a result, it is also widely
employed for the realization of microlasers [68].
An interesting approach for obtaining nanolasers multiple colors is to vary the
plasmonic nanocavity’s morphology and dimension such as the substrate type, the
width, and the length of semiconductors. Ma et al. demonstrated a room-tempera-
ture, multicolor plasmonic laser that consisted of a single CdS nanobelt and five Ag
strips [69]. The nanobelt was crosswise placed onto the Ag strips. The nanobelt’s
width and thickness were ~620 nm and 100 nm, respectively. Each Ag strip had a
width of 1 μm and a thickness of 250 nm. A 5-nm-thick MgF2 gap was also imple-
mented as an isolator. Additionally, six In/Au (10/120 nm) ohmic contact electrodes
were integrated into the laser structure for additional electrical modulation, as shown
in Figure 4.9(a)–(b) [69].
The overlap areas of Ag and CdS were optically pumped by an fs laser. It was
shown that the square-shaped overlap areas formed a local plasmonic laser cavity by
the total internal surface plasmonic reflection. It was attributed to the high refractive
index contrast between the surface plasmonic mode and the surroundings. As shown
in Figure 4.9(c), the evolution from broad spontaneous emission to lasing can be seen
clearly and lasing threshold is determined to be around 6 GW·cm−2. By varying the
width of the CdS waveguide, the lasing emission was tuned from 490.2 to 502.7 nm
[Figure 4.9(d)]. Interestingly, a small linear peak shift (below 0.3 nm) was observed
by applying an electric field of 4V. It was referred to as the density enhancement of
excited carriers. This helped to vary the wavelength in a very precise and dynamic
way [69].

4.3.4 Particle and Hole Arrays-Based Plasmonic Nanolasers

A single particle-based nanolaser or spaser can be extremely small and can contain
only several hundred atoms, thus the energy emission from such a small volume is
also very small. The use of a periodic structure of those separate nanoparticles or
nanoholes can significantly increase the generated energy due to the constructive
interference between adjacent nanoparticles/nanoholes. This can support direct las-
ing emission and enhance the efficient out-coupling of plasmonic nanolasers.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp04.indd 114 9/26/23 4:57 PM

 Plasmonic Nanolasers 115

FIGURE 4.9 Multiple color plasmonic nanolasers based Whispering Gallery Mode. (a)
Schematic diagram of a waveguide-embedded plasmonic laser, including a CdS nanobelt
waveguide crossing over a Ag strip and thin MgF2 gap layer. (b) SEM image of a plasmonic
nanolaser including a CdS nanobelt crossing five Ag strips. (c) Lasing spectra with different
pump intensities from 3.3 GW·cm−2 (black) to 7.4 GW·cm−2 (gray). The inset demonstrates
integrated light output as a function of pump intensity. (d) Lasing emission was tunable in
the range of 490.2–502.7 nm. (Reprinted with permission from [69]. Copyright © 2012,
American Chemical Society.)

4.3.4.1 Plasmonic Nanolasers Based on Nanoparticle Arrays

The possibility of obtaining lasing from a periodic structure formed by metal nanopar-
ticle arrays was first demonstrated by Stehr et al. [70]. Herein, arrays of Au nanodiscs
with a diameter of 110 nm and a height of 30 nm were embedded into a 460-nm organic
polymer film (methyl-substituted ladder-type poly para phenylene). The organic poly-
mer functioned as a gain medium while the nanodisc arrays acted as a DFB grating. The
laser was optically pumped by a 130-fs laser at a wavelength of 400 nm. In this work,
the laser did not require two end mirrors for cavity feedback since the periodic structure
of the nanodisc array provided optical feedback for the laser. As a result, directional las-
ing emission was supported at a very low threshold of only 2 nJ, and the emission line
width was narrow of 0.44 nm at the wavelength of 492 nm. The constructive interfer-
ence between adjacent nanodisc arrays also enhanced the efficient out-coupling.
The IR lasing emission was also obtained with a similar nanoparticle array-based
nanolaser configuration. The nanolasers contained either Ag or Au nanoparticles with
an average diameter of 130 nm surrounded by IR – 140 dye molecules. The laser could
function at room temperature with directional emission at a wavelength of 913 nm and
a spectral width of 1.3 nm [71]. Interestingly, the plasmonic character of lasing was
also confirmed by performing measurements with periodic nanoparticle arrays of
non-plasmon materials such as Ti and TiO2. No lasing was observed in this case.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp04.indd 115 9/26/23 4:57 PM

 116 Photonics Elements for Sensing and Optical Conversions

FIGURE 4.10 Tunable nanoparticle arrays based on plasmonic nanolasers. (a) The sche-
matic diagram demonstrates a tunable laser with a microfluidic channel. (b) Continuous tun-
ing of the lasing emission was obtained due to the difference in the refractive index of liquid
gain materials. Reprinted with permission [73]. Copyright © 2015, The Authors. (c) and (d)
Schematic diagram of the plasmonic nanolasers consisting of Au nanoparticle arrays based
embedded within a stretchable dye-doped PDMS gain material. (e) and (f) Schematic of a
tunable nanolaser based on the difference of applied strain ε on the elastomeric substrate.
(Reprinted with permission from [74]. Copyright © 2018, American Chemical Society.)

Later, tunable lasing of nanoparticle array-based nanolasers was observed in

[72–74]. The laser was constructed by embedding Au nanoparticle arrays within
microfluidic channels [Figure 4.10(a)–(b)]. When liquid gain materials with differ-
ent refractive indices flew in those channels, a plasmonic lasing emission was tuned
dynamically in a wavelength range of 862 to 891 nm with an emission line width of
1.5 nm [72, 73]. The laser tuning could also be obtained by changing lattice spacing
between the metal nanoparticle arrays [Figure 4.10(c)–(f)] [74]. In this approach,
Au nanoparticle arrays were patterned onto a flexible substrate that contained an
elastomeric slab of polydimethylsiloxane (PDMS) and liquid dye molecules. By
changing the interparticle distances, the cavity resonance was modulated, and thus
a continuous tuning emission over 31 nm was obtained by stretching and releasing
the flexible substrate.

4.3.4.2 Plasmonic Lasers Based on Nanohole Arrays

Lasing in a periodic structure of nanoholes on a thin film embedded within an active
medium was also demonstrated [75–78]. Theoretical study of a system consisting
of a metal film perforated by a periodic array of nanoholes predicted that when the

BK-TandF-KAZANSKIY_9781032572949-230642-Chp04.indd 116 9/26/23 4:57 PM

 Plasmonic Nanolasers 117

spasers were pumped, they exhibited a mutual synchronization between each other,
which led to the generation of a superradiant lasing. The radiation intensity of the
periodic nanohole arrays-based nanolasers was considerably higher than that of a
single spaser, by two orders of magnitude [78].
Experimental studies on the plasmonic lasing from metal hole arrays are shown
in Figure 4.11 (a) and (b) [75]. A stacked layer of 5 nm SiN, 15 nm InP, and 105 nm
InGaAs was subsequently deposited on an InP substrate. Afterward, a 100-nm Au
film was coated on the stack layer, and the nanohole arrays were fabricated by lithog-
raphy. The nanoholes had a diameter of 160 nm each. The InGaAs acted as an active
medium. The nanolaser was pumped at the wavelength of 1,064 nm at cryogenic
temperatures. Below the threshold, emission at three frequencies was observed,
which was attributed to the resonances of the hole arrays. Above the threshold, the
resonance at 1,480 nm became dominant, with emission power increased by more
than three orders of magnitude. The dominant resonance was denoted as lasing.
Later, lasing from periodic hole arrays-based nanolasers was also observed at
room temperature [77, 79]. The nanolasers consisted of an array of nanoholes on

FIGURE 4.11 Nanohole arrays-based plasmonic nanolasers. (a) Schematic diagram of a

nanohole arrays-based plasmonic nanolaser including a Au metal hole array deposited on
an InGaAs gain medium. (b) Below the threshold, emission spectra exhibited three frequen-
cies, and above the threshold, the wavelength at 1,480 nm became dominant and denoted as
lasing. Reprinted with permission [75]. Copyright © 2013, American Physical Society. (c)
Planar plasmonic laser. The laser apparatus uses a perforated silver film. (d) The spectra of
a plasmonic laser used a liquid medium as an active medium at below and above the lasing
threshold. The insets show the SEM image of the nanohole arrays. (Reprinted with permis-
sion from [77]. Copyright © 2017, AIP Publishing.)

BK-TandF-KAZANSKIY_9781032572949-230642-Chp04.indd 117 9/26/23 4:57 PM

 118 Photonics Elements for Sensing and Optical Conversions

Ag thin film deposited on a silica substrate [Figure 4.11(c)] [77]. An active medium
could be either dye molecules in a liquid solution [77] or a polymer film [79]. A
single mode at 610 nm with a narrow width of 1.5 nm was obtained when a rhoda-
mine 101 dye-polymer film was implemented as a gain medium [79]. The nanolasers
utilized a solution of R101 dye mixed with dimethyl sulfoxide (DMSO) as an active
medium. The liquid active gain medium was supposed to limit the photobleaching
of the nanolasers. The nanolaser exhibited a lasing emission at 628-nm wavelength
with a spectral line width of 1.7 nm [Figure 4.11(d)] [77].

4.3.5 Summary of Laser Characteristics

Summaries of some important characteristics of typical nanolasers based on the
mechanism for light amplification are shown in Table 4.1.

4.4 APPLICATIONS OF NANOLASERS

Plasmonic nanolasers have demonstrated very interesting properties, such as small
physical volume and the ability to deliver optical energy to even smaller modes with
simultaneously spectral and temporal optical localization. These unique properties
open up a wide range of applications for plasmonic nanolasers, from integrated pho-
tonic circuits, to sensing, to biological applications.

4.4.1 Plasmonic Nanolasers for Integrated Photonic Circuits

The main challenge in the development of photonic integration is the generation
of efficient radiation sources on a chip. Plasmonic nanolasers can be considered
as potential sources for photonic integration due to their unique properties such as
their ability to be integrated with other electronic and optical components. Their
subwavelength size and low-quality cavities can offer frequency modulations with
bandwidths of over hundreds of gigahertz. Furthermore, in photonic devices the
capacitance load effect inherent, which is a typical characteristic of electric circuits,
is absent. Therefore, photonic integration systems based on plasmonic nanolasers
can significantly overcome the current problems of modern electronic devices.
In 1996, Miller et al. first proposed the possible application of nanolasers in
photonic integrated circuits [83]. Based on this proposal, Kim et al. simultaneously
demonstrated a metal-clad plasmonic nanolaser integrated onto a bidirectional or
unidirectional, single-mode Si on an insulator waveguide [84, 85]. The actual nano-
cavity consists of an InGaAsP bulk semiconductor cuboid cladded by Si, SiO2, and
Ag (Figure 4.12) [84].
By adjusting the nanocavity size (l parameter), and the cladding thickness (a, g
parameters), the optimum coupling efficiency of the plasmonic laser with a single-
mode on an insulator Si waveguide was obtained. The maximum coupling efficiency
of 90% could be achieved with a unidirectional waveguide [85]. Depending on the
structure, the Q-factor of the nanolasers varied from 600 to 1,200 [84, 85].
Recently, Ma et al. experimentally demonstrated subwave guide-embedded plas-
monic nanolasers with over 70% coupling efficiency into a semiconductor nanobelt

BK-TandF-KAZANSKIY_9781032572949-230642-Chp04.indd 118 9/26/23 4:57 PM

 TABLE 4.1
Summaries of Some Important Characteristics of Typical Nanolasers Based on the Mechanism for Light Amplification
Gain
Cavity Structure Size Medium Emission Pumping and Lasing Conditions Threshold Refs
Localized SP Spaser (core/shell) Core: Au Diameter: 44 nm Dye 531 nm Optical, pulsed Room temperature ~4 nJ [15]
nanoparticle/Sheel: dye-doped silica
Spaser (core/shell) Core: Au Diameter: 22 nm Dye 550 nm Optical, pulsed Room temperature ~26 mJ·cm–2 [80]
Plasmonic Nanolasers

nanoparticle/Shell: dye-doped silica

Fabry–Perot Metal-Insulator-Metal waveguides Thickness: ~310 nm InGaAs 1500 nm Electrical, continuous wave Room ~5 mA [56]
Length: 6 μm temperature
Metallic-coated nanopillar Diameter: 260 nm InGaAs 1400 nm Electrical, continuous wave ~5 μA [14]

BK-TandF-KAZANSKIY_9781032572949-230642-Chp04.indd 119
High: 300 nm 77 K
Fabry–Perot Plasmonic coaxial Diameter: 200 nm InGaAsP 1380 nm Optical, continuous wave 4.5 K Thresholdless [67]
Height: 210 nm MQWs
Plasmonic nanowire Diameter: 60 nm InGaN 510 nm Optical, continuous wave 78 K ~3 kW·cm−2 [62]
Length: 480 nm
Plasmonic nanowire Diameter: 100 nm CdS 490 nm Optical, pulsed 10 K ~100 MW·cm–2 [16]
Length: ~10 μm
Plasmonic nanowire Diameter: ~100 nm GaN 375 nm Optical, pulsed Room temperature ~3.5 MW·cm–2 [61]
Length: ~15 μm
Plasmonic pseudo-wedge Wedge length: 80 nm ZnO 370 nm Optical, pulsed 77 K ~55 MW·cm–2 [81]
Whispering Metallic-coated disc Thickness: 480 nm InGaAsP 1430 nm Optical, pulsed Room temperature ~70 kW·cm−2 [82]
Gallery Mode Diameter: 490 nm
Plasmonic nanodisc Thickness: 235 nm InAsP 1300 nm Optical, pulsed 8 K ~120 kW·cm−2 [66]
Diameter: ~1 μm MQWs
Whispering Plasmonic nanosquare Thickness: 45 nm CdS 495 nm Optical, pulsed Room temperature ~2 GW·cm−2 [65]
Gallery Mode Length: 1 μm
Particle and Arrays of Au nanoparticles – Dye 913 nm Optical, pulsed Room temperature ~0.23 mJ·cm−2 [71]
Hole Array Array of holes on an ITO substrate – Dye 620 nm Optical, pulsed Room temperature 12 mW [79]
119

9/26/23 4:57 PM
 120 Photonics Elements for Sensing and Optical Conversions

FIGURE 4.12 Plasmonic nanolasers for integrated photonic circuits: (a) schematic diagram,
(b) the side view, and (c) the top view of a Si-waveguide coupled metal-clad nanolaser cav-
ity. (d) Electric field |E| distribution in the x-z and y-z plane. (e) Far field radiation patterns
to the substrate direction of the plasmonic nanolaser. (Reprinted with permission from [84].
Copyright © 2011, Optical Society of America.)

waveguide [69]. The plasmonic nanolaser-based integrated circuit had four key prop-
erties: multicolored plasmonic light source, wavelength multiplexing, efficient wave-
guide collection and out-coupling, and direct electrical modulation. The compact
and novel configuration of the nanolaser opens up a promising application in large-
scale, on-chip, integrated hybrid optoelectronic circuitry [24, 28].

4.4.2 Plasmonic Nanolasers for Sensing

Surface plasmon resonance sensors are now widely used in biochemical sensing
and detection [86]. However, their strong radiative and nonradiative damping, which
leads to the weakness of the plasmonic resonance, are their main drawbacks. As a
result, the plasmonic resonances in the visible and near-infrared range exhibit broad
line widths of tens to hundreds of nanometers. This strongly limits the sensor’s sen-
sitivity, and increase the sensor’s limit of detection. However, this drawback can be
fully solved by injecting a gain medium in a plasmonic nanolaser cavity. Due to the
amplification of the plasmonic resonance, the line width can be narrowed by two
orders of magnitude, thus, the sensor’s sensitivity can be significantly enhanced [87].

BK-TandF-KAZANSKIY_9781032572949-230642-Chp04.indd 120 9/26/23 4:57 PM

 Plasmonic Nanolasers 121

The application of a spaser in surface-enhanced Raman scattering (SERS) was

numerically investigated [55]. It was shown that the spaser exhibited an extremely
high local field intensity with an enhancement factor reached 108, while the enhance-
ment factor of SERS was on the order of 1016 –1017. These enhancement factors are
sufficient for single-molecule detection [55]. The surface plasmon resonance peak in
the spasers can be tuned from visible to NIR range by controlling the edge length
and the thickness of the shell. This can facilitate single-molecule SERS detection
using different laser wavelengths. The strong coherent high-intensity lasing emission
from the spaser will lead to the sufficiently high efficiency of single-molecule SERS
signal observation [55].
Significant enhancement in sensor’s performance with ultrahigh sensitivity has
been obtained with plasmonic nanolaser-based sensors. Ma et al. demonstrated a
nanolaser that consists of a single crystalline semiconductor CdS nanoslab (50 nm
in thickness, and 600 nm in length) placed on top of an Ag surface, and an 8-nm
gap layer of MgF2 was inserted in between [88]. Here, the CdS semiconductor slab
with atomically smooth surface functions at the same time as an active medium, a
feedback geometry, and a sensing medium. The sensor’s working principle is based
on lasing-enhanced surface plasmon resonance in which the electromagnetic field
was localized at the interface between the metal and the semiconductor by the sur-
face plasmon. This allowed both the device’s physical size and mode confinement
to shrink down to the nm scale in the vertical dimension. As a proof of concept,
the plasmonic nanolaser-based sensor was utilized for the detection of three differ-
ent explosive molecules: 2,4-dinitrotoluene, ammonium nitrate, and nitrobenzene. It
was shown that when the concentration of the analyte of interest was increased, it led
to a decrease in laser emission intensity, while there was no noticeable change in the
peak wavelength. The sensing mechanism is referred to as the surface recombina-
tion velocity modification of the gas molecules. Due to the large surface area to vol-
ume ratio, the sensor’s sensitivity was ultrahigh, which could reach down to subpart
per billion. It ensured that any modulation on the atomically smooth surface would
strongly influence the laser emission intensity.
A similar configuration of plasmonic nanolasers with CdSe semiconductor layer
on top of Au substrate separated by a few nanometers of MgF2 was also imple-
mented for the refractive index sensor [87, 89]. Wang et al. demonstrated that the
sensor’s stability was improved by thin film deposition of Al2O3 on top of CdSe
layer [89]. Here, the sensors represented a significant wavelength shift of about 22
nm per refractive index unit at the wavelength of 718 nm. Due to the high spectral
coherence of lasing surface plasmon (with a line width of less than 0.3 nm at λ =
700 nm), the sensor’s sensitivity and the Figure of Merit were greatly improved.
Additionally, the high signal-to-noise ratio, which was due to the considerably sup-
pressed background emission, and the sharp tails of the Gaussian shaped laser spec-
trum also contributed to the significant enhancement of the sensors’ Figure of Merit.
When sensing the refractive index of a mixed solution of water and glycerin, the
nanolasers presented a Figure of Merit of about 40 times higher than state-of-the-
art surface plasmon resonance-based sensor [89]. In addition, the Figure of Merit
reached up to 400 times when sensing the refractive index change of ethanol and
propyl alcohol mixed solution [88]. For a demonstration, as shown in Figure 4.13a,

BK-TandF-KAZANSKIY_9781032572949-230642-Chp04.indd 121 9/26/23 4:57 PM

 122 Photonics Elements for Sensing and Optical Conversions

FIGURE 4.13 Plasmonic nanolaser for sensing. (a) A laser device’s emission peak wave-
length evolution in varied refractive index environments. From top to down, the device is in
propyl alcohol and ethanol, respectively. (b) The spectrum shifts with the refractive index
at the lasing state. From left to right, the device in ethanol and propyl alcohol, respectively.
(Reprinted with permission from from [87]. Copyright © 2017, the Author(s).)

the lasing peak wavelength of a plasmonic nanolaser shifts with the refractive index
in two different environments [87]. Interestingly, the performance of the device is
quite stable because the value of the wavelength shift is independent of the pumping
power. The spontaneous emission and lasing spectra in ethanol and propyl alcohol
are plotted in Figure 4.13b, which indicates a clear redshift with the increase of the
refractive index of the solution [87].

4.4.3 Plasmonic Nanolasers for Biological Applications

Plasmonic nanolasers have been attracting a lot of attention for biological applica-
tions, including biological probing, cellular labeling and tracking, and super resolu-
tion imaging.
E.I. Galanzha et al. demonstrated an application of a biocompatible spaser as a
biological probe [80]. The spaser with nanoparticle configuration had a gold nano-
sphere core and a silica shell doped with dye molecules (see Figure 4.14). The spasers
were then loaded into complex biological backgrounds from cellular cytoplasm in
vitro to a mouse tissue in vivo. The proposed spasers can be considered as one of the
smallest nanolasers, with an average size of only 22 nm (including a 10-nm diameter
Au core and a 6-nm-thick silica shell embedded with organic dye molecules), which
is approximately 100-fold smaller than microlasers. Under the excitation of a single
ns laser pulse, the spasers functioned as low-toxicity ultrafast probes with narrow
emission spectra of ~1 nm and exhibited a super bright stimulated emission. This
can be clearly seen above the threshold. The image of human cancer cells with the
integrated spasers has a high image contrast exhibiting many individual “hot spots.”
Furthermore, the spasers also exhibited strong photoacoustic effects when the opti-
cal pump was increased. This opened up a promising possibility of using spasers as
a useful tool for theranostic diagnosis and photothermal destruction of cancer cells.
Due to the sub-resolution size, nanolasers can be essentially used for super-
resolution nanoscopy. Cho et. al. represented a super high resolution optical

BK-TandF-KAZANSKIY_9781032572949-230642-Chp04.indd 122 9/26/23 4:57 PM

 Plasmonic Nanolasers 123

FIGURE 4.14 Spaser for biomedical application. (a) Schematic of a spaser as a multimodal
nanoprobe loaded in a single cell. (b) Fluorescence image of a breast cancer cell loaded with
a single spacer. (c) Fluorescent image of a breast cancer cell loaded with multiple spacers.
(d) Transverse electromagnetic mode (TEM) image of spacers within a breast cancer cell. (e)
Photothermal image of a breast cancer cell with spacers. (f) Absorption spectra of uranine
and Au NPs and normalized photoacoustic (PhA) spectra of spasers in linear and nonlinear
modes with respect to the increased optical pump. (Reprinted with permission from [80].
Copyright © 2017, the authors.)

microscopy technique, known as Laser Particle Stimulated Emission (LASE),

that uses the nonlinear effect of nanolasers [90]. The technique utilized iodide
perovskite nanowires laser embedded in a sample as an imaging probe. At the
focal point, the excitation beam was focused into a small region which is defined
by the diffraction limit.
Around the lasing threshold, the nanolasers exhibited nonlinear light output, indi-
cating that at the focal point, lasing was achieved at a very narrow region below the
diffraction limit (as seen as the beam profile along the z-axis in Figure 4.15). Outside
the focal point, the stimulated emission intensity can be considered as negligible,
which yielded a low out-of-focus background, and thus led to an increase in signal to
background contrast. This enabled the enhancement in imaging depths in scattering
samples. Since the nonlinear effect just appeared at the focal point, the technique
does not require a tight pinhole to remain high resolution as it does in confocal
microscopy. The stimulated emission from the nanolasers led to an enhancement of
optical resolution that is around five times higher than that of the same device oper-
ated below lasing threshold [90].

BK-TandF-KAZANSKIY_9781032572949-230642-Chp04.indd 123 9/26/23 4:57 PM

 124 Photonics Elements for Sensing and Optical Conversions

FIGURE 4.15 The principle of LASE microscopy. (a) Laser particles were excited by a
tightly focused optical pump beam. (b) Emission lasing energy is a function of optical pump.
The lasing emission increased sharply when the optical pump energy was above the thresh-
old. (c) The comparison of point spread function between the laser emission (gray line) and
the traditional diffraction-limited fluorescence detection (black line). (Reprinted with per-
mission from [90]. Copyright © 2016, American Physical Society.)

4.5 SUMMARY AND OUTLOOK

We have shown the overview and advances of nanolasers. In the last decade, research
works on nanolasers have obtained great achievements in both theoretical studies
and experimental demonstrations. Nowadays, nanolasers are exhibited in many
forms, such as single nanoparticles, an array of nanoparticles, nanowires, nanopil-
lars, and nanodiscs.
With unique properties such as small size and low power consumption, nanolas-
ers can be important for a wide range of applications. For instance, optically pumped
nanolasers have already shown great potential for biosensing and cell tracking. More
importantly, nanolasers can be employed in photonic-integrated circuits, which pro-
vide ultrafast optical communications. However, to attain the level of performance
required for this application, two issues need to be addressed. First, the performance
of plasmonic nanolasers has to improve to achieve electrically injected nanolasers.
To solve this problem, both cavity configuration and metal structure are crucial, thus
novel fabrication techniques and materials should be developed. Second, the com-
patibility of nanolasers within currently existing systems that mainly rely on a much
larger size of nanolasers must be considered [26]. It is expected that this issue will be
solved in the near future by research investment and research collaboration between
industry and academic communities.

REFERENCES
1. Maiman, T. H. 1960. Stimulated optical radiation in ruby. Nature. 187(4736):493–4.
https://doi.org/10.1038/187493a0.
2. Vahala, K. J. 2003. Optical microcavities. Nature. 424(6950):839–46. https://doi.
org/10.1038/nature01939.
3. Humar, M., and Yun, S. H. 2015. Intracellular microlasers. Nat. Photon. 9(9):572–6.
https://doi.org/10.1038/nphoton.2015.129.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp04.indd 124 9/26/23 4:57 PM

 Plasmonic Nanolasers 125

4. He, L., Ozdemir, S. K., Zhu, J., Kim, W., and Yang, L. 2011. Detecting single vi-ruses
and nanoparticles using whispering gallery microlasers. Nat. Nanotech. 6(7):428–32.
https://doi.org/10.1038/nnano.2011.99.
5. Hall, R. N., Fenner, G. E., Kingsley, J. D., Soltys, T. J., and Carlson, R. O. 1962.
Coherent light emission from GaAs junctions. Phys. Rev. Lett. 9(9):366–8. https://doi.
org/10.1103/PhysRevLett.9.366.
6. Lee, Y. H., Jewell, J. L., Scherer, A., McCall, S. L., Harbison, J. P., and Florez, L. T.
1989. Room-temperature continuous-wave vertical-cavity single-quantum-well micro-
laser diodes. Electron. Lett. 25(20):1377–8. https://doi.org/10.1049/el:19890921.
7. McCall, S. L., Levi, A. F. J., Slusher, R. E., Pearton, S. J., and Logan, R. A. 1992.
Whispering-gallery mode microdisk lasers. Appl. Phys. Lett. 60(3):289–91. https://doi.
org/10.1063/1.106688.
8. Spillane, S. M., Kippenberg, T. J., and Vahala, K. J. 2002. Ultralow-threshold Raman
laser using a spherical dielectric microcavity. Nature. 415(6872):621–3. https://doi.
org/10.1038/415621a.
9. Painter, O., Lee, R. K., Scherer, A., Yariv, A., O’Brien, J. D., Dapkus, P. D., and
Kim, I. 1999. Two-dimensional photonic band-gap defect mode laser. Science.
284(5421):1819–21.
10. Huang, M. H., Mao, S., Feick, H., Yan, H., Wu, Y., Kind, H., Weber, E., Russo, R.,
and Yang, P. 2001. Room-temperature ultraviolet nanowire nanolasers. Science.
292(5523):1897–9. https://doi.org/10.1126/science.1060367.
11. Hill, M. T., and Gather, M. C. 2014. Advances in small lasers. Nat. Photon. 8(12):
908–18. https://doi.org/10.1038/nphoton.2014.239.
12. Barnes, W. L., Dereux, A., and Ebbesen, T. W. 2003. Surface plasmon subwavelength
optics. Nature. 424(6950):824–30. https://doi.org/10.1038/nature01937.
13. Brongersma, M. L., and Shalaev, V. M. 2010. The case for plasmonics. Science.
328(5977):440–1. https://doi.org/10.1126/science.1186905.
14. Hill, M. T., Oei, Y.-S., Smalbrugge, B., Zhu, Y., de Vries, T., van Veldhoven, P. J., van
Otten, F. W. M., Eijkemans, T. J., Turkiewicz, J. P., de Waardt, H., Geluk, E. J., Kwon,
S.-H., Lee, Y.-H., Nötzel, R., and Smit, M. K. 2007. Lasing in metallic-coated nano-
cavities. Nat. Photon. 1:589–94. https://doi.org/10.1038/nphoton.2007.171.
15. Noginov, M. A., Zhu, G., Belgrave, A. M., Bakker, R., Shalaev, V. M., Narimanov, E.
E., Stout, S., Herz, E., Suteewong, T., and Wiesner, U. 2009. Demonstration of a spaser-
based nanolaser. Nature. 460(7259):1110–2. https://doi.org/10.1038/nature08318.
16. Oulton, R. F., Sorger, V. J., Zentgraf, T., Ma, R.-M., Gladden, C., Dai, L., Bartal, G., and
Zhang, X. 2009. Plasmon lasers at deep subwavelength scale. Nature. 461(7264):629–
32. https://doi.org/10.1038/nature08364.
17. Ding, K., and Ning, C. Z. 2012. Metallic subwavelength-cavity semiconductor nanolas-
ers. Light Sci. Appl. 1(7):e20. https://doi.org/10.1038/lsa.2012.20.
18. Ma, R.-M., Oulton, R. F., Sorger, V. J., and Zhang, X. 2013. Plasmon lasers: Coherent
light source at molecular scales. Laser Photon. Rev. 7(1):1–21. https://doi.org/10.1002/
lpor.201100040.
19. Premaratne, M., and Stockman, M. I. 2017. Theory and technology of SPASERs. Adv.
Opt. Photon. 9(1):79–128. https://doi.org/10.1364/AOP.9.000079.
20. Wang, Z., Meng, X., Kildishev, A. V., Boltasseva, A., and Shalaev, V. M. 2017.
Nanolasers enabled by metallic nanoparticles: From spasers to random lasers. Laser
Photon. Rev. 11(6):1700212. https://doi.org/10.1002/lpor.201700212.
21. Balykin, V. I. 2018. Plasmon nanolaser: Current state and prospects. Physics-Uspekhi.
61(9):846–70. https://doi.org/10.3367/UFNe.2017.09.038206.
22. Xu, L., Li, F., Liu, Y., Yao, F., and Liu, S. 2019. Surface plasmon nanolaser: Principle,
structure, characteristics and applications. Appl. Sci. 9(5):861. https://doi.org/10.3390/
app9050861.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp04.indd 125 9/26/23 4:57 PM

 126 Photonics Elements for Sensing and Optical Conversions

23. Wu, H., Gao, Y., Xu, P., Guo, X., Wang, P., Dai, D., and Tong, L. 2019. Plasmonic
nanolasers: Pursuing extreme lasing conditions on nanoscale. Adv. Optical Mater.
7(17):1900334. https://doi.org/10.1002/adom.201900334.
24. Chun, L., Liu, Z., Chen, J., Gao, Y., Li, M., and Zhang, Q. 2019. Semiconductor
nanowire plasmonic lasers. Nanophotonics. 8(12):2091–110. https://doi.org/10.1515/
nanoph-2019-0206.
25. Wang, D., Wang, W., Knudson, M. P., Schatz, G. C., and Odom, T. W. 2018. Structural
engineering in plasmon nanolasers. Chem. Rev. 118(6):2865–81. https://doi.org/10.1021/
acs.chemrev.7b00424.
26. Ma, R.-M., and Oulton, R. F. 2019. Applications of nanolasers. Nat. Nanotech. 14(1):
12–22. https://doi.org/10.1038/s41565-018-0320-y.
27. Sambles, J. R., Bradbery, G. W., and Yang, F. 1991. Optical excitation of surface plasmons:
An introduction. Contemp. Phys. 32(3):173–83. https://doi.org/10.1080/00107519108211048.
28. Johnson, P. B., and Christy, R. W. 1972. Optical constants of the noble metals. Phys.
Rev. B. 6(12):4370–9. https://doi.org/10.1103/PhysRevB.6.4370.
29. McPeak, K. M., Jayanti, S. V., Kress, S. J. P., Meyer, S., Iotti, S., Rossinelli, A., and
Norris, D. J. 2015. Plasmonic films can easily be better: Rules and recipes. ACS
Photonics. 2(3):326–33. https://doi.org/10.1021/ph5004237.
30. West, P. R., Ishii, S., Naik, G. V., Emani, N. K., Shalaev, V. M., and Boltasseva, A. 2010.
Searching for better plasmonic materials. Laser Photon. Rev. 4(6):795–808. https://doi.
org/10.1002/lpor.200900055.
31. Naik, G. V., Shalaev, V. M., and Boltasseva, A. 2013. Alternative plasmonic mate-
rials: Beyond gold and silver. Adv. Mater. 25(24):3264–94. https://doi.org/10.1002/
adma.201205076.
32. Naik, G. V., Kim, J., and Boltasseva, A. 2011. Oxides and nitrides as alternative plas-
monic materials in the optical range [Invited]. Opt. Mater. Express. 1(6):1090–9.
https://doi.org/10.1364/OME.1.001090.
33. Naik, G. V., Schroeder, J. L., Ni, X., Kildishev, A. V., Sands, T. D., and Boltasseva,
A. 2012. Titanium nitride as a plasmonic material for visible and near-infrared wave-
lengths. Opt. Mater. Express. 2(4):478–89. https://doi.org/10.1364/OME.2.000478.
34. Ma, Y., Guo, X., Wu, X., Dai, L., and Tong, L. 2013. Semiconductor nanowire lasers.
Adv. Opt. Photon. 5(3):216–73. https://doi.org/10.1364/AOP.5.000216.
35. Lu, Y.-J., Wang, C.-Y., Kim, J., Chen, H.-Y., Lu, M.-Y., Chen, Y.-C., Chang, W.-H.,
Chen, L.-J., Stockman, M. I., Shih, C.-K., and Gwo, S. 2014. All-color plasmonic nano-
lasers with ultralow thresholds: Autotuning mechanism for single-mode lasing. Nano
Lett. 14(8):4381–8. https://doi.org/10.1021/nl501273u.
36. Klimov, V. I., Mikhailovsky, A. A., Xu, S., Malko, A., Hollingsworth, J. A., Leatherdale,
C. A., Eisler, H.-J., and Bawendi, M. G. 2000. Optical gain and stimulated emission in
nanocrystal quantum dots. Science. 290(5490):314–17.
37. Duarte, F. J., and Hillman, L. W., eds. 1990. Dye laser principles: With applications.
New York: Academic Press. ISBN: 978-0-12-396073-3.
38. Ta, V. D., Wang, Y., and Sun, H. 2019. Microlasers enabled by soft-matter technology.
Adv. Optical Mater. 7(17):1900057. https://doi.org/10.1002/adom.201900057.
39. Liao, Q., Jin, X., and Fu, H. 2019. Tunable halide perovskites for miniaturized solid-
state laser applications. Adv. Optical Mater. 7(17):1900099. https://doi.org/10.1002/
ADOM.201900099.
40. Wei, Q., Li, X., Liang, C., Zhang, Z., Guo, J., Hong, G., Xing, G., and Huang, W. 2019.
Recent progress in metal halide perovskite micro- and nanolasers. Adv. Optical Mater.
7(17):1900080. https://doi.org/10.1002/adom.201900080.
41. Zhang, Q., Su, R., Du, W., Liu, X., Zhao, L., Ha, S. T., and Xiong, Q. 2017. Advances
in small perovskite-based lasers. Small Methods. 1(9):1700163. https://doi.org/10.1002/
smtd.201700163.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp04.indd 126 9/26/23 4:57 PM

 Plasmonic Nanolasers 127

42. Stockman, M. 2008. Spasers explained. Nat. Photon. 2:327–9. https://doi.org/10.1038/

nphoton.2008.85.
43. Purcell, E. M. 1995. Spontaneous emission probabilities at radio frequencies. In
Confined electrons and photons: New physics and applications, ed. E. Burstein and C.
Weisbuch, 839–839. Boston, MA: Springer Science+Business Media LLC. https://doi.
org/10.1007/978-1-4615-1963-8_40.
44. Oxborrow, M. 2007. Traceable 2-d finite-element simulation of the whispering-gallery
modes of axisymmetric electromagnetic resonators. IEEE Trans. Microw. Theory Tech.
55(6):1209–18. https://doi.org/10.1109/TMTT.2007.897850.
45. Ta, V. D., Chen, R., and Sun, H. D. 2011. Wide-range coupling between surface plasmon
polariton and cylindrical dielectric waveguide mode. Opt. Express. 19(14):13598–603.
https://doi.org/10.1364/OE.19.013598.
46. Oulton, R. F., Sorger, V. J., Genov, D. A., Pile, D. F. P., and Zhang, X. 2008. A Hybrid
plasmonic waveguide for subwavelength confinement and long-range propagation. Nat.
Photon. 2(8):496–500. https://doi.org/10.1038/nphoton.2008.131.
47. Bergman, D. J., and Stockman, M. I. 2003. Surface plasmon amplification by stimulated
emission of radiation: Quantum generation of coherent surface plasmons in nanosys-
tems. Phys. Rev. Lett. 90(2):027402. https://doi.org/10.1103/PhysRevLett.90.027402.
48. Stockman, M. I. 2010. The spaser as a nanoscale quantum generator and ultrafast
amplifier. J. Opt. 12(2):024004. https://doi.org/10.1088/2040-8978/12/2/024004.
49. Baranov, D. G., Andrianov, E. S., Vinogradov, A. P., and Lisyansky, A. A. 2013. Exactly
solvable toy model for surface plasmon amplification by stimulated emission of radia-
tion. Opt. Express. 21(9):10779–91. https://doi.org/10.1364/OE.21.010779.
50. Gordon, J. A., and Ziolkowski, R. W. 2007. The design and simulated performance
of a coated nano-particle laser. Opt. Express. 15(5):2622–53. https://doi.org/10.1364/
OE.15.002622.
51. Meng, X., Guler, U., Kildishev, A. V., Fujita, K., Tanaka, K., and Shalaev, V. M. 2013.
Unidirectional spaser in symmetry-broken plasmonic core-shell nanocavity. Sci. Rep.
3(1):1241. https://doi.org/10.1038/srep01241.
52. Bordo, V. G. 2017. Cooperative effects in spherical spasers: ab initio analytical model.
Phys. Rev. B. 95(23):235412. https://doi.org/10.1103/PhysRevB.95.235412.
53. Arnold, N., Piglmayer, K., Kildishev, A. V., and Klar, T. A. 2015. Spasers with retarda-
tion and gain saturation: Electrodynamic description of fields and optical cross-sec-
tions. Opt. Mater. Express. 5(11):2546–77. https://doi.org/10.1364/OME.5.002546.
54. Plum, E., Fedotov, V. A., Kuo, P., Tsai, D. P., and Zheludev, N. I. 2009. Towards the
lasing spaser: Controlling metamaterial optical response with semiconductor quantum
dots. Opt. Express. 17(10):8548–51. https://doi.org/10.1364/OE.17.008548.
55. Li, Z.-Y., and Xia, Y. 2010. Metal nanoparticles with gain toward single-molecule
detection by surface-enhanced Raman scattering. Nano Lett. 10(1):243–9. https://doi.
org/10.1021/nl903409x.
56. Hill, M. T., Marell, M., Leong, E. S. P., Smalbrugge, B., Zhu, Y., Sun, M., van Veldhoven,
P. J., Geluk, E. J., Karouta, F., Oei, Y.-S., Nötzel, R., Ning, C.-Z., and Smit, M. K. 2009.
Lasing in metal-insulator-metal sub-wavelength plasmonic waveguides. Opt. Express.
17(13):11107–12. https://doi.org/10.1364/OE.17.011107.
57. Marell, M. J. H., Smalbrugge, B., Geluk, E. J., van Veldhoven, P. J., Barcones, B.,
Koopmans, B., Nötzel, R., Smit, M. K., and Hill, M. T. 2011. Plasmonic distributed
feedback lasers at telecommunications wavelengths. Opt. Express. 19(16):15109–18.
https://doi.org/10.1364/OE.19.015109.
58. Flynn, R. A., Kim, C. S., Vurgaftman, I., Kim, M., Meyer, J. R., Mäkinen, A. J.,
Bussmann, K., Cheng, L., Choa, F. S., and Long, J. P. 2011. A room-temperature
semiconductor spaser operating near 1.5 μm. Opt. Express 19(9):8954–61. https://doi.
org/10.1364/OE.19.008954.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp04.indd 127 9/26/23 4:57 PM

 128 Photonics Elements for Sensing and Optical Conversions

59. Wu, C.-Y., Kuo, C.-T., Wang, C.-Y., He, C.-L., Lin, M.-H., Ahn, H., and Gwo, S. 2011.
Plasmonic green nanolaser based on a metal–oxide–semiconductor structure. Nano
Lett. 11(10):4256–60. https://doi.org/10.1021/nl2022477.
60. Ho, J., Tatebayashi, J., Sergent, S., Fong, C. F., Iwamoto, S., and Arakawa, Y. 2015.
Low-threshold near-infrared GaAs–AlGaAs core–shell nanowire plasmon laser. ACS
Photonics. 2(1):165–71. https://doi.org/10.1021/ph5003945.
61. Zhang, Q., Li, G., Liu, X., Qian, F., Li, Y., Sum, T. C., Lieber, C. M., and Xiong,
Q. 2014. A Room temperature low-threshold ultraviolet plasmonic nanolaser. Nat.
Commun. 5(1):4953. https://doi.org/10.1038/ncomms5953.
62. Lu, Y.-J., Kim, J., Chen, H.-Y., Wu, C., Dabidian, N., Sanders, C. E., Wang, C.-Y., Lu,
M.-Y., Li, B.-H., Qiu, X., Chang, W.-H., Chen, L.-J., Shvets, G., Shih, C.-K., and Gwo, S.
2012. Plasmonic nanolaser using epitaxially grown silver film. Science. 337(6093):450–
3. https://doi.org/10.1126/science.1223504.
63. Liao, Y.-J., Cheng, C.-W., Wu, B.-H., Wang, C.-Y., Chen, C.-Y., Gwo, S., and Chen, L.-J.
2019. Low threshold room-temperature UV surface plasmon polariton lasers with ZnO
nanowires on single-crystal aluminum films with Al2O3 interlayers. RSC Advances.
9(24):13600–07. https://doi.org/10.1039/C9RA01484E.
64. Wu, X., Xiao, Y., Meng, C., Zhang, X., Yu, S., Wang, Y., Yang, C., Guo, X., Ning, C. Z.,
and Tong, L. 2013. Hybrid photon-plasmon nanowire lasers. Nano Lett. 13(11):5654–9.
https://doi.org/10.1021/nl403325j.
65. Ma, R.-M., Oulton, R. F., Sorger, V. J., Bartal, G., and Zhang, X. 2011. Room-
temperature sub-diffraction-limited plasmon laser by total internal reflection. Nat.
Mater. 10(2):110–3. https://doi.org/10.1038/nmat2919.
66. Kwon, S.-H., Kang, J.-H., Seassal, C., Kim, S.-K., Regreny, P., Lee, Y.-H., Lieber, C.
M., and Park, H.-G. 2010. Subwavelength plasmonic lasing from a semiconductor
nanodisk with silver nanopan cavity. Nano Lett. 10(9):3679–83. https://doi.org/10.1021/
nl1021706.
67. Khajavikhan, M., Simic, A., Katz, M., Lee, J. H., Slutsky, B., Mizrahi, A., Lomakin,
V., and Fainman, Y. 2012. Thresholdless nanoscale coaxial lasers. Nature. 482(7384):
204–7. https://doi.org/10.1038/nature10840.
68. Nguyen, T. V., Pham, N. V., Mai, H. H., Duong, D. C., Le, H. H., Sapienza, R., and Ta,
V.-D. 2019. Protein-based microsphere biolasers fabricated by dehydration. Soft Matter.
15(47):9721–6. https://doi.org/10.1039/c9sm01610d.
69. Ma, R.-M., Yin, X., Oulton, R. F., Sorger, V. J., and Zhang, X. 2012. Multiplexed and
electrically modulated plasmon laser circuit. Nano Lett. 12(10):5396–402. https://doi.
org/10.1021/nl302809a.
70. Stehr, J., Crewett, J., Schindler, F., Sperling, R., von Plessen, G., Lemmer, U., Lupton,
J. M., Klar, T. A., Feldmann, J., Holleitner, A. W., Forster, M., and Scherf, U. 2003.
A low threshold polymer laser based on metallic nanoparticle gratings. Adv. Mater.
15(20):1726–9. https://doi.org/10.1002/adma.200305221.
71. Zhou, W., Dridi, M., Suh, J. Y., Kim, C. H., Co, D. T., Wasielewski, M. R., Schatz, G.
C., and Odom, T. W. 2013. Lasing action in strongly coupled plasmonic nanocavity
arrays. Nat. Nanotech. 8(7):506–11.
72. Yang, A., Li, Z., Knudson, M. P., Hryn, A. J., Wang, W., Aydin, K., and Odom, T. W.
2015. Unidirectional lasing from template-stripped two-dimensional plasmonic crys-
tals. ACS Nano. 9(12):11582–8. https://doi.org/10.1021/acsnano.5b05419.
73. Yang, A., Hoang, T. B., Dridi, M., Deeb, C., Mikkelsen, M. H., Schatz, G. C., and
Odom, T. W. 2015. Real-time tunable lasing from plasmonic nanocavity arrays. Nat.
Commun. 6(1):6939. https://doi.org/10.1038/ncomms7939.
74. Wang, D., Bourgeois, M. R., Lee, W.-K., Li, R., Trivedi, D., Knudson, M. P., Wang, W.,
Schatz, G. C., and Odom, T. W. 2018. Stretchable nanolasing from hybrid quadrupole
plasmons. Nano Lett. 18(7):4549–55. https://doi.org/10.1021/acs.nanolett.8b01774.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp04.indd 128 9/26/23 4:57 PM

 Plasmonic Nanolasers 129

75. Beijnum, F., Veldhoven, P. J., Geluk, E., de Dood, M., Hooft, G., and Exter, M. 2013.
Surface plasmon lasing observed in metal hole arrays. Phys. Rew. Lett. 110(20):206802.
https://doi.org/10.1103/PhysRevLett.110.206802.
76. Wu, C., Khanal, S., Reno, J., and Kumar, S. 2016. Terahertz plasmonic laser radiating in
an ultra-narrow beam. Optica 3(7):734–40. https://doi.org/10.1364/OPTICA.3.000734.
77. Melentiev, P., Kalmykov, A., Gritchenko, A., Afanasiev, A., Balykin, V., Baburin, A.
S., Ryzhova, E., Filippov, I., Rodionov, I. A., Nechepurenko, I. A., Dorofeenko, A. V.,
Ryzhikov, I., Vinogradov, A. P., Zyablovsky, A. A., Andrianov, E. S., and Lisyansky, A.
A. 2017. Plasmonic nanolaser for intracavity spectroscopy and sensorics. Appl. Phys.
Lett. 111(21):213104. https://doi.org/10.1063/1.5003655.
78. Dorofeenko, A. V., Zyablovsky, A. A., Vinogradov, A. P., Andrianov, E. S., Pukhov, A.
A., and Lisyansky, A. A. 2013. Steady state superradiance of a 2D-spaser array. Opt.
Express. 21(12):14539–47. https://doi.org/10.1364/OE.21.014539m.
79. Meng, X., Liu, J., Kildishev, A. V., and Shalaev, V. M. 2014. Highly directional spaser
array for the red wavelength region. Laser Photon. Rev. 8(6):896–903. https://doi.
org/10.1002/lpor.201400056.
80. Galanzha, E. I., Weingold, R., Nedosekin, D. A., Sarimollaoglu, M., Nolan, J.,
Harrington, W., Kuchyanov, A. S., Parkhomenko, R. G., Watanabe, F., Nima, Z., Biris,
A. S., Plekhanov, A. I., Stockman, M. I., and Zharov, V. P. 2017. Spaser as a biological
probe. Nat. Commun. 8(1):15528. https://doi.org/10.1038/ncomms15528.
81. Chou, Y.-H., Hong, K.-B., Chang, C.-T., Chang, T.-C., Huang, Z.-T., Cheng, P.-J., Yang,
J.-H., Lin, M.-H., Lin, T.-R., Chen, K.-P., Gwo, S., and Lu, T.-C. 2018. Ultracompact
pseudowedge plasmonic lasers and laser arrays. Nano Lett. 18(2):747–53. https://doi.
org/10.1021/acs.nanolett.7b03956.
82. Nezhad, M. P., Simic, A., Bondarenko, O., Slutsky, B., Mizrahi, A., Feng, L., Lomakin,
V., and Fainman, Y. 2010. Room-temperature subwavelength metallo-dielectric lasers.
Nat. Photon. 4(6):395–9. https://doi.org/10.1038/nphoton.2010.88.
83. Miller, S. E. 1969. Integrated optics: An introduction. The Bell System Technical
Journal 48(7):2059–69. https://doi.org/10.1002/j.1538-7305.1969.tb01165.x.
84. Kim, M.-K., Lakhani, A. M., and Wu, M. C. 2011. Efficient waveguide-coupling of
metal-clad nanolaser cavities. Opt. Express. 19(23):23504–12. https://doi.org/10.1364/
OE.19.023504.
85. Kim, M.-K., Li, Z., Huang, K., Going, R., Wu, M. C., and Choo, H. 2013. Engineering
of metal-clad optical nanocavity to optimize coupling with integrated waveguides. Opt.
Express. 21(22):25796–804. https://doi.org/10.1364/OE.21.025796.
86. Prabowo, B. A., Purwidyantri, A., and Liu, K.-C. 2018. Surface plasmon resonance
optical sensor: A review on light source technology. Biosensors 8(3):80. https://doi.
org/10.3390/bios8030080.
87. Wang, X.-Y., Wang, Y.-L., Wang, S., Li, B., Zhang, X.-W., Dai, L., and Ma, R.-M. 2017.
Lasing enhanced surface plasmon resonance sensing. Nanophotonics. 6(2):6. https://
doi.org/10.1515/nanoph-2016-0006.
88. Ma, R.-M., Ota, S., Li, Y., Yang, S., and Zhang, X. 2014. Explosives detection in a
lasing plasmon nanocavity. Nat. Nanotech. 9(8):600–4. https://doi.org/10.1038/
nnano.2014.135.
89. Wang, S., Li, B., Wang, X.-Y., Chen, H.-Z., Wang, Y.-L., Zhang, X.-W., Dai, L., and
Ma, R.-M. 2017. High-yield plasmonic nanolasers with superior stability for sens-
ing in aqueous solution. ACS Photonics 4(6):1355–60. https://doi.org/10.1021/
acsphotonics.7b00438.
90. Cho, S., Humar, M., Martino, N., and Yun, S. H. 2016. Laser particle stimulated
emission microscopy. Phys. Rev. Lett. 117(19):193902. https://doi.org/10.1103/
PhysRevLett.117.193902.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp04.indd 129 9/26/23 4:57 PM

 5 Metasurfaces and
Several Well-Studied
Applications
Muhammad Ali Butt1,2, Nikolay Lvovich Kazanskiy1,3,
and Svetlana Nikolaevna Khonina1,3
1Samara National Research University, Samara, Russia
2 Warsaw University of Technology, Institute of
Microelectronics and Optoelectronics, Warszawa, Poland
3Image Processing Systems Institute – Branch of the

Federal Scientific Research Centre “Crystallography and

Photonics” of Russian Academy of Sciences, Samara, Russia

5.1 INTRODUCTION
Metasurfaces (hereafter represented as MSs) are a major research subject and are
used in different utilizations because of their exceptional capability to manipu-
late electromagnetic (EM) waves at microwave (MWv) and optical frequencies.
Since the theoretical calculation of their fascinating capabilities such as super-
radiance and perfect absorbance or invisibility cloak configuration, they have
received a lot of attention. Due to negative diffraction, these can be obtained
experimentally with the simultaneous induction of strong electric (E) and mag-
netic (H) dipolar moments. Nevertheless, this condition supposes significant
technological limitations connected to the accessibility of materials operating at
frequencies of technological interest.
Such artificial sheet materials, usually consisting of metallic patches or dielec-
tric etchings in planar or multilayer structures of subwavelength thickness, have
the benefits of a light weight, ease of processing, and the ability to control the wave
propagation on the substratum and in the neighboring free space. This chapter
introduces several important applications based on MSs. MSs are subwavelength
thickness two-dimensional (2D) or planar versions of metamaterials (hereafter
represented as MMs). Due to their light weight and straightforward manufactur-
ing, they are widely investigated and implemented in EM implementations. MSs
have the distinctive capacity to block, absorb, concentrate, disperse, or guide
waves from the MWv through visible (VIS) frequencies, both on the surface at
grazing incidence and in space at normal and oblique incidence. The importance
of this field can be understood by the number of publications in the last 12 years
(2008–2019). The data is collected from Scopus indexed database as shown in
Figure 5.1.

130 DOI: 10.1201/9781003439165-5

BK-TandF-KAZANSKIY_9781032572949-230642-Chp05.indd 130 9/26/23 4:58 PM

 Metasurfaces and Several Well-Studied Applications 131

FIGURE 5.1 Number of publications containing the word “metamaterials” over the years
2008–2019. The data is taken from the Scopus database.

5.2 WHAT ARE METASURFACES?

The prefix “meta” suggests that the material’s features go beyond what is apparent
in nature. MMs are structural materials that are deliberately constructed and receive
their characteristics from the internal microstructure instead of the chemical confor-
mation present in normal materials. The core model of MMs is to create materials by
means of artificially designed and engineered structural units to realize the preferred
characteristics and features. Such structural units can be customized in shapes and
size, the lattice constant and interatomic interface can be synthetically adjusted, and
defects can be designed and positioned at the anticipated positions. By integrating
these nanoscales (n-scales) unit cells into the preferred construction or geometry, the
refractive index (hereafter represented as RI) of the MM can be adjusted to positive,
near-zero, or negative values.
As a result, MMs may be endowed with characteristics and functions that those
normal materials are not. For example, for invisibility cloak technology to conceal
an item or, on the other hand, to avoid refraction by a perfect lens and permit direct
viewing of an individual protein in a light microscope, the material must be able to
regulate the direction of light similarly. MMs provide this potential. While MMs
have already revolutionized optics, their execution has been hampered by their
inability to operate over large light bandwidths (BWs). Modeling an MM that func-
tions throughout the whole VIS spectrum remains a big issue. Such a material must
be able to interact with both E- and H-light fields.
At VIS and infrared (IR) wavelengths, some natural materials are immune to
the H-field of light. Previous attempts at MM have resulted in the creation of artifi-
cial atoms comprised of two components: one that interacts with the E-field and the
other with the H-field. The advantage of this method is that the separate components

BK-TandF-KAZANSKIY_9781032572949-230642-Chp05.indd 131 9/26/23 4:58 PM

 132 Photonics Elements for Sensing and Optical Conversions

interact with different colors of light, making it difficult to overlap them throughout
a large range of wavelengths.
Three-dimensional (3D) MMs, which are primarily periodic artificial materials
consisting of metal and dielectric, have received a lot of attention in the last decade
because of their extraordinary interaction with EM waves, which outperform natural
or homogeneous materials. Their exceptional wave management ability stems from
their high interaction with E- and/or H-fields, which are generally given via reso-
nant effects regulated by unit cell geometry. Such properties enable a wide range
of implementations, including antenna efficacy increase, perfect absorbers (PAs),
superlenses, cloaking and energy harvesting, among others. On the other hand,
significant losses, and problems in 3D manufacturing, particularly at micro- and
n-scales, exclude many MM uses.

5.3 NEGATIVE REFRACTION

The negative RI of light and other radiation is one of the most demanding properties
of MMs. Negative refraction is founded on the equations established by the Scottish
physicist James Maxwell in 1861. All identified natural materials have a positive
RI so that light flowing from one medium to another becomes slightly bent in the
transmission direction. For example, air has the lowest RI in nature at standard con-
ditions, which is just above 1. Water’s RI is 1.33. That’s about 2.4 of the diamond.
The higher the RI of the material, the more it distorts the original direction of light.
Negative refraction occurs in some MMs, though, such that light and other radiations
bends backwards as it approaches the system, as shown in Figure 5.2.
As early as the middle of the 20th century, the existence of substances with a
negative RI was anticipated. In 1976, Soviet physicist V.G. Veselago published a
scientific paper that described their properties, theoretically, as comprising unusual
light refraction. In 1999, Roger Walser suggested the term MMs for such substances.
But researchers were only able to determine how materials of any sort could be
created to achieve negative refraction in the early 2000s. The first MM specimens
were made from thin wire formations and operated only with MWv radiation. In
electronics, lithography (litho), biomedicine, isolation coatings, heat transfer, space

FIGURE 5.2 Schematic of negative refraction of light. In negative RI materials, the light
gets bent the opposite way.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp05.indd 132 9/26/23 4:58 PM

 Metasurfaces and Several Well-Studied Applications 133

implementations, and perhaps even novel methods to optical computation and energy
harvesting are possible with such negative refraction materials.

5.4 METASURFACES
Traditional optical components use refractive, diffractive, and reflective principles to
direct light. These devices work by using transmission inside the component’s mate-
rial to change the characteristics of an incident wavefront (e.g. phase, amplitude, and
polarization). On the other hand, others work by creating interference patterns that
result in Huygens–Fresnel sources that create far-field distributions. The propagation
characteristics of a component are determined by differences in component shape
and material properties. The volumetric management of the RI on a subwavelength
scale allows for extraordinary control of light transmission in MMs, allowing for
unique functionalities like negative refraction.
Metasurfaces (MSs, MTSs) are thin artificial layers comprising regular place-
ments of tiny inclusions in a dielectric host medium. They can be customized to
attain unusual reflection/transmission of space waves and/or to change the dispersion
properties of surface/guided waves. The scientific community has given significant
attention to MTS technology and has been employed in several implementations.
MTSs are usually formed in MW antenna implementations by a standard texture
of small units that are printed on a grounded slab with or without shorting vias.
Additionally, MTSs can be made up of a concentrated formation of metallic pins on
a ground plane at sub-millimeter wavelengths.
Surface waves (SWs) can be controlled by the design and management of the
phase or groups velocity of MS unit cells with the correct impedance. These are
designed to guide or separate waves in certain directions and are employed in disper-
sal control utilizations. Different effective surface refractive indices can be obtained
by controlling the sizes and shapes of the MS unit cells and designing the surface to
deliver numerous purposes. These can be utilized for the development of 2D MWvs/
optical lenses such as Luneburg and fish-eye lenses, which are employed in surface
WGs for antenna systems and planar MWv sources.
When the SW interacts with the modulated, locally periodic boundary conditions
(BCs) imposed by the MTS, it helps to replicate the desired aperture field that radiates
through a leaky-wave (LW) effect. The conventional MTS configuration has a low pro-
file, is light in weight and is simple to put together using traditional circuit board tech-
niques. Furthermore, the feeding unit is incorporated in the MTS plane, allowing for the
avoidance of external protrusion and (sub-) reflectors. Because of these attractive quali-
ties, as well as the possibility of precise control and shape matching of the aperture field,
various novel antenna configurations have recently been presented in the literature.
Challenging configurations were also conceivable from the perspective of beam
scanning through the evolution of MTSs modeling. Huygens MTSs were suggested to
acquire extra degrees of freedom in radiation field control. The prospect of receiving
multi-beam radiation from a single aperture has also been shown by the coupling of
the MTS and Luneburg lenses. MTSs use units whose geometries gradually change
from cell to cell. At the operating frequency, the components that enforce the imped-
ance BCs are small in terms of wavelength (usually between λ/5 and λ/10, with λ

BK-TandF-KAZANSKIY_9781032572949-230642-Chp05.indd 133 9/26/23 4:58 PM

 134 Photonics Elements for Sensing and Optical Conversions

being the free-space wavelength). Therefore, the interacting SW sees the impedance
BC interface as a continuum. The constitutive units of MTS act like pixels in a
printed black and white image, the grayscale of which is implemented by modifying
the proportions of the units within a regular lattice. In a regular lattice, the pixels are
arranged with elementary cell size in the range λ/10–λ/5. The lattice is typical of the
Cartesian type, but hexagonal lattices could also be manipulated.

5.4.1 MS Design Process

While EM, or effective medium theory, is superior at describing the scattering prop-
erties of individual subwavelength units, Fourier optics can explain the combined
activity of all the units and their spatial transfer function. Both phase and amplitude
modulation can be utilized in MS configuration.
Let us consider a plane wave I = Ae –j(kxx+kyy+kzz) is incident upon a phase mask in the
z = 0 plane described by a transfer function H = ejФ(x,y), as indicated in Figure 5.3. This
system provides an MS model with unit magnitude and a phase distribution Ф(x, y).
If the phase distribution about the point r0 = (x0, y0) is certified as a Taylor expansion
then to first order it can be expressed as follows:

∂Φ ∂Φ
Φ ( x , y ) ≈ Φ ( x o , yo ) + ( x − x o ) + ( y − yo ) . (5.1)
∂ x r = ro ∂ y r = ro

If the transfer function H is then applied to the plane wave I, the coefficients of
the linear terms in the expansion are added along with the x and y wave vector com-
ponents of the incident wave to produce the result where all the constant phase terms
are merged into θ:

 ∂Φ   ∂Φ  
− j  + kx  x + + k y  y + k zz +θ 
 ∂ x r = ro    ∂ y 
 
T = IH ≈ Ae  r = ro 
. (5.2)

FIGURE 5.3 Schematic representation portraying the functioning of a MS. An incident

plane wave I is acted upon by the spatial transfer function of the MS to produce a transmitted
wave T.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp05.indd 134 9/26/23 4:58 PM

 Metasurfaces and Several Well-Studied Applications 135

The gradient of the phase distribution has the effect of altering the wave vector
of the incident wave, as indicated in (5.2). This change in wave vector caused by a
mask’s spatial phase function gives a fundamental understanding of how MSs may
alter wavefront unpredictably. To perform an optical transformation, it is necessary
to first determine how the incident wave vector must change, and then utilize this
knowledge to calculate the appropriate spatial phase function.
The first step in the MS configuration process is to establish the necessary spatial
phase function. This step provides a continuous function of position with the ideal
transfer characteristics, yet such a phase contour cannot be manufactured in terms of
implementation because the infinitesimal resolution would be necessary. MSs phase
contours are divided into a grid of finite unit cells representing a spatially discrete
function, comparable to signals in discrete-time systems with specific sampling peri-
ods. The MS must be designed to meet the criterion of the Nyquist–Shannon sam-
pling theorem for the spatial sampling period. A suitable scattering unit is positioned
in each unit cell to accomplish the required phase shift at that point. To physically
implement the discrete phase function on the MS, it is necessary to achieve arbitrary
phase functions by scattering units that can induce phase shifts from 0 to 2π. Due
to configuration constraints, appropriate selection of scattering units is often a chal-
lenge. The primary restriction is the requirement to be subwavelength for the period
of the scattering units. This ensures that only the zero-order diffraction mode is
transmitted by the MS. A second constraint is that the patterned units are of the same
height across the surface, allowing the MS to be manufactured with a single litho
stage. Due to the uniform height of the scattering structures, researchers are left with
limited options to alter the phase. By changing the geometry or orientation of the
scattering formations, it is usually possible to achieve phase shifts ranging from 0 to
2π with uniform magnitude. For instance, most MSs use Pancharatnam–Berry (PB)
phase shift, which is proportional to the orientation angle of the unit within its cell.
Other phase tuning methods can also be utilized, including modifying the widths,
diameters, or periodicity of the scattering units.

5.4.2 State of the Art Metasurfaces and Their Limitations

In the last few years, MSs have experienced several major developments. Many of
the initial MS-based optics have been using surface plasmon based-scattering units
that permit the MS to be made of metal. These systems are limited to working in
reflection mode, with metals being so lossy in the optical regime. MSs based on sili-
con helped to address this setback, allowing MSs that can work in the near-IR mode
with high efficacy in both reflection and transmission. More recent developments
have revealed that low RI materials have been employed to operate efficiently in the
VIS system [1, 2].
Various generic optical components were replicated with the assistance of MS
implementations and were fruitfully tested. Many of these replicated generic optical
components in their MS-based implementations have been considerably miniatur-
ized. Further advanced features were also accomplished. Polarization-multiplexed
formations were introduced allowing various optical makeovers to be performed
depending on the polarization state of the input light; on the other hand, these

BK-TandF-KAZANSKIY_9781032572949-230642-Chp05.indd 135 9/26/23 4:58 PM

 136 Photonics Elements for Sensing and Optical Conversions

systems were principally limited to operating at the IR region. MSs also enables
novel functionalities as well as imitating functionalities of traditional refractive and
diffractive optics.
Although MSs offer an impressive range of capabilities, these devices remain
subject to substantial limits that remain open to research. While some researchers
suggested some devices working in the VIS region, most of them are limited to
functioning in the IR region. Since the materials utilized are in general too lossy,
with high band gaps or opaque at VIS wavelengths. The scope of the implemen-
tations for which MSs can work is significantly reduced without the capability to
operate at VIS wavelengths. Such formations often face several aberrations, perhaps
most notably chromatic aberrations. Like any system that works based on diffractive
principles, as the wavelength varies, the functioning of MS-based devices deviates
greatly from their intended functioning. Attempts to accomplish achromatic MSs
have been made, but these methods have only validated multi-wavelength function-
ing instead of delivering a truly broadband achromatic apparatus.
Besides the aberrations and material constraints of MSs that limit the VIS
regime functioning, one of the major problems in MS research is the development
of a method for achieving phase units that can be dynamically changed to attain
reconfigurable phase contours. Attaining this capability would permit the optical
equivalent of a field-programmable gate array (FPGA) to be implemented and repro-
grammed as preferred. Many approaches to achieving reconfigurable MSs consist of
local tuning of the RI of the phase-shifting units. Regrettably, it is difficult to adjust
the RI to achieve arbitrary changes from 0 to 2π phase shift. The range over the RI
can be modulated is limited, and MS n-scale size constraints restrict the transmis-
sion distance and change in phase accumulation that could result from a shift in RI.
Some methods are available for a substantial change of the RI, for example, germa-
nium-antimony tellurium (Ge2Sb2Te5, GST). However, depending on the operating
wavelength, these materials can show a substantial loss. A stable and reconfigurable
MS system would need phase units that can easily be modified without significant
loss in performance.

5.5 APPLICATIONS OF METASURFACES

Although MM-based optical devices allow distinctive functionality, they are
restricted by their reliance on light having to transmit through a bulk medium. Signal
loss in the bulk medium is a significant restriction on execution, and the resolution
of 3D manufacturing techniques limits the range of viable devices from a practi-
cal point of view. Because of these limitations, the 2D analogue of bulk MMs has
attracted considerable interest in recent years. Instead of modulating the RI over a
volume, the MSs are ultra-thin formations patterned on a subwavelength scale with
quasi-periodic arrangements of scattering components. MSs implement spatial trans-
fer functions that transform the input wavefronts over a wavelength-scale range by
correctly designing and patterning the scattering components. The designed transfer
functions spatially modulate the phase and or amplitude. This practice has facilitated
MSs antennas, lenses, PAs, sensors, WG, cloaking, modulators, frequency selective
surfaces, polarizers, etc. This section presents several implementations based on

BK-TandF-KAZANSKIY_9781032572949-230642-Chp05.indd 136 9/26/23 4:58 PM

 Metasurfaces and Several Well-Studied Applications 137

MSs discussed one by one along with examples taken from the recent publications
in this field.

5.5.1 Metasurface Antennas

MSs, apart from their wide implementations in surface and free-space wave manage-
ment, have exclusive ability and distinctive advantages as media for radiating EM
waves into free space. For many implementations, antenna systems require a large gain
to meet the requirements of communications links. Because of their large aperture size
and narrow beam width, large planar formations are implemented in several antenna
systems. Nevertheless, feeding systems are typically complex to diminish loss and
deliver phase control for each antenna component to determine the radiation pattern.
MSs can be employed in antenna implementations to boost antenna gain, improve
BW and execution, minimize backward radiation, transform polarization, and steer
antenna beams. They can be utilized to create a Fabry–Perot cavity to stimulate the
LW mode of the formation for antenna gain enhancement. Moreover, MSs can also be
tailored to act as a lens to produce an exceedingly focused beam and acts as a spatial
filter to transform and transfer wave vectors parallel to the transmission direction. BW
enhancement of antennas by MSs is typically credited to the increased impedance
matching or parasitic resonance generated by the MS superstrate, and the lessening
of the back radiation is generally due to the stopband of the EM band-gap structure
formed by the MS. LW antennas allow for the removal of feeding networks for certain
implementations, by using a traveling wave feed technique. When an LW antenna is
fed by a point source, the field of the surface can be characterized as follows:

ψ ( ρ, z ) = ψ o exp ( − jγρ − jk z z ) , (5.3)

where γ = α + jβ represents the propagation constant on the surface of the LW

antenna and k z = ko2 − β 2 is the propagation constant normal to the antenna sur-
face. If β < ko, kz is strictly real in the fast wave regime, it means that the energy is
coupled as radiation into free-space. It is possible to configure the path of radiation
by manipulating the value β. The unit cells of the MS are specifically calculated to
match the phase delay on the surface with that of the chosen direction of the beam
in the free-space. These also have frequency-dependent beam scanning properties
(beam squint), which, depending on the implementation, can be an advantage or a
shortcoming.
In MWv antenna implementations, MSs are typically designed by a regular
texture of small units printed on a grounded slab with or without shorting vias.
Conversely, MSs can be made up of a dense formation of metallic pins on a ground
plane at submillimeter wavelengths. When the SW interacts with an MS imposed
modulated locally periodic boundary conditions (BCs), it permits one to replicate
the anticipated aperture field that radiates through an LW effect. The traditional MS
configuration has a low profile, weight, and complexity. Furthermore, it can be fab-
ricated utilizing standard printed circuit board techniques. In addition, the feeding
unit is fixed in the MS plane, so that an external protruding or backfeeding arrange-
ment or (sub-) reflector can be avoided.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp05.indd 137 9/26/23 4:58 PM

 138 Photonics Elements for Sensing and Optical Conversions

These appealing characteristics, together with the potential of fine control and
shape adjustment of the aperture field, result in a few interesting antenna devices.
It was suggested that Huygens MSs gain extra degrees of freedom in the control
of the radiating field. In contrast, the probability of receiving multi-beam radiation
from a single aperture was shown by combining MSs and Luneburg lenses. The MSs
use units whose geometries progressively transform from cell to cell. At the oper-
ating frequency, the units implementing the impedance BCs are small in terms of
the wavelength (typically, between λ/5 and λ/10, with λ being the free-space wave-
length). As a result, the interacting SW considers the interface impedance BC to
be a continuum. The MS constituent units function similarly to pixels in a black
and white printed picture, with grayscale achieved by varying the dimension of the
components inside a regular lattice. The pixels are organized in a consistent lattice
with primary cell sizes ranging from λ/10–λ/5. Characteristically, the lattice is of
Cartesian type, but hexagonal lattices could also be utilized.
High gain holographic antennas are successfully designed and produced with MSs
[3, 4]. The SW is the primary incident wave for holographic antennas. Radiation occurs
for such antennas when a phase matches an LW to a reverse one. Forward LW emerges
from the grid, which has a greater periodicity, while the shorter period of grating cor-
responds to a backward leaky wave. The source with elliptical grating is surrounded
by both periodicities. LW modes for such a grating are matched, leading to directive
radiation [5]. Circularly polarized leaky wave antennas with a gain of 26 dB are sug-
gested [3] based on this strategy. The grating effect was achieved through the modu-
lation of the surface impedance of the MS. The major benefit of this method is that
the MS modulation, i.e. surface impedance, is optimized instead of modifying the
antenna shape to model a particular response. An MS antenna was presented [6] that
offers the opportunity for both right and left circular polarization. Both TE and TM
SWs were released and for their independent control, the polarization decoupling of
the two modes was employed. Such modes are phase-matched, and the modulated MS
was rotationally symmetric. A circular WG fed the antenna, and a corrugated hat was
located on top of the WG to block the space-wave radiation. The phase-matching is
important for LW MS antennas, and in case of mismatching, the gain of the structure
decreases. Moreover, the BW of such antennas is restricted by the dispersion of the
MS. For these antennas, a BW gain expression can be found in [7].
Marco Faenzi et al. [8] suggested novel MS (MTS) antenna models, implemen-
tation, and experiments made up of subwavelength components reproduced on a
grounded dielectric slab. To generate an essentially arbitrary aperture field, these
antennas employ the interaction of a cylindrical SW wavefront and an anisotropic
impedance BC. They are very thin and may be stimulated by a simple in-plane mono-
pole. Such antennas may be substantially modified in terms of beam shape, BW, and
polarization by constructing the printed components. For the first time, testing results
demonstrate that these antennas can have an aperture efficacy of up to 70 percent, a
BW of up to 30 percent, create two distinct high-gain direction beams and identical
beams at two distinct frequencies, and display outcomes not ever attained before.
MS antenna model for beam steering implementations is described here. The MS
superstrate is made up of identical square rings (SRs) that are evenly spaced. The
anticipation of SRs in series with varying phase delays is the basic mechanism of

BK-TandF-KAZANSKIY_9781032572949-230642-Chp05.indd 138 9/26/23 4:58 PM

 Metasurfaces and Several Well-Studied Applications 139

FIGURE 5.4 The manufactured single-slit MS antenna: (a) top view of the MS, (b) bottom
view of the slit radiator for P = 30, (c) standpoint view when the slit radiator and MS are
gathered, (d) photo of the MS setup [9].

radiation. The entire emitted beam may be directed in distinct directions based on
the primary excitation of SR components. To attain beam steering, either a single-
slit radiator with mechanical drive or a double-slit model with fixed slit locations but
variable feeding phase differences between two slits can be utilized as the excitation
source. A beam-steering angle of −35 to 35 degrees can be reached with a single-slit
arrangement, and −30 to 30 degrees can be achieved with a double-slit configuration.
The constructed single-slit MS antenna is presented in Figure 5.4.

5.5.2 Metasurface Lenses

MS has fascinated major research interests in the field of MMs as originally presented
to generalize Snell’s law, which is, in fact, a distinct case of the grating equation. It
allows management of the transmission of EM waves with a geometric gradient phase
contour. The 2D artificial structure consisting of subwavelength scatter formations
may create anomalous phenomena of reflection and refraction, which is technically
more attractive in comparison with the 3D phenomena. The MS lens is mostly studied
with the hope that the outdated bulky lens will be replaced by a piece of structured
surface layer, which may optimistically lead to noteworthy advances in modern opti-
cal systems. The MS lens can accommodate both transmission magnitude and phase
as well as polarization in subwavelength scale resolutions compared to the present
thin-film based Fresnel’s lens which is mostly to produce a specified phase contour.
The increased freedoms would expand the lens functions to include, for example, the
removal of unwanted diffractions or reflections and further enhance focusing/imaging

BK-TandF-KAZANSKIY_9781032572949-230642-Chp05.indd 139 9/26/23 4:58 PM

 140 Photonics Elements for Sensing and Optical Conversions

capabilities. A technological benefit is that the manufacturing process can be simpli-

fied by single-step litho. Dielectric materials with a high RI have been developed to
reduce losses to improve the efficacy of the MS lenses. By putting all these develop-
ments into viewpoint, a bright future could be envisaged for MS lenses. However,
some practical problems still need to be managed for implementations, which mostly
involve device strength and BW. A MS should have a phase contour given by Eq. (5.4)
to focus the EM energy at a distance d:

ΦL ( x, y) =
2π
λ ( )
x 2 + y2 + d 2 − d . (5.4)

This contour also transforms the wavefront’s shape from planar to spherical,
which is required for focusing. A high numerical aperture (NA) efficacy may be
attained if the EM wave reaches the surface at a normal angle. When the angle of
incidence (AOI) is not perpendicular, however, a phenomenon known as “coma”
develops, which can significantly reduce the NA’s effectiveness. It is possible to
reduce the effect of “coma” by employing the surface on a curved dielectric unit
[10]. It has been shown that the MS composed of V-shaped antennas focuses power at
telecommunication frequencies [11]. Another reflected array-based MS lens is intro-
duced [12].
Previous solutions to the development of miniature optical components included
gradient RI materials, such as graded-index (GRIN) lenses, traditional Fresnel-type
lenses, and binary gratings. These are presumably still bulky due to the wavelength of
the light for which they are constructed and depend on classical optical phenomena,
i.e. refractive optics, involving the bulk RI to shape the light’s wavefront. However,
MSs are real miniature optical units, where the thickness is typically less than one
wavelength of light. These MSs are typically an area made up of periodic scatter-
ers (either metallic or dielectric) that are themselves subwavelength in the in-plane
direction. Nonetheless, to avoid diffraction effects, MSs must have their periodicity
less than the wavelength of interest.
The development of lensing MSs, namely metalenses or ultra-thin lenses, has
been a major focus in recent years. The early execution of these lenses takes advan-
tage of plasmonic resonance along with the well-known Pancharatnam–Berry phase
(Geometric phase) [13]. However, a problem with these lenses is the very low efficacy
of less than 15% – generally inherent in plasmonic devices – along with the small
lens areas in the VIS to near-IR regimes. Also, the plasmonic structures that often
rely on the use of gold (Au), prevent these lenses from operating below the wave-
length of 550 nm and thus could not be employed to operate across the whole VIS
spectrum. The problem of wavelength can be addressed by selecting other plasmonic
metals such as silver (Ag) or aluminum (Al), but the problem of low efficacy remains.
Here are some basics of MS-based lenses.

5.5.2.1 Control of the Phase Profile

Refractive lenses are commonly utilized in a variety of optical systems, for instance
telescopes and microscopes. Despite having upright characteristics in phase control
and polarization, traditional high-numerical-aperture refractive lenses are frequently

BK-TandF-KAZANSKIY_9781032572949-230642-Chp05.indd 140 9/26/23 4:58 PM

 Metasurfaces and Several Well-Studied Applications 141

bulky and pricy. Moreover, the complicated process of macro or mesoscale produc-
tion also depends on conventional methods of optics manufacturing established over
100 years ago. Refractive lenses are typically intended with distinct shapes to satisfy
optical specifications. A metalens, however, offers novel prospects to overcome these
constraints. For example, by varying the meta building blocks (MBBs), the phase
contour can be changed. The hyperbolic phase contour mandatory to focus a normal
incident beam inside the substratum that remains collimated can be articulated as:

ϕ (r ) = −
2π
λ ( )
r2 + f 2 − f , (5.5)

where f is the focal length of the incident light and r is the radial coordinate. The MS
is designed to create a phase contour to transform the incident planar wavefront into
a spherical shape at focal length f from the lenses.

5.5.2.2 Plasmonic Metasurface-Based Lenses

Typically, lenses established on MSs employ MBBs to adjust their optical features.
One of the most common methods is to generate plasmonic effects on the surface.
By means of advanced electron beam litho (EBL), and a straightforward lift-off pro-
cess, a plasmonic antenna can be easily micro-machined. It is possible to transform
the concentrated incident light into a smaller region that matches its wavelength and
produces oscillations. The metalens has drawn considerable interest in the field of
optics by getting the plasmonic effect on their MS. For instance, the experimental
results stated in [14] showed that the micro-structures have a plasmonic effect on
the surface of Ag thin film and effectively formed a focal point at the focusing
plane. Alternative research establishes that the focusing could be accomplished and
tuned assuming a variety of n-antenna shapes for example elliptical and circular
blocks [15].

5.5.2.3 Lenses Rely on All-Dielectric Metasurfaces

The energy absorption loss of the incident light could possibly be reduced meaning-
fully by utilizing dielectric phase shifters MBBs. Researchers have benefited from
this and involved dielectric phase shifters in several novel optical configurations.
A polarization-insensitive lens with dielectric building blocks in the form of cir-
cular silicon formations has been suggested, as shown in Figure 5.5(a–f). With the
incident light at a wavelength of 850 nm, a high transmission efficacy of 70% was
achieved [16].
It has been shown that a single dielectric n-antenna could be built as an effective
building block that could offer maximum phase coverage. The spatial image reso-
lution, based on engineered n-structures, has excellent quality and relatively high
transmission efficacy [17]. Moreover, dielectric metalens has greater execution in
VIS range spectral implementations. μ-fabricated titanium dioxide (TiO2) n-pillars
have been employed to develop polarization-insensitive metalens. For incident wave-
lengths of 532 nm and 660 nm, the metalens could accomplish a comparatively high
NA of 0.85 with an efficacy of more than 60% [18].

BK-TandF-KAZANSKIY_9781032572949-230642-Chp05.indd 141 9/26/23 4:58 PM

 142 Photonics Elements for Sensing and Optical Conversions

FIGURE 5.5 High-contrast gratings (HCG) lenses. (a) l = 0 focusing lens. (b) Vortex HCG
lens. (c) and (d) SEM images show an expanded view of the fabricated lens. (e) Measured
image intensity contour from the l = 0 focusing lens. (f) Measured image intensity contour
from the l = 1 vortex lens [16].

5.5.3 Metasurface Perfect Absorbers

MMs have made it possible to demonstrate several exotic EM phenomena and have
stimulated some interesting potential utilizations. Although bulk MMs present
severe manufacturing challenges, mainly in the optical regime, MSs offer alterna-
tive ways to achieve desirable functionalities, including wavefront management, the
transformation of polarization, and absorption/emission engineering. In various
implementations, for example sensing [19], compressive imaging [20], and thermal
management [21], MS PAs with a thickness much less than the operating wavelength
are desirable.
Also of great interest are broadband PAs covering the whole solar spectrum in the
harvesting of solar energy. Using dense n-rods and n-tube films, multilayer planar
photonic structures, and photonic crystals, several examples of material structures
as high-execution solar PAs have been created. The complex multi-resonator unit

BK-TandF-KAZANSKIY_9781032572949-230642-Chp05.indd 142 9/26/23 4:58 PM

 Metasurfaces and Several Well-Studied Applications 143

cell MS has evolved into a powerful and versatile platform for achieving flawless
multiband and broadband absorption, particularly in the MWvs, terahertz, and IR
regimes.
Here, we are presenting an example of MS broadband solar PA demonstrated by
[22] based on a metallic MS architecture, which achieves an absorptance of more
than 90% in the VIS and near-IR range of the solar spectrum and shows low absorp-
tivity at mid and far-IR wavelengths. The complex unit cell of the MS solar PA is
made up of eight pairs of Au n-resonators isolated by a thin silicon dioxide spacer
from a Au ground plane. The absorbed power by the 2D sheet (resonator formation)
is given by:

1
∫
sheet
Pabs = Re K .E * d A,
2 (5.6)
A

where K = σE is the surface current density, E is the electric field at the sheet, and A
is the area. The absorptance of the sheet is then given by:

 Pabs
sheet

 A 
Re ( σ ) 1 + r
2
Asheet = = , (5.7)
Io cε o

where Io is the incident intensity, r is the reflection coefficient of the multilayered

system, c is the speed of light in a vacuum, and εo is the vacuum permittivity. The
power degenerated within the ground plane is expressed as:

1 εω
gp
Pabs =
2 ∫
Re J . E * d A = o Im nAu
2
2
( )∫ E 2
dV ,
(5.8)
A Vgp

where nAu is the complex RI of Au, and the integration extends across the entire vol-
ume of the ground plane. However, since the ground plane acts as an almost perfect
mirror, we can approximate the field by its value at the interface spacer-ground plane
and restrict the integral to the volume Aδ/2, where δ/2 = c/[2ω Im(nAu)] is the energy
penetration depth. The absorptance of the ground plane is then expressed as:

 Pabs
gp


Agp =
 A 
=
2
Im nAu( )
t
2

, (5.9)
Io 2 Im ( nAu )

where t is the field transmission into the ground plane.

Figure 5.6(a) depicts the schematic diagram of the MS PA, which is based on the
metal-insulator-metal architecture. It is made up of a formation of 50-nm-thick Au
n-resonators parted by a 60-nm-thick SiO2 and a 200-nm-thick ground plane. To
achieve broadband absorption, a supercell having 16 resonators of various sizes and
shapes was employed, and they all had four-fold symmetry to give a polarization-
insensitive response.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp05.indd 143 9/26/23 4:58 PM

 144 Photonics Elements for Sensing and Optical Conversions

FIGURE 5.6 (a) Graphic depiction of the supercell of the broadband MS PA consisting of 16
resonant units forming a square formation. (b) SEM image of a portion of the fabricated PA.
(c) Experimentally measured absorptance for s- and p-polarizations at 20° AOI [22].

Figure 5.6(b) depicts a SEM image of the sample, with an inset displaying a mag-
nified view of the supercell. Figure 5.6(c) depicts the observed absorptance at 20°
AOI for both s- and p-polarizations, demonstrating the polarization-insensitive strong
absorption throughout roughly the full solar spectrum. MS PA achieves absorptance
of more than 90% in the wavelength range 450–920 nm.

5.5.4 Metasurfaces for Sensor Applications

Currently, MMs and MSs are increasingly being considered, due to their anomalous
and highly tunable light dispersion characteristics. Their execution is reliant on their
subwavelength spacing and geometric variables of each meta-atom (MA). MAs are
made up of one or more subwavelength n-structures made of noble metals or high
RI dielectrics. They are intended to display desired effective local optical responses
in terms of amplitude and phase, as well as E- and H- polarization. Various possible
implementations with unusual characteristics, such as negative RI, perfect absorp-
tion, and transmissivity, have been identified. The most practical and typical imple-
mentation of many possible applications, such as superlens, slow light, and cloaking
technologies, is RI biosensing. Biomolecular interactions in analyte layers induce RI
alterations. The RI sensor offers distinguishing capabilities for sensitive and label-
free biochemical tests, allowing it to play a vital role in a wide range of chemical and
biological implementations.
Resonant EM spectra, which are dominated by the ambient environment, can
be largely tuned by designing individual MAs and their arrangements. The RI
of the neighboring bimolecular analytes can be checked with a variety of scat-
tered output spectra due to such resonant properties. The configuration of suitable

BK-TandF-KAZANSKIY_9781032572949-230642-Chp05.indd 144 9/26/23 4:58 PM

 Metasurfaces and Several Well-Studied Applications 145

MAs at the target wavelengths and certain configurations is therefore essential.

Furthermore, sensing systems based on MM and MS offer numerous benefits over
traditional SPP-based biosensors in terms of manufacturing tolerance and read-
out signal stability. Second, periodic MA configurations enable reduced radiative
damping and greater quality factor through intriguing physical phenomena such as
plasmonically generated transparency and Fano resonances. MS might eventually
be added to the capability of a single n-photonic RI sensor. Several resonances and
broadband slow-light phenomena can be produced by delicate and intricate com-
positions of multiple distinct MAs in a unit-cell or supercell, which are difficult to
permit in SPP-based sensors.
Sensing light through light–matter interactions, on the other hand, has been a
well-studied issue in optical society. Fundamental characteristics of any incident EM
waves may be detected and studied via interference between multiple beams, aniso-
tropic scattering, and dispersive responses from specific EM devices. Polarization
and spectral composition are essential aspects of interest, and they are examined
using polarimeters and spectrometers, which are types of industrial bulk-optic equip-
ment. MSs are projected to be a viable platform for implementing these characteris-
tics with small volume and light weight, as integration and downsizing requirements
for these light-sensing components have grown considerably for electronic devices.

5.5.4.1 Sensors Based on Surface-Enhanced Raman Scattering

The inelastic scattering of photons by molecules driven to higher vibrational or rota-
tional energy levels is known as Raman scattering. Because it is based on molecular
states, the Raman scattered light spectrum is widely utilized to categorize and study
materials. It appears to have worth in that it can be utilized to identify and evaluate
not just fundamental things like solids, liquids, and gases but also more complicated
materials, including human tissues, biological species, and chemical compounds.
Raman scattering has been an essential way of assessing the structural fingerprint by
which molecules may be recognized in many chemical processes since the discovery
and development of lasers in the 1960s.
Despite its flexibility, applying Raman scattering directly to the sensor requires
a lengthy time due to its tiny scattering cross section. When matter and light scat-
ter, most photons are elastically dispersed, which is known as Rayleigh scattering,
and the scattered photons have the same frequency as the original photons. Only
approximately 1 in 10 million scattered photons are dispersed due to the expectation
of scattered photons with a frequency differing from that of input photons. In low
sample concentration settings, such as thin film covered with diluted solutions or
monolayers, sufficient Raman spectra are difficult to obtain due to the low chance
of molecule Raman scattering. Another significant difficulty is that, unlike pure or
colorless samples, materials having color or fluorescence are difficult to study using
Raman spectra. Also, when a laser in a VIS area is lit, reliable findings are extremely
difficult to acquire in the event of a sample that emits greater background-fluorescent
signals than Raman scattering of the sample itself.
In 1974, Fleischmann et al. found that electrochemically roughed Ag improved
the pyridine-adsorbed Raman signals. A series of experiments in 1977 verified that
Raman signals could be significantly enhanced when molecules are adjacent to rough

BK-TandF-KAZANSKIY_9781032572949-230642-Chp05.indd 145 9/26/23 4:58 PM

 146 Photonics Elements for Sensing and Optical Conversions

metal surfaces or n-structures. It was referred to as surface-enhanced Raman scat-

tering (SERS) or surface-enhanced Raman spectroscopy [23]. Scientists have been
essentially employing noble metals such as Au or Ag and distinct n-structures such as
plasmonic antennas for several decades to obtain more amplified signals. This is due
to the ability of metal n-antennas to support nonpropagating excitations of the con-
duction electrons tied to the EM field, known as localized surface plasmons (LSPs).
MMs and MSs are ideal candidates for SERS substratum because they can excite
surface plasmon polaritons (SPPs) which helps to reinforce Raman signals, where
EM responses can be regulated by self-engineered MAs in the whole range of VIS
to near-IR spectrum to detect biomolecules and proteins. Recently, researchers have
developed SERS substratum by utilizing split ring resonators MMs and attained
chemical sensing [24]. The tunable VIS to near-IR MMs based n-sensor with SERS
molecular detection has also been demonstrated [25].

5.5.5 Metasurface Waveguides

Impedance surfaces are metal surfaces that are subwavelength in form and printed
on dielectric media. Because of the substratum’s small electric thickness, basic 2D
surfaces may be modeled. Because it is frequently periodic or pseudo-periodic, even
the structures may be precisely and efficiently described using periodic boundary
conditions and eigenmode models. The surface impedance and dispersion properties
are determined by the size and form of the periodic patterns. If the wave vector along
the surface (k) is larger than its free-space counterpart (ko), the impedance surface
supports and binds the propagating wave along the surface, resulting in a planar WG.
For guiding SWs along a constrained route, simple WG structures entailing a thin
strip of high impedance surrounded by two low impedance sheets have been proven
[26]. The most popular approach for creating materials with variable SW indices is
to employ a grid of varying size metallic patches on a dielectric substratum. The
use of asymmetric patches yields a tensor SW index. SW WG formation interacts
with SWs similarly to how an optical fiber transmission line interacts with light. The
physical phenomenon is the same; the wave travels in an ideal high RI zone bordered
by a low RI region. In the case of dielectric WG, the high RI and low RI areas are
achieved with high- and low-permittivity materials. In the case of SW WG, the high
and low RI areas can be achieved by variable size and/or shape metallic patches on
a dielectric substratum.

5.5.5.1 Surface Wave Media

Surface wave medium (SWM) refers to any structure that may sustain a SW. They
are a subsection of the larger category of MMs identified as artificial impedance
surfaces and may serve SWs polarized in either transverse electric (TE) or transverse
magnetic (TM) modes. The SW index (nSW) or the SW impedance (ZTE and ZTM) is
used to describe the SWM. The SW index is equivalent to an optical RI, and it is
associated to SW impedance by:
1 1
 Z2  2  n2  2
nSW =  1 + TM and nSW =  1 + 2o  ,
 no2   Z TE 

BK-TandF-KAZANSKIY_9781032572949-230642-Chp05.indd 146 9/26/23 4:58 PM

 Metasurfaces and Several Well-Studied Applications 147

for TM and TE modes, respectively, where,

1/ 2
µ 
no =  o  .
 εo 

A grounded dielectric sheet is the simplest form of an SWM. The transverse reso-
nance method can be utilized to analytically calculate its SW RI. The grounded
dielectric is not practical at frequencies less than 10 or 20 GHz because it must be
very thick or use a substratum with excessively high permittivity to support SWs
effectively. The most common way to prevail over this limitation is to distribute a
grid of metallic patches (in any possible shape and distributes in any pattern) on a
dielectric substratum. Nevertheless, the distribution on a uniform, periodic forma-
tion makes simulations, calculations, and development easier. A uniform periodic
formation of square patches is the simplest grid. A frequency-independent capaci-
tance defines a grid of squares or any convex shape. The SW RI of SWMs gener-
ated with these forms is insufficient due to the substratum’s RI. Grids constructed
out of filamentary forms, such as the Jerusalem cross, have equivalent capacitance,
inductance, and resonance [27]. SWMs with these forms have a greater SW RI than
square-grid SWMs and are not subject to the same SW RI restrictions. They have
lower BW, though, since the grid’s resonant structure decreases the SW band-gap
frequency. Generally, SWM constructed of metallic patches of any shape and pattern
obey the following rules: SW RI rises monotonically with SW frequency, forma-
tion period, patch size relative to the formation period, substratum permittivity, and
thickness. SWMs are classified as either scalar or tensor. In a scalar SWM, the SW
RI is independent of the transmission direction, but in a tensor SWM, the RI is direc-
tion dependent, and tensor quantities represent the RI and the impedance.

5.5.5.2 Surface Waves Waveguide Theory

Here three methods for manipulating the transmission of SWs are presented: (1) use
abrupt variations in SW RI as confinement boundaries, (2) guide the phase fronts
along curved paths deprived of alteration by grading the SW RI across the subwave-
length gratings (SWG) width, and (3) align the lower-RI principle axis of a tensor
SWM with transmission direction. The easiest SWG consists of a strip of constant
SW RI encircled by a low-RI SWM, which is a 2D counterpart of a 3D dielectric WG.
It can be viewed as a high RI 2D fiber encircled by a low RI medium. The simula-
tion results for straight and curved SWGs are presented in [26]. The SW is excited
by a dipole feed. The high-impedance channel is one free-space wavelength wide.
The corresponding SW indices are n1 = 1.44, and n2 = 2.24. Simulation is done
using FastScat, a high-order integral-equation solver. Higher field values are repre-
sented as brighter colors. Power outside the SWG is more than 20 dB below the peak
of the SWG. The phase front is approximately perpendicular to the transmission
path. Some phase front falsification arises as the SWG width is 2.24 times the SW
wavelength, and more than one transmission mode is existing. A curved SWG guides
the SW around a curvature radius of 6.7 λsw, where λsw is the wavelength of the SW.
The SW is confined to the SWG, but some power is leaked as it negotiates the curve.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp05.indd 147 9/26/23 4:59 PM

 148 Photonics Elements for Sensing and Optical Conversions

The leaked power is 15–20 dB below the peak levels at the SWG center. The leakage
of power is due to the path length, which differs across the width of the curve, which
triggers the wavefront to diverge from being perpendicular to the WG axis. By clas-
sifying the SW RI across the width of the SWG, this problem can be solved so that
the electrical path length is constant across the width. This results in the progressive
rotation of the phase front to match the curvature of the transmission path.

5.5.6 Cloaking
One of the most fascinating and well-known characteristics of MS is the potential to
create an invisibility cloak. It has been theoretically demonstrated and experimen-
tally confirmed in several scenarios that the anomalous wave interaction of artificial
materials and MMs can be tailored to significantly reduce an object’s overall vis-
ibility in distinct frequency regions spanning radio frequencies (RF), IR, and VIS
light [28].
The presence of anomalous localized resonance near superlenses has been dem-
onstrated to contribute to cloaking phenomena. This happens when a polarizable
line or point dipole-generated resonant field acts back on the polarizable line or
point dipole, cancelling out the field operating on it from outside sources. Cloaking
is proven in the quasi-static limit for finite collections of polarizable line dipoles
that are all within a certain distance of a coated cylinder with a shell permittivity
εs ≈ –εm ≈ εc, where εm is the permittivity of the neighboring matrix, and εc is the core
permittivity [29].
The first suggested models employed 3D anisotropic MMs and were based on
the notion of transformation optics [30]. Such cloaks, on the other hand, were overly
heavy and narrowband, resulting in enormous losses. An alternate technique called
“mantle cloaking” was suggested, which involves placing ultra-thin MS screens at a
distance over the item to be cloaked.
Although the theoretical basis of all these MMs-related cloaking methods is well
established, the practical implementation of these systems, based on existing MM
technologies, is far from optimal. At the moment, experimental demonstrations on
cloaking technology have resulted in very modest scattering reduction, which falls
well short of theoretical projections. This is largely owing to the intrinsic difficul-
ties of establishing bulk MMs as collections of tiny inclusions with the necessary
unusual bulk characteristics that may be handled as a continuum. These constraints
are especially apparent for transformation-based cloaks, which need complicated
and accurate inhomogeneity contours that might be difficult to implement with finite
granularity formations. Furthermore, some of these cloaking approaches need an
average thickness equal to the size of the cloaked zone, which may result in BW
limits and greater sensitivity to realistic material loss.

5.5.6.1 The Challenges Faced by the First Cloaking Device

On October 19, 2006, a tiny cloaking device was utilized to illustrate the idea of
cloaking at frequencies in the MWv radiation range for the first time. It was smaller
than 13 mm in height and 125 mm in diameter, and it effectively directed MWvs
around it. The object to be hidden, a tiny cylinder, has been put in the device’s center.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp05.indd 148 9/26/23 4:59 PM

 Metasurfaces and Several Well-Studied Applications 149

FIGURE 5.7 (a) The cross section of a PEC cylinder subject to a plane wave (only the
E-field component of the wave is shown). The field is dispersed. (b) A circular cloak, designed
using transformation optics methods, is employed to cloak the cylinder. In this case, the field
remains unaffected outside the cloak and the cylinder is unseen electromagnetically.

The invisibility cloak deflected MWv beams, causing them to flow around the cylin-
der inside with only a tiny alteration, almost as if nothing was there at all. Typically,
such a device entails enclosing the item to be cloaked with a shell that alters the pas-
sage of light close to it. The object’s reflection of EM waves was decreased. Unlike a
homogeneous natural substance with the same material characteristics everywhere,
the cloak’s material properties vary from point to point, with each point being tai-
lored for specific EM interactions (inhomogeneity) and varying in various direc-
tions (anisotropy). This achieves a gradient in the properties of the material. This
can be implemented in an MM in which split-ring resonators known to deliver an
H-response that can be custom-made are located with their axes along the radial
direction [31]. The interaction of plane wave around the cloaked and uncloaked cyl-
inder is revealed in Figure 5.7. Three notable limitations can be shown, despite a
successful demonstration.

1. Since the experimental demonstration was successful only in the MWv

range, the small object is somewhat transparent only at MWv frequen-
cies. It implies that invisibility had not been achieved for the VIS spec-
trum since the VIS spectrum’s wavelengths are significantly shorter than
MWvs. However, this was the first step toward a cloaking device for VIS
light, although more advanced nanotechnology-related techniques would
be needed due to the short wavelengths of light.
2. According to the experiment conducted in 2006, only small objects can be
made to appear as the neighboring air.
3. For any given demonstration, cloaking can only take place over a narrow
frequency band. This means that there is no broadband cloak, on hand cur-
rently, that works across the EM spectrum, from radio frequencies to MWv
to VIS spectrum to X-ray. This is because of the dispersive nature of today’s
MMs. The transformation optics requires exceptional material variables
that are merely accessible using resonant units that are essentially narrow-
band, and dispersive in resonance.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp05.indd 149 9/26/23 4:59 PM

 150 Photonics Elements for Sensing and Optical Conversions

5.5.7 All-dielectric Metasurfaces Visible Light Modulator

Although the field of MSs has started to evolve based on metallic materials, the poor
execution of these devices for short wavelengths, mainly due to ohmic dissipation and
dispersion losses, has prompted the scientific community to shift its attention to all-
dielectric alternatives. In the case of MSs utilizing high RI dielectric n-antennas, they
benefit not only from the lower ohmic losses but also from the naturally rich phenom-
enology of optical modes, including both E- and H-multipolar Mie resonances, which
permit a great deal of flexibility in adapting their scattering directivity.
Most of the MSs that have been shown to date are static. Nevertheless, imple-
mentations in the display and other optoelectronic device industries require compact
and electrically tuned optical components. Several strategies have been explored to
dynamically control light modulation by an MS. By using diodes, one can easily
achieve that at MW frequencies. For optical frequencies where this strategy does not
work, mechanically deformable MAs or substratum may be introduced and MEMS
can be used to tune their resonances or alter the dielectric MS environment using
birefringence crystals, graphene, gated semiconductors, phase change materials, or
electro-optical polymers.
Nonetheless, due to the small change in permittivity, mainly at high frequencies,
the tuning range for the cases of charge injection or electro-optical effects is inad-
equate, while phase change materials appear to be lossy in the VIS spectrum. At
room temperature, the liquid crystal has a large birefringence and is almost lossless
in the VIS range. Together with the innovative liquid crystal technology, it makes
them the ideal platform for active devices for the VIS spectral range, which can be
addressed via electrical, optical, H-, and thermal approaches.
Several works have recently been reported based on tunable devices through the
combination of liquid crystals with plasmonic and dielectric MSs. To date, most of
the liquid crystal–based high RI dielectric MSs have been documented to have used
silicon as MAs and show broad tuning of the supported E- and H-dipole resonances
under external and electrical control in the near-IR frequency range. Nevertheless,
given the low band-gap of silicon, which is approximately equal to 1.1 eV, signifi-
cant losses are likely for such MSs within the VIS spectral range. Meanwhile, the
methodical discussion with the liquid crystal birefringence on the evolution of EM
resonance within MS n-antennas is still inadequate.
As shown in Figure 5.8(a and b), a novel EO-polymer-based metallic MS mod-
ulator operating at 1.55-μm wavelength that employs bimodal resonance is sug-
gested [32]. EO polymer has several advantages among various active materials,
such as it exhibits a large EO coefficient above 200 pm/V, ultrahigh-speed modula-
tion capability over 40 GHz, simple and low-cost spin-coating-based manufactur-
ing that allows flexible device configuration, and high material reliability proven
in commercial products. The device configuration is modeled to excite two metal-
insulator-metal (MIM) resonant modes inside the EO polymer. A sharp dip arises in
the reflected spectrum by employing the strong coupling between these two modes,
with a meaningfully higher Q-factor than that attained when using a single-mode
Fabry–Perot (FP) resonance, as demonstrated in Figure 5.8(c). A transmission-type,
liquid crystal-based, tunable MS that efficiently operates in the VIS spectral range

BK-TandF-KAZANSKIY_9781032572949-230642-Chp05.indd 150 9/26/23 4:59 PM

 Metasurfaces and Several Well-Studied Applications 151

FIGURE 5.8 (a) Diagram of an active MS modulator with electro-optic polymer. (b) Cross-
sectional view. (c) Modulation principle of the surface normal MS modulator [32]. (d) Plan of
an electrically tunable transmission type TiO2 MS with liquid crystal infiltration sandwiched
by ITO-coated glass substratum. (e) Assessment of calculated and measured transmittance
spectra of the TiO2 n-discs MS. The inset gives SEM image of TiO2 MS [33].

is demonstrated where titanium dioxide (TiO2) n-antennas are used as demonstrated

in Figure 5.8(d). TiO2 has a wide band-gap and is lossless throughout the VIS spec-
trum. The TiO2 n-discs are mounted in a thin liquid crystal cell to achieve resonance
tuning. The E- and H-resonances are dynamically controlled by switching the liquid
crystal alignment under electrical voltages [33]. When modulation voltage is put on
by means of the same electrodes, it is possible to change the RI of the EO polymer
using the Pockels effect and shift the resonant wavelength (λ0) so that the reflected
light intensity is modulated at (λ0), as demonstrated in Figure 5.8(e).

5.5.8 Frequency Selective Surfaces Based on Metasurfaces

A frequency selective surface (FSS) is a periodic arrangement with each component
resonating at the resonance frequency. Usually, the FSSs have a periodicity equal to
half the wavelength of the resonant frequency. These have often been used to deliver
spectral filtering for signal communications. They are also utilized as diplexers, res-
onant beam splitters, and antenna radomes. Due to the resonant nature of MAs, MSs

BK-TandF-KAZANSKIY_9781032572949-230642-Chp05.indd 151 9/26/23 4:59 PM

 152 Photonics Elements for Sensing and Optical Conversions

can replace FSS. Because the building blocks of an MS are subwavelength in nature,
they often bring additional benefits compared to a conventional FSS. The subwave-
length periodicity allows many unit cells to be packed in a restricted space that is
highly useful for space-limited radomes. The tiny size of the unit cell also enables
the MSs to have a balanced response to variations in the AOI of the incoming EM
wave. FSSs have now been used in dichroic subreflectors, radio frequency identifica-
tion (RFID), lenses antennas, and shield from EM interference.
FSS is a vigorously researched subject of EM science, 2D periodic structures with
planar metal formation units (apertures) on a nonconductor substratum that display
transmission and reflection at a particular resonant wavelength. Subject to the layout
of the formation component, the inbound plane wave will be either transmitted or
reflected, in whole or in part. This happens when the frequency of the plane wave cor-
responds to the resonant frequency of the FSS units. An FSS can, therefore, transfer
or block EM waves with certain frequencies, so they are best known as space filters.
According to the circuit theory, FSS structures (capacitive and inductive), also
known as spatial filters, are like MWv filters. It is possible to categorize FSS filtering
characteristics into four groups, including high pass, low pass, stopband and band-
pass, as demonstrated in Figure 5.9. High-pass FSS filters allow a higher frequency
range to pass through the structure while bypassing the lower-frequency range. Using
the Babinet principle, low-pass FSS filter functioning is a counterpart of the high-
pass filter function. Likewise, the stopband FSS filter blocks unwanted frequencies
while the band-pass FSS filter only allows frequency ranges. FSSs are devised by
periodic arrangement of metal patches and/or slits etched on a dielectric material for
a desired resonant functioning. The highly crucial part of the configuration develop-
ment is the correct choice of FSSs formation units, shape, size, and substratum mate-
rial. Munk has clarified the functional concept of FSS based structures in detail [34].
The functionality of FSS is attained by a corresponding self-resonating network.
Electric currents are stimulated in the formation units when EM waves strike the FSS
structure. The amplitude of the currents created is determined by the amount of energy
coupling. Induced currents, on the other hand, frequently operate as EM sources, gen-
erating further dispersed fields. The resultant field in the neighboring FSS is formed

FIGURE 5.9 The frequency response characteristics of the FSS: (a) high pass, (b) low pass,
(c) stopband and (d) band-pass.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp05.indd 152 9/26/23 4:59 PM

 Metasurfaces and Several Well-Studied Applications 153

by incident EM fields coupled with these dispersed fields. As a result, accurately con-
structed components may get the required currents and field characteristics, and a filter
response can be created. The patch components (dipole FSSs) are meant to function as
stopband filters for inbound plane waves and to operate as completely reflecting sur-
faces in a narrow band of frequencies. The aperture units, on the other hand, have band-
pass properties, which means they operate as transparent surfaces for the incidence of
EM waves within the operating frequency range. Traditional FSSs, on the other hand,
are limited in their usage to fulfill the practical requirements for a range of EM utiliza-
tions due to their poor filter response, low angular stability, and small BW.
FSS patches produce resistance (R) and inductance (L), whereas capacitance (C)
is generated by gaps between the FSS units. The basic electrostatic assumption holds
the physical meaning of these passive values for various FSS components, e.g. L of
two parallel wires and C produced by a parallel plate capacitor. The essential filter
response is therefore created by combining these capacitive and inductive compo-
nents. Any adjustment in the dimensional variables of FSS, however, results in an
analogous difference in the values of L and C. Physically, when the unit cell of the
FSS is lit by the EM pulse, it can be transformed into an analogous resonance circuit.
The resonance frequency (fres) can be expressed in which L and C indicate the cor-
responding inductance and the capacitance:

1
fres = . (5.10)
2π LC

The choice of a suitable formation component is very critical when designing

FSS. Although distinct unit cell geometries have been introduced, some of them
can be easily controlled and are therefore more common in the FSS community.
For instance, a nonresonant unit (patch, wire grid) can be modeled by a capacitance,
while a single resonant component (loop, cross, dipole) can be characterized by a
series of capacitors and inductor combinations. It should be noted that there is a
direct relation between the number of resonances and the number of lumped units.

5.5.9 Polarizers Based on Metasurfaces

Polarization (POL) is an inherent characteristic of EM waves, and transition between
POL states is very often suitable for several modern EM and photonic utilizations.
For example, the transformation of linear POL to circular POL in advanced com-
munication and sensing renders a beam immune to environmental variation, scat-
tering, and diffraction in the atmosphere. Thanks to the configuration stability and
compactness, POL transformation using MSs has drawn increased interest in recent
years. The subsequent ability to tune a phase delay covering the whole 2π spec-
trum across a broad BW and with a deep subwavelength resolution may theoretically
resolve some critical issues in the development of flat optics.
Simple, substantially symmetrical MAs can be beneficial for maintaining POL
states. However, violating the symmetry can give extra flexibility for gaining cus-
tomized usage, allowing POL states to be utilized. It is possible to have an identi-
cal transmission magnitude but a relative phase delay (∆φ) at a given frequency by

BK-TandF-KAZANSKIY_9781032572949-230642-Chp05.indd 153 9/26/23 4:59 PM

 154 Photonics Elements for Sensing and Optical Conversions

FIGURE 5.10 The individual unit cells for (a) the electric split-ring resonator quarter-wave
plate and optical microscope picture (left to right), and (b) the meander line quarter-wave
plate and optical microscope picture (left to right) [35].

customizing the two eigenmodes leading to orthogonal linear POLs. Single-layer

MSs [35] or multilayer cascading MSs operating from MWv to optical frequencies
have achieved narrowband POL transformation between linear and circular POL
states. While the conventional meander line has a better BW and transmission mag-
nitude, the electric split-ring resonator is simpler to make since it has only one active
Au layer, a more constant phase shift, and a higher peak POL percentage. Figure 5.10
depicts the configuration of an electric split-ring resonator quarter-wave plate and a
meander line quarter-wave plate.
Dielectric MSs containing distinct half-waveplate-like hydrogenated amorphous
silicon n-posts and are shown to control the wavefront of transmitted VIS light with
controllable linear POL angles. The cross-polarized scheme shows a high-quality
anomalous beam deflector configuration with an absolute deflection efficacy of 82%,
a POL conversion efficacy of 96%, and an extinction ratio of 37 dB. The anomalously
deflected light could also hold a high degree of linear POL (>0.96), which can be
continuously rotated by adjusting the POL angle of the incident [36].
The low level of POL conversion efficacy can be resolved by implementing MSs
with few layers. Using the definition suggested by Jones matrix linearly polarized
incident fields (Ex, Ey) through an MS can be described as:

 E xt   Txx Txy   E xi   E xi 
 t  =   i  = Tˆlinear  i  . (5.11)
 E y   Tyx Tyy   E y   E y 

BK-TandF-KAZANSKIY_9781032572949-230642-Chp05.indd 154 9/26/23 4:59 PM

 Metasurfaces and Several Well-Studied Applications 155

For circularly polarized incident fields, it becomes:

 E+t   T++ T+−   E+i 

 t  =  = Tˆcircular  i  , (5.12)
 E−   T−+ T−−   E− 

1 i 1 i
where T± ± =
2
( Txx + Tyy ) ± ( Txy − Tyx ) and T±  = ( Txx − Tyy )  ( Txy + Tyx ).
2 2 2

REFERENCES
1. Zhan, A. et al. 2016. Low-contrast dielectric metasurface optics. ACS Photonics
3(2):209–14. https://doi.org/10.1021/acsphotonics.5b00660.
2. Astilean, S., Lalanne, P., Chavel, P., Cambril, E., and Launois, H. H. 1998. High-
efficiency subwavelength diffractive element patterned in a high-refractive-index mate-
rial for 633 nm. Opt. Lett. 23(7):552–4. https://doi.org/10.1364/OL.23.000552.
3. Minatti, G., Caminita, F., Casaletti, M., and Maci, S. 2011. Spiral leaky-wave antennas
based on modulated surface impedance. IEEE Trans. Antennas Propag. 59(12):4436–
44. https://doi.org/10.1109/TAP.2011.2165691.
4. Gregoire, D. J. 2013. 3-D conformal metasurfaces. IEEE Antennas Wirel. Propag. Lett.
12:233–6. https://doi.org/10.1109/LAWP.2013.2247017.
5. Nannetti, M., Caminita, F., and Maci, S. 2007. Leaky-wave based interpretation of
the radiation from holographic surfaces. Proc. 2007 IEEE Antennas and Propagation
Society International Symposium:5813–6. https://doi.org/10.1109/APS.2007.4396873.
6. Pereda, A. T., Caminita, F., Martini, E., Ederra, I., Iriarte, J. C., Gonzalo, R., and Maci,
S. 2016. Dual circularly-polarized broadside beam metasurface antenna. IEEE Trans.
Antennas Propag. 64(7):2944–53. https://doi.org/10.1109/TAP.2016.2562662.
7. Minatti, G., Faenzi, M., Sabbadini, M., and Maci, S. 2017. Bandwidth of gain in
metasurface antennas. IEEE Trans. Antennas Propag. 65(6):2836–42. https://doi.
org/10.1109/TAP.2017.2694769.
8. Faenzi, M. et al. 2019. Metasurface antennas: New models, applications and realiza-
tions. Sci. Rep. 9:10178. https://doi.org/10.1038/s41598-019-46522-z.
9. Hongnara, T., Chaimool, S., Akkaraekthalin, P., and Zhao, Y. 2018. Design of compact
beam-steering antennas using a metasurface formed by uniform square rings. IEEE
Access 6:9420–9. https://doi.org/10.1109/ACCESS.2018.2799551.
10. Yu, N., and Capasso, F. 2014. Flat optics with designer metasurfaces. Nat. Mater.
13:139–50. https://doi.org/10.1038/nmat3839.
11. Aieta, F., Genevet, P., Kats, M. A., Yu, N., Blanchard, R., Gaburro, Z., and Capasso, F.
2012. Aberration-free ultrathin flat lenses and axicons at telecom wavelengths based on
plasmonic metasurfaces. Nano Lett. 12(9):4932–6. https://doi.org/10.1021/nl302516v.
12. Pors, A., Nielsen, M. G., Eriksen, R. L., and Bozhevolnyi, S. I. 2013. Broadband focus-
ing flat mirrors based on plasmonic gradient metasurfaces. Nano Lett. 13(2):829–34.
https://doi.org/10.1021/nl304761m.
13. Chen, X., Huang, L., Mühlenbernd, H., Li, G., Bai, B., Tan, Q., Jin, G., Qiu, C.-W.,
Zentgraf, T., and Zhang, S. 2013. Reversible three-dimensional focusing of visible light
with ultrathin plasmonic flat lens. Adv. Opt. Mater. 1(7):517–21. https://doi.org/10.1002/
ADOM.201300102.
14. Yin, L., Vlasko-Vlasov, V. K., Pearson, J., Hiller, J. M., Hua, J., Welp, U., Brown, D. E.,
and Kimball, C. W. 2005. Subwavelength focusing and guiding of surface plasmons.
Nano Lett. 5(7):1399–402. https://doi.org/10.1021/nl050723m.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp05.indd 155 9/26/23 4:59 PM

 156 Photonics Elements for Sensing and Optical Conversions

15. Liu, Z., Steele, J. M., Srituravanich, W., Pikus, Y., Sun, C., and Zhang, X. 2005.
Focusing surface plasmons with a plasmonic lens. Nano Lett. 5:1726–9.
16. Vo, S., Fattal, D., and Wayne, V. et al. 2014. Sub-wavelength grating lenses with a
twist. IEEE Photonics Technol. Lett. 26(13):1375–8. https://doi.org/10.1109/LPT.2014.
2325947.
17. Arbabi, A., Horie, Y., Ball, A. J., Bagheri, M., and Faraon, A. 2015. Subwavelength-
thick lenses with high numerical apertures and large efficiency based on high-contrast
transmit arrays. Nat. Commun. 6:7069. https://doi.org/10.1038/ncomms8069.
18. Khorasaninejad, M., Zhu, A. Y., Roques-Carmes, C., Chen, W. T., Oh, J., Mishra,
I., Devlin, R. C., and Capasso, F. 2016. Polarization-insensitive metalenses at visible
wavelengths. Nano Lett. 16:7229–34. https://doi.org/10.1021/acs.nanolett.6b03626.
19. Liu, N., Mesch, M., Weiss, T., Hentschel, M., and Giessen, H. 2010. Infrared perfect
absorber and its application as plasmonic sensor. Nano Lett. 10(7):2342–8. https://doi.
org/10.1021/nl9041033.
20. Watts, C. M., Shrekenhamer, D., Montoya, J., Lipworth, G., Hunt, J., Sleasman, T.,
Krishna, S., Smith, D. R., and Padilla, W. J. 2014. Terahertz compressive imaging with
metamaterial spatial light modulators. Nat. Photon. 8:605–9. https://doi.org/10.1038/
nphoton.2014.139.
21. Shi, N. N., Tsai, C.-C., Camino, F., Bernard, G. D., Yu, N., and Wehner, R. 2015.
Keeping cool: Enhanced optical reflection and radiative heat dissipation in Saharan
silver ants. Science 349(6245):298–301. https://doi.org/10.1126/science.aab3564.
22. Azad, A. K. et al. 2016. Metasurface broadband solar absorber. Sci. Rep. 6:20347.
https://doi.org/10.1038/srep20347.
23. Goul, R., Das, S., Liu, Q., Xin, M., Lu, R., Hui, R., and Wu, J. Z. 2017. Quantitative
analysis of surface enhanced raman spectroscopy of rhodamine 6G using a compos-
ite graphene and plasmonic au nanoparticle substrate. Carbon 111:386–92. https://doi.
org/10.1016/j.carbon.2016.10.019.
24. Xu, X., Peng, B., Li, D., Zhang, J., Wong, L. M., Zhang, Q., Wang, S., and Xiong, Q.
2011. Flexible visible-infrared metamaterials and their applications in highly sensi-
tive chemical and biological sensing. Nano Lett. 11(8):3232–8. https://doi.org/10.1021/
nl2014982.
25. Cao, C., Zhang, J., Wen, X., Dodson, S. L., Dao, N. T., Wong, L. M., Wang, S., Li,
S., Phan, A. T., and Xiong, Q. 2013. Metamaterials-based label-free nanosensor for
conformation and affinity biosensing. ACS Nano 7(9):7583–91. https://doi.org/10.1021/
nn401645t.
26. Gregoire, D. J., and Kabakian, A. V. 2011. Surface-wave waveguides. IEEE Antennas
Wireless Propag. Lett. 10:1512–5. https://doi.org/10.1109/LAWP.2011.2181476.
27. Simovski, C., de Maagt, P., and Melchakova, I. V. 2005. High-impedance surfaces hav-
ing stable resonance with respect to polarization and incidence angle. IEEE Trans.
Antennas Propag. 53(3):908–14. https://doi.org/10.1109/TAP.2004.842598.
28. Chen, X., Luo, Y., Zhang, J., Jiang, K., Pendry, J. B., and Zhang, S. 2011. Macroscopic
invisibility cloaking of visible light. Nat. Commun. 2:176. https://doi.org/10.1038/
ncomms1176.
29. Milton, G. W., and Nicorovici, N.-A. P. 2006. On the cloaking effects associated with
anomalous localized resonance. Proc. Math. Phys. Eng. Sci. 462(2074):3027–59.
https://doi.org/10.1098/rspa.2006.1715.
30. Fleury, R., Monticone, F., and Alù, A. Invisibility and cloaking: Origins, present,
and future perspectives. Phys. Rev. Appl. 2015; 4(3):037001. https://doi.org/10.1103/
PhysRevApplied.4.037001.
31. Schurig, D. et al. 2006. Metamaterial electromagnetic cloak at microwave frequencies.
Science 314(5801):977–80. https://doi.org/10.1126/science.1133628.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp05.indd 156 9/26/23 4:59 PM

 Metasurfaces and Several Well-Studied Applications 157

32. Zhang, J., Kosugi, Y., Otomo, A., Nakano, Y., and Tanemura, T. 2017. Active metasur-
face modulator with electro-optic polymer using bimodal plasmonic resonance. Opt.
Express 25(24):30304–11. https://doi.org/10.1364/OE.25.030304.
33. Sun, M. et al. 2019. Efficient visible light modulation based on electrically tunable
all dielectric metasurfaces embedded in thin-layer nematic liquid crystals. Sci. Rep.
9:8673. https://doi.org/10.1038/s41598-019-45091-5.
34. Munk, B. A. 2000. Frequency selective surfaces: Theory and design. Hoboken, NJ:
Wiley Online Library. ISBN: 978-0-471-37047-5.
35. Strikwerda, A. C., Fan, K., Tao, H., Pilon, D. V., Zhang, X., and Averitt, R. D. 2008.
Comparison of birefringent electric split-ring resonator and meanderline structures as
quarter-wave plates at terahertz frequencies. Opt. Express 17(1):136–49. https://doi.
org/10.1364/OE.17.000136.
36. Gao, S., Park, C.-S., Lee, S.-S., and Choi, D.-Y. 2019. All-dielectric metasurfaces for
simultaneously realizing polarization rotation and wavefront shaping of visible light.
Nanoscale 11(9):4083–90. https://doi.org/10.1039/C9NR00187E.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp05.indd 157 9/26/23 4:59 PM

 6 Optical Fiber Sensors
Based on Diffractive
and Fiber Periodic
Microstructures
Sergey Vladimirovich Karpeev1,2 and
Svetlana Nikolaevna Khonina1,2
1I mage Processing Systems Institute –
Branch of the Federal Scientific Research Centre,
“Crystallography and Photonics” of Russian
Academy of Sciences, Samara, Russia
2 Samara National Research University,

Samara, Russia

6.1 INTRODUCTION
Optical fiber sensors are now widely used as parts of measuring systems for moni-
toring various technical and natural systems. Their main advantages are compatibil-
ity with existing fiber-optic data transmission networks, high resistance to external
chemical and radiation influences and insensitivity to electromagnetic fields, high
sensitivity to selected parameters, and ease of integration into measuring systems.
Depending on the type of optical fiber sensor, the output signal can be either the
amplitude of the transmitted light, or the phase or the frequency shift. In the case
of amplitude modulation, multimode optical fibers can be used, and in this case it is
possible to increase the sensitivity and adjust the measurement ranges using the dif-
fractive optical elements (DOEs) incorporated into the measuring system. However,
amplitude-type sensors are susceptible to noise when using coherent radiation and
can be used with small lengths of coupling fibers, especially in the case of filter-
ing individual modes. Optical fiber sensors that implement the frequency shift of
the transmitted radiation are much more resistant to interference due to the fun-
damental nature of the wavelength. They are also easily combined into measuring
networks with the identification of each individual sensor and allow maintaining the
measurement mode at a sufficiently large distance from the object of study. These
sensors are based on variously microstructured fibers for creating Bragg fiber grat-
ings. Structuring methods are now actively developing and have therefore found their
place in this chapter.

158 DOI: 10.1201/9781003439165-6

BK-TandF-KAZANSKIY_9781032572949-230642-Chp06.indd 158 9/26/23 5:04 PM

 Optical Fiber Sensors 159

6.2 AMPLITUDE OPTICAL FIBER SENSORS BASED

ON DIFFRACTIVE OPTICAL ELEMENTS
MATCHED WITH TRANSVERSE MODES
6.2.1 Design Principles of Fiber-Optical Converters Based
on Mode Filtering

Depending on the underlying physical processes, optical fiber converters can be cat-
egorized as interference or amplitude [1]. Interference-aided converters commonly uti-
lize a pair of single-mode optical fibers, with one fiber exposed to an external physical
influence and the other used as a reference. By deciphering the resulting interference
pattern, the magnitude of the physical influence is evaluated. Amplitude-aided con-
verters utilize a single multimode optical fiber, measuring variations in the amplitude
of light in the fiber caused by the external influence. The amplitude variations occur
because the energy of a guided mode is partly transferred to unguided modes owing to
the physical influence. With different converters utilizing different types of influence,
irrespective of the type, the energy is transferred into the unguided modes differently
for different modes.
For higher-order modes, which are nearer to the cutoff frequency, the light energy
is confined closer to the fiber core boundaries, which means that the transfer to the
unguided modes takes place at smaller external influences, compared to the lower-
order modes. For the amplitude optical fiber converters, it is the difference in sensi-
tivity of different modes to the external influence that gives a gain in the converter
parameters when an analyzer and a transverse mode shaper are inserted into the beam.
In terms of the scalar theory of diffraction, this is equivalent to evaluating variations
in the optical properties of a wave medium through variations in the transverse mode
spectrum. It is worth noting that thanks to the use of mode filtering, characteristics of
the interference optical fiber converters can also be improved. Here, with the interfer-
ence taking place between the same fiber modes, the design of the converter can be
simplified compared with the known designs [1]. The same principle also applies when
dealing with an important task of controlling the refractive index profile in the fiber.
Among the optical fiber converters of physical magnitudes, a significant place
belongs to micro-displacement sensors that work by measuring power losses on fiber
microbendings. A physical principle behind the design of an optical fiber converter
is mode coupling due to fiber deformation, with the energy redistributed between the
modes and partly transferred to unguided modes [2]. As a result, the energy of indi-
vidual modes changes and the total energy flux in the fiber is decreased. Let us analyze
a converter from [3, 4], which utilizes an optical fiber that supports guided modes with
scalar-approximated complex amplitudes {ψp(x)} and the mode numbers p = (p, l).
The modes of interest may include both classical (e.g. Laguerre–Gaussian,
Hermite–Gaussian, and Bessel) optical fiber modes [5–8] and more general types
of laser beams [9–11]. Mode excitation occurs under illumination by a laser beam
[12–14] with complex amplitude:

F (x) = ∑F ψ (x),
p
p p x ∈G.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp06.indd 159 9/26/23 5:04 PM

 160 Photonics Elements for Sensing and Optical Conversions

The propagation of the beam in a fiber can be described using a linear operator
T̂ acting on the F(x) function. The complex amplitude F1(x) of the beam at the fiber
output is described by the equation:

F 1 = T̂F .

In the output beam, the mode spectrum {F1}:

F1 ( x) = ∑F
p′
1
p′ ψ p ′ ( x) = ∑F T̂ψ (x),
p
p p (6.1)

is defined by the input beam mode content {Fp} and the fiber properties described by
the operator T̂. The form of the operator T̂ depends on the fiber state, which depends
on the perturbations of fiber physical parameters, characterized by measurable
parameters V = (V1,V2,…,Vn). Hence, we can write T̂ = T̂ ( V ) , Fp1 = Fp1 ( V ).
For an unperturbed fiber, V = 0, T̂ = T̂ (0), where the operator T̂ (0) describes
the eigen-medium propagation of the mode. Using the definition of the propagation
constant βp, we can write:

T̂ (0)ψ p ( x) = exp(i β p z )ψ p ( x), (6.2)

where z is the mode travel distance in the fiber. We note that in an unperturbed fiber,
the modes are uncoupled:
2 2
Fp1′ = exp(i β p′ z ) Fp′ , Fp1′ = Fp′ . (6.3)

This is because in an unperturbed fiber, modes are the eigen-functions of the propa-
gation operator. Values of p correspond to the guided modes numbers. If the optical
fiber has a non-zero perturbation V, instead of Eq. (6.2), we introduce coefficients
Tpp and obtain:

T̂ ( V ) ψ p ( x ) = ∑Tp′
pp ′ ( V ) ψ p′ ( x ) . (6.4)

The coefficients Tpp′ (V) form a matrix of operator T̂ components, being equal to
the values of mode overlap integrals. Because the mode basis ψp is orthogonal, we
obtain:

∫
Tpp′ ( V) = ψ *p′ ( x) T̂ψ p ( x) d 2 x.
G
(6.5)

The coefficients Fp1′ can be derived from:

∫
Fp1' ( V ) = F 1 ( x ) ψ p′ ( x ) d 2 x,
G
(6.6)

BK-TandF-KAZANSKIY_9781032572949-230642-Chp06.indd 160 9/26/23 5:04 PM

 Optical Fiber Sensors 161

Fp1′ ( V ) = ∑T
p
pp ′ ( V ) Fp . (6.7)

Considering that the Parseval equality holds for orthogonal bases, the beam
power at the fiber input and output is defined by series:

∑F ∑F
2 2
E= p , E′(V) = 1
p′ (V) . (6.8)
p p′

The majority of familiar fiber-optic sensors operate by measuring the total output
beam energy E′(V) as a function of perturbations V. From (6.8), the value of E′(V) is
seen to be averaged over the individual-mode power |Fp′(V)|2 for the entire ensemble
of the excited modes with numbers p′. A converter proposed in [3, 4] performs selec-
2
tive measurements of the power Fp1′ ( V) of individual modes. We note that in a special
case of single-mode excitation, the Fp coefficient is non-zero only at one value of p.
We term the quantity |Tp′p(V)|2 a mode excitation coefficient because it determines the
power of the fiber mode with the number p′ excited by the mode with the number p.

6.2.2 Analysis of Mode Coupling Due to Periodic

Microbendings of a Graded-Index Fiber
The physical principle behind the design of an optical fiber converter is mode coupling
due to fiber deformations, which leads to the energy being redistributed between the
modes and partially transferred to non-propagating modes [15]. This causes changes
in the energy of individual modes and a decrease in the total light flux in the fiber.
Many well-known converter designs are based on measuring the total light flux. But
this effect is secondary and as such is essentially less sensitive (four to five times for a
gradient fiber) to fiber deformations compared to when measuring individual modes.
The sensitivity of the total light flux can be increased by using converter designs in
which the fiber has multiple periodic bendings with a period close to the light wave-
length, causing resonance phenomena contributing to the increase of losses.
It was in an optical fiber converter with microbendings that the mode selection
was first demonstrated [3]. For this purpose, the device for transverse mode excita-
tion and analysis was complemented with a deformation device, as shown in a sche-
matic diagram in Figure 6.1.
The deformation device was put 0.3 m away from the input end of a 1.5-m long
fiber. The periodic microbendings were implemented via pulling them through a pair
of profiled plates, with the fiber-facing plates’ surface relief forming triangular teeth
of period Λ = 1.5 mm and the number of periods being k = 5. The bending radii were
varied through applying an external force to the upper plate and controlled by means

FIGURE 6.1 A schematic diagram of a fiber-optic converter.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp06.indd 161 9/26/23 5:04 PM

 162 Photonics Elements for Sensing and Optical Conversions

FIGURE 6.2 Mode power versus microbending radii (solid lines – theory, dashed lines –
experiment).

of a microscope and an eyepiece micrometer. Experimentally derived curves for the

relative mode powers:
2
ξ op (h)
Pop ( h ) = 2 , (6.13)
ξ oo (h) h = 0

are shown as dashed lines in Figure 6.2.

From Figure 6.2, the curves are seen to behave in agreement with the theory, but
if the deformations exceed 5–6 μm, the rate of mode power variations is somewhat
higher (at lower deformations) than theoretical predictions. Arguably, here, reasons
are the same as those leading to the variation of mode excitation conditions, because
these processes have a similar mechanism.
For comparison with the earlier measured characteristics of the converters under
analysis, the integrated light flux was measured as a function of the microbending
radius. The corresponding plot is presented in Figure 6.3 in comparison with that for
the principal mode power.
From Figure 6.3, the sensitivity for the principal mode is seen to be twice as high
as that for the total light flux. Note that although the curve for the total light flux fea-
tures a better linearity, after the proper correction the advantage will not be essential.

6.2.3 Study of the Output Mode Power for a Graded-Index

Fiber against the Microbending Magnitude
Along with being complex in design, the just-discussed converter with multiple peri-
odic fiber bendings is rather massive, which deteriorates its mechanical frequency

BK-TandF-KAZANSKIY_9781032572949-230642-Chp06.indd 162 9/26/23 5:04 PM

 Optical Fiber Sensors 163

FIGURE 6.3 Variations in the total light flux and principal mode power versus the micro-
bending radius.

characteristics. Besides, the entire range of measurable deformations falls within

10–15 μm. On the other hand, mass-produced optical fibers endure 100-μm-deep
bendings without being damaged. In [4], a simpler mechanical design of the defor-
mation device was proposed, producing a single 50- to 60-μm-deep microbending.
The device has the advantage of lighter weight and smaller dimensions, which also
results in better frequency characteristics. At the same time, thanks to mode filter-
ing, the sensitivity of the converter is close to that of “multi-bending” converters
measuring the total flux. Moreover, the proposed design [4] enables the converter
sensitivity to be readjusted in real time by measuring the energy of different-order
modes. Corresponding experiments with multimode gradient optical fibers were
described in [4], and a feasibility of the implementation of such types of converters
was demonstrated. However, the approach has a drawback, namely, the power of
higher-order modes was found to have a fairly large constant component (~0.3–0.4
of maximum) independent of the bending magnitude. At the same time, in the theo-
retical model of an ideal parabolic profile such a phenomenon is not expected to
occur. Hence, we may infer that its explanation should be sought for in the deviation
of the refractive index profile from the theoretical model and the large number of
higher-order modes in such fibers, leading to noise, in addition to non-ideal DOEs
utilized. On the other hand, higher-selective excitation and more accurate analysis of
transverse modes can be achieved in a few-mode fiber with stepped refractive index
profile [15]. Experimental measurements of the mode power in a stepped-index fiber
against the microbending depth have been reported [16, 17].
For a single, specific-depth bending to be realized, a dedicated deformation
device was designed and fabricated, as schematically shown in Figure 6.4.
The device comprises a base (1) (Figure 6.4), mounted on which are a lever
(2) with a return spring (3), a pressure plate (4), and an adjustable support unit.
The lever axis rests on prisms (5). Lever movements are controlled by means of

BK-TandF-KAZANSKIY_9781032572949-230642-Chp06.indd 163 9/26/23 5:04 PM

 164 Photonics Elements for Sensing and Optical Conversions

FIGURE 6.4 An outline of the deformation device.

an indicator (6), which sits on the same platform (7) as a pressure bolt (8). The
adjustable support unit (Figure 6.4) comprises two supports (9) for springs (10) that
thrust supports (11) for the fiber under study. The gap between the supports (11) is
adjusted with a taper-shank bolt (12), which passes through them limiting the gap.
For a comparison with earlier reported results [3], the inter-support gap was chosen
to be ~1 mm. The fiber is being controllably deformed by rotating the pressure bolt,
which displaces the lever and, according to a lever arm ratio of 1:7.5, the pressure
plate. In the experiment, a commercial 2-m-long single mode fiber (SMF) was used.
Some differences in techniques for analyzing the transverse mode powers are due to
the use in the experiment of a DOE for laser mode generation (MODAN), intended
for a larger number of modes. These phase elements are described in detail in [15].
The mode content of laser light at the fiber output can be analyzed using multi-
order and multi-channel diffractive optical elements (DOEs) [18–20]. The analysis
is easiest to conduct with the focal length of microlens M2 set equal to 13.9 mm and
the focal length of the Fourier-transform lens at 300 mm. With the principal mode
radius being about five times larger compared to other modes, for a desired scale
picture to be obtained in the CCD camera plane, the phase element should be put
into a diverging beam, ensuring an ~1-m distance between the Fourier-transform
objective and the CCD camera. In the CCD camera field of view, about nine differ-
ent modes are observed, including all modes of the fiber under analysis. As we show
later in the chapter, such a scale provides an acceptable (noise level) systematic error
of measurements. A distinguishing feature of the filter that is utilized is the absence
of the principal mode. However, the adjustment criterion of the minimal intensity at
the centers of all diffraction orders still holds. Figure 6.5 shows an intensity pattern

BK-TandF-KAZANSKIY_9781032572949-230642-Chp06.indd 164 9/26/23 5:04 PM

 Optical Fiber Sensors 165

FIGURE 6.5 Intensity pattern at the output plane of a Fourier stage when coupling the
principal mode to the phase filter.

at the output plane of a Fourier stage with a spatial filter (correlation field) when
coupling the principal mode to the fiber.
The pattern in Figure 6.5 needs to be explained. The Fourier transform of the
binary phase element employed in the scheme is centrally symmetric relative to the
zero order, which is clearly seen in Figure 6.5. The frame in the pattern is a bit
displaced to show a larger number of modes on one side. We note that the image in
Figure 6.5 can be processed using software described in [17]. The result of process-
ing an experimental image is depicted in Figure 6.6.
Table 6.1 connects the diffraction order numbers with particular modes.
The next step was measuring the mode power against the fiber deformation
magnitude. Figure 6.7 depicts a correlation field upon the excitation of the mode
LP11.

FIGURE 6.6 The result of processing the intensity distribution aimed at finding the diffrac-
tion order centers.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp06.indd 165 9/26/23 5:04 PM

 166 Photonics Elements for Sensing and Optical Conversions

TABLE 6.1
Diffraction Order Numbers versus Mode Indices
Order Numbers 1 2 3 4 6 7 8 9 10 11
Modes LP03 LP12 LP02 LP11 LP02 LP11 LP12 LP03 LP21 LP31

FIGURE 6.7 Intensity distribution at the output plane of the Fourier stage when coupling
the mode LP11 to the phase filter.

A corresponding correlation peak is clearly seen, with no intensity peaks being

observed in other diffraction orders. From Figure 6.8, the correlation peak is seen
to occupy 10–15 pixels on either axis, allowing it to be measured with an error not
higher than the noise level.
When the fiber is being deformed, the energy is redistributed between the modes,
leading to changes in the correlation peak. Simultaneously, the intensity distribution
between diffraction orders ceases to be symmetric. Figure 6.9 depicts a correlation
field for a bending depth of 60 μm, with the rest parameters of the optical setup being
the same as in Figure 6.7.
Shown in Figure 6.10 is an experimental curve of the LP11 mode power against the
fiber microbending depth, which is seen to be nonlinear. The permanent component
is seen to be as low as 0.17, unlike results previously derived for a GRIN fiber [3,
4] where it was found to be 0.3–0.4, with the power stabilized at the same bending

FIGURE 6.8 Intensity distribution in the correlation peak region.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp06.indd 166 9/26/23 5:04 PM

 Optical Fiber Sensors 167

FIGURE 6.9 Diffraction pattern for a 60-μm deep bending.

depth of ~75 μm. The latter result agrees better with theory, enabling the dynamic
range to be extended and the accuracy of the bending-based converter enhanced.
In a similar way to Figure 6.3, Figure 6.11 plots the power of the principal mode
and, for comparison, the total light flux against the microbending depth.
The curves in Figure 6.11 are seen to be somewhat different from those for a multi-
bending GRIN-fiber converter of Figure 6.3. The curves in Figure 6.11 have similar-type
nonlinearity, with a twofold increase in sensitivity being observed over the entire range
of the analyzed deformations. The two curves have a larger proportion of the perma-
nent component when compared with the multi-bending converter of Figure 6.3, which
is apparently due to the fiber’s single bending. Besides, the sensitivity over the total
light flux is more than twice as high as that for a single-bending GRIN-fiber converter
[4] (with the total flux decreasing to a level of 0.8 at half as large fiber deformation),

FIGURE 6.10 Power of the LP11 mode against the fiber bending.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp06.indd 167 9/26/23 5:04 PM

 168 Photonics Elements for Sensing and Optical Conversions

FIGURE 6.11 Variations in the total light flux and the power of the principal mode as a
function of the microbending magnitude in a stepped-index optical fiber.

whereas the sensitivity for the principal mode increases considerably less, by ~20%.
The explanation can be found in different core diameters, with the stepped-index fiber
core being significantly thinner, meaning that as a result of a smaller bending, light
will leave the stepped-index fiber sooner than it will the multimode GRIN fiber. Hence,
in such converters it is preferable to use higher-order modes for which the proportion
of the permanent component is essentially less (see Figure 6.10).
Thus, we have derived experimental curves for the dependence of the LP-modes
power and the total light flux on the magnitude of a stepped-index fiber bending.
A twofold decrease in the mode power permanent component has been attained,
enabling a dynamic range of the converter to be extended.

6.3 SPECTRAL OPTICAL FIBER SENSORS BASED ON

PERIODICALLY MICROSTRUCTURED FIBERS
6.3.1 Approaches to Measuring Physical Quantities
External physical influences introduce perturbations in the surrounding medium of
periodic fiber-optic microstructures, leading to a shift in the Bragg resonance posi-
tion in the optical reflection and transmission spectra. This effect underlies the use of
fiber Bragg gratings (FBG) as optical fiber sensors. A variety of methods for measur-
ing the Bragg wavelength shift have been proposed.

6.3.1.1 Measuring the Intensity

One method for measuring the physical quantities of interest utilizes a narrowband
laser and a photodetector, which provides higher processing speed. In this case, the
Bragg wavelength shift in the grating is transformed into intensity variations of the
optical signal on the photodetector.
The magnitude of the external force can be evaluated from a magnitude of the
frequency shift for the reflection spectrum or a shift of the intensity curve dip for the

BK-TandF-KAZANSKIY_9781032572949-230642-Chp06.indd 168 9/26/23 5:04 PM

 Optical Fiber Sensors 169

FIGURE 6.12 Spectrum analyzer for the reflection spectrum of a Bragg grating.

transmission spectrum. Accordingly, spectral devices need to be used for the spec-
tral analysis. A schematic diagram of a bandpass-filter-aided spectrum analyzer for
the reflection spectrum of a Bragg grating is depicted in Figure 6.12.
In this setup, one part of light reflected from the Bragg grating serves as a refer-
ence signal, being fed to the photodetector. The other part of light comes to the sec-
ond photodetector through a bandpass filter, with the filter transmission depending
on the wavelength. With a varying external force exerted on the grating structure, the
reference signal intensity at the first photodetector remains unchanged; meanwhile
the intensity of the light coming to the second photodetector varies due to changing
conditions of the Bragg reflection in the fiber grating. A signal processing unit cal-
culates the ratio of the intensities coming to the photodetectors, producing the infor-
mation on the force magnitude in a desired form. Disadvantages of this scheme stem
from the mechanical instability of the optical setup and instability of characteristics
of the wideband light source.
The bandpass filters may be in the form of scanned Fabry–Perot interferometers,
acoustic-optical filters, and filters based on tunable Bragg gratings.
In the scheme with a scanned interferometer in Figure 6.13, the reflected light
comes to the input of a filter that has the near-unit transmission on resonance

FIGURE 6.13 Schematic of a scanned Fabry–Perot interferometer.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp06.indd 169 9/26/23 5:04 PM

 170 Photonics Elements for Sensing and Optical Conversions

FIGURE 6.14 Schematic of a tunable Bragg grating filter.

wavelengths. In the FP interferometer, the position of the maximum transmission

depends on the inter-mirror distance. With one of the mirrors being mounted on a
piezoelectric substrate, its position is adjusted by varying the command voltage. With
a sawtooth command voltage, the interferometer resonance frequency is chirped
during the forward stroke, enabling the reflected signal spectrum to be obtained
within a single sawtooth period.
The fiber-optic Bragg grating may also be used as a tunable spectral filter
(Figure 6.14). In this case, the wideband radiation is fed to a measuring Bragg grat-
ing, after which the reflected radiation comes to the second grating. The optical fiber
and the second grating are located on a piezoceramic substrate.
In this grating, the resonance wavelength is specified by the command voltage
applied across the piezoelectric element electrodes from a command voltage genera-
tor (CFG). If both gratings have the same resonance wavelengths, after reflection at
the second grating the signal comes to the photodetector. In this way, it is also pos-
sible to analyze spectral characteristics of the radiation reflected at the first grating.

6.3.1.2 Reflectometry Analysis

In the practical use of fiber-optic Bragg gratings (FOBG), major shortcomings are
linked with their complex design and costly equipment for spectral measurements of
the wavelength shift. The majority of spectral systems provide a redundant accuracy
when measuring physical influences. Because of this, simplified measurement setups
can be utilized, where some physical parameters are measured with lower accuracy.
A simplified scheme can be implemented on the principle of detecting the reflected
signal amplitude at a fiber grating.
An optical time-domain reflectometry (OTDR) method enables distributed mea-
surements to be performed over the entire length of the fiber-optic path. Equipment for
optical reflectometry comprises a pulse generator, a sensitive photodiode, and a signal
processing unit. The parameter under measurement is the power of the backscattered
light due to effects of Rayleigh, Mandelstam–Brillouin, or Raman scattering. An opti-
cal reflectometer measures the power of backscattered light against distance.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp06.indd 170 9/26/23 5:04 PM

 Optical Fiber Sensors 171

FIGURE 6.15 Optical setup of a CCD spectrometer.

6.3.1.3 Spectral Analysis

Measurements of reflection/transmission spectra in fiber-optic structures using a
wideband light source and a spectrum analyzer are most comprehensive. The method
is insensitive to possible losses in an optical channel. Disadvantages include the high
cost of equipment and its limited performance.
With this method, it is possible to determine in more detail spectral character-
istics linked with mode interference, energy exchange, polarization effects, and
absorption.
The scheme of an optical spectrum analyzer comprises a monochromator, a dif-
fraction grating, and a CCD array (Figure 6.15). After being reflected at the Bragg
grating, the light goes to the diffraction grating, which sends the reflected light to the
CCD array. At the diffraction grating, the angle of diffraction depends on the inci-
dent wavelength, meaning that different wavelengths come to different CCD array
areas. Considering these properties, CCD spectrometers can be utilized for detect-
ing a large number of fiber-grating-based sensors. With the sensitivity threshold of
CCD arrays being essentially higher compared to photodiodes, the former can detect
lower-energy light.
For several fiber-optic channels to be sensed, a spectral arrangement with a
curved diffraction grating can be used. Radiation from different fiber-optic chan-
nels hits different CCD array columns, thus enabling a two-dimensional analysis
(Figure 6.16).

6.3.2 Optical Fiber Sensors

Modes traveling in an optical fiber cladding are used in a wide class of sensors
for measuring physical quantities. For a cladding mode to be excited as a result of
matching with the principal mode, the fiber needs to have an inhomogeneity, which
may be introduced by a fiber Bragg grating, a long-period fiber grating, or coupling
fibers with unmatched refractive index profiles.
Sensors based on in-fiber Bragg gratings have been widely used in both quasi-
distributed and point sensing [21–26]. Spectral characteristics of Bragg gratings,
such as the dynamic range and sensitivity to physical influences, essentially depend
on the fiber material, doping agent concentration, and techniques of fabricating the
grating structure and processing the fiber core/cladding material.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp06.indd 171 9/26/23 5:04 PM

 172 Photonics Elements for Sensing and Optical Conversions

FIGURE 6.16 Scheme of a two-dimensional spectral analysis.

A sensor on the basis of a sapphire Bragg grating implemented using a Talbot

interferometer was described in [21] and utilized for high temperature measure-
ments. The sensor benefits an extended dynamic range of measured temperatures
up to 1,900°С.
The said structures can be synthesized in a standard single-mode fiber using fem-
tosecond laser pulses [22] and remain stable at temperatures of up to 1,000°С.
In the standard gratings, the temperature sensitivity is 8 pm/°C. In [23], the tem-
perature sensitivity was increased by placing the fiber grating into a thermoplastic
cylinder with high thermal expansion coefficient, which was fabricated using 3D
printing. The exposure to high temperatures leads to a small decrease in the reflec-
tivity and a redshift of the resonance wavelength owing to the material deformation.
The temperature sensitivity was reported to be 139 pm/°С.
The sensitivity of standard fiber Bragg gratings to the on-axis deformation is
about 0.44 pm/με and can be improved by modifying the design of the optic fiber
Bragg grating (OFBG)-aided sensors. For instance, by etching the grating cladding,
the deformation sensitivity can be increased to 5.5 pm/με [24].
Unlike in-fiber Bragg gratings, OFBG-based superstructures [27] have higher
sensitivity to mechanical and temperature influences. The fiber optic superstruc-
tures are long-period gratings, which are formed by alternating Bragg grating sec-
tions [27–36]. However, such long-period gratings have a limited sensitivity because
changes in the cladding properties due to external influences are compensated for
by similar changes in the fiber core. The deformation sensitivity of photo-induced
superstructures is about 1.7 and 0.44 pm/με, whereas the temperature sensitivity is
9 and 64 pm/°С [29]. A structure that has the axial-strain sensitivity of 2.8 and 1.06
pm/με and temperature sensitivity of 80 and 10 pm/°С has also been described [30].
The effect of microbending strain on the Bragg grating may be used for multi-
parameter measurements of external physical factors like pressure and temperature.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp06.indd 172 9/26/23 5:04 PM

 Optical Fiber Sensors 173

In this case, a shift in the Bragg-grating-aided resonance peaks compensates for

thermal effects that occur when applying pressure to an optical fiber.
In a superstructure formed by acoustically induced microbendings in a homoge-
neous Bragg grating [37, 38], the reflection spectrum has more resonance side-lobes
with an increased distance between them. Longitudinal acoustic vibrations produce
radial fiber deformations in the form of periodic diameter widening and narrowing.
This acoustic-optical effect causes changes in the effective refractive index for the
principal and the opposite modes, also changing the grating period [37–39]. For low-
frequency vibrations, bending acoustic modes are excited, with longitudinal modes
prevailing in the high-frequency range. A noticeable decrease in the reflectivity for
the main peak and an increase in the amplitude of the side-lobes are the result of
coupling between the fiber principal mode and higher-order cladding modes [40].
When classical Bragg gratings are exposed to high temperatures, short-period
fiber structures get erased. Because of this, prior to generating thermo-induced
superstructures [41–44], segments of an optical fiber are first preheated, after which
Bragg gratings are written. Besides, phase discontinuity in the Bragg grating can be
generated by means of an electric arc discharge, which will partly erase the recorded
periodic structure [41].
A technique for producing a helical superstructure by means of an electric arc
and UV light has been reported [42]. Thanks to an increased twist period of the
long-period grating section at the unchanged period of the Bragg grating, the inter-
peak distance in the reflection spectrum is decreased. If the Bragg grating period is
increased but the helical structure period is fixed, the reflection spectrum peaks are
redshifted with the inter-peak intervals remaining unchanged. The shift resonance
wavelength of the helical superstructure was measured with an accuracy of 64.6
nm·mm·rad−1 in the interval from –0.075 to +0.075 rad·mm−1 and was found to be
9.36 nm. Unlike the shift of a wideband dip, FBG resonances remain practically
unchanged upon twisting. The fundamental mode of the fiber is practically insensi-
tive to the helical structure twisting. The temperature sensitivity is 60.51 pm/°C for
a broadband dip generated by long-period sections, being 10.12 pm/°C for a narrow-
band peak that the Brag grating generates for temperatures ranging from 26°C to
75°C. In this case, the FBG spectral response allows obtaining information on tem-
perature-related effects.
A superstructure has been reported to be fabricated by heating the Bragg grat-
ing surface with an IR laser light [44]. In a two-stage procedure, a Bragg grating is
first written with 800-nm 120-fs laser pulses passing through a phase mask, before
using a 0.5-W CO2 laser for the pointwise heating of the surface of a short-period
structure. In the process, one fiber end remains rigid and the other has a 5-g weight
attached to it to exclude the appearance of microbendings in the course of heating.
When increasing the length of the long-period grating section, intervals between the
reflection spectrum peaks also increase. When increasing the duration of the Bragg
grating heating, the intensity of reflection peaks decreases.
Another type of fiber-optic superstructure is realized by etching the Bragg grating
cladding. The loss bands are highly sensitive to twisting because the accompanying
strain is higher on the etched sections compared to the non-etched sections. Stretching
a corrugated fiber grating leads to a higher attenuation in the transmission spectrum.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp06.indd 173 9/26/23 5:04 PM

 174 Photonics Elements for Sensing and Optical Conversions

However, a lower mechanical strength of etched structures imposes limits on the defor-
mation measurements. If the fiber Bragg grating has a corrugated surface [45], changes
in the propagation and coupling of the cladding mode will be insignificant, weakly
affecting the reflection spectrum of the fiber Bragg grating as the principal mode is
mainly confined to the fiber core. When applying an axial strain to the corrugated
superstructure, the resonance wavelength in the reflection spectrum is redshifted. With
increasing tension stress, the side-lobes in the reflection spectrum increase, which can
be explained by the interaction of the principal mode with the cladding modes due to a
greater stretching of the etched segments compared to the non-etched ones.
However, alongside fiber-optic superstructures, sensing elements with a simpler
design may also be utilized. For instance, we may recall a fiber Mach–Zehnder inter-
ferometer built by connecting conventional single-mode fibers with nonmatching
profiles. The insert may be in the form of a multimode fiber segment, a W-shaped
fiber, or a photonic crystal fiber segment.
In [46], a fiber interferometer with an insert of a multimode stepped-index fiber
with periodic twisted inhomogeneities on either side of the input and output ends
was proposed. In the experimental bending tests in the range from 0 to 1.739 m−1 the
sensitivity was found to be 2.42 nm/m−1. In the temperature tests in the range from
21 to 120°C, the sensitivity was shown to be 10 pm/°C.
In [47], an interferometer with a two-layered optical fiber segment was proposed.
Inter-mode coupling between the fundamental mode and cladding modes at the
interface of fibers SM630 and SMF-28 with different refractive indices was ana-
lyzed. Such structures are formed by welding together standard single-mode fibers
SMF-28. After passing through the first welded joint, the energy from the core is
partly redistributed to the insert-fiber cladding, leading to the excitation of cladding
modes. When the energy is coupled into the core of the third fiber at the second joint,
there is interference between the fundamental mode and cladding modes. For such a
structure, the sensitivity of the transmission spectrum to stretching and temperature
variations is, respectively, 0.3 pm/με and 10 pm/°C.
Fiber-optic interferometers on the basis of photonic crystal fiber segments are also
feasible. Such sensors are fabricated by welding the end of a photonic crystal fiber
with standard single-mode optical fibers. Another approach utilizes the collapse of
photonic crystal holes at two sections. Such a sensor with an insert of the dual-core
photonic crystal fiber has been described [48]. In the transmission spectrum of a
composite fiber structure, multiple interference intensity peaks were observed. The
temperature and deformation sensitivity was 25 pm/°C and 1.26 pm/με.

REFERENCES
1. Fang, Z., Chin, K., Qu, R., and Cai, H. 2012. Fundamentals of optical fiber sensors.
Hoboken, NJ: John Wiley and Sons Ltd. ISBN: 978-0-470-57540-6.
2. Soifer, V. A., and Golub, M. A. 1994. Laser beam mode selection by computer gener-
ated holograms. Boca Raton, FL: CRC Press. ISBN: 978-0-8493-2476-5.
3. Garitchev, V. P., Golub, M. A., Karpeev, S. V., Krivoshlykov, S. G., Petrov, N. I., Sissakian,
I. N., Soifer, V. A., Haubenreisser, W., Jahn, J.-U., and Willsch, R. 1985. Experimental
investigation of mode coupling in a multimode graded-index fiber, caused by periodic
microbends using computer-generated spatial filters. Opt. Commun. 55(6):403–5.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp06.indd 174 9/26/23 5:04 PM

 Optical Fiber Sensors 175

4. Golub, M. A., Sisakyan, I. N., Soifer, V. A., and Uvarov, G. V. 1991. Mode-selective
fiber sensor operating with computer generated optical elements. Proc. SPIE 1572:101.
https://doi.org/10.1117/12.50022.
5. Kotlyar, V. V., Soifer, V. A., and Khonina, S. N. 1998. Rotation of multimodal Gauss-
Laguerre light beams in free space and in a fiber. Opt. Lasers Eng. 29(4–5):343–50.
https://doi.org/10.1016/S0143-8166(97)00121-8.
6. Khonina, S. N., Kotlyar, V. V., and Soifer, V. A. 1998. Diffraction optical elements
matched to the Gauss-Laguerre modes. Opt. Spectrosc. 85(4):636–44.
7. Khonina, S. N., Kotlyar, V. V., and Soifer, V. A. 1999. Self-reproduction of multimode
Hermite-Gaussian beams. Tech. Phys. Lett. 25(6):489–91. https://doi.org/10.1134/1.1262525.
8. Kharitonov, S. I., and Khonina, S. N. 2018. Conversion of a conical wave with circular
polarization into a vortex cylindrically polarized beam in a metal waveguide. Comput.
Opt. 42(2):197–211. https://doi.org/10.18287/2412-6179-2018-42-2-197-211.
9. Khonina, S. N., Striletz, A. S., Kovalev, A. A., and Kotlyar, V. V. 2010. Propagation
of laser vortex beams in a parabolic optical fiber. Proc. SPIE 7523:75230B. https://doi.
org/10.1117/12.854883.
10. Kirilenko, M. S., Mossoulina, O. A., and Khonina, S. N. 2016. Propagation of vortex
eigenfunctions of bounded Hankel transform in a parabolic fiber. IEEE Proc. Int. Conf.
on Laser Optics:R4–32. https://doi.org/10.1109/LO.2016.7549787.
11. Monin, E. O., Ustinov, A. V., and Khonina, S. N. 2018. Propagation modeling of vortex
generalized Airy beams in parabolic fiber. Proc. Progress in Electromagnetics Research
Symposium (PIERS) F134321:583–9. https://doi.org/10.1109/PIERS.2017.8261809.
12. Khonina, S. N., and Karpeev, S. V. 2004. Excitation and detection of angular harmon-
ics in an optical fiber using DOE. Comput. Opt. 26:16–26.
13. Karpeev, S. V., and Khonina, S. N. 2007. Experimental excitation and detection of
angular harmonics in a step-index optical fiber. Opt. Mem. Neural Netw. 16(4):295–
300. https://doi.org/10.3103/S1060992X07040133.
14. Savelyev, D. A., and Khonina, S. N. 2017. Investigation of vortex laser beam
injection into an optical fiber. J. Phys. Conf. Ser. 917:062035. https://doi.
org/10.1088/1742-6596/917/6/062035.
15. Karpeev, S. V., Pavelyev, V. S., and Kazanskiy, N. L. 2005. Steplike fiber modes excite-
ment with binary phase DOEs. Opt. Mem. Neural Netw. 14(4):223–8.
16. Karpeev, S. V., Pavelyev, V. S., and Khonina, S. N. 2005. High-effective fiber sensors based
on transversal mode selection. Proc. SPIE 5854:163–9. https://doi.org/10.1117/12.634603.
17. Karpeev, S. V., Pavelyev, V. S., Khonina, S. N., Kazanskiy, N. L., Gavrilov, A. V., and
Eropolov, V. A. 2007. Fibre sensors based on transverse mode selection. J. Mod. Opt.
54(6):833–44. https://doi.org/10.1080/09500340601066125.
18. Khonina, S. N., Skidanov, R. V., Kotlyar, V. V., Jefimovs, K., and Turunen, J. 2003.
Phase diffractive filter to analyze an output step-index fiber beam. Opt. Mem. Neural
Netw. 12(4):317–24. https://doi.org/10.1117/12.509468.
19. Lyubopytov, V. S., Tlyavlin, A. Z., Sultanov, A. K., Bagmanov, V. K., Khonina, S. N.,
Karpeev, S. V., and Kazanskiy, N. L. 2013. Mathematical model of completely optical
system for detection of mode propagation parameters in an optical fiber with few-mode
operation for adaptive compensation of mode coupling. Comput. Opt. 37(3):352–9.
20. Khonina, S. N., Kazanskiy, N. L., and Soifer, V. A. 2012. Optical vortices in a fiber:
Mode division multiplexing and multimode self-imaging, Ch. 15. In Recent progress
in optical fiber research, eds. M. Yasin, S. W. Harun, and H. Arof. Croatia: INTECH
Publisher. https://doi.org/10.5772/28067.
21. Anania, S., Unnikrishnan, A., Aparna, A., Parvathi, G. R., Baby Sreeja, S. D., and
Mohan, P. 2018. Analytical study of FBG spectrum for temperature sensing appli-
cations. Proc. 2nd Int. Conf. on Inventive Communication and Computational
Technologies (ICICCT 2018):1109–13. https://doi.org/10.1109/ICICCT.2018.8473283.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp06.indd 175 9/26/23 5:04 PM

 176 Photonics Elements for Sensing and Optical Conversions

22. Habisreuther, T., Elsmann, T., Pan, Z., Graf, A., Willsch, R., and Schmidt, M.
2015. Sapphire fiber Bragg gratings for high temperature and dynamic temperature
diagnostics. Appl. Therm. Eng. 91:860–5. https://doi.org/10.1016/j.applthermaleng.
2015.08.096.
23. Leal-Junior, A., Casas, J., Marques, C., José Pontes, M., and Frizera, A. 2018.
Application of additive layer manufacturing technique on the development of high
sensitive fiber Bragg grating temperature sensors. Sensors. 18(12):4120. https://doi.
org/10.3390/s18124120.
24. Sridevi, S., Vasu, K. S., Asokan, S., and Sood, A. K. 2016. Enhanced strain and tem-
perature sensing by reduced graphene oxide coated etched fiber Bragg gratings. Opt.
Lett. 41(11):2604–07. https://doi.org/10.1364/OL.41.002604.
25. Yang, S., Hu, D., and Wang, A. 2017. Point-by-point fabrication and characterization
of sapphire fiber Bragg gratings. Opt. Lett. 42(20):4219–22. https://doi.org/10.1364/
OL.42.004219.
26. Zhang, C., Yang, Y., Wang, C., Liao, C., and Wang, Y. 2016. Femtosecond-laser-
inscribed sampled fiber Bragg grating with ultrahigh thermal stability. Opt. Express.
24(4):3981–8. https://doi.org/10.1364/OE.24.003981.
27. Tan, R. X., Yap, K., Tan, Y. C., Tjin, S. C., Ibsen, M., Yong, K. T., and Lai, W. J. 2018.
Functionalized fiber end superstructure fiber Bragg grating refractive index sensor for
heavy metal ion detection. Sensors. 18(6):1821. https://doi.org/10.3390/s18061821.
28. Chi, H., Tao, X.-M., and Yang, D. 2001. Simultaneous measurement of axial strain, tem-
perature, and transverse load by a superstructure fiber grating. Opt. Lett. 26(24):1949–
51. https://doi.org/10.1364/OL.26.001949.
29. Sengupta, S., Ghorai, S. K., and Biswas, P. 2016. Design of superstructure fiber Bragg
grating with efficient mode coupling for simultaneous strain and temperature measure-
ment with low cross-sensitivity. IEEE Sens. J. 16(22):7941–9. https://doi.org/10.1109/
JSEN.2016.2611002.
30. Guan, B.-O., Tam, H.-Y., Tao, X.-M., and Dong, X.-Y. 2000. Simultaneous strain and
temperature measurement using superstructure fiber Bragg grating. IEEE Photonics
Technol. Lett. 12(6):675–7. https://doi.org/10.1109/68.849081.
31. Zhang, A.-P., Guan, B.-O., Tao, X.-M., and Tam, H.-Y. 2002. Mode couplings in super-
structure fiber Bragg gratings. IEEE Photonics Technol. Lett. 14(4):489–91. https://doi.
org/10.1109/68.992587.
32. Zeng, X., and Liang, K. 2011. Analytic solutions for spectral properties of superstruc-
ture, Gaussian-apodized and phase shift gratings with short- or long-period. Opt.
Express 19(23):22797–808. https://doi.org/10.1364/OE.19.022797.
33. He, Y.-J., Hung, W.-C., and Lai, Z.-P. 2016. Using finite element and eigenmode expan-
sion methods to investigate the periodic and spectral characteristic of superstructure
fiber Bragg gratings. Sensors. 16(192):192. https://doi.org/10.3390/s16020192.
34. Rodriguez-Cobo, L., Cobo, A., and Lopez-Higuera, J. M. 2012. Sampled fiber Bragg
grating spectral synthesis. Opt. Express. 20(20):22429–41. https://doi.org/10.1364/
OE.20.022429.
35. Luo, Z., Ye, C., Cai, Z., Dai, X., Kang, Y., and Xu, H. 2007. Numerical analysis and
optimization of optical spectral characteristics of fiber Bragg gratings modulated
by a transverse acoustic wave. Appl. Opt. 46(28):6959–65. https://doi.org/10.1364/
ao.46.006959.
36. Idrisov, R. F. 2020. A review on fiber-optical superstructures based on in-fiber bragg
grids. Appl. Phys. Math. 3:15–27.
37. Pohl, A., Oliveira, R., Da Silva, R., Marques, C., de Tarso, P., Cook, K., Canning,
J., and Nogueira, R. 2013. Advances and new applications using the acousto-optic
effect in optical fibers. Photonic Sens. 3(1):1–25. https://doi.org/10.1007/s13320-013-
0100-0.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp06.indd 176 9/26/23 5:04 PM

 Optical Fiber Sensors 177

38. Oliveira, R. A., Marques, C. A. F., Mayer, C. E. N., Pereira, J. T., Nogueira, R. N.,
and Pohl, A. A. P. 2009. Single device for excitation of both flexural and longitudi-
nal acousto-optic effects in fiber Bragg gratings. SBMO/IEEE MTT-S International
Microwave & Optoelectronics Conference (IMOC):546–9. https://doi.org/10.1109/
IMOC.2009.5427526.
39. Silva, R., Tiess, T., Becker, M., Eschrich, T., Rothhardt, M., Jäger, M., Pohl, A., and
Bartelt, H. 2015. All-fiber 10 MHz acousto-optic modulator of a fiber Bragg grat-
ing at 1060 nm wavelength. Opt. Express. 23(20):25972–8. https://doi.org/10.1364/
OE.23.025972.
40. Gao, Z., Chang, P., Huang, L., Gao, F., Mao, D., Zhang, W., and Mei, T. 2019. All-fiber
frequency shifter consists of fiber Bragg grating modulated via acoustic flexural wave
for optical heterodyne measurement. Opt. Lett. 44(15):3725–8. https://doi.org/10.1364/
OL.44.003725.
41. Ouyang, Y., Liu, J., Xu, X., Zhao, Y., and Zhou, A. 2018. Phase-shifted eccentric core
fiber Bragg grating fabricated by electric arc discharge for directional bending mea-
surement. Sensors 18(6):1168. https://doi.org/10.3390/s18041168.
42. Li, J., Chen, G., Ma, P., Sun, L.-P., Wu, C., and Guan, B. 2018. Sampled Bragg gratings
formed in helically twisted fibers and their potential application for the simultaneous
measurement of mechanical torsion and temperature. Opt. Express. 26(10):12903–11.
https://doi.org/10.1364/OE.26.012903.
43. Guo, G. 2019. Superstructure fiber Bragg gratings for simultaneous temperature and
strain measurement. Optik. 182:331–40. https://doi.org/10.1016/j.ijleo.2019.01.011.
44. Fang, X., He, X. Y., Liao, C. R., Yang, M., Wang, D. N., and Wang, Y. 2010. A new
method for sampled fiber Bragg grating fabrication by use of both femtosecond laser
and CO2 laser. Opt. Express. 18(3):2646–54. https://doi.org/10.1364/OE.18.002646.
45. Lin, C., Chern, G., and Wang, L. 2001. Periodical corrugated structure for form-
ing sampled fiber Bragg grating and long-period fiber grating with tunable coupling
strength. J. Lightw. Technol. 19(8):1212–20. https://doi.org/10.1109/50.939803.
46. Tian, K., Xin, Y., Yang, W., Geng, T., Ren, J., Fan, Y.-X., and Farrell, G. 2017. A curva-
ture sensor based on twisted single-mode–multimode–single-mode hybrid optical fiber
structure. J. Lightw. Technol. 35(9):1725–31. https://doi.org/10.1109/JLT.2017.2650941.
47. Zlodeev, I. V., and Ivanov, O. V. 2013. Transmission spectra of a double-clad fibre
structure under bending. Quantum Electron. 43(6):535–41. https://doi.org/10.1070/
QE2013v043n06ABEH014971.
48. Chen, W., Lou, S., and Wang, L. 2011. In-fiber modal interferometer based on dual-
concentric-core photonic crystal fiber and its strain, temperature and refractive
index characteristics. Opt. Commun. 284(12):2829–34. https://doi.org/10.1016/j.
optcom.2011.02.019.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp06.indd 177 9/26/23 5:04 PM

 7 Wave Front Aberration
Sensors Based on
Optical Expansion by
the Zernike Basis
Svetlana Nikolaevna Khonina1,2, Pavel Alexeyevich
Khorin2, and Alexey Petrovich Porfirev1
1I mage Processing Systems Institute – Branch of the
Federal Scientific Research Centre “Crystallography and
Photonics” of Russian Academy of Sciences, Samara, Russia
2 Samara National Research University, Samara, Russia

7.1 INTRODUCTION
Studying the deviation of a wave front from the desired shape is one of the most
significant problems in optics. There are many well-known methods for solving this
problem, and new techniques are being constantly developed. The most common
and versatile among them is interferometry [1, 2], which has unsurpassed accu-
racy and allows one to directly obtain a pattern of wave front deviations at very large
apertures. The accuracy of interferometers, especially heterodyne interferometers,
exceeds λ/100. The disadvantages of interferometry are well known and include
complexity of decoding interferograms, sensitivity of measuring equipment to vibra-
tions, and a need for the physical presence of a reference wave front. At the initial
stages of the development of optical production, the schlieren (or shadow) method
was used to control spherical surfaces; [3] however, shadow patterns are difficult to
quantify, and schlieren systems, like interferometers, must have high rigidity and be
vibration-proof.
The Hartmann method [4], which appeared later, differs from the previous
techniques in the fact that the wave front deviations are calculated from a set of
sub-apertures, with some step covering the full size of the region to be studied.
Wave front deviations are calculated using ray tracing data, with the rays pass-
ing through sub-apertures. A further development of the Hartmann method was a
Shack–Hartmann wave front sensor [5]. In this version of the sensor, the data on the
wave front deviations are transferred to the photodetector plane by installing a lens
raster. Each lens forms a sub-aperture for which an average wave front deviation is
calculated. Information about the wave front phase within the sub-aperture is con-
tained in the coordinates of the focused light spot. The main advantage of both the
Hartmann method and the Shack–Hartmann wave front sensor consists in the fact
that there is no need to use a reference wave front in calculations. However, these
178 DOI: 10.1201/9781003439165-7

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 178 9/26/23 5:08 PM

 Wave Front Aberration Sensors 179

techniques are not exempt from disadvantages, such as sensitivity to vibrations and
initially discrete nature of the measurements. This disadvantage arises if too much
distorted wave front incidents on a sensor matrix. If aberrations are too much, a
local light spot leaves its own sub-area and jumps into the neighboring section of
the sensor matrix, in the Shack–Hartmann scheme. In this case, data on a part of the
surface of the wave front are inevitably lost. In addition, a false spot can be wrongly
detected. In order to eliminate errors due to cross talk, fairly complex algorithms
are used. Many Shack–Hartmann sensors currently being manufactured have no
more than 103 sub-apertures, which does not satisfy the requirements of many tasks.
A new approach to the formation of microlens arrays based on mesoscale square
cubic dielectric particles [6] was proposed, which will significantly increase the
dimension of the lens raster.
In the methods just described, to visualize the phase, some additional digital
operations are required to process the obtained measurements. Direct methods of
phase visualization are also known. The Zernike phase contrast method [7] is a pow-
erful tool for transforming the spatial phase information of an optical beam into a
spatial intensity distribution without absorbing light. The basic principle is to split a
light beam into its Fourier components using a lens and a filter. The introduced phase
shift creates an intensity distribution in accordance with the phase information car-
ried by the higher spatial frequencies. This method has been successfully applied to
analyze aberrations and improve resolution in telescopes, in decoding phase-coded
information, as well as in biological tissue microscopy.
Another approach to phase visualization is the optical knife method, as well as
the radial Hilbert transform, which is implemented on the basis of a spiral phase
plate. The radial Hilbert transform is used both for the selection of the edges of the
image and for phase contrast [8].
One of the tools for describing wave fronts, in addition to the deviation patterns,
is the decomposition of aberrations over various bases. The most famous decomposi-
tion bases include Zernike polynomials [9], as well as Seidel aberrations. Note that
the generally accepted representation of wave front aberrations [10–12], including
in the individual optical system of the human eye, is a series of Zernike polynomi-
als [7, 13–16]. The human eye can be described as a system of lenses consisting of
three basic components: the cornea, pupil and lens. The cornea (including the tear
film) structure is the dominant eye optical power (it provides about 70% of the opti-
cal power of the eye). Accordingly, it is a major source of aberrations in the eye.
The front surface of the cornea has an elongated profile; that is, the central region
is steeper than at the periphery. This shape helps reduce the amount of spherical
aberration in the whole eye. However, the cornea shape may be significantly differ-
ent for different people, and this results in asymmetric astigmatism and high-order
aberration (for example, coma). Optical elements of the human eye optical systems
work concertedly to create an image on the retina. However, the image in the real
system of a human eye is never perfect. The emergence of additional optical aberra-
tions associated with aging or disease degrades the image quality significantly. To
compensate the wave front distortions, it is necessary to determine exactly which
aberration led to the distortions. Wave front analysis based on an expansion in the
Zernike basis facilitates diagnosis in clinical ophthalmology and allows the creation

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 179 9/26/23 5:08 PM

 180 Photonics Elements for Sensing and Optical Conversions

of the most advanced lenses, presently providing the highest quality vision. If the
wave front is represented by a linear combination of Zernike polynomials, it has a
number of useful properties. First, Zernike polynomials are easy to relate to classi-
cal aberrations. Second, polynomials are usually defined by procession of values at
the points using ordinary least squares. Consequently, since Zernike polynomials
are orthogonal on the unit circle, any of the terms of the expansion are also the best
approximations of the ordinary least squares. Thus, to prevent the shift of focus or
tilt of the wave front, it is necessary that the corresponding coefficients are equal to
zero. The average value of the aberration is determined by the value of each respec-
tive member, so it is not necessary to perform a new approximation using ordinary
least squares.
Aberrational representations are more efficient [17] in terms of data volumes and
also allow one to make use of the wave front features that are important for solving
specific problems. Direct measurement of aberration coefficients is possible only for
some types of aberrations. Calculation of the Zernike aberration coefficients [18–21]
on the basis of a two-dimensional (2D) array of measured values of the wave front
deviations in each of the sub-apertures is provided in the data processing programs
supplied with Shack–Hartmann sensors, as well as with ophthalmic aberrometers.
However, it should be noted that due to the rather rough discretization of wave front
data, the calculation of high-order aberrations is difficult.
Active employment of Zernike polynomials for representation of wave aberra-
tions stimulates the development of new sensors, including for direct measurements
of expansion coefficients by the Zernike basis. In this work, we propose a new sen-
sor for measuring aberration coefficients based on a special multichannel diffractive
optical element [22–24]. The developed sensor provides a sensitivity to wave front
deviations no worse than λ/20, is resistant to vibrations, and does not require the use
of reference optical elements.
Diffractive optical elements (DOEs) for the integral calculation of the expan-
sion coefficients of amplitude–phase distributions of light fields over various bases
[25–28], including the basis of Zernike functions [29–31], have been developed and
used in fiber-optic sensors [32–34] for measuring the angular momentum of laser
beams [35–37], for optical communication using mode and polarization (de)multi-
plexing [38–44], and in testing problems [45, 46]. These elements make it possible
to simultaneously obtain the values of the decomposition coefficients in the given
elements of the photodetector matrix. In contrast to the Shack–Hartmann sensor, in
which the calculation of aberration coefficients requires mathematical processing
of a 2D data array, the values of aberration coefficients in multichannel diffractive
optical sensors are proportional to the intensities of diffraction maxima located at
the photodetector matrix points with constant coordinates. Thus, the entire area of
the tested beam is simultaneously involved in the formation of the values of each
coefficient, while in Shack–Hartmann sensors and especially in Hartmann sensors,
information about part of the wave front area is not involved in the measurements
and remains unknown. It should also be noted that the calibration function of the
proposed sensor is substantially nonlinear, which leads to a decrease in the dynamic
range of aberrations being measured. However, this is quite enough for most practi-
cally significant cases of certification of optical systems; for example, it is believed

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 180 9/26/23 5:08 PM

 Wave Front Aberration Sensors 181

that the average aberration should not exceed λ/10 for budget imaging systems and
λ/100 for high-end systems.

7.2 METHOD OF OPTICAL EXPANSION BY THE ZERNIKE BASIS

7.2.1 Theoretical and Numerical Investigations
In the problem of analysis of laser light, it is advisable to use spatial filters separating
the analyzed beam diffraction on individual components corresponding to angu-
lar harmonics. Angular harmonics of different orders m are the complex harmonic
functions with a single module and a linear dependence on the polar angle exp(imϕ).
Such harmonics appear in the optical implementation of the higher-order Hankel
transform based on the spiral phase plates [47, 48]. The complex amplitude of mono-
chromatic light wave E(r, ϕ) represented in polar coordinates (r, ϕ) can always be
expanded in a functional Fourier series of angular harmonics. Such functional series
is written as follows:
∞

E (r , φ) = ∑E
m =−∞
m ( r ) exp ( imφ ) , (7.1)

2π
1
Em (r ) =
2π ∫
E ( r , φ ) exp ( −imφ ) dφ.
0
(7.2)

Similarly, the expansion (7.1) of the angular harmonics of the spatial spectrum
of the light field can be performed. Indeed, if we take the Fourier transform of the
function E(r, ϕ), we find that:
∞

FT  E ( r , φ )  = F ( ρ, θ ) = ∑ F (ρ) exp (im θ) ,

m =−∞
m (7.3)

∞
4π 2i m  2π 
Fm ( ρ) =
λf ∫
Em ( r ) J m 
0
rρ r dr ,
 λf 
(7.4)

where (ρ, θ) is the polar coordinates in the focal plane, λ is the radiation wavelength,
f is a focal length of the optical system, and Jm(x) is the m-th order Bessel function of
the first order. It is clearly seen that the expressions (Eq. 7.1) and (7.3) are similar in
structure, while the function Fm(ρ) and Em(r) are associated with the Hankel trans-
form of m-th order (see Eq. (7.4)). Optical expansion in Eq. (7.3) can be performed
using tandem “lens+DOE,” where the spherical lens with a focal length f performs
the Fourier transform, and the DOE provides a decomposition of the incident radia-
tion with a wavelength λ on the set of “vortex” basis.
The expansion in Eq. (7.3) is useful for the invariant to the rotation analysis and
detection, since the rotation of the analyzed field F(ρ, θ) on the angle Δθ will not
change the experimentally measured modulus of the coefficients |Fm(ρ)|. There is

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 181 9/26/23 5:08 PM

 182 Photonics Elements for Sensing and Optical Conversions

a complete set of orthogonal functions with angular harmonics in a circle of radius

R. These are the circular Zernike polynomials corresponding to a complete set of
orthogonal functions in polar coordinates (r, ϕ) in a circle of radius r0 [13]:

 cos(mφ) 
Z nm (r , φ) = An Rnm (r )  , (7.5)
sin (mφ) 
 

where An = ( n + 1) / π and Rnm (r ) are the radial Zernike polynomials:

( n − m )/ 2

R (r ) =
m
n ∑ (−1) (n − p)! ×
p= 0
p

(7.6)
−1 n−2 p
 n+m n−m  r
×  p!  − p !  − p !   .
  2   2    r0 

Examples of Zernike polynomials are shown in Figure 7.1.

The expansion of the light field with complex amplitude E(r, ϕ) into a series in
terms of the functions in Eq. (7.5) is given by:
∞ n

E (r , φ) = ∑∑C
n = 0 m =− n
nm Z nm ( r , φ ) . (7.7)

The expansion coefficients for the wave front in the Zernike orthogonal functions
(Eq. 7.5) allow one to determine the deviations (aberrations) from the ideal wave
front [13–16, 18, 45, 46]. Taking into account expression (Eq. 7.4), the spatial spec-
trum of the field E(r, ϕ) can be represented as an expansion in Zernike polynomials:
∞ n

F (ρ, θ) = −
2i π
λfR ∑∑
n = 0 m =− n
n + 1 ( − i )m ×

R
(7.8)
 2π 
× Cnm exp ( im θ )
∫
o
Z nm ( r ) Jm  rρ r dr .
 λf 

The integral in expression (Eq. 7.8) may be taken explicitly [1]:

 2π 
R J n +1  Rρ
 2π   λf 
∫
(n− m)/ 2 3
Wnm (ρ) = Z n ( r ) Jm 
m
rρ r dr = ( −1) R . (7.9)
 λf   2π 
o
 λf R ρ

Let us consider an aberrated wave front in the form of a field:

g(r , φ) = exp [ iψ (r , φ) ] , (7.10)

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 182 9/26/23 5:08 PM

 Wave Front Aberration Sensors 183

Trigonometric
N n m Representation Aberration Type Amplitude and Phase
1 0 0 1 Constant

2 1 –1 2r sin(θ) Tilt

3 1 1 2r cos(θ) Tilt

4 2 –2 6 r 2 sin(2θ) Astigmatism

5 2 0 3(2r 2 − 1) Defocus

6 2 2 6 r 2 cos(2θ) Astigmatism

7 3 –3 2 2 r 3 sin(3θ) Zero curvature coma

(trefoil)

8 3 –1 2 2 (3r 3 − 2r )sin(θ) Pure coma

9 3 1 2 2 (3r 3 − 2r ) cos(θ) Pure coma

10 3 3 2 2 r 3 cos(3θ) Zero curvature coma

(trefoil)

FIGURE 7.1 Examples of Zernike polynomials.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 183 9/26/23 5:08 PM

 184 Photonics Elements for Sensing and Optical Conversions

whose phase is a superposition of Zernike functions:

ψ (r , φ) = 2πα ∑b
n ,m
nm Z nm (r , φ), (7.11)

where bnm are coefficients of the superposition.

The expansion coefficients of field (7.10) in basis (7.5) are calculated as follows:
r0 2 π

c pq =
∫∫g(r , φ)Z (r , φ) r dr dφ.
0 0
q
p (7.12)

We represent field (7.10) as the following expansion:

g(r , φ) = exp [ iψ (r , φ) ] =
1 2 i
= 1 + iψ (r , φ) − ψ (r , φ) − ψ 3 (r , φ) + ... =
2 6 (7.13)
2
 

n ,m
∑  n ,m 
∑
= 1 + i 2πα bnm Z nm (r , φ) − 2 ( πα )2  bnm Z nm (r , φ)  + ....

For small aberrations (i.e. the value of α), expression (Eq. 7.13) can be signifi-
cantly simplified:

g(r , φ) ≈ 1 + i 2πα
α→ 0 ∑b n ,m
nm Z nm (r , φ). (7.14)

Thus, if aberrations are small enough to leave only the first two terms in expan-
sion (7.13), then the field can be considered as a superposition of the functions them-
selves. In this case, the expansion coefficients of field (7.12) will be proportional to
the coefficients in superposition (Eq. 7.11):

r0 2 π
 
c pq ≈
α→ 0 ∫∫ 
n ,m
∑
 1 + i 2πα bnm Z n (r , φ) Z p (r , φ) r dr dφ =
m

q

0 0 (7.15)
= Ap + i 2πα ⋅ δ pq ,nm ∑b
n ,m
nm = Ap + i 2παb pq ,

where Ap is the normalization constant, and δ pq ,nm is the Kronecker delta.

In this case, the type and magnitude of aberrations can be detected using a mul-
tichannel filter matched with the Zernike functions [29, 31, 49]. Approximation
error of decomposition in Eq. (7.14) increases with increasing of meaning of aber-
rations. It happens because of a finite quantity of Zernike polynomials in the sum
(Eq. 7.14). Figure 7.2 shows simulation results of wave front aberrations analysis

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 184 9/26/23 5:08 PM

 Wave Front Aberration Sensors 185

Zernike Coefficients
Expansion Approximation
Spatial
Incident Field (Intensity Spectrum Expansion Coefficients, gˆ ( r , φ ) = ∑b nm , Ψ nm ( r , φ )
and Phase Distribution) (Amplitude) |bnm|2 n,m ∈ Ω

g(r , φ) = Z31 (r , φ) + Z3−1 (r , φ)

{
g(r , φ) = exp i 0.2π ×

}
× ( Z31 (r , φ) + Z3−1 (r , φ))

{
g(r , φ) = exp i 0.4 π ×

}
× ( Z31 (r , φ) + Z3−1 (r , φ))

{
g(r , φ) = exp i 0.6π ×
× ( Z (r , φ) + Z3−1 (r , φ))
1
3 }

{
g(r , φ) = exp i 0.8π ×

}
× ( Z31 (r , φ) + Z3−1 (r , φ))

FIGURE 7.2 Wave front aberration analysis with Zernike polynomials expansion.

using a 20-order optical filter that is matched with 20 Zernike polynomials: n ∈ 0,5
m ∈− n, n .
Physically, this corresponds to an average path difference between an ideal wave
front and a wave front with an aberration coefficient of 0.3λ. For visualization of the
expansion coefficients, it is convenient to use a single index, rather than a double one.
Table 7.1 shows the correspondence of Zernike functions (7.5) to a single index l.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 185 9/26/23 5:08 PM

 186 Photonics Elements for Sensing and Optical Conversions

TABLE 7.1
Correspondence of Zernike Functions (1) to a Single Index l
l 0 1 2 3 4
Z (r , φ)
m
n
0
R (r )
0 R (r ) cos φ
1
1 R (r )sin φ
1
1 R (r ) cos 2φ
2
2
0
R (r )
2

l 5 6 7 8 9
Z nm (r , φ) R22 (r )sin 2φ R33 (r ) cos3φ R31 (r ) cos φ R31 (r )sin φ R33 (r )sin 3φ
l 10 11 12 13 14
Z nm (r , φ) R44 (r ) cos 4φ R42 (r ) cos 2φ R40 (r ) R42 (r )sin 2φ R44 (r )sin 4φ

It should be noted that the task of detecting small aberrations is topical, because
the point scattering function (PScF) in this case slightly differs from the Airy pattern
(diffraction spot) in the absence of aberrations (see Table 7.2).
Table 7.2 shows the phase distribution (wave front) and PScF patterns in the pres-
ence of various aberrations corresponding to Zernike polynomials with different
values of α. One can see that at α = 0.4 (Table 7.2a2–7.2h2), the intensity distribu-
tions in the focal plane of the lens (PScF) look approximately the same regardless
of the aberration type and differ little from the diffraction spot. This fact makes it
possible to establish the applicability criterion for a multichannel filter, matched with
the Zernike functions, during the measurement process. The first column of Table
7.2 (see Table 7.2a1–7.2a4) shows the distribution of the coefficients at α = 0.4; the
correspondence of the index l to decomposition coefficients (Eq. 7.12) is shown in
Table 7.1. It follows from Table 7.2 (see Table 7.2a2–7.2h4) that only at α ≤ 0.4 (which
corresponds to the average aberration coefficient of ≤0.4λ) can we detect (recognize)
with confidence the aberration structure.
Approximations (Eq. 7.14) and (Eq. 7.15) become invalid with increasing α, and
in expanding the field expands in the Zernike basis, other coefficients will appear,
except for those present in superposition (Eq. 7.11).
Knowing this fact, we can determine to some extent how significant the level of
aberration is. However, this can be done by measuring a sufficiently large number of
factors. Given the need to do this using a single multichannel diffractive sensor, it is
desirable to optimize the number of necessary coefficients.
Note that the basis of the Zernike functions with trigonometric functions on angle
(Eq. 7.5) is not invariant to rotation, which is inconvenient in practical applications.
Another representation of the Zernike functions is also well known:

Z n ,m (r , φ) = Bn Rnm (r ) exp ( imφ ) , (7.16)

where Bn = An at m = 0 and Bn = An /2 at m ≠ 0.
Obviously, functions (Eq. 7.5) can be represented via a superposition of functions
(Eq. 7.16) and vice versa. Compared to expression (Eq. 7.5), representation (Eq. 7.16)
is more convenient due to its in invariance to rotation. The Zernike basis in form
expression (Eq. 7.16) cannot be used in superposition (Eq. 7.11); however, it can be
conveniently used for the expansion of the optical field [31]. Amplitude and phase

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 186 9/26/23 5:08 PM

 Wave Front Aberration Sensors 187

TABLE 7.2
The Type of Phase (Wave Front) and PScF in the Presence of
Various Aberrations Corresponding to Zernike Polynomials

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 187 9/26/23 5:08 PM

 188 Photonics Elements for Sensing and Optical Conversions

Function Standard Function Standard

Indices, Vortex Zernike Zernike Indices, Vortex Zernike Zernike
(n, m) Functions (5) Functions (11) (n, m) Functions (5) Functions (11)
1 (0,0) 7 (3,1)

2 (1,1) 8 (3,–1)

3 (1,–1) 9 (3,3)

4 (2,0) 10 (3,–3)

5 (2,2) 11 (4,0)

6 (2,–2)

FIGURE 7.3 Amplitude and phase pictures of several first Zernike functions in representa-
tion in Eqs. (7.5) and (7.16).

pictures of several first Zernike functions in representation in expression (7.5) and

expression (Eq. 7.16) are shown in Figure 7.3. The functions having a zero angular
index m coincide in both representations. It should be noted that the Zernike func-
tions in standard representation in expression (Eq. 7.5) are all real-valued. However,
when we use for the analysis of wave front multi-order diffractive filters [29, 50–53]
matched with Zernike functions, it can be a more convenient vortex representation
in expression (Eq. 7.16).

7.2.2 Multi-Order Diffractive Optical Elements

for Optical Decomposition

A multichannel diffractive filter matched with Zernike functions has the following
representation [53]:
∞ n

τ(r , φ) = ∑∑ Z
n = 0 m =− n
*
n ,m (r , φ) exp isn,m (r , φ)  , (7.17)

where sn,m(r, φ) are phase functions corresponding to definite diffractive orders. It is

obvious, if they are chosen as follows:

sn , − m ( r , φ ) = − sn , m ( r , φ ) , (7.18)

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 188 9/26/23 5:08 PM

 Wave Front Aberration Sensors 189

then for vortex functions in expression (7.16) we shall receive the multi-order filter
with the real-valued transmission function:
∞ n

τb (r , φ) = ∑∑
n=0 m=0
n +1 m
πR 2
Z n ( r ) cos  sn ,m ( r , φ ) + mφ  . (7.19)

A phase of the filter in expression (7.19) is binary, which is convenient for

manufacturing.
Figure 7.4 shows phases of 4-channel filters matched with functions 2, 3, 5, 6 (see
corresponding lines in Figure 7.3) for vortex and standard functions, accordingly.
Intensity and phase distributions formed in a focal plane are also shown in Figure
7.4 (we use a lens added to the filter). Apparently from Figure 7.4 the filter’s phase
of the vortex basis has binary structure (black color corresponds to zero value of a
phase, and grey color corresponds to π value of a phase). Distributions in the focal
plane show preservation of corresponding vortex or trigonometric structure of basic
functions. Note, when we add into a filter Zernike functions with zero angular index
m, they should be duplicated to keep binary phase structure of the filter.
Another feature of the vortex basis in expression (Eq. 7.16) compared with the
standard one in expression (Eq. 7.5) is: the intensity of expansion coefficients of
the analyzed fields are invariant to rotation of the field. This property is illustrated

Intensity (Negative) and Phase Distributions

Filter’s Phase Formed in a Local Plane
Vortex
Zernike
Basis (5)

Standard
Zernike
Basis (11)

FIGURE 7.4 Type and action of 4-channel filters matched with functions 2, 3, 5, 6 (see cor-
responding lines in Figure 7.3).

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 189 9/26/23 5:08 PM

 190 Photonics Elements for Sensing and Optical Conversions

Intensity Distributions (Negative) Formed

in a Local Plane
Analyzed Field (Amplitude
and Phase) Vortex Zernike Basis (5) Standard Zernike Basis (5)
E (r , φ) = 1, r ≤ R

E (r , φ) = r cos(φ + φ0 ), φ0 = 0

E (r , φ) = r cos(φ + φ0 ), φ0 = −π /3

E (r , φ) = r cos(φ + φ0 ), φ0 = −π /2

E (r , φ) = r 2 cos(r + 2φ + φ0 ), φ0 = −π /4

FIGURE 7.5 Results of optical decomposition of light fields by means of 6-channel filters.

in the images in Figure 7.5, which show the results of the expansion of the optical
light field E(r, ϕ) = rcos(ϕ + ϕ0), where ϕ0 is the angle of rotation of the field. We
use 6-channel filters, consistent with the functions 1–6 of Figure 7.3 for two types
of Zernike functions.
Diffraction orders corresponding to the functions (0, 0) and (2, 0) are vertically
arranged: (0, 0) is at the top and (2, 0) is at the bottom. Other functions are arranged as
shown in Figure 7.4. Note that we can observe the Fourier transform of the analyzed

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 190 9/26/23 5:08 PM

 Wave Front Aberration Sensors 191

TABLE 7.3
Correspondence of the Considered Zernike Functions (9) to a Single Index l
l 0 1 2 3 4 5 6 7 8
(n, m) (0,0) (1,1) (1,–1) (2,2) (2,0) (2,–2) (3,3) (3,1) (3,–1)
l 9 10 11 12 13 14 15 16 17
(n, m) (3,–3) (4,4) (4,2) (4,0) (4,–2) (4,–4) (5,5) (5,3) (5,1)
l 18 19 20 21 22 23 24 25 26
(n, m) (5,–1) (5,–3) (5,–5) (6,6) (6,4) (6,2) (6,0) (6,–2) (6,–4)
l 27 28 29 30 31 32 33 34 35
(n, m) (6,–6) (7,7) (7,5) (7,3) (7,1) (7,–1) (7,–3) (7,–5) (7,–7)
l 36 37 38 39 40 41 42 43 44
(n, m) (8,8) (8,6) (8,4) (8,2) (8,0) (8,–2) (8,–4) (8,–6) (8,–8)

field in the diffraction order of the function (0, 0). The presence of non-zero intensity
in the center of a diffraction order indicates the presence of the corresponding basis
function in the analyzed beam with a factor that is proportional to the intensity. The
first row of Figure 7.5 shows results of the expansion of the uniform field E(r, ϕ) = 1,
r ≤ R. In this case, only the single function from six Zernike functions written in these
filters will have a non-zero coefficient, namely, the center of the top diffraction order
(0,0) has a correlation peak. The second row of Figure 7.5 shows results of expansion
of the cosine E(r, ϕ) = rcos(ϕ + ϕ0), ϕ0 = 0. In this case, the vortex basis gives correla-
tion peaks in the left (1,1) and the right (1,–1) diffraction orders, while standard basis
gives the response only in the cosine order. In the third and fourth rows of Figure 7.5,
results of decomposition of E(r, ϕ) = rcos(ϕ + ϕ0) with ϕ0 = π/6 and ϕ0 = π/2 are shown,
correspondingly. Obviously, the vortex basis shows invariance to rotation, in differ-
ence to the standard basis. However, the advantage of the standard basis is the ability
to unequivocally respond to standard aberration distributions as it is apparently seen,
from rows 2, 4, and 5 of Figure 7.5. Therefore we are planning to use these two types
of Zernike functions together. Besides, the additional information can be used for
measurements of not only intensity of expansion coefficients but also their relative
phases [29, 52, 53].
Basis (15) implies positive and negative values of the index m. The correspon-
dence of a single index l to the pair indices (n, m) is shown in Table 7.3.

7.2.3 Analysis of the Influence of the Level of Aberration

on the Detection Accuracy

Next we will consider some examples. For the convenience of further analysis, we
write out explicit expressions for several Zernike polynomials in Table 7.4.

1. Defocusing Z 20 (r , φ) .

( )
g(r , φ) = exp i 2παZ 20 (r , φ)  = exp i 2παA2 2r 2 − 1  . (7.20)

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 191 9/26/23 5:08 PM

 192 Photonics Elements for Sensing and Optical Conversions

TABLE 7.4
Explicit Expressions for Some Zernike
Polynomials
(n, m) Rnm ( r ) (n, m) Rnm ( r )
(0,0) 1 (4,0) 6r 4 − 6r 2 + 1
(1,1) r (4,2) 4r 4 − 3r 2
(2,0) 2r 2 − 1 (4,4) r4
(2,2) r 2
(5,1) 10r − 12r 3 + 3r
5

(3,1) 3r − 2r
3
(5,3) 5r 5 − 4r 3
3
(3,3) r (5,5) r5

Taking into account expansion (7.13), field (7.17) can be represented as:

exp i 2παZ 20 (r , φ)  =

(
= 1 + iπαZ 2,0 (r , φ) − ( παA2 ) 4r 4 − 4r 2 + 1 + ... =
2
) (7.21)
= D0 + iαD1 Z 2,0 (r , φ) − α D2 Z 4,0 (r , φ) + ... ,
2

where D j are the reduced constants.

It can be seen from expression (Eq. 7.21) that for large values of the
parameter α, when field (7.20) is expanded in the Zernike basis, in addition
to coefficients with indices (0,0) (l = 0) and (2,0) (l = 4) there will appear a
coefficient with the index (4,0) (l = 12), as well as higher-order coefficients
(see Figure 7.6).

FIGURE 7.6 Wave front expansion coefficients with defocusing Z 20 (r , φ) of various levels:
(a) α = 0.4, (b) α = 0.6, and (c) α = 1.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 192 9/26/23 5:08 PM

 Wave Front Aberration Sensors 193

Figure 7.6 shows the results of calculating the expansion coefficients of field
(7.20) in basis (7.16) at different values of the parameter α. As can be seen, at
small values of α, the coefficient (2,0) (l = 4) is the largest [Figure 7.6(a)]. With
increasing α, the weight of the coefficient (4,0) (l = 12), as well as that of the
coefficient (6,0) (l = 24), increases [Figure 7.6(b)]. At a high level of defocus-
ing, the weight of the coefficient (8,0) (l = 40) is significantly enhanced [Figure
7.6(c)]. Thus, the appearance of energy in high-order aberrations corresponds
to a large level of available low-order aberration, and this effect can be detected
by optical expansion of the analyzed wave front in the Zernike basis.
2. Astigmatism Z 22 (r , φ) .

g(r , φ) = exp i 2παZ 22 (r , φ)  = exp i 2παA2 r 2 cos 2φ  , (7.22)

Taking into account expansion (7.13), field (7.19) can be represented as:
exp i 2παZ 22 (r , φ)  =
1 + iπαZ 2,±2 (r , φ) − 2 ( παA2 ) r 4 (1 + cos 4φ ) + ... =
2
(7.23)
= D0 + iαD1 Z 2,±2 (r , φ) −
− α 2 [ D2 Z 2,0 (r , φ) + D3 Z 4,0 (r , φ) + D4 Z 4,±4 (r , φ) ] + ... .

The presence of r4cos4ϕ in (Eq. 7.23) leads to the appearance of

Z4,±4(r, ϕ), with r4 resulting in the defocusing of various orders, in particular,
( )
r 4 = R40 (r ) + 3 R20 (r ) + 2 /6 .
Thus, expression (Eq. 7.23) shows that for large values of the parameter α,
in addition to coefficients with indices (0,0) (l = 0) and (2,±2) (l = 3,5), there
will appear coefficients with indices (2,0) (l = 4), (4,0) (l = 12), and (4,±4)
(l = 10,14), as well as higher-order coefficients (see Figure 7.7).

FIGURE 7.7 Wave front expansion coefficients with astigmatism Z 22 (r , φ) of various levels:
(a) α = 0.4, (b) α = 0.6, and (c) α = 1.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 193 9/26/23 5:08 PM

 194 Photonics Elements for Sensing and Optical Conversions

As can be seen from the simulation, for small values of α, the coefficients
(0,0) (l = 0) and (2,±2) (l = 3,5) predicted in (19) are the largest [Figure 7.7(a)].
With increasing α, the coefficients (2,0) (l = 4) and (4,±4) (l = 10,14) become
more significant [Figure 7.7(b)]. It should be noted that with a high level of
astigmatism, the field energy is distributed over a large number of coefficients,
with defocusing (2,0) (l = 4) being the most noticeable [Figure 7.7(c)].
3. Coma Z31 (r , φ) .

( )
g(r , φ) = exp i 2παZ31 (r , φ)  = exp i 2παA3 3r 3 − 2r cos φ  . (7.24)

Field (7.21) can be represented as:

exp i 2παZ31 (r , φ)  =

( )
= 1 + iπαZ3,±1 (r , φ) − ( παA3 ) 9r 6 − 6r 4 + 4r 2 (1 + cos 2φ ) + ... =
2
(7.25)
= D0 + iαD1 Z3,±1 (r , φ) −
− α 2 [ D2 Z6,0 (r , φ) + D3 Z6,±2 (r , φ) + D4 Z 2,±2 (r , φ) + D5 Z 4,±2 (r , φ) ] + ... .

Expression (Eq. 7.25) is quite complex, but one can clearly see that aberra-
tions with even n = 2,4,6 and m = ±2 appear additionally. Obviously, if we also
take the cubic term into account, then additional odd aberrations with m = ±3
should appear.
As can be seen from the simulation, for small values of α, the coefficients
(0,0) (l = 0) and (3,±1) (l = 7,8) predicted in (24) are the largest [Figure 7.8(a)].
With increasing α, the coefficients (6,0) (l = 24), (6,±2) (l = 23,25), and (2,±2)
(l = 3,5), also predicted in (24), become significant [Figure 7.8(b)]. At a high

FIGURE 7.8 Wave front expansion coefficients with coma Z 31 (r , φ) of various levels:
(a) α = 0.4, (b) α = 0.6, and (c) α = 1.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 194 9/26/23 5:08 PM

 Wave Front Aberration Sensors 195

coma level, the field energy is distributed over a large number of coefficients
[Figure 7.8(c)]. As expected, aberrations with higher angular multiplicity also
appear in this case, namely, (5,±3) (l = 16,19) and (7,±3) (l = 30,33). However,
the coefficient (6,0) (l = 24), corresponding to high-order defocusing, becomes
the most noticeable [Figure 7.8(c)].
4. Coma (trefoil) Z33 (r , φ).

g(r , φ) = exp i 2παZ33 (r , φ)  = exp i 2παA3r 3 cos3φ  . (7.26)

Field (7.23) can be represented as:

exp i 2παZ33 (r , φ)  =

= 1 + iπαZ3,±3 (r , φ) − ( παA3 ) r 6 (1 + cos6φ ) + ... =
2
(7.27)
= D0 + iαD1 Z3,±3 (r , φ) −
− α 2 [ D2 Z 2,0 (r , φ) + D3 Z 4,0 (r , φ) + D4 Z6,0 (r , φ) + D5 Z6,±6 (r , φ) ] + ... .

The presence of r 6 cos6φ in Eq. (7.27) leads to the appearance of

Z6,±6 (r , φ), i.e. a multiple increase in the angular dependence. The dependence
r 6 can be described by superposition R60 (r ), R40 (r ), R20 (r ) (i.e. defocusing of
various orders). In addition, aberrations with the same angular dependence
m = 3, but with a higher degree of radial polynomials (n > 3), may appear.
The simulation confirms the theoretical analysis: for small values of α, the
coefficients (0,0) (l = 0) and (3,±3) (l = 6,9) are the largest [Figure 7.9(a)]. With
increasing α, the coefficients (5,±3) (l = 16,19) and (6,±6) (l = 21,27) increase,
and the weight of the coefficients associated with defocusing (2,0) (l = 4) also

FIGURE 7.9 Wave front expansion coefficients with coma (trefoil) Z 33 (r , φ) of various levels:
(a) α = 0.4, (b) α = 0.6, and (c) α = 1.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 195 9/26/23 5:08 PM

 196 Photonics Elements for Sensing and Optical Conversions

grows [Figure 7.9(b)]. At a high level of aberrations, there also appear aberra-
tions with a higher multiplicity [Figure 7.9(c)].
5. Defocusing (quatrefoil) Z 44 (r , φ) .

g(r , φ) = exp i 2παZ 44 (r , φ)  = exp i 2παA4 r 4 cos 4φ  . (7.28)

Field (7.23) can be represented as:

exp i 2παZ 44 (r , φ)  =

= 1 + iπαZ 4,±4 (r , φ) − ( παA4 ) r 8 (1 + cos8φ ) + ... =
2
(7.29)
= D0 + iαD1 Z 4,±4 (r , φ) − α 2 [ D2 Z 2,0 (r , φ) + D3 Z 4,0 (r , φ) +
+ D4 Z6,0 (r , φ) + D5 Z8,0 (r , φ) + D6 Z8,±8 (r , φ) ] + ... .

Expression (Eq. 7.29) is obtained similarly to the previous example.

The simulation results are shown in Figure 7.10: for small values of α,
the coefficients (0,0) (l = 0), (2,0) (l = 4), and (4,±4) (l = 10,14) are the larg-
est [Figure 7.10(a)]. With increasing α, the coefficients (6,±4) (l = 22,26) and
(8,±8) (l = 36,44) increase, and the weight of the coefficients (2,0) (l = 4) asso-
ciated with defocusing also grows [Figure 7.10(b)]. At a large aberration level,
aberrations with the same angular dependence but a higher degree of radial
polynomials (6,±4) (l = 22,26) and (8,±4) (l = 38,42) become more significant
and the effect of defocusing (2,0) (l = 4), (4,0) (l = 12), (6,0) (l = 24) is enhanced
[Figure 7.10(c)].

FIGURE 7.10 Wave front expansion coefficients with defocusing (quatrefoil) Z 44 (r , φ) of

various levels: (a) α = 0.4, (b) α = 0.6, and (c) α = 1.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 196 9/26/23 5:08 PM

 Wave Front Aberration Sensors 197

The preceding examples allow us to identify the main trends associated

with an increase in the level of aberrations. If the initial aberration has an
angular dependence on the order of m, then aberrations appear with a multiple
angular dependence of 2m, 3m. In addition, as a rule, there appears defocusing
of various orders. The detection of defocusing is associated with an increase
in the PScF area, which is always observed with an increase in the aberration
level (see Table 7.2).
6. Superposition Z31 (r , φ) + Z 42 (r , φ) .

{ ( ) ( ) }
g(r , φ) = exp i 2πα  A3 3r 3 − 2r cos φ + A4 4r 4 − 3r 2 cos 2φ  . (7.30)

A theoretical analysis of expression (Eq. 7.30) is rather difficult; therefore,

we consider only the results of numerical simulation shown in Figure 7.11.
For small values of α, the expected coefficients (0,0) (l = 0), (3,±1) (l = 7,8),
and (4,±2) (l = 11,13), as well as additional defocusing (6,0) (l = 24), are the
largest [Figure 7.11(a)]. With an increase in α, the weight of the coefficients
(2,0) (l = 4) and (6,0) (l = 24) corresponding to defocusing increases [Figure
7.11(b)], which is explained by an increase in the PScF area.
Further enhancement of aberration leads to an almost uniform distribution
of field energy over all coefficients, which should serve as a signal of a high
level of wave front distortion in measurements. In this case, other methods,
including neural networks, need to be used to recognize and compensate for
aberrations [54–56].
Listing of software codes written in the C++ programming language for the
Visual Studio development environment for calculating the sensor matched
with the Zernike basis functions is provided in Appendix A.

FIGURE 7.11 Wave front expansion coefficients with aberration superposition

Z 31 (r , φ) + Z 42 (r , φ) of various levels: (a) α = 0.4, (b) α = 0.6, and (c) α = 1.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 197 9/26/23 5:08 PM

 198 Photonics Elements for Sensing and Optical Conversions

7.2.4 Experimental Results
7.2.4.1 Experimental Investigation of Multi-Order Diffractive
Optical Elements Matched with Two Types
of Zernike Functions
Figure 7.12 shows the optical setup used in our experiments. A system composed of
a lens L1 (f1 = 350 mm) and a pinhole (40 μm aperture) was utilized for expansion
and filtration of input laser beam. Then the laser beam incident on the diffractive
optical element DOE1, generating a Hermite–Gaussian mode TEM10. The gener-
ated beam illuminated the diffractive optical element DOE2. In the first series of
experiments, we used 20-th order diffractive optical element generating 10 pairs of
Zernike functions as the DOE2; in the second series of experiments we used 25-th
order diffractive optical element generating nine pairs of Zernike functions and their
combinations. Figure 7.13 shows the phase transmission functions of diffractive opti-
cal elements used in experiments. Transmission function of DOE1 matched with the
basis of first type (see Eq. 7.16)). Transmission function of DOE2 matched with the
basis of second type (see Eq. (7.5)). The video camera Cam LOMO TC-1000 (pixel
size is 1.67 × 1.67 μm) was used for shooting the intensity distributions generated in
the focal plane of the lens L2 (f2 = 150 mm).
In experiments, an element DOE2 was stationary, while an element DOE1 rotated
about its center located on the axis of the beam propagation. Thus, we investigated
the influence of the orientation of the illuminating beam relative to the Zernike fil-
ter. In the first series of experiments, the 25-th order diffractive optical element with
the transmission function shown in Figure 7.13(b) was used as DOE2. The intensity
distribution obtained at different rotation angles DOE1 relative to this element is
shown in Figure 7.14. From these images, you can see that at rotation angles equal
to 0 and 180 degrees, peaks of intensity appear in the central parts of some of the
diffraction orders. This indicates the presence of the components corresponding
Zernike functions encoded in these diffraction orders in the illuminating beam. At
angles of 90 degrees and 270, peaks of intensity are not observed in the respective
orders. This indicates that the 25-th order Zernike filter is sensitive to rotation. The
intensity distribution generated in the zero order allows you to see the rotation of
the illuminating beam.

FIGURE 7.12 Experimental optical setup: L is a linearly polarized laser (λ = 633 nm), L1
and L2 are lenses (f1 = 150 mm, f2 = 150 mm), PH is a pinhole (40 μm), DOE1 is a diffractive
optical element to generate the Hermite–Gaussian mode TEM10, DOE2 is diffractive opti-
cal element matched with the basis of Zernike polynomials, Cam is a video camera LOMO
TC-1000 (the pixel size is 1.67 × 1.67 μm).

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 198 9/26/23 5:08 PM

 Wave Front Aberration Sensors 199

FIGURE 7.13 The phase transmission functions of the diffractive optical elements utilized
in experiments (top row), and generated intensity distributions in far-field (negative): (a, d)
DOE-generated Hermite–Gaussian mode TEM10; (b, e) 25-th order phase Zernike filter; (c, f)
20-th order phase Zernike filter.

FIGURE 7.14 Intensity distributions (negative) experimentally generated by 25-th order

phase Zernike filter illuminated by plane wave (a) and by Hermite–Gaussian mode TEM10 at
different angles of rotation of the Hermite–Gaussian mode TEM10 relative to the propagation
axis of the beam: (b) 0 degrees, (c) 90 degrees, (d) 180 degrees, (e) 270 degrees. Only the
central part of the generated diffraction pattern is shown.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 199 9/26/23 5:08 PM

 200 Photonics Elements for Sensing and Optical Conversions

FIGURE 7.15 Intensity distributions (negative) experimentally generated by 20-th order

phase Zernike filter illuminated by plane wave (a) and by Hermite–Gaussian mode TEM10 at
different angles of rotation of the Hermite–Gaussian mode TEM10 relative to the propagation
axis of the beam: (b) 0 degrees, (c) 90 degrees, (d) 180 degrees, (e) 270 degrees. Only the
central part of the generated diffraction pattern is shown.

Figure 7.15 shows the intensity distribution generated in the second series of
experiments, when we used 20-th order diffractive optical element with transmission
function, as shown in Figure 7.13(c), as an element DOE2. It is clearly seen in this
case, that regardless of the rotation angle of the generated Hermite–Gaussian mode
TEM10, distribution of the peaks of intensity in generated diffraction pattern does not
change. This indicates that such Zernike filter is not sensitive to rotation.

7.2.4.2 Experiments on Detection of Various Wave Front

Aberrations Using a Zernike Filter
Figure 7.16(a) shows the optical scheme used in the experiment. The output from a
solid-state laser (λ = 532 nm) was collimated using a system consisting of a pinhole
(PH) with a hole diameter of 40 μm and a spherical lens (L1) (f1 = 250 mm). In this
case, the lens (L1) was mounted on a linear translation stage and could be moved
along the beam propagation axis with a step of 10 μm, which was done for subsequent
experiments on measuring the defocusing of the collimated initial beam using our
wave front analyzers. Then, the expanded laser beam passed through a HOLOEYE
LC 2012 transmissive spatial light modulator (SLM1) with a 1024 × 768 pixel resolu-
tion and a pixel size of 36 μm, which was used to form a wave front with a required
set of aberrations. Lenses (L2) (f2 = 150 mm), L3 (f3 = 150 mm), and a diaphragm
(D) were used together to ensure spatial filtering of the aberration-distorted beam
formed by the first modulator. A HOLOEYE PLUTO VIS reflective spatial light
modulator (SLM2) with a 1,920 × 1,080 pixel resolution and a pixel size of 8 μm
was used to implement a phase mask of a multi-order analyzing diffractive optical

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 200 9/26/23 5:08 PM

 Wave Front Aberration Sensors 201

FIGURE 7.16 Detection of wave front aberrations using a multichannel analyzing DOE.
(a) Schematic of the experimental setup: laser is a solid-state laser (λ = 532 nm); PH is a
pinhole (hole size of 40 μm); L1, L2, L3, L 4, L5, and L6 are spherical lenses (f1 = 250 mm, f 2 =
150 mm, f3 = 150 mm, f4 = 150 mm, f5 = 150 mm, and f6 = 350 mm); SLM1 is a transmissive
spatial light modulator (HOLOEYE LC 2012); SLM2 is a reflective spatial light modulator
(HOLOEYE PLUTO VIS); D is an aperture; BS is a beam splitter; M is a mirror; Cam is a
ToupCam UCMOS08000KPB video camera. (b) Phase mask of a 25-order analyzing DOE,
which displays the incident light field in terms of the Zernike polynomials.

element (DOE), which served to decompose the studied light field in terms of the
Zernike polynomial basis. The laser beam reflected from the reflective light modula-
tor using a beam splitter (BS), a 4-f optical system of lenses L 4 and L5 (f4 = 150 mm,
f5 = 150 mm) and a mirror (M) was directed to lens L6 (f 6 = 350 mm), which focused
it on the matrix of a ToupCam UCMOS08000KPB camera with a 3,264 × 2,448
pixel resolution and a pixel size of 1.67 μm. Part of the scheme, including light
modulator SLM2, lens L6, and video camera is, in fact, a sensor.
Other elements of the optical system are designed to simulate the studied beam
and match the light modulator operating in reflective regime with the rest of the
system. The described optical system, in addition to checking the operability of the
wave front sensor, is used to calibrate the sensor by forming the studied beams with
different aberrations and different α values using a controlled light modulator SLM1.
In the industrial version of the sensor, light modulator SLM2 will be replaced by a
classical transmissive phase DOE made of a transparent material. This version of
the sensor is easier to manufacture than the Shack–Hartmann sensor and does not
require vibration isolation. In the process of measurements, the intensity patterns
were stable both in coordinates and in measured intensity, although the optical sys-
tem was not vibration-proof.
Phase mask of a 25-order analyzing DOE, which decomposes the incident light
field in terms of the basis of Zernike functions (Eq. 7.16) with numbers (n, m) =
{(0,0), (1,1), (2,0), (2,2), (3,1), (3,3), (4,0), (4,2), (4,4), (5,1), (5,3), (5,5), (6,0), (6,2),
(6,4), (6,6), (7,1), (7,3), (7,5), (7,7), (8,0), (8,2), (8,4), (8,6), (8,8)} is shown in Figure
7.16(b). Although the complete basis (7.10) implies positive and negative values of
the index m, taking into account a certain duplication of information in complex con-
jugate coefficients (with ±m), we used basis (7.10) only with positive m index values
to reduce the number of diffraction orders (channels).
Figure 7.17 shows the experimental and numerical (using the Fourier transform)
intensity distributions formed in the focal plane of lens L6 when modulator SLM2 is
illuminated by a laser beam with an aberration less plane wave front.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 201 9/26/23 5:08 PM

 202 Photonics Elements for Sensing and Optical Conversions

FIGURE 7.17 (a) Numerically and (b) experimentally obtained intensity distributions formed
in the focal plane of lens L6 when SLM2 is illuminated by a laser beam with a plane wave front.
In the images of numerically obtained distributions, the zero diffraction order is cut out.

The absence of any light peaks in the formed diffraction orders is clearly seen
in Figure 7.17, which confirms the absence of aberrations in the illuminating beam.
Figures 7.18, 7.19, and 7.20 show the experimentally obtained intensity distribu-
tions for various aberrations at α = 0.4, 0.6, and 1, respectively. One can see the
emergence of correlation peaks in diffraction orders corresponding to additional
theoretically predicted aberrations. In all cases, the presence of defocusing (2,0) is
clearly visible.
It is convenient to evaluate the value of α in the wave front under study and,
accordingly, the applicability of the sensor by the intensity distribution in the zero
diffraction order corresponding to the PScF of the beam in question. Using the
numerical simulation (see Table 7.2) we showed that for α ≤ 0.4 corresponding to the
range of the sensor applicability, the PScF is close to the Airy pattern of a diffrac-
tion-limited system. With α > 0.4, the size of the PScF begins to increase due to the
appearance of additional petals and rings, and the intensity in the center decreases
(see Figure 7.19 and especially Figure 7.20). This serves as a criterion for exceeding
the level of aberrations acceptable for the sensor.
Thus, we have experimentally confirmed that at α ≤ 0.4, the aberration structure
can be confidently detected (recognized); however, with a further increase in α, rec-
ognition becomes problematic.

7.2.4.3 Experiments on the Collimator Fine-Tuning

One of the most important applications of the developed technique – analysis of
wave front aberrations – is the testing and accurate adjustment of various optical
components, for example, collimators.
Figure 7.21 shows the intensity distributions obtained in the focal plane of lens L6
at various displacements of lens L1 forming the collimator along the beam propaga-
tion axis. One can see that when the lens is displaced from the initial plane z = 0,
which corresponds to an ideal position of the lens (i.e. the position at which a laser

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 202 9/26/23 5:08 PM

 Wave Front Aberration Sensors 203

FIGURE 7.18 Experimentally obtained intensity distributions at α = 0.4 for (a) Z 22 (r , φ)

{(0,0), (2,0), (2,2)}; (b) Z 33 (r , φ) {(0,0), (2,0), (3,3)}; (c) Z 44 (r , φ) {(0,0), (2,0), (4,4)}; and
(d) Z 31 (r , φ) + Z 42 (r , φ) {(0,0), (2,0), (3,1), (4,2)}.

FIGURE 7.19 Experimentally obtained intensity distributions at α = 0.6 for (a) Z 31 (r , φ)

{(0,0), (2,0), (2,2), (3,1), (4,0), (6,0)} and (b) Z 33 (r , φ) {(0,0), (2,0), (3,3), (5,3)}.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 203 9/26/23 5:09 PM

 204 Photonics Elements for Sensing and Optical Conversions

FIGURE 7.20 Experimentally obtained intensity distributions at α = 1 for (a) Z 33 (r , φ) (0,0),

(2,0), (5,3), (6,0), (7,3) and (b) Z 44 (r , φ) (0,0), (2,0), (6,0), (6,4), (8,4).

FIGURE 7.21 Intensity distributions obtained in the focal plane of lens L6 at different dis-
placements of lens L1 forming the collimator. The dotted circle highlights the diffraction
order corresponding to defocusing (Zernike polynomial Z 20).

beam with a plane wave front is formed behind the lens) at the center of the diffrac-
tion order responsible for the Zernike polynomial with numbers (n, m) equal to (2,0)
(i.e. defocusing aberrations), an intensity peak appears.
The intensity of this peak increases equally with distance from the original plane
z = 0 in both directions. Calculations show that for the used lens with a focal length
of 250 mm and the illuminating beam with a diameter of about 5.4 mm, the devia-
tion of the wave front from the plane at the edge of the lens aperture at z = 0.5 mm
is about λ/20. Thus, we can conclude that the sensitivity of the proposed sensor to
local deviations of the wave front is no worse than λ/20. The average aberration
over the entire beam is much less. This sensitivity of the sensor allows one to detect
defocusing of less than 0.5 mm for a laser radiation collimator with a relative aper-
ture of about 1:50. Thus, in terms of sensitivity to aberrations, the proposed sensor
is quite competitive with the Shack–Hartmann sensor, but the proposed sensor does

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 204 9/26/23 5:09 PM

 Wave Front Aberration Sensors 205

not require vibration isolation. The main advantage of the proposed sensor is that,
unlike the Shack–Hartmann sensor, where the aberration coefficients are calculated
by processing mathematically a 2D data array, the values of the coefficients in mul-
tichannel DOE sensors are directly proportional to the light intensities measured in
individual pixels of the photodetector matrix with fixed coordinates. The entire area
of the tested beam is simultaneously involved in the formation of the values of each
coefficient, which reduces errors.

7.3 APPLICATION OF WAVE FRONT SENSORS

IN OPHTHALMOLOGY
7.3.1 Wave Front Sensors in Ophthalmic Research
Over the last decade, aberrometers have been widely employed in many areas of
ophthalmology and optometry [57, 58], e.g. for detecting refraction anomalies, diag-
nosing a dry-eye symptom [59] and keratoconus [60], as well as in refractive eye
surgery [61, 62]. Wave front aberrations may be in the form of defocus, astigma-
tism, and higher-order aberrations. Higher-order aberrations (HOA) occur when
the eye surface has minor roughness or imperfections that cannot be compensated
for using standard spectacles [63]. As a rule, cornea topography is performed using
dedicated algorithms that rely on height data. Intraocular aberrations can be identi-
fied by deducting eye aberrations from cornea aberrations. Principles of aberrometer
operation may be based on a Shack–Hartmann method, a ray tracing technique, and
a Churning principle.
Ophthalmology practices now employ a new technology of ocular pyramid aber-
rometer. In [64] it was demonstrated that using new pyramid aberrometers Osiris,
repeatable and consistent measurement results of ocular aberrometry of healthy eyes
could be attained.
A wave front sensor with a pyramid-shaped optical element and expanded source
has been successfully used to measure human eye aberrations [65]. In this sensor, a
significant benefit for the eye is that it can easily be adapted to variations in the aber-
ration range, which may be expected of a human eye, varying from healthy eyes with
slight aberration to cornea pathologies causing extreme aberration. The dynamic
range of a sensor can be changed by simply changing the length of a light source
projected onto the retina. These capabilities may prove useful in practical designs of
devices that operate by measuring the human eye wave aberration. A shortcoming of
the sensor is that it also collects the light from false reflections from the eye surface.
In addition, pyramid wave front sensors (PWFS) are widely employed in adap-
tive optics systems, including astronomy, ophthalmology [66], and microscopy.
Wave front reconstruction algorithms for pyramid wave front sensors are in high
demand. With few exceptions, the majority of currently existing algorithms rely on
an assumption of a linear pyramid sensor model. Nonetheless, the use of a simplified
nonlinear pyramid operator enables an acceptable quality of wave front reconstruc-
tion to be achieved.
Also known is a crossed sinusoidal wave front sensor that is based on a gradi-
ent transfer filter placed near the image plane of the system under testing [67]. The

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 205 9/26/23 5:09 PM

 206 Photonics Elements for Sensing and Optical Conversions

theoretical principle behind the sensor operation is comprehensively described in

the framework of Fourier optics. Numerical simulation confirms that a root mean
square (RMS) measurement accuracy of λ/80 can be attained, which is much more
accurate compared to other types of wave front sensors (WFS). A crossed-sine WFS
also offers a benefit of being quasi-сhromatic and able to operate with small-length,
natural, or artificial light sources.
The crossed sinusoidal WFSs belong to a class of WFSs in which the pupil of the
system under testing is imaged on a sensor array, thus providing high spatial resolu-
tion. In this aspect, it is closer to a pyramid WFS or an optical differentiation sensor
(ODS) [68].
Interesting studies were conducted using Shack-Hartmann aberrometers [69].
Measurements of higher-order aberrations (HOA) conducted using a tracing aber-
rometer iTrace and Shack–Hartmann aberrometer Topcon KR-1 W were evaluated
for the accuracy and consistency provided. The results arrived at demonstrated
excellent repeatability but less reliable reproducibility when measuring HOA. Thus,
it may be concluded that due to essentially different results of HOA measurements,
the two aberrometers should not be mutually replaced when applied clinically.
A hand-held wave front aberrometer with post-cycloplegic autorefraction (AR)
and cycloplegic refraction (CRf) has shown good agreement of measurement results
with post-cyclopegic AR and CRf in spherical equivalents but tended to produce
slightly myopic results. Considering that the wave front aberrometer has also been
shown to overestimate astigmatism, it is recommended that a device for refraction
anomaly evaluation be additionally used, which may also be helpful when testing
individuals with abnormal posture and bedridden patients [70].
Also popular in ophthalmology are Scheimpflug sensors [71]. Accurate cornea
analysis is important for pre-surgery risk evaluation and early detection of ectasia after
the refractive surgery [72]. The analyzer DRS (dual rotating Scheimpflug) (Galilei®;
Ziemer Ophthalmology, Port, Switzerland) utilizes two rotating Scheimpflug cam-
eras in combination with a Placido topographical system. The analyzer utilizes a
Placido disc in order to obtain more accurate topographical data on the anterior
cornea curvature as addendum to the data from Scheimpflug cameras. Moreover,
the double system generates images on both sides, thus minimizing the decentering
effect due to eye movements on the cornea pachymetry and posterior cornea curva-
ture measurements [73]. On the whole, the DRS analyzer enables the anterior cornea
curvature for both a healthy eye and an eye after refractive surgery to be measured
with good repeatability and reproducibility [74].
To achieve a higher sensitivity and specificity in keratoconus (KC) detection,
ophthalmologist practices utilize Scheimpflug-tomograph-type aberrometers, for
instance, an Oculyzer (Alcon, USA). The device is intended for doing computer
tomography of the cornea and analyzing the anterior segment of the eyeball. With
the Oculyzer, the examination results can be presented in a visual form convenient
for further use, for instance, making it possible to move in space a 3D cornea model
together with all its parameters. Other examination results are visualized as color
diagrams, maps, and 3D images, providing an ophthalmologist with a comprehen-
sive picture, making it possible to make a maximally accurate diagnosis, draw up a
plan and propose treatment techniques, and prognosticate the treatment outcome.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 206 9/26/23 5:09 PM

 Wave Front Aberration Sensors 207

7.3.2 Medical Ophthalmological Measurement Devices

Among a multitude of ophthalmological measurement devices, we may single out
several devices that are most widely used clinically and described in the specialist
literature. The detectors presented here are discussed in high-impact journals and
analyzed in research theses, with data from those detectors underlying a decision-
making process on the diagnosis and the need for surgical intervention.

7.3.2.1 ALLEGRO Topolyzer VARIO

ALLEGRO Topolyzer VARIO is a medical measurement instrument for the human
eye examination, a diagnostic medical device of IIb class according to the Medical
Device Directive 93/42/EEC classification. It is used for measuring the cornea
topography and was specifically designed for ophthalmologists. The measurement
data can be exported and used for refractive surgery with laser systems WaveLight
AG (with topographic guidance).
The illumination system, containing a custom reflector, produces concentric
intensity rings on the backside of a transparent dome. The dome image is reflected
from the eye under examination and a virtual image is captured with a precision
lens and an objective of a digital CCD camera. In this way, all distortions in the cur-
vature radii of the eye become accessible for the measurement data processing. An
analogous image is then processed with a measuring device, becoming accessible
for further processing as a transformed compressed digital image. After getting the
relevant data from a diagram, the software uses this information to produce a topo-
graphic image of the cornea surface. Finally, the measurement results are displayed
on the screen not only in colors but also as a diagram and a 3D image.
This device allows a wave front analysis based on expansion in the Zernike basis
to be conducted using data on height measurements by calculating a coefficient that
describes the contribution of each Zernike polynomial to the heights measured.
The device operates by using index calculus data: data on the curvature and height,
Fourier analysis, and an output Zernike analysis. Such a comprehensive use of the
key cornea-related data enable an early diagnosis of a wide variety of eye and sight
impairments.

7.3.2.2 WaveLight Oculyzer

WaveLight Oculyzer II is a diagnostic instrument intended for the identification and
examination of the anterior eye segment, which is specially manufactured for clini-
cal uses in ophthalmologist practices, as well as for optometrist and optician offices.
The device measures the topography of the anterior chamber of the eyeball
(including the cornea, aqueous humor, iris, and the front surface of the intraocular
lens). The optical power of the cornea is calculated using a Gullstrand model. Other
functionalities of the device include the estimation of the cornea shape, the analysis
of the intraocular lens condition (lenticular opacity), analysis of the anterior cham-
ber angle, analysis of the anterior chamber depth, analysis of the anterior chamber
volume, analysis of the anterior and posterior cortical opacity, analysis of cataract
localization (kernel, subcapsular, or cortical), the use of cross images with densitom-
etry, and cornea thickness measurements.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 207 9/26/23 5:09 PM

 208 Photonics Elements for Sensing and Optical Conversions

The device also performs cornea measurements and is custom designed for oph-
thalmologists. The data measured can be exported and utilized jointly with a laser
system WAVE ALLEGRETTO (OcuLink) for refractive surgery. The device per-
forms a Zernike analysis of the front and rear cornea surfaces based on height mea-
surements, calculating for each Zernike polynomial a coefficient that describes its
contribution to the heights measured.
Using Zernike polynomials, the user can also characterize the front and rear
surface of the cornea independently. While mathematically calculating the Zernike
coefficients, the polynomials used are being corrected in an optimal way with respect
to the heights measured. The larger the number of the polynomials employed, the
higher the accuracy of height calculations. If the abnormal Zernike coefficient is
absent, the aberration coefficient is 0.0. Values exceeding 1.0 indicate that the cor-
nea surface contains untypical wave components, which may result in an impaired
vision. When surgically correcting keratoconus, it is only possible to reform the front
cornea surface, with the rear surface remaining astigmatic and thus affecting the
retinal image clarity. Moreover, a change in the cornea thickness leads to a change in
the entire chamber of the eyeball due to the aqueous humor pressure.
All the factors just mentioned should be accounted for when doing refractive laser
surgery. The apparatus WaveLight Oculyzer II makes it possible to represent the
front and rear cornea surfaces as Zernike polynomials, also allowing the polynomial
weights and deviations from standard “ideal” values to be measured.

7.3.2.3 HD Analyzer
Considering that an objective measurement of image quality is an indispensable step
in eye diagnostics, the firm VISIOMETRICS has designed an HD Analyzer – a new
instrument that relies on a double passage approach providing an objective clinical
evaluation of eyes’ optical quality.
The device operates on a principle of refraction of a point light source from the
retina. As a result, light passes through the eye medium twice. The HD Analyzer
analyzes the size and shape of the reflected light spot. The design of a dual-passage
aberrometer is based on a Shack–Hartmann sensor. The device contains all informa-
tion pertaining to optical properties of the eye, including all higher-order aberrations
and scattered light, which are usually overlooked by the majority of aberrometry
methods. The said higher-order aberrations may have a significant impact on the
refractive surgery outcome due to light scatter in an older eye. Key data provided by
the HD Analyzer include OSI (objective scattering index), MTF (modulation transfer
function), and PScF (point scattering function).
OSI (objective scattering index) is a parameter that enables the intraocular scat-
tered light to be objectively estimated and calculated by evaluating the amount of
light on the periphery of a dual-passage image relative to the amount of light at the
center. Thus, the higher the OSI, the higher the intraocular scattering level.
This is the only parameter that allows the intraocular scattered light to be evalu-
ated quantitatively. It is useful in all clinical situations when the amount of scat-
tered light may be of clinical significance, such as the development of a cataract
and surgical intervention, refractive surgery, intraocular lens, aging process, dry eye
syndrome, and the like.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 208 9/26/23 5:09 PM

 Wave Front Aberration Sensors 209

MTF (modulation transfer function) is a function that allows evaluating the level
of detail in the image following the passage through an optical system, evaluating
the ratio of the contrast between a system-generated image and the source image. In
the human eye, the MTF represents the loss of contrast following the eye passage.
Like in any optical system, a decrease in contrast in the human eye is higher
for high spatial frequencies (fine details in the image). Thus, MTF is a function of
spatial frequency. It is possible to simulate an ordinary scene imaged in the patient’s
retina. In particular, the program presents an image of an infant put 1 m away from
the viewer. The modeling is conducted through the convolution of the original scene
with the eye PScF measured by the device. Thus, it shows how aberrations and intra-
ocular scattering in the patient’s optical system affect the image acquired. This does
not mean, however, that an individual sees the same image as that shown on the
screen, because the modeling takes into account only an optical image quality rather
than neural network processing of the retinal image.
Thus, using the HD Analyzer it is possible to obtain quantitative and objective
estimates of (1) the intraocular scattered light, (2) optical quality of the eye, (3) opti-
cal quality loss due to tear film destruction, (4) accommodation and pseudo-accom-
modation, and (5) maps of a dual-passage retinal image.

7.3.2.4 Tomey TMS-4

Tomey TMS (Japan) is a diagnostic instrument based on Placido discs and intended
for kerato topography measurements, allowing one to obtain a high-quality photo-
graph of the cornea surface that is undistorted with shadows cast by superciliary
arches and nose. With 25 discs used in the optical scheme and 256 data points being
analyzed on average on each disc, the device makes it possible to evaluate the cornea
refractivity at 6,400 points (TMS-4, User manual).
With the kerato topograph TMS-4, abnormal astigmatism can be diagnosed using
different types of maps – sagittal and tangential ones, which use different curva-
ture radii when measuring the refractive power of the cornea. The sagittal map is
calculated using a cornea radius related to the vision (sagittal) axis, whereas the
tangential map is calculated using an instantaneous curvature radius unrelated to the
vision axis. The sagittal map offers more information relating to the cornea refrac-
tivity, while the tangential map provides more data on the cornea shape and local
irregularities.

7.3.2.5 Pentacam HR
The diagnostic instrument Pentacam HR manufactured by the company Oculus
(Germany) belongs to a class of projection mappers or kerato mappers and is intended
for analyzing the anterior eye segment with a Scheimpflug camera.
Unlike Placido disc-based topography operating in reflection and only measuring
parameters of the front cornea surface, this instrument forms an optical cross sec-
tion of the front eye segment by projecting a light slit in a similar way to a narrow
light beam in a slit lamp. Images of the structures under analysis (cornea, anterior
chamber of the eyeball, iris, intraocular lens) are acquired based on a Scheimpflug
principle.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 209 9/26/23 5:09 PM

 210 Photonics Elements for Sensing and Optical Conversions

Using the resulting Scheimpflug images, the instrument software then outlines
contours of the front and back cornea surfaces. With this approach, the cornea shape
can be directly determined by measuring the elevation of its points, based on which
all other parameters are then calculated (tilt, curvature, and optical power). This
approach also allows the reconstructed 3D image of the front eye segment to be
mapped, and data on the cornea thickness to be obtained at each point. The differen-
tial maps can also be used for a comparison of the cornea refractive power (sagittal
map) and corneal pachymetry.

7.3.2.6 OPD-Scan ARK-10000 (Nidek)

The OPD-Scan aberrometer comprises a wave front analyzer (seven measurement
zones, 2,520 points at the eye pupil projection), a Placido disc-based topographer
with a keratometry interval ranging from 33.75 to 67.5 dioptres (D), 11,880 mea-
surement points, and a refraction analyzer. Unlike a standard refractometry method
employed in autorefractometers, which measure refraction in a 3-mm region, the
OPD-Scan aberrometer measures refraction in 3-mm and 5-mm regions. Besides,
the OPD-Scan features a wider range of measurements of the cylindrical refrac-
tion component from 0 to ±16D, as compared with a standard autorefractometer
(from 0 to ±10D), which is a significant benefit for patients with irregular corneal
astigmatism.
Among the multitude of parameters the aberrometer measures, the most impor-
tant and informative parameter of image quality in an eye optical system is the point
spread function (PSF). The size and spatial distribution of the light energy from
a point light source focused on the retina after light has traveled through the eye
refractive medium determine the PSF.
In addition, the instrument calculates an important parameter enabling the evalu-
ation of eye optical quality, namely, the Strehl coefficient [75–76]. It describes the
ratio of the PSF maximum in a diffraction-free (ideal) optical system to the PSF
maximum in a system with aberrations. The Strehl coefficient takes values in the
interval 0 < St < 1. If an optical system is aberration-free, the Strehl coefficient
equals one. For a healthy human eye, the Strehl coefficient with regard to higher-
order aberration is about 0.05.

7.3.3 Analysis of Wave Front Sensors in Ophthalmological Studies

Potentialities of wave front sensors for solving tasks in applied ophthalmology were
experimentally studied in eye clinics around the world. Among most recent stud-
ies (2018–2020), some publications concerned with an analysis of different types of
aberrometers are worth special mentioning.
Moorfields Eye Hospital (UK, London) reported results of a study that examined
human eyes following treatment under a wave front sensor control that involved the
use of a pyramid aberrometer (SCHWIND eye-tech-solutions GmbH) [77]. Similar
studies were conducted at the eye clinic Maja Clinic (Serbia, Nish), where the instru-
ments used included a WaveLight Allegro Oculyzer, a WaveLight Allergo Biograph,
and an ultrasound pachymeter DGH Pachette 3. Average values of the central corneal
thickness measured using different instruments were put under comparison [78].

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 210 9/26/23 5:09 PM

 Wave Front Aberration Sensors 211

Studies conducted at the eye clinic Al Vatani (Egypt, Cairo) looked into detecting
keratoconus (KC) with higher sensitivity and specificity by means of a Scheimpflug
sensor called Oculyzer [79]. At Beijing Tongren Eye Center (China, Beijing),
researchers conducted a comparative study of aberrometers Oculyzer and Topolyzer
Vario for eye measurements prior and after a corneal refractive surgery [80]. In addi-
tion, Branchevsky Eye clinic (Samara, Russia) conducted a comparative analysis of
optical instruments based on Placido discs, Scheimpflug camera, and optical coher-
ence tomography (OCT) for kerotometry measurements in patients after laser eye-
sight correction [81].

7.3.3.1 Study of a Pyramid Aberrometer Peramis

Moorfields Eye Hospital (UK, London) published results [77] of the examination of
eyes following myopia laser-assisted in situ keratomileusis (LASIK) treatment, under
wave front control, executed with an excimer laser Amaris 1050RS and a pyramid
aberrometer Peramis (SCHWIND eye-tech-solutions GmbH). Repeatability limits of
the results were calculated for spherical refraction and higher-order aberration through
consecutive eye scans made prior and after the surgical treatment for the first 100 mea-
surements. The repeatability limits were found to be 95% for pyramid aberration mea-
surements. This data suggests good repeatability of measurements, alongside safety
and effectiveness of pyramid aberrometry upon a routine myopic LASIK procedure.

7.3.3.2 Analysis of the Use of Aberrometers Oculyzer and Biograp

Measurements of the central corneal thickness (CCT) is an important step of the eye
examination preceding any corneal refractive surgery procedures and post-surgery
patient follow-up, observation of patients with corneal diseases including kerato-
conus and Fuchs corneal dystrophy, as well as in CCT-dependent intraocular pres-
sure control. Such studies were conducted at a specialist eye clinic “Maja Clinic”
(Nish) using instruments such as WaveLight Allegro Oculyzer, WaveLight Allergo
Biograph, and ultrasound pachymeter DGH Pachette 3. A comparison was conducted
of average CCT values measured with the different instruments [78].
With no statistically meaningful differences revealed between CCT measure-
ments conducted using an Oculyzer, BioGraph, and ultrasound pachymetry, the CCT
measurement results from any of the three instruments can be considered reliable.
The authors of the research came to a conclusion that any of the measurement
results obtained by the instruments may independently be considered as reliable and
form a basis for selecting therapy tactics. Each of the instruments may find its place
in ophthalmology depending on the patient’s eye condition and required treatment
procedures.

7.3.3.3 Analysis of the Use of the Aberrometer Oculyzer

Research by ophthalmologists at the eye clinic Al-Vatani (Egypt, Cairo) [79] was
aimed at identifying indices for detecting keratoconus (KC) with higher sensitivity
and specificity compared to those currently available with the use of the software for
rotational Scheimpflug visualization (Oculyzer I, Pentacam). Note that while Placido
disc-based systems provide accurate data in many cases of KC, they have essential
limitations associated with the curvature map. The cornea coverage area is limited to

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 211 9/26/23 5:09 PM

 212 Photonics Elements for Sensing and Optical Conversions

about 60% of the entire cornea surface, thus excluding critical data relating to many
peripheral or paracentral pathologies and providing no information relating to the
rear corneal surface. Moreover, it appears impossible to construct pachymetry maps
that present the cornea thickness distribution. In addition, there are limited possibili-
ties to reconstruct the cornea topography from curvature measurements data [82].
In the study at the eye clinic Al-Vatani (Cairo, Egypt), patients’ eye visualiza-
tion was conducted by the Pentacam company under a trademark Allegro Oculyzer
(WaveLight, GmbH, Erlangen, Germany) with the aid of the software 1.16r12. The
limitations of the said study are that it was conducted using just a base Pentacam
model (Allegro Oculyzer I), which is currently utilized by many eye centers. Further
studies of Pentacam HR are recommended because when conducting refractive sur-
gery screening, conventional high-resolution Scheimpflug visualization instruments
produce different objective values and the two instruments are not mutually replace-
able. It is also recommended that the accuracy of the indices should be compared
with a new biochemical index that is based on deformation parameters and cornea
thickness profile because it is shown to be highly accurate in distinguishing between
a normal cornea and a KC-affected cornea.

7.3.3.4 Comparative Analysis of Sensors Oculyzer and Topolyzer

A study conducted at Beijing Tongren Eye Center (China) was concerned with a
consecutive comparison of data from aberrometers Oculyzer and Topolyzer Vario
[80] prior and during the corneal refractive surgery conducted with an excimer laser
WaveLight EX500.
The Oculyzer and Topolyzer Vario map the correlation between the Cartesian
coordinates XY of the apex and the eye pupil center. When the patient’s eye focuses
on a target, the (reproducible) cornea apex is the highest point most close to the vision
ray, irrespective of the pupil size. Should a human-eye optical system be coaxial, the
optical axis and the cornea would intersect exactly at the cornea apex. For the correct
ablation adjustment, the surgeon needs to analyze the pre-surgery corneal topography
in combination with intra-surgery condition. It was experimentally revealed that the
effect of a manual adjustment of Kapp’s tilt angle based on the Oculyzer topography
manual is analogous to the automated tilt angle adjustment based on the Topolyzer
Vario topography when using an excimer laser WaveLight EX500. The manual adjust-
ment based on the Oculyzer topography is recommended in the case of discrepancy
between the Topolyzer Vario topography and a live image during surgery.
Ophthalmologists managed to demonstrate a high level of similarity between pat-
terns of Oculyzer topography and Topolyzer Vario topography. The data of Oculyzer
topography becomes significant when Topolyzer Vario topography images do not coin-
cide with a live image in the course of surgery with an excimer laser WaveLight EX500.

7.3.3.5 Analysis of the Diagnostic Instruments Oculyzer,

Tomey AO-2000, and IOL Master
Ophthalmologists at Branchevsky Eye Clinic (Samara, Russia) conducted a com-
parison [81] of topography maps of a group of patients who underwent excimer-
laser-aided myopia correction. The parameters were measured using three different
instruments: IOL Master 700, Tomey AO-2000, and Oculyzer.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 212 9/26/23 5:09 PM

 Wave Front Aberration Sensors 213

Schleimpflug tomography was implemented using an Oculyzer (Alcon, USA)

by making 25 meridian scans, with the analysis conducted using SimK indices.
Keratometry data obtained using Placido disc-based topography were recorded on
an optical biometer with built-in Placido topographer AO-2000 (Tomey, Japan). With
this instrument, keratometry of corneal areas of different diameters can be done.
For a comparative analysis, keratometry data from a 2.5-mm zone were utilized. An
optical biometer IOL Master 700 (Carl Zeiss Meditec AG, Germany) operates by
using an OCT principle with a tunable wavelength.
No statistically meaningful differences between the average values of keratom-
etry of a steep meridian, keratometry of a flat meridian, and an average value of
keratometry obtained from the three instruments analyzed were revealed. However,
statistically meaningful differences were revealed between the instruments Tomey
AO-2000 (Placido) and a Schleimpflug tomograph Oculyzer when measuring astig-
matism. Although the study did not intend to identify the instrument providing the
most reliable keratometry data, finding essentially different keratometry patterns
could be useful for the interpretation of the examination data.

7.3.4 Attempts of an Analysis of Eye Aberration

Our studies aimed to solve a problem of differential diagnostics of eye cornea optical
structures through an analysis of focusing system aberrations. In [20, 83], an analysis
of eye cornea aberrations was conducted using data presented by Branchevsky Eye
Clinic. As a result of the analysis, basis Zernike functions most characteristic of
some particular eye pathologies were revealed.
An analysis of weight coefficients most informative in terms of the classification
of patient diagnoses was conducted. As a result, basis Zernike functions most infor-
mative in terms of diagnosing certain eye pathologies were identified.

7.3.4.1 Analysis of a Human Eye Corneal Aberrations

People are known to obtain a predominant proportion of information about the out-
side world with their eyes; hence, the quality of life deteriorates with deteriorat-
ing vision. Science has accumulated enormous expertise fighting eye diseases, from
which it became clear that a major problem is age-related changes in the eye optical
system, which can be evaluated on the basis of wave aberrations.
The human eye can be described as a lens array composed of three major com-
ponents: a cornea, a pupil, and a crystalline lens [84]. All optical characteristics of a
healthy eye are defined by a combination of aberrations in the cornea and intraocular
optics.
Studies [85–87] have shown that the total amount of aberration in the whole eye
is always lower than that in the front corneal segment or internal optics. This is
explained by the fact that aberrations of the cornea and the crystalline lens compen-
sate for each other [88]. There is convincing proof that the aberration compensation
between the cornea and intraocular optics occurs in the case of astigmatism (along
horizontal/vertical axis), horizontal coma, and spherical aberration [89]. The general
result of aberration compensation consists in a decrease of aberration amount in the
retina plane, potentially leading to a higher-quality optical image in the foveola. In

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 213 9/26/23 5:09 PM

 214 Photonics Elements for Sensing and Optical Conversions

young people, aberrations of the cornea and crystalline lens are normally higher than
those in the whole eye. This fact lends support to the hypothesis that the crystalline
lens of the human eye plays an important role in the compensation of the cornea
aberration, resulting in an improved retina image.
Wave front analysis using the Zernike basis expansion facilitates making a diag-
nosis in clinical ophthalmology, also making it possible to design almost perfect
lenses enabling the best eyesight correction possible to date.
Data on aberration of the human eye optical system was obtained using
Branchevsky Eye Clinic aberrometers WaveLight Oculyzer II and HD Analyzer (a
fragment of the data is shown in Figure 7.22).
For a better visual representation, Figure 7.23 depicts a “Zernike pyramid” com-
posed of several first Zernike functions. From top to bottom, radial numbers are varying
from n = 0 to 4 and from left to right – azimuthal numbers vary from m = –n to m = n.
Patterns of the point spread function (PSF) in the presence of typical aberra-
tions were numerically constructed using a simplest Fourier correlator scheme in the
Zemax programming environment [90].
The optical scheme in Figure 7.24 consists of two identical lenses made from BK8
glass and an aperture of radius 3 mm at the center, with the numerical simulation

ID Age Eye Doctor’s Diagnosis Z00 Z11 Z1-1 Z22 Z20

1 27 OS Low myopia 1.34E-01 −4.18E-04 −1.33E-03 6.35E-04 7.83E-02
2 27 OD Low myopia 1.33E-01 −1.52E-03 8.90E-04 6.94E-04 7.78E-02
3 20 OS Mediate myopia 1.23E-01 3.75E-04 −2.10E-04 9.59E-04 7.25E-02
4 20 OD Mediate myopia 1.24E-01 −7.31E-04 6.93E-04 3.21E-04 7.28E-02
5 24 OS Mediate myopia 1.26E-01 −9.28E-04 6.60E-04 4.99E-04 7.40E-02
6 24 OD Mediate myopia 1.26E-01 −8.69E-04 2.50E-05 5.42E-04 7.39E-02
7 27 OS Low myopia with 1.31E-01 −1.33E-03 −1.36E-03 5.34E-03 7.73E-02
compound myopic
astigmatism
8 27 OD Low myopia with 1.30E-01 −8.83E-04 1.77E-03 1.15E-03 7.65E-02
compound myopic
astigmatism
9 23 OS Low myopia 1.41E-01 2.25E-04 −3.55E-04 7.54E-04 8.33E-02
10 24 OD Low myopia 1.42E-01 1.82E-03 −4.10E-05 8.81E-04 8.41E-02
11 24 OS Mediate myopia with 1.37E-01 1.07E-03 −1.07E-03 8.68E-04 8.06E-02
compound myopic
astigmatism
12 24 OD Mediate myopia with 1.38E-01 4.81E-04 5.31E-04 1.13E-03 8.09E-02
compound myopic
astigmatism
13 22 OS Mediate myopia with 1.25E-01 –1.38E-03 1.05E-03 1.01E-03 7.33E-02
compound myopic
astigmatism

FIGURE 7.22 A fragment of the data obtained in the studies (Zernike weight coefficients).

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 214 9/26/23 5:09 PM

 Wave Front Aberration Sensors 215

FIGURE 7.23 Several examples of Zernike functions.

conducted for a 780-nm wavelength. Rather than using the lens radius, the Zernike
polynomials were normalized relative to the output pupil radius calculated using
Zemax, which was found to be 8.37 mm. The first surface of the second lens was
consecutively subjected to third-order coma aberrations (Figure 7.25) corresponding
to the seventh Zernike function (the coefficient assumed to equal 1).
The system just described can be numerically simulated using the following algo-
rithm. The wave front can be constructed as superposition of Zernike polynomials
using relevant formulae before calculating the PSF as a Fourier transform of the
derived distribution. For a more visual representation, Figure 7.27 depicts a PSF
pyramid corresponding to the Zernike pyramid. The results obtained in this way

FIGURE 7.24 An optical scheme of a Fourier correlator.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 215 9/26/23 5:09 PM

 216 Photonics Elements for Sensing and Optical Conversions

FIGURE 7.25 PSF (a) without aberration, (b) with coma, and (c, d) corresponding test images.

agree well with the numerical simulation performed in the Zemax programming
environment. The possibility of rotating the pattern by angle φ 0 (Figure 7.26) is
made possible through the use of the transform:

ψ ( r , ϕ ) = Rnm ( r )  a sin ( mϕ ) + b cos ( mϕ )  , (7.31)

where

a = sin ( m ϕ 0 ) , b = cos ( m ϕ 0 ) , ϕ 0 ∈[0; 2π].

Let us perform the numerical simulation of image distortion following the intro-
duction of aberration into a wave front incident on an ideal lens of an imaging system.

FIGURE 7.26 Rotation of the PSF of the Zernike function at n = 3, m = 1, φ 0 = {0, π/6, π/4,
π/3, π/2, π, 3π/2, 2π}.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 216 9/26/23 5:09 PM

 Wave Front Aberration Sensors 217

FIGURE 7.27 Patterns illustrate PSFs corresponding to typical aberrations.

For the acquisition of an ideal image f(x, y), an imaging optical system containing an
ideal lens was utilized. A distorted image fa(x, y) was numerically simulated using
the following algorithm: a Fourier image F(u, v) of an ideal image was multiplied
by an aberration-distorted wave front w(x, y), before performing an inverse Fourier
transform fa ( x , y) = ℑ−1  ℑ  f ( x , y)] w ( x , y)]. Thus, convolution of an ideal image
with a distorted PSF was numerically simulated.
The aforementioned algorithm was utilized to study the impact of all Zernike
polynomials up to a fourth order inclusively on image distortion (as shown in Figure
7.27). Some types of the same-order aberrations with the same weight coefficients
but with different signs of the meridional index m were found to introduce a different
degree of distortion into the same test image (a cross, see Figures 7.28 and 7.29). The
reason is that some types of aberration are orthogonal to the test image utilized. This

FIGURE 7.28 A test image is distorted by defocus-type aberration.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 217 9/26/23 5:09 PM

 218 Photonics Elements for Sensing and Optical Conversions

FIGURE 7.29 Distortion of a test image “cross.”

property may be useful when specific phase patterns of DOEs intended to compen-
sate for some aberrations need to be generated.
By way of illustration, aberration in the form of astigmatism Z22 Z44 deteriorates
the image quality in a greater degree when compared with analogous aberration Z2–2
Z4–4. In the meantime, aberration in the form of defocus (Z20) and spherical aberra-
tion (Z40) introduce similar image distortion but differ in the impact degree.
It is worth noting that apart from numbering and normalization, Zernike func-
tions may have different angular dependence. In particular, multichannel diffractive
optical elements [31, 53, 91, 92] utilized for wave front analysis operate by using an
exponential and trigonometric angular dependence. Note that in the former case the
Zernike basis is rotation invariant [31]. In the latter case, the rotation of the pattern
leads to a change in coefficients by the basis functions.
The human eye can be described as a lens array composed of three major com-
ponents: a cornea, a pupil, and a crystalline lens. For the image to be formed on
the retina, the performance of the optical components is being mutually coordi-
nated. The improper functioning of the eye system leads to vision defects, ranging
from a minor blur of the image to complete blindness. Being the predominant
component of an eye’s optical power (~70%), the cornea (tear film included) is the
main source of optical aberrations. The anterior corneal surface has an elongated

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 218 9/26/23 5:09 PM

 Wave Front Aberration Sensors 219

profile with the steeper central region compared to the periphery. This shape
serves to reduce the number of spherical aberrations in the whole eye. However,
corneas of different individuals significantly differ in shape, possibly causing
astigmatism and higher-order asymmetric aberrations (e.g. coma). An ideal eye
focuses an image of the outside world specifically in the foveola (a region contain-
ing only photoreceptors), irrespective of the field angle. However, a perfect image
cannot be formed in a real eye as some amount of aberration deteriorates optical
characteristics of the eye.
Lower-order aberrations are predominant in the human eye, accounting for up
to 90% of all wave aberrations [91]. At the top of the aberration list is defocus, with
positive defocus causing long-sightedness and negative – short-sightedness. The sec-
ond from the top is astigmatism. Correction of low-order aberration with the aid of
spectacles, contact lenses, or laser surgery essentially improves visual acuity in the
majority of cases. On the other hand, lower-order aberrations known to occur in
the human eye currently are unable to be corrected by means of popular optical or
surgical techniques. The eye also has chromatic aberration, which is the result of dis-
persion in optical elements with particular values of the refractive index. Different
eye parts are affected by different wavelengths. Thus, the incident white light is
decomposed into a color spectrum in the eye. Putting it simply, chromatic aberration
is wavelength-dependent spherical refraction. Types of chromatic aberration (disper-
sion) are traditionally divided into longitudinal and transverse ones. The former is
characterized by varying on-axis optical power and wavelength and is relatively the
same in different individuals. Transverse aberration causes the image to be shifted
across the entire image plane depending on the wavelength and is significantly vary-
ing in different individuals. With the real world being polychromatic, chromatic
aberration undoubtedly poses limitations on the image quality.
Diffraction is a fundamental property of light waves, which occurs when light
passes through a pinhole comparable in size with its wavelength. In the eye, diffrac-
tion takes place when light passing through the eye interacts with the iris edge. Even
with no aberrations, an infinitesimally small point cannot be imaged in the retina
due to diffraction. In general, considering that all optical systems (including the eye)
have a rigid diaphragm, it is impossible to design a diffraction-free optical device.
Theoretically, it is possible to improve the image quality in an optical system by
minimizing aberrations. Nonetheless, diffraction-related limits on the image quality
cannot be overcome, and the highest image quality attainable is called diffraction-
limited as it is limited solely by diffraction. Thus, even “perfect” optical systems
cause unavoidable blur due to diffraction. With no aberration, a perfect eye would
transform incoming wave fronts into converging spherical waves. Thus, due to dif-
fraction, an image of a point has a finite size. The image of a point in a diffraction-
limited optical system is called an Airy spot.
Scattered light also hampers the performance of an optical system (of an eye) in
terms of image processing. This results in a lower image contrast, causing halos and
flares. Scattering in the eye medium mostly occurs as a result of diffusion of light
and deterioration of the transparency of the cornea (including the tear film) and the
crystalline lens [53]. Retinal scattering may depend on the fraction of light traveling
along photoreceptors and the retina layer from which the light is reflected (depending

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 219 9/26/23 5:09 PM

 220 Photonics Elements for Sensing and Optical Conversions

on the wavelength used). Higher-order aberrations are comparatively low, with pre-
dominant aberrations represented by third-order comatic aberrations (vertical coma,
horizontal coma, slanted trefoil, horizontal trefoil) and spherical aberration.
The study discussed herein was based on data sets presented by Branchevsky
Eye Clinic, which contained data on conditionally healthy eyes (with no pronounced
pathologies in either the pupil or the crystalline lens) as well as those with patholo-
gies, like low and medium myopia. A reference model of the cornea can be described
by a sphere or an ellipse of eccentricity 0.75. The reference model is always axisym-
metric, with its shape affecting only the coefficients of axisymmetric Zernike func-
tions (1) with the numbers (n, 0). The central radius of the reference model is always
taken to be equal to an average central radius of measurements. In the absence of a
reference model, the largest contribution to the cornea shape comes from the com-
ponent corresponding to the Zernike function (2,0) (a paraboloid), because this com-
ponent is closest to the cornea in shape. Shown in Figure 7.30 is an optical setup
enabling the simplest human eye model to be implemented in Zemax [90]. The
numerical simulation was conducted for a wavelength of 780 nm (the wavelength
used by the ophthalmological instrument HD Analyzer).
A number of polynomials were consecutively applied to the cornea surface.
Figure 7.31 depicts the performance of an optical system after coma described by the
Zernike function (3,1) with the coefficient 1 introduced to the anterior cornea surface.
When specifying a normalization radius for the Zernike polynomials, instead of
using the lens radius, an output pupil radius calculated using Zemax was utilized,
amounting to 100 mm.
Within this study, a simplest statistical analysis was conducted, with the (anterior
and posterior) cornea surfaces of individuals with healthy eyes being independently
tested. The individuals were grouped by age (20–29 years and 30–39 years) and
diagnosis (weak myopia and medium myopia).
Using an aberrometer WaveLight Oculyzer II, Zernike coefficients for the ante-
rior and posterior cornea surface can be measured separately. The averaged values
are given in Figures 7.32–7.35. For a better visualization of the Zernike coefficients,
coefficients with the numbers (n, 0), n = 0, 2, 4 were excluded from consideration (by
putting them equal to zero). Although they are by an order of magnitude higher than

FIGURE 7.30 Schematic view of key eye refractive surfaces: 1 is an anterior cornea sur-
face, 2 is a posterior cornea surface, 3 is a front crystalline lens surface, 4 is a back crystalline
lens surface, and 5 is retina.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 220 9/26/23 5:09 PM

 Wave Front Aberration Sensors 221

FIGURE 7.31 PSF (a) without aberration and (b) with coma and (c, d) corresponding test images.

FIGURE 7.32 Values of weight coefficients of the Zernike polynomials for the anterior
cornea surface of healthy eyes.

FIGURE 7.33 Values of weight coefficients for the Zernike polynomials for the posterior
cornea surface of healthy eyes.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 221 9/26/23 5:09 PM

 222 Photonics Elements for Sensing and Optical Conversions

FIGURE 7.34 Values of weight coefficients for the Zernike polynomials for the anterior
cornea surface in patients with weak myopia.

FIGURE 7.35 Values of weight coefficients for the Zernike polynomials for the posterior
cornea surface in patients with weak myopia.

all the other coefficients, weight coefficients for these functions carry no information
whatsoever on any deviations.
Figure 7.36 depicts patterns corresponding to averaged aberrations in the anterior
cornea surface in patients with weak myopia diagnosis and corresponding PSFs.
Similar patterns for the posterior cornea surface are shown in Figure 7.37.

FIGURE 7.36 (а) Averaged aberrations over the anterior cornea surface in patients with
weak myopia and (b) PSF.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 222 9/26/23 5:09 PM

 Wave Front Aberration Sensors 223

(b)

FIGURE 7.37 (a) Averaged aberrations of the posterior cornea surface in patients with weak
myopia and (b) PSF.

(b)

FIGURE 7.38 Combined averaged aberrations (of the anterior and posterior cornea surface)
in patients with weak myopia: (а) wave front and (b) PSF.

To evaluate the impact of whole-cornea aberrations, wave front aberrations were

summed up, with the result shown in Figure 7.38.
In a similar way, results for medium myopia have been derived (Figure 7.39).
Thus, the most characteristic Zernike functions for particular eye pathologies
have been revealed (see for details Figure 7.40). In particular, in a sampling for
patients with weak myopia, a tilt both on the anterior and posterior cornea surfaces
was observed, but astigmatism was revealed on a larger scale (especially for the

FIGURE 7.39 Combined averaged aberrations (of the anterior and posterior cornea surface)
in patients with medium myopia: (a) wave front and (b) PSF.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 223 9/26/23 5:09 PM

 224 Photonics Elements for Sensing and Optical Conversions

FIGURE 7.40 Summary table for the anterior and posterior cornea surface, correlation
between the Zernike polynomials, most pronounced weight coefficients, and the diagnosis.

posterior surface). An analysis of samplings for medium-myopic eyes has shown that
the influence of the third-order coma comes to the forefront. Though the aberrations
typical of low myopia are still retained.

7.3.4.2 Extracting Informative Features Based on Zernike Polynomial

Coefficients for Different Pathologies of the Human Eye
Research conducted in [83] was concerned with an analysis of wave front aberra-
tions linked with changes in the curvature of the human eye cornea. The analysis
was conducted by representing the aberration of the anterior and posterior cornea
surface as a superposition of Zernike functions. The study aimed to examine weight
coefficients of the Zernike polynomials, with the data collected in the course of

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 224 9/26/23 5:09 PM

 Wave Front Aberration Sensors 225

clinical tests at Branchevsky Eye Clinic. Most informative weight coefficients were
analyzed in terms of their use for grouping patients according to particular diagno-
ses. A comparison was conducted of the classification results obtained on the basis
of 30 features derived both on the anterior and posterior cornea surface and using
most informative features. The features were ranked in terms of their information
content for a concrete classification task. The information content was evaluated
based on the values of a separability criterion and additionally evaluated by calculat-
ing a classification error with the aid of a K-means method. As a result of the statisti-
cal analysis, basis Zernike functions most informative for particular eye pathologies
were revealed.
Then, the most informative weight coefficients needed to be extracted from the
presented set in order to reduce the false classification probability. The task was
implemented using a programming language R intended for intelligent data analysis,
with built-in mathematical tools of the language used in the study. The input data
for the analysis included n objects, each being characterized by 30 features, and
classes defined by the following binary relations: weak myopia/medium myopia (I),
weak myopia/weak myopia + astigmatism (II), medium myopia/medium myopia +
astigmatism (III), weak myopia/medium myopia + astigmatism (IV), and medium
myopia/weak myopia + astigmatism (V).
When analyzing the classification quality, a learning sample was constructed
using a new classifier that was calibrated based on a reference vector method [95].
Within an approach called a U-method [96], the classifier was synthesized using
just those objects of the learning sampling that were not found in the test sampling.
True distribution objects can be replaced by objects not used for the classifier syn-
thesis and independent of the objects involved in the classifier synthesis. While the
U-method can be implemented in a number of ways, the probability of false classi-
fication was evaluated using a method of excluding a single object. Objects from the
test sampling were classified using object-class related data from the training sam-
pling using a nearest-neighbor algorithm. In the course of experimental studies, an
average rate of classification error was found to be 0.319 when using the total set of
features. To conduct an analysis of weight coefficients, binary relations for the aver-
age vectors for each of the aforementioned diagnoses were chosen. For convenience
sake, a continuous numbering of Zernike basis functions was proposed, where N =
1:15 are basis functions describing the curvature of the anterior cornea surface and
N = 16:30 are basis functions describing the curvature of the posterior cornea surface
(the correlation between the ordinal number N and the basis function Znm is shown on
the abscissa of the plots in Figures 7.41–7.45).
Let a vector аavg(x1,x2,…,x30) be the center of class a of a 30-dimensional feature
space and a vector bavg(x1,x2,…,x30) be the center of class b. Then, the most informa-
tive features in terms of the classification of the two classes can be revealed from the
difference of coordinates of the two vectors, c = |аavg – bavg|, enabling the proximity
of each weight coefficient to be revealed independently for different pathologies.
Results for the binary relation of interest are presented in Figure 7.41.
Bar charts in the plots enable the most informative features to be revealed.
The larger is the absolute difference for each feature ci, the more informative it is.
Afterwards, the most informative features among the vectors аavg and bavg were

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 225 9/26/23 5:09 PM

 226 Photonics Elements for Sensing and Optical Conversions

FIGURE 7.41 The most informative weight coefficients for the binary relation – weak myo-
pia/medium myopia.

clustered, with corresponding error rates shown in Table 7.5. Table 7.5 suggests that
the most informative features that occur in three and more binary relations (I–V)
are represented by polynomials with the following indices (N): 1, 5, 15, 16, 17, 18,
19, and 20. They describe aberrations including constant (Z00), defocus (Z20), qua-
trefoil (Z4–4) on the anterior cornea surface and constant (Z00), distortion (Z11, Z1–1),

TABLE 7.5
Informative Weight Coefficients for Binary Relations (I–V)
Binary Relations I II III IV V
Numbers of Zernike Weight 1 – 1 1 1
Coefficients (N) 2 2 – – 2
– – 4 4 –
5 – 5 5 –
– – – – 6
– – – – 8
15 15 – 15 –
– – 16 16 16
17 17 – 17 17
18 18 18 18 –
– – 19 19 19
– – 20 20 20
– – – – 21
– – – – 22
– – 23 – 23
– – – – 24
– – – – 25
– – – – 26
– – – – 27
– – – – 28
– – – – 30
Error 0.233 0.160 0.241 0.142 0.239

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 226 9/26/23 5:09 PM

 Wave Front Aberration Sensors 227

FIGURE 7.42 The most informative Zernike weight coefficients for the binary relation
weak myopia/weak myopia + astigmatism.

astigmatism (Z22), and defocus (Z20) on the posterior cornea surface. Table 7.5 con-
tains only features that were found to be informative for distinguishing between at
least one pair of classes.
Similarly, an information content of the coefficients for the binary relations II–V
can be evaluated (Figures 7.42–7.45).

FIGURE 7.43 The most informative Zernike weight coefficients for the binary relation
medium myopia/medium myopia + astigmatism.

FIGURE 7.44 The most informative Zernike weight coefficients for the binary relation
weak myopia/medium myopia + astigmatism.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 227 9/26/23 5:09 PM

 228 Photonics Elements for Sensing and Optical Conversions

FIGURE 7.45 The most informative Zernike weight coefficients for the binary relation
medium myopia/weak myopia + astigmatism.

From Table 7.5, the clustering error rates are seen to exceed values acceptable for
making a diagnosis [97], therefore a decision was made to use a more complex pro-
cedure for selecting most effective features for illness diagnosis [98–100].
The approach relies on statistical analysis techniques when constructing an infor-
mative feature space. In [99, 100], similar approaches to selecting informative fea-
tures for the analysis of medical-biological objects were described.
Thus, the selection procedure aims to find the most informative features for a par-
ticular classification task. In the course of study, the information content was evalu-
ated based on a separability criterion [96, 97] J1 = tr ((T) −1 B), where T = B + W, B
is an intergroup scattering matrix whose elements are calculated using the formula:
g

bij = ∑n (x
k =1
k ik − xi )( x jk − x j ) , i, j = 1, p, (7.32)

where n is the number of elements of a sampling containing p features divided into g

classes. The intergroup scattering matrix W is described by its elements:
g

wij = ∑( x
k =1
ikm − xi )( x jkm − x j ) , i, j = 1, p,
(7.33)

where xikm is the value of the i-th feature of the m-th element in the class
k, xik = 1/nk ∑ nmk=1 xikm is the average value of the i-th feature in the class k,
xi = (1/n ) ∑ gk =1 nk xik is the average value of the i-th feature for all classes, and nk is
the number of elements in the class k. The larger the value of the criterion, the better
the separability of the classes. The information content of features was additionally
evaluated by calculating a clustering error based on a K-means technique. When per-
forming clustering, К equals the number of the classes under analysis, and the error
rate is defined as a proportion of wrongly clustered sampling vectors.
As far as the task under analysis is concerned, evaluating the information content
based on the values of an individual separability criterion would be of low efficiency
because the value of the criterion is low. For instance, if all four classes of interest
are evaluated simultaneously, the criterion value is below 0.1, meanwhile when ana-
lyzing each pair of classes, the criterion value is not higher than 0.2.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 228 9/26/23 5:09 PM

 Wave Front Aberration Sensors 229

TABLE 7.6
Values of Group Separability Criteria for Each
Pair of Classes
Values of Separability Criteria
Binary Relation All Features Informative Features
I 0.909 0.891
II 0.880 0.840
III 0.765 0.728
IV 0.810 0.762
V 0.820 0.778

Because of this, when seeking to discern each pair of classes, the information con-
tent was evaluated by evaluating the contribution of each feature into the separability
of the entire feature space. Thus, features with the lowest contribution to the total infor-
mation content were excluded from the set of the features used to discern two definite
classes (binary relations I, II, III, IV). The features continued to be excluded until the
value of the all-group separability criterion was reduced by more than 0.05. Finally, we
extracted five sets of features sufficiently informative to discern each pair of classes.
Values of the group separability criterion and the clustering error obtained for all 30
initial features and resulting sets of informative features are shown in Table 7.6.
The analysis of Table 7.6 suggests that all features chosen are informative enough
to discern different pathology classes. Importantly, the feature space dimensionality
has been essentially reduced.
In an experimental study, we established a number of comparisons for which the
preceding features were informative, evaluating their efficiency for discerning the
aforementioned classes. The features were broken down as follows: recommended
ones – those utilized in at least five pairs of comparisons; non-recommended ones –
those utilized in no more than two pairs of comparison; with the remaining features
termed as conventional.
Using the derived features as a basis, clustering was performed, with the cluster-
ing errors shown in Table 7.7. From an analysis of Table 7.7, it can be inferred that the
most informative features that are found in all the binary relations of interest (I–V)
are polynomials with indices N: 1, 3, 7, 8, 13, 20, 21, 22, 23, 26, 28. These polynomi-
als describe aberrations, which include constant (Z00), distortion (Z1-1), trefoil (Z33),
coma (Z31), and spherical aberration (Z40) for the anterior cornea surface, and defocus
(Z20), astigmatism (Z2-2), trefoil (Z33), coma (Z31), quatrefoil (Z44), and spherical aber-
ration (Z40) for the posterior surface.
Hence, the conclusion is that for the classes under analysis to be discerned, the use
of the following features may be recommended (Table 7.8): distortion (Z1-1) and coma
(Z31) – the anterior cornea surface; defocus (Z20), astigmatism (Z2-2), trefoil (Z33), qua-
trefoil (Z44), and spherical aberration (Z40) – the posterior surface.
It is interesting to note that highly informative features contain higher-order poly-
nomials Z31, Z33, Z40, Z44 (higher-than-second order). In classical ophthalmology,

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 229 9/26/23 5:09 PM

 230 Photonics Elements for Sensing and Optical Conversions

TABLE 7.7
Summary Table Containing the Most Informative Features on
the Anterior and Posterior Cornea Surfaces
Binary Relations I II III IV V
Numbers of Zernike Weight 1 1 1 – 1
Coefficients (N) – 2 2 2 –
3 3 3 3 –
– – 4 4 4
5 5 – – –
– 6 6 – 6
7 7 7 7 –
8 8 8 8 –
9 9 – – 9
10 – – – –
11 11 – 11 –
– 12 – – 12
13 – 13 13 13
14 14 14 – –
15 – – – –
– 16 16 – –
17 – – 17 –
18 18 – – 18
– 19 – – –
20 20 20 20 20
21 21 21 21 21
22 22 – 22 22
23 23 23 – 23
24 24 – – 24
25 – – 25 –
26 26 26 26 –
27 27 – 27 –
– 28 28 28 28
29 – – – 29
30 – 30 – 30
Error 0.030 0.024 0.049 0.047 0.026

TABLE 7.8
Recommended Features for the Anterior
and Posterior Cornea Surface
Recommended Features
Anterior Cornea Surface Posterior Cornea Surface
Z1–1 Z31 Z20 Z2–2 Z33 Z44 Z40

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 230 9/26/23 5:09 PM

 Wave Front Aberration Sensors 231

specialists take into consideration just the first- and second-order polynomials, which
account for the so-called “cylinder” and “sphere.” However, the study presented here
has shown that the account of higher-order aberrations (third-order coma, trefoil,
quatrefoil, and fourth-order spherical aberration) has enabled a threefold reduction
in the classification error, leading to a more accurate computer-aided diagnosis.
A listing of software codes written in the C++ programming language for the
Visual Studio development environment for calculating a distorted image is provided
in Appendix B.

7.4 CONCLUSION
A review of recent studies conducted by professional ophthalmologists has shown
that there are no wave front sensor designs, currently utilized or proposed, that would
be capable of simultaneously providing high spatial resolution in the pupil of the
optical components under testing and a measurement precision comparable with that
of laser interferometers. Over the last decade, aberrometers have found applications
in many areas of ophthalmology, including detection of refraction anomalies, kera-
toconus diagnosis, and refractive surgery. Practicing ophthalmologists use a wide
variety of instruments for the eye examination, such as ocular pyramid and tracing
aberrometers, Shack–Hartmann sensors, and Scheimpflug cameras in combination
with Placido disc-based topographers.
For the ophthalmologist to be able to make a maximally accurate diagnosis, draw
up an adequate plan, select treatment procedures, and prognosticate the treatment
outcome, the patient’s eyesight needs to be thoroughly examined. Independent stud-
ies [77–82] conducted in a number of eye clinics show that popular wave front sen-
sors, such as ALLEGRO Topolyzer VARIO, WaveLight Oculyzer, HD Analyzer,
Tomey TMS-4, Pentacam HR, OPD-Scan ARK-10000, and others, generally pro-
vide consistent data. However, in specific cases, measurements of the human eye
optical system may produce different results, which may be due to both techno-
logical reasons (different indirect technologies and data processing algorithms) and
varying approaches to data interpretation.
The analysis of the human eye aberrations was conducted using measurements
data provided by Branchevsky Eye Clinic [20]. Aberrations were represented as a
superposition of Zernike functions. As a result of the analysis [which excluded from
consideration coefficients corresponding to the functions with the numbers (n,0), n =
0,2,4], Zernike basis functions most typical of some eye pathologies were identified.
In particular, a sampling for weakly myopic eyes revealed a tilt for both anterior and
posterior cornea surfaces, but astigmatism was found to be prevailing (especially for
the posterior surface). In patients with medium myopia, the sampling analysis found
that the third-order coma had an essential effect, meanwhile aberrations typical of
weak myopia were also retained.
A study reported in [83] looked into wave front aberrations caused by variations
in the human eye corneal topography. The analysis relied on the representation of
anterior and posterior cornea surface aberrations as a superposition of Zernike func-
tions. The study focused on weight coefficients of the Zernike functions, with the
data obtained in a series of clinical studies conducted at Branchevsky Eye Clinic.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 231 9/26/23 5:09 PM

 232 Photonics Elements for Sensing and Optical Conversions

The study analyzed most informative weight coefficients in terms of the classifica-
tion of patients by definite diagnoses. We proposed two approaches to extracting
informative coefficients, with the feature extraction technique based on separabil-
ity criteria shown to be most efficient for solving a particular classification task.
Interestingly, among highly informative features there are higher-than-second order
polynomials Z31, Z33, Z40, Z44. In classical ophthalmology, specialists take into con-
sideration just the first- and second-order polynomials, which account for the so-
called “cylinder” and “sphere.” However, the study presented here has shown that
the account of higher-order aberrations (third-order coma, trefoil, quatrefoil, and
fourth-order spherical aberration) has enabled a threefold reduction in the classi-
fication error, leading to a more accurate computer-aided diagnosis. Therefore, the
said features need to be used when correcting the eye cornea surface curvature to
improve vision acuity.
Thus, based on the experience gained when studying wave front aberrations in
the human eye, we have demonstrated the relevance of the development of new types
of aberrometers and wave front sensors for professional ophthalmology, enabling
the human eye optical system to be better evaluated and the accuracy of computer-
aided diagnosis to be enhanced, making possible the transition to personified hi-tech
medicine.

REFERENCES
1. David, F. 2015. Buscher practical optical interferometry. Cambridge: Cambridge
University Press. ISBN: 978-1-107-04217-9.
2. Malacara, D. 2007. Optical shop testing. Hoboken, NJ: John Wiley & Sons Inc. ISBN:
978-0-471-48404-2.
3. Vasil’ev, L. A. 1971. Schlieren methods. New York, Jerusalem, London: Israel Program
for Scientific Translations. ISBN: 978-0-7065-1100-0.
4. Hartmann, J. 1900. Bemerkungen über den Bau und die Justierung von Spektrographen.
Zeitschrift für Instrumentenkunde. 20:17–27, 47–58.
5. Artzner, G. 1992. Microlens arrays for Shack-Hartmann wavefront sensors. Opt. Eng.
31(6):1311–22. https://doi.org/10.1117/12.56178.
6. Minin, I., and Minin, O. 2015. Diffractive optics and nanophotonics. Resolution
below the diffraction limit. Cham: Springer International Publishing. ISBN:
978-3-319-24251-4.
7. Zernike, F. 1955. How I discovered phase contrast. Science. 121(3141):345–9. https://
doi.org/10.1126/science.121.3141.345.
8. Situ, G., Warber, M., Pedrini, G., and Osten, W. 2010. Phase contrast enhancement
in microscopy using spiral phase filtering. Opt. Commun. 283(7):1273–7. https://doi.
org/10.1016/j.optcom.2009.11.084.
9. Zernike, F. 1934. Beugungstheorie des schneidenver-fahrens und seiner verbesserten
form, der phasenkontrastmethode. Physica. 1(7–12):689–704. https://doi.org/10.1016/
S0031-8914(34)80259-5.
10. Nijboer, B. R. A. 1943. The diffraction theory of optical aberrations: Part i: General dis-
cussion of the geometrical aberrations. Physica. 10(8):679–92. https://doi.org/10.1016/
S0031-8914(43)80016-1.
11. Nijboer, B. R. A. 1947. The diffraction theory of optical aberrations: Part II: Diffraction
pattern in the presence of small aberrations. Physica. 13(10):605–20. https://doi.
org/10.1016/0031-8914(47)90052-9.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 232 9/26/23 5:09 PM

 Wave Front Aberration Sensors 233

12. Nienhuis, K., and Nijboer, B. R. A. 1949. The diffraction theory of optical aberrations:
Part III: General formulae for small aberrations; Experimental verification of the theo-
retical results. Physica. 14(9):590–604. https://doi.org/10.1016/0031-8914(49)90002-6.
13. Born, M., and Wolf, E. 1999. Principles of optics: Electromagnetic theory of propaga-
tion, interference and diffraction of light. 7th ed. Cambridge: Cambridge University
Press. ISBN: 978-0-521-64222-4.
14. Roddier, N. 1990. Atmospheric wavefront simulation using Zernike polynomials. Opt.
Eng. 29(10):1174–80. https://doi.org/10.1117/12.55712.
15. Neil, M. A. A., Booth, M. J., and Wilson, T. 2000. New modal wave-front sensor:
A theoretical analysis. J. Opt. Soc. Am. A. 17(6):1098–107. https://doi.org/10.1364/
JOSAA.17.001098.
16. Thibos, L. N., Applegate, R. A., Schwiegerling, J. T., and Webb, R. 2002. Standards for
reporting the optical aberrations of eyes. J. Refract. Surg. 18:S652–S660. https://doi.org/
10.3928/1081-597X-20020901-30.
17. Evans, C. J., Parks, R. E., Sullivan, P. J., and Taylor, J. S. 1995. Visualization of sur-
face figure by the use of Zernike polynomials. Appl. Opt. 34(34):7815–9. https://doi.
org/10.1364/AO.34.007815.
18. ANSI Z80.28. 2004. Methods for reporting optical aberrations of eyes. American
National Standards Institute Inc.
19. Lombardo, M., and Lombardo, G. 2010. Wave aberration of human eyes and new
descriptors of image optical quality and visual performance. J. Cataract Refr. Surg.
36(2):313–31. https://doi.org/10.1016/j.jcrs.2009.09.026.
20. Khorin, P. A., Khonina, S. N., Karsakov, A. V., and Branchevskiy, S. L. 2016. Analysis
of corneal aberration of the human eye. Comput. Opt. 40(6):810–7. https://doi.org/
10.18287/0134-2452-2016-40-6-810-817.
21. Martins, A. C., and Vohnsen, B. 2019. Measuring ocular aberrations sequentially
using a digital micromirror device. Micromachines. 10(2):117. https://doi.org/10.3390/
mi10020117.
22. Soifer, V. A., ed. 2002. Methods for computer design of diffractive optical elements.
New York: John Wiley & Sons Inc. ISBN: 978-0-471-09533-0.
23. Soifer, V. A., ed. 2013. Computer design of diffractive optics. Cambridge: Cambridge
International Science Publishing Limited & Woodhead Publishing Ltd. ISBN:
978-1-84569-635-1.
24. Picart, P. 2015. New techniques in digital holography. John Wiley & Sons Inc. ISBN:
978-1-84821-773-7.
25. Golub, M. A., Karpeev, S. V., Krivoshlykov, S. G., Prokhorov, A. M., Sisakyan, I. N.,
and Soifer, V. A. 1984. Spatial filter investigation of the distribution of power between
transverse modes in a fiber waveguide. Sov. J. Quantum Electron. 14(9):1255–6. https://
doi.org/10.1070/QE1984v014n09ABEH006201.
26. Golub, M. A., Karpeev, S. V., Kazanskiĭ, N. L., Mirzov, A. V., Sisakyan, I. N., Soifer,
V. A., and Uvarov, G. V. 1988. Spatial phase filters matched to transverse modes. Sov. J.
Quantum Electron. 18(3):392–3. https://doi.org/10.1070/QE1988v018n03ABEH011528.
27. Almazov, A. A., Khonina, S. N., and Kotlyar, V. V. 2005. Using phase diffraction opti-
cal elements to shape and select laser beams consisting of a superposition of an arbitrary
number of angular harmonics. J. Opt. Technol. 72(5): 391–9. https://doi.org/10.1364/
JOT.72.000391.
28. Khonina, S. N., and Ustinov, A. V. 2019. Binary multi-order diffraction optical ele-
ments with variable fill factor for the formation and detection of optical vortices of
arbitrary order. Appl. Opt. 58(30):8227–36. https://doi.org/10.1364/AO.58.008227.
29. Ha, Y., Zhao, D., Wang, Y., Kotlyar, V. V., Khonina, S. N., and Soifer, V. A. 1998.
Diffractive optical element for Zernike decomposition. Proc. SPIE. 3557:191–7. https://
doi.org/10.1117/12.318300.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 233 9/26/23 5:09 PM

 234 Photonics Elements for Sensing and Optical Conversions

30. Booth, M. J. 2003. Direct measurement of Zernike aberration modes with a modal
wavefront sensor. Proc. SPIE. 5162:79–90. https://doi.org/10.1117/12.503695.
31. Porfirev, A. P., and Khonina, S. N. 2016. Experimental investigation of multi-order
diffractive optical elements matched with two types of Zernike functions. Proc. SPIE.
9807:98070E. https://doi.org/10.1117/12.2231378.
32. Garitchev, V. P., Go1ub, M. A., Karpeev, S. V., Krivoshlykov, S. G., Petrov, N. I.,
Sissakian, I. N., Soifer, V. A., Haubenreisser, W., Jahn, J.-U., and Willsch, R. 1985.
Experimental investigation of mode coupling in a multimode graded-index fiber,
caused by periodic microbends using computer-generated spatial filters. Opt. Commun.
55(6):403–5. https://doi.org/10.1016/0030-4018(85)90140-3.
33. Karpeev, S. V., Pavelyev, V. S., Khonina, S. N., and Kazanskiy, N. L. 2005. High-
effective fiber sensors based on transversal mode selection. Proc. SPIE. 5854:163–9.
https://doi.org/10.1117/12.634603.
34. Karpeev, S. V., Pavelyev, V. S., Khonina, S. N., Kazanskiy, N. L., Gavrilov, A. V., and
Eropolov, V. A. 2007. Fibre sensors based on transverse mode selection. J. Mod. Opt.
54(6):833–44. https://doi.org/10.1080/09500340601066125.
35. Khonina, S. N., Kotlyar, V. V., Soifer, V. A., Paakkonen, P., Simonen, J., and Turunen, J.
2001. An analysis of the angular momentum of a light field in terms of angular harmon-
ics. J. Mod. Opt. 48(10):1543–57. https://doi.org/10.1080/09500340108231783.
36. Moreno, I., Davis, J. A., Pascoguin, B. M. L., Mitry, M. J., and Cottrell, D. M. 2009.
Vortex sensing diffraction gratings. Opt. Lett. 34(19):2927–9. https://doi.org/10.1364/
OL.34.002927.
37. Lei, T., Zhang, M., Li, Y., Jia, P., Liu, G. N., Xu, X., Li, Z., Min, C., Lin, J., Yu, C., and Niu,
H. 2015. Massive individual orbital angular momentum channels for multiplexing enabled
by Dammann gratings. Light Sci. Appl. 4(3):e257. https://doi.org/10.1038/lsa.2015.30.
38. Karpeev, S. V., Pavelyev, V. S., Soifer, V. A., Khonina, S. N., Duparre, M., Luedge, B.,
and Turunen, J. 2005. Transverse mode multiplexing by diffractive optical elements.
Proc. SPIE. 5854: 163–9. https://doi.org/10.1117/12.634547.
39. Moreno, I., Davis, J. A., Ruiz, I., and Cottrell, D. M. 2010. Decomposition of radially
and azimuthally polarized beams using a circular-polarization and vortex-sensing dif-
fraction grating. Opt. Express 18(7):7173–83. https://doi.org/10.1364/OE.18.007173.
40. García-Martínez, P., Sánchez-López, M. M., Davis, J. A., Cottrell, D. M., Sand, D., and
Moreno, I. 2012. Generation of Bessel beam arrays through Dammann gratings. Appl.
Opt. 51(9):1375–81. https://doi.org/10.1364/AO.51.001375.
41. Lyubopytov, V. S., Tlyavlin, A. Z., Sultanov, A. K., Bagmanov, V. K., Khonina, S. N.,
Karpeev, S. V., and Kazanskiy, N. L. 2013. Mathematical model of completely optical
system for detection of mode propagation parameters in an optical fiber with few-mode
operation for adaptive compensation of mode coupling. Comput. Opt. 37(3):352–9.
42. Ni, B., Guo, L., Yue, C., and Tang, Z. 2017. A Novel measuring method for arbi-
trary optical vortex by three spiral spectra. Phys. Lett. A. 381(8):817–20. https://doi.
org/10.1016/j.physleta.2016.12.050.
43. Khonina, S. N., Karpeev, S. V., and Paranin, V. D. 2018. A technique for simulta-
neous detection of individual vortex states of Laguerre–Gaussian beams transmitted
through an aqueous suspension of microparticles. Opt. Laser Eng. 105:68–74. https://
doi.org/10.1016/j.optlaseng.2018.01.006.
44. Abderrahmen, T., Park, K.-H., Zghal, M., Ooi, B. S., and Alouini, M.-S. 2019.
Communicating using spatial mode multiplexing: Potentials, challenges, and perspectives.
IEEE Commun. Surv. Tut. 21(4):3175–203. https://doi.org/10.1109/COMST.2019.2915981.
45. Wilby, M. J., Keller, C. U., Haert, S., Korkiakoski, V., Snik, F., and Pietrow, A. G. M.
2016. Designing and testing the coronagraphic Modal Wavefront Sensor: a fast non-
common path error sensor for high-contrast imaging. Proc. SPIE. 9909:990921. https://
doi.org/10.1117/12.2231303.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 234 9/26/23 5:09 PM

 Wave Front Aberration Sensors 235

46. Lyu, H., Huang, Y., Sheng, B., and Ni, Z. 2018. Absolute optical flatness testing by sur-
face shape reconstruction using Zernike polynomials. Opt. Eng. 57(9):094103. https://
doi.org/10.1117/1.OE.57.9.094103.
47. Bereznyi, A. E., Prokhorov, A. M., Sisakyan, I. N., and Soifer, V. A. 1984. Bessel-
optics. Soviet Phys Dokl. 29(2):115–7.
48. Khonina, S. N., Kotlyar, V. V., Shinkaryev, M. V., Soifer, V. A., and Uspleniev,
G. V. 1992. The phase rotor filter. J. Mod. Opt. 39(5):1147–54. https://doi.
org/10.1080/09500349214551151.
49. Degtyarev, S. A., Porfirev, A. P., and Khonina, S. N. 2017. Zernike basis-matched
multi-order diffractive optical elements for wavefront weak aberrations analysis. Proc.
SPIE. 10337:103370Q. https://doi.org/10.1117/12.2269218.
50. Koltyar, V. V., and Khonina, S. N. 2003. Multi-order diffractive optical elements to
process data. In Perspectives in engineering optics. Delhi: Anita Publications. 47–56.
51. Khonina, S. N., Kazanskiy, N. L., and Soifer, V. A. 2012. Optical vortices in a fiber:
Mode division multiplexing and multimode self-imaging. In Recent progress in optical
fiber research. London: IntechOpen Limited. 327–52. ISBN: 978-953-307-823-6.
52. Kotlyar, V. V., Khonina, S. N., Soifer, V. A., Wang, Y., and Zhao, D. 1997. Coherent
field phase retrieval using a phase Zernike filter. Comput. Opt. 17:43–8.
53. Khonina, S. N., Kotlyar, V. V., and Kirsh, D. V. 2015. Zernike phase spatial filter for
measuring the aberrations of the optical structures of the eye. J. Biomed. Photonics
Eng. 1(2):146–53. https://doi.org/10.18287/JBPE-2015-1-2-146.
54. Guo, H., Korablinova, N., Ren, Q., and Bille, J. 2006. Wavefront reconstruction with artifi-
cial neural networks. Opt. Express. 14(14):6456–62. https://doi.org/10.1364/OE.14.006456.
55. Paine, S. W., and Fienup, J. R. 2018. Machine learning for improved image-based wave-
front sensing. Opt. Lett. 43(6):1235–8. https://doi.org/10.1364/OL.43.001235.
56. Nishizaki, Y., Valdivia, M., Horisaki, R., Kitaguchi, K., Saito, M., Tanida, J., and Vera,
E. 2019. Deep learning wavefront sensing. Opt. Express. 27(1):240–51. https://doi.
org/10.1364/OE.27.000240.
57. Maeda, N. 2009. Clinical applications of wavefront aberrometry – A review. Clin. Exp.
Ophthalmol. 37(1):118–29. https://doi.org/10.1111/j.1442-9071.2009.02005.x.
58. Mello, G. R., Rocha, K. M., Santhiago, M. R., Smadja, D., and Krueger, R. R. 2012.
Applications of wavefront technology. J. Cataract. Refract. Surg. 38(9):1671–83.
https://doi.org/10.1016/j.jcrs.2012.07.004.
59. Montés-Micó, R., Cáliz, A., and Alió, J. L. 2004. Wavefront analysis of higher
order aberrations in dry eye patients. J. Refract. Surg. 20(3):243–7. https://doi.
org/10.3928/1081-597X-20040501-08.
60. Maeda, N., Fujikado, T., Kuroda, T., Mihashi, T., Hirohara, Y., Nishida, K., Watanabe,
H., and Tano, Y. 2002. Wavefront aberrations measured with Hartmann-shack sensor in
patients with keratoconus. Ophthalmology. 109(11):1996–2003. https://doi.org/10.1016/
s0161-6420(02)01279-4.
61. Gobbe, M., Reinstein, D. Z., and Archer, T. J. 2015. LASIK-induced aberrations:
Comparing corneal and whole-eye measurements. Optom. Vis. Sci. 92(4):447–55.
https://doi.org/10.1097/OPX.0000000000000557.
62. Denoyer, A., Ricaud, X., Van Went, C., Labbé, A., and Baudouin, C. 2013. Influence of
corneal biomechanical properties on surgically induced astigmatism in cataract surgery.
J. Cataract. Refract. Surg. 39(8):1204–10. https://doi.org/10.1016/j.jcrs.2013.02.052.
63. Cerviño, A., Hosking, S. L., Montes-Mico, R., and Bates, K. 2007. Clinical ocular
wavefront analyzers. J. Refract. Surg. 23(6):603–16. https://doi.org/10.3928/1081-
597X-20070601-12.
64. Plaza-Puche, A. B., Salerno, L. C., Versaci, F., Romero, D., and Alio, J. L. 2019. Clinical
evaluation of the repeatability of ocular aberrometry obtained with a new pyramid wave-
front sensor. Eur. J. Ophthalmol. 29(6):585–92. https://doi.org/10.1177/1120672118816060.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 235 9/26/23 5:09 PM

 236 Photonics Elements for Sensing and Optical Conversions

65. Iglesias, I., Ragazzoni, R., Julien, Y., and Artal, P. 2002. Extended source pyra-
mid wave-front sensor for the human eye. Opt. Express. 10(9):419–28. https://doi.
org/10.1364/OE.10.000419.
66. Hutterer, V. et al. 2019. Real-time adaptive optics with pyramid wavefront sensors: Part i.
A theoretical analysis of the pyramid sensor model. Inverse Probl. 35(4):045007. https://
doi.org/10.1088/1361-6420/ab0656.
67. Hénault, F., Spang, A., Feng, Y., and Schreiber, L. 2020. Crossed-sine wavefront sen-
sor for adaptive optics, metrology and ophthalmology applications. Eng. Res. Express.
2(1):015042. https://doi.org/10.1088/2631-8695/ab78c5.
68. Oti, J. E., Canales, V. F., and Cagigal, M. P. 2003. Analysis of the signal-to-noise ratio
in the optical differentiation wavefront sensor. Opt. Express. 11(21):2783–90. https://
doi.org/10.1364/OE.11.002783.
69. Xu, Z., Hua, Y., and Qiu, W. et al. 2018. Precision and agreement of higher order aberra-
tions measured with ray tracing and Hartmann-Shack aberrometers. BMC Ophthalmol.
18:18. https://doi.org/10.1186/s12886-018-0683-8.
70. Han, J. Y., Yoon, S., Brown, N. S., Han, S.-H., and Han, J. 2020. Accuracy of the hand-
held wavefront aberrometer in measurement of refractive error. Korean J. Ophthalmol.
34(3):227–34. https://doi.org/10.3341/kjo.2019.0132.
71. Kim, B. K., Mun, S. J., and Yang, Y. H. et al. 2019. Comparison of anterior segment
changes after femtosecond laser LASIK and SMILE using a dual rotating Scheimpflug
analyzer. BMC Ophthalmol. 19:251. https://doi.org/10.1186/s12886-019-1257-0.
72. Randleman, J. B., Woodward, M., Lynn, M. J., and Stulting, R. D. 2008. Risk assess-
ment for ectasia after corneal refractive surgery. Ophthalmology. 115(1):37–50. https://
doi.org/10.1016/j.ophtha.2007.03.073.
73. Salouti, R., Nowroozzadeh, M. H., Zamani, M., Fard, A. H., and Niknam, S. 2009.
Comparison of anterior and posterior elevation map measurements between 2
Scheimpflug imaging systems. J. Cataract. Refract. Surg. 35:856–62. https://doi.
org/10.1016/j.jcrs.2009.01.008.
74. Kim, E. J., de Oca, I. M., Wang, L., Weikert, M. P., Koch, D. D., and Khandelwal, S. S.
2015. Repeatability of posterior and total corneal curvature measurements with a dual
Scheimpflug–Placido tomographer. J. Cataract. Refract. Surg. 41(12):2731–8. https://
doi.org/10.1016/j.jcrs.2015.07.035.
75. Güell, J. L., Pujol, J., and Arjona, M. et al. 2004. Optical quality analysis system:
Instrument for objective clinical evaluation of ocular optical quality. J. Cataract.
Refract. Surg. 30(7): 1598–9. https://doi.org/10.1016/j.jcrs.2004.04.031.
76. Santamaría, J., Artal, P., and Bescós, J. 1987. Determination of the point-spread
function of human eyes using a hybrid optical-digital method. J. Opt. Soc. Am. A.
4(6):1109–14. https://doi.org/10.1364/JOSAA.4.001109.
77. Frings, A., Hassan, H., and Allan, B. D. 2020. Pyramidal aberrometry in wavefront-
guided myopic LASIK. J. Refract. Surg. 36(7):442–8. https://doi.org/10.3928/10815
97X-20200519-03.
78. Zlatanović, M., Živković, M., Hristov, A., Stojković, V., Novak, S., Zlatanović, N., and
Brzaković, M. 2019. Central corneal thickness measured by the Oculyzer, BioGraph,
and ultrasound pachymetry. Acta Medica Medianae. 58(2):33–7. https://doi.org/10.5633/
AMM.2019.0206.
79. Roshdy, M. M., Wahba, S. S., and Fikry, R. R. 2018. New corneal assessment index from
the rational thickness and other OCULUS values (CAIRO index). Clin. Ophthalmol.
12:1527–32. https://doi.org/10.2147/OPTH.S171827.
80. Sun, M.-S., Zhang, L., Guo, N., Song, Y.-Z., and Zhang, F.-J. 2018. Consistent com-
parison of angle Kappa adjustment between Oculyzer and Topolyzer Vario topography
guided LASIK for myopia by EX500 excimer laser. Int. J. Ophthalmol. 11(4):662–7.
https://doi.org/10.18240/ijo.2018.04.21.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 236 9/26/23 5:09 PM

 Wave Front Aberration Sensors 237

81. Branchevskiy, S. L., and Branchevskaya, E. S. 2018. Comparative analysis of devices

based on Plasido, Scheimpflug and OCT for measuring keratometry in patients
after laser vision correction. Modern Technologies Ophthalmol. 5:185–7. https://doi.
org/10.25276/2312-4911-2018-5-185-187.
82. Belin, M. W., and Zloty, P. 1993. Accuracy of the PAR corneal topography system
with spatial misalignment. CLAO J. 19(1):64–8. https://doi.org/10.1097/00140068-
199301000-00012.
83. Khorin, P. A., Ilyasova, N. Y., and Paringer, R. A. 2018. Informative feature selection
based on the Zernike polynomial coefficients for various pathologies of the human
eye cornea. Comput. Opt. 42(1):159–66. https://doi.org/10.18287/2412-6179-2018-42-1-
159-166.
84. Lombardo, M., and Lombardo, G.. 2010. Wave aberration of human eyes and new
descriptors of image optical quality and visual performance. J. Cataract. Refract. Surg.
36:313–31. https://doi.org/10.1016/j.jcrs.2009.09.026.
85. Artal, P., Guirao, A., Berrio, E., and Williams, D. R. 2001. Compensation of corneal
aberrations by the internal optics in the human eye. J. Vis. 1(1):1–8. https://doi.org/
10.1167/1.1.1.
86. Artal, P., Berrio, E., Guirao, A., and Piers, P. 2002. Contribution of the cornea and
internal surfaces to the change of ocular aberrations with age. J. Opt. Soc. Am. A.
19(1):137–43. https://doi.org/10.1364/josaa.19.000137.
87. He, J. C., Gwiazda, J., Thorn, F., and Held, R. 2003. Wave-front aberrations in the ante-
rior corneal surface and the whole eye. J. Opt. Soc. Am. A. 20(7):1155–63. https://doi.
org/10.1364/JOSAA.20.001155.
88. Mrochen, M., Jankov, M., Bueeler, M., and Seiler, T. 2003. Correlation between corneal
and total wavefront aberrations in myopic eyes. J. Refract. Surg. 19(2):104–12. https://
doi.org/10.1117/12.470593.
89. Kelly, J. E., Mihashi, T., and Howland, H. C. 2004. Compensation of corneal horizon-
tal/vertical astigmatism, lateral coma, and spherical aberration by internal optics of the
eye. J. Vis. 4(4):262–71. https://doi.org/10.1167/4.4.2.
90. ZEMAX®. Optical design program. User’s manual. July 8, 2011. https://neurophysics.
ucsd.edu/Manuals/Zemax/ZemaxManual.pdf.
91. Khonina, S. N., Kotlyar, V. V., and Wang, Y. 2001. Diffractive optical element matched
with Zernike basis. Pattern Recognit. Image Anal. 11(2):442–5.
92. Kirilenko, M. S., Khorin, P. A., and Porfirev, A. P. 2016. Wavefront analysis
based on Zernike polynomials. CEUR Workshop Proc. 1638:66–75. https://doi.
org/10.18287/1613-0073-2016-1638-66-75.
93. Tocci, M. How to model the human eye in Zemax. Zemax knowledgebase. 2007. http://
www.zemax.com/support/resource-center/knowledgebase/how-to-model-the-human-
eye-in-zemax.
94. Guirao, A., Redondo, M., and Artal, P. 2000. Optical aberrations of the human cor-
nea as a function of age. J. Opt. Soc. Am. A. 17(10):1697–702. https://doi.org/10.1364/
josaa.17.001697.
95. Vapnik, V. N. 2000. The nature of statistical learning theory. 2nd ed. New York:
Springer Science+Business Media. ISBN: 978-1-4419-3160-3.
96. Fukunaga, K. 1990. Introduction to statistical pattern recognition. San Diego:
Academic Press. ISBN: 978-0-12-269851-4.
97. Ilyasova, N. Y. 2013. Methods for digital analysis of human vascular system. Literature
review. Comput. Opt. 37(4):517–41.
98. Ilyasova, N., Paringer, R., and Kupriyanov, A. 2016. Regions of interest in a fundus image
selection technique using the discriminative analysis methods, 408–417. In Computer
vision and graphics, ed. L. J. Chmielewski, A. Datta, R. Kozera, K. Wojciechowski.
Springer International Publishing AG. https://doi.org/10.1007/978-3-319-46418-3_36.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 237 9/26/23 5:09 PM

 238 Photonics Elements for Sensing and Optical Conversions

99. Kutikova, V. V., and Gaidel, A. V. 2015. Study of informative feature selection
approaches for the texture image recognition problem using Laws’ masks. Comput.
Opt. 39(5):744–50. https://doi.org/10.18287/0134-2452-2015-39-5-744-750.
100. Gaidel, A. V., and Krasheninnikov, V. R. 2016. Feature selection for diagnosing the
osteoporosis by femoral neck X-ray images. Comput. Opt. 40(6):939–46. https://doi.
org/10.18287/2412-6179-2016-40-6-939-946.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp07.indd 238 9/26/23 5:09 PM

 8 Key Problems, Achievements,
Optical Computing

and Perspectives
Nikolay Lvovich Kazanskiy1,2, Muhammad Ali Butt1,3,
and Svetlana Nikolaevna Khonina1,2
1I mage Processing Systems Institute – Branch of the
Federal Scientific Research Centre “Crystallography and
Photonics” of Russian Academy of Sciences, Samara, Russia
2 Samara National Research University, Samara, Russia
3Warsaw University of Technology, Institute of

Microelectronics and Optoelectronics, Warszawa, Poland

8.1 INTRODUCTION
The topic of optical computing (OC) dates to the early 1960s, or perhaps before,
when the military became interested in utilizing the Fourier transform (FT) rela-
tionships essential in coherent optical imaging systems to perform processes like
convolution and correlation. On data supplied in imaging optical format to a bulk
optical system, it was easily demonstrated that these processes could be performed
with considerable speed. Such processors were fundamentally analog in operation,
and, as a result, they constantly struggled to maintain adequate dynamic range and
signal-to-noise ratios, severely restricting their use. Despite several spectacular dem-
onstrations, silicon digital electronic processing appears to have nearly always been
chosen for final manufacturing equipment [1]. Nonetheless, there is still a lot of inter-
est in such specialized machines, and it is certainly true that they can attain very high
comparable digital processing speeds. The use of nonlinear optical tools to establish
the basic digital processing functions of AND, OR, NAND, NOR, etc., sparked even
more interest in the early 1980s, this time in imaging optical format, where a single
lens could, in theory, image a very large number of parallel channels from a two-
dimensional (2D) array of devices. As a result, assertions have been made that those
future high-speed computers would use vastly parallel digital-optical processing to
attain speeds much above those conceivable with electronics [2].
Following that, a slew of large R&D initiatives were aimed at capitalizing on the
potential. Such statements are established on basic expectations about prospective
digital throughput, but easily disregard the enormous practical issues that come with
putting them into practice. Because the analog light wave level represents the digital
state, practically all optical logic systems addressed in the literature use threshold
logic, meaning highly tight control of the optical power level across a complicated
multichannel system. Given how readily unanticipated 3-dB insertion losses may be

DOI: 10.1201/9781003439165-8 239

BK-TandF-KAZANSKIY_9781032572949-230642-Chp08.indd 239 9/26/23 1:12 PM

 240 Photonics Elements for Sensing and Optical Conversions

obtained in complicated systems of mirrors, holograms, lenses, and other components,

one would wonder if this had any chance of being realized. Nevertheless, employing
a dual-rail optical signaling system, the symmetric-SEED technique established at
AT&T Bell Labs and explored elsewhere does give a sophisticated resolution to this
problem [3]. Other important concerns include the fact that the finest optical logic sys-
tems are optically activated electrical equipment, and huge ones at that because light
must enter them, and optical wavelengths are extremely long by electronic norms. The
practical concerns of compiling high-resolution imaging optical systems, the shal-
low depth of field, the accuracy inferred in the lenses (in terms of focal length), and
the mechanical tolerances and massive dimensions of constructed structures are all
simple issues that are conveniently overlooked by many. Nevertheless, none of these
disadvantages indicates that it cannot be done; rather, they raise the challenge, and as
some of the results will demonstrate, genuinely spectacular experimental procedures
established on free-space optics may be created in the research laboratory.
The switches, optical logic gates (OLGs), and memory components that regulate
the flow of electrons in an electronic computer would be replaced by optical systems
in a digital optical computer. Optical modulators, which come in a variety of shapes
and sizes, may perform the roles of the OLGs and switches. A 1 × 1 switch is a simple
on–off switch with the ability to link two lines. A 1 × 2 switch connects one line to one
of two lines. Two lines are connected by a 2 × 2 switch. It can be in one of two states:
input (I/P) 1 linked to output (O/P) 1 and I/P 2 connected to O/P 2, or I/P 1 connected
to O/P 2 and I/P 2 connected to O/P 1, known as the bar state, or I/P 1 connected to
O/P 2 and I/P 2 connected to O/P 1, known as the cross-state. As a result, it’s known
as a crossbar switch. A crossbar switch may be stretched to n × n arrangement, which
allows any of n I/P lines to be linked to any of n O/P lines at any time without causing
interference.
Optical switches can be made from electro-optic modulators, acousto-optic mod-
ulators, magneto-optic modulators, and liquid crystals, among other forms of optical
modulators. A crossbar switch can also be made from an array of modulators. An
array of n × n light valves, for instance, might be used to create n × n crossbar switch.
The light source would be a vertical linear array of n laser diodes. The light from the
diodes is dispersed horizontally by a cylindrical lens, such that each diode illumines
one row of the n × n array. Then, one lens per row of the modulator array, a set of n
cylindrical lenses, positioned perpendicular to the first lens, is employed to direct the
signal emitted by the array onto a horizontally oriented linear array of n photodetec-
tors. This setup allows light from any laser diode to be connected to any detector
without interruption. Switches that are operated both electrically and optically have
been designed. Optically regulated apparatuses would be used in an all-optical com-
puter. The switches should also be tiny, high-speed apparatuses that can be produced
in vast arrays and consume very little switching energy. Several alternative technolo-
gies have been used to illustrate the needed switching functionalities. However, mas-
sive, high-density arrays of OLGs are currently being developed.
A multinational research team lead by Russia’s Skolkovo Institute of Science and
Technology has developed an exceptionally energy-efficient optical switch in con-
junction with IBM [4]. The switch is extremely rapid and does not involve any cool-
ing. It might serve as the foundation for a next generation of computers that manage

BK-TandF-KAZANSKIY_9781032572949-230642-Chp08.indd 240 9/26/23 1:12 PM

 Optical Computing 241

photons instead of electrons. A 35-nanometer semi-conducting polymer consisting

of organic material is placed between highly reflecting surfaces to form the switch.
This results in the formation of a tiny chamber that traps light beam. The gadget is
powered by two lasers: a pump laser and a seed laser. When the pump laser shines on
the switch, thousands of indistinguishable quasiparticles develop in the same area,
generating a Bose–Einstein condensate – a collection of particles that individually
act like a single atom. The seed beam is utilized to alter this condensate between two
measured states that act as binary codes “0” and “1.” The new technology can do
1012 calculations per second, which is 1,000 times quicker than today’s finest com-
mercial transistors. Furthermore, it requires significantly less energy to transition
states than transistors. This is due to the optical switch’s ability to be actuated by a
single photon of light. Equivalent electrical transistors that employ single electrons
typically need enormous quantities of cooling device, which uses a lot of electric-
ity. The new switch, on the other hand, will function at room temperature. Before it
can be deployed, the technology has a long way to go. The first electronic transistor
literally takes years to make its way into a personal computer. The researchers face
a hurdle in that, despite requiring relatively little energy to switch, the device still
requires continual input from the pump laser. The team is researching ways to get
around this by employing superfluorescent perovskite supercrystals to aid in reduc-
ing power usage. Despite the difficulties, the researchers expect that the novel switch
will be employed in various types of optical computing systems in the near future,
maybe as a method to ramp up supercomputer processing [4].
The urgent demand for optical technology relies on the fact that the time response
of electrical circuits limits today’s computers [5]. A solid transmission medium
restricts signal speed and volume while also generating heat that destroys compo-
nents. A 1-foot length of wire, for instance, creates roughly 1 nanosecond (billionth
of a second) of time delay. The extreme downsizing of microscopic electronic appa-
ratuses also causes “cross talk,” or signal mistakes that compromise the system’s
dependability. These and other challenges have prompted researchers to look for
solutions in light itself. Light doesn’t even have the time domain restrictions of elec-
tronics, does not require insulators, and may even deliver dozens or hundreds of pho-
ton communication streams at distinct color frequencies at the same time, those that
are not susceptible to electromagnetic (EM) interference and do not experience elec-
trical short circuits. They feature low-loss transmission and a broad bandwidth; that
is, they can communicate with numerous channels simultaneously without interfer-
ence. They may transmit signals inside the same or neighboring optical fibers (OFs)
with little to no interference or cross talk. They are smaller, lighter, and less costly
to produce, and they work better with stored data than magnetic materials. Scientists
want to construct a new class of computers that run 100 million times quicker than
today’s apparatuses by restoring electrons and cables with photons, OFs, crystals,
thin films, and mirrors [6].
Optical processing hits at a critical juncture in computing history. The need for
AI is growing at the same time as Moore’s Law is breaking down in silicon-based
computing. Optalysys is a game-changing technology startup that uses ground-
breaking optical processing techniques to address this challenge, allowing new
levels of AI capability for high-resolution image and video-based applications [7].

BK-TandF-KAZANSKIY_9781032572949-230642-Chp08.indd 241 9/26/23 1:12 PM

 242 Photonics Elements for Sensing and Optical Conversions

The technology is emerging to speed up several of the most demanding processor-

intensive operations while using a fraction of the energy that silicon processors con-
sume. It may be configured to execute optical correlation and convolution operations
using the API or Tensorflow interface to unleash new levels of AI and pattern rec-
ognition potential. Optalysys employs photons instead of electrons, but it also does
high-resolution computations, allowing huge image/pattern-based data to be handled
at speeds significantly quicker than silicon. The Optalysys technology can deliver
something distinctive in the AI space: a scalable processor that can accomplish end-
to-end, full-resolution processing of multi-megapixel image and video data, or con-
textually pre-process data for enhancing the effectiveness of existing Convolutional
Neural Network (CNN)-type models for high-resolution data applications, thanks to
the characteristics of diffractive optics.

8.2 TYPES OF COMPUTING AND NECESSITY

OF OPTICS FOR COMPUTING
OC may be divided into two types: digital optical computing (DOC) and analog opti-
cal computing (AOC). For more than 30 years, DOC based on Boole logics has been
established, employing a technique comparable to general-purpose computing real-
ized on transistors. Nonetheless, given reduced optical device integration density, it
is impossible to beat traditional digital computing. AOC, on the other hand, makes
use of the physical characteristics of light, including amplitude and phase, as well as
the interactions between light and optical apparatuses, to perform specific computa-
tions. Due to the obvious unique mathematical portrayal of computational processes
in one AOC system, it is considered dedicated computing. AOC can achieve higher
data processing speed in specialized applications, including pattern recognition and
numerical calculation, as compared to standard digital computing. As a result, AOC
has received a lot of academic attention as one of the most potential pervasive com-
puting systems in the post-Moore age.
Optical technology offers tremendous improvements in computational efficiency
and speed, along with considerable size and cost reductions [8, 9]. Since many pro-
cesses may be conducted at the same time, an optical desktop computer might process
data 100,000 times quicker than existing versions. Low production costs, resilience to
EM interference, low-loss transmission tolerance, independence from short electrical
connections, and the capacity to deliver wide bandwidth and transport signals with-
out interference within the same or neighboring OFs are all features of optics. The
distinction between optical and electrical synchrony may be shown with a simplistic
example. Imagine an imaging system with 1,000 × 1,000 distinct units per mm2 in
the object plane that are optically coupled to a comparable number of points per mm2
in the image plane; the lens essentially conducts a real-time FFT (fast Fourier trans-
form) of the image plane. A million processes are necessary to do this electronically.
If parallelism is combined with rapid switching speeds, startling computing speeds
are possible. Consider there are only 100 million OLGs on a chip, which is signifi-
cantly fewer than the figure indicated earlier. Furthermore, imagine that each OLG
has a switching period of just 1 nanosecond (organic optical switches may switch at
sub-picosecond rates, but electronic switching has a maximum picosecond switching

BK-TandF-KAZANSKIY_9781032572949-230642-Chp08.indd 242 9/26/23 1:12 PM

 Optical Computing 243

time). Well over 1,017-bit operations per second may be performed by such a device.
When compared to the gigabits (109) or terabits (1,012) per second speeds that elec-
tronics are now restricted to or aiming for, this is a significant difference. In other
words, an operation that would take a normal computer a hundred thousand hours
(>11 years) might just be completed in less than an hour by an optical computer.
As photons are uncharged and do not react with each other as easily as electrons,
light does have another benefit. As a result, in full-duplex functioning, beams of light
can travel across each other without altering the information transmitted. Loops in
electronics produce noisy voltage spikes anytime the EM fields pass through the
loop change. Furthermore, switching pulses with a high frequency or a short dura-
tion will generate cross talk in surrounding lines. Signals in nearby OFs or optical-
integrated channels are unaffected by each other, and they do not pick up noise
from loops. Ultimately, optical materials outperform magnetic materials in terms
of storage density and availability. The subject of OC is growing quickly and offers
many exciting possibilities for transcending the limits of today’s electrical comput-
ers. Optical apparatuses are already being integrated into a variety of computing
systems. Resulting in mass manufacture, the cost of laser diodes as coherent light
sources has reduced dramatically. Optical CD-ROM discs have also been widely
used in both home and office PCs [10].

8.3 OPTICAL PROCESSING ARCHITECTURE

Optical information processing is introduced and established by utilizing all of light’s
attributes of speed and parallelism to handle data at a high rate. The data is represented
as a light wave or graphic. The inherent parallel processing of OC was frequently noted
as a fundamental benefit of optical processing over electronic processing employing
predominantly serial processors. As a result, optics holds significant promise for inter-
preting enormous amounts of data in real time. OC is established on the FT feature of
a lens. When employing coherent light, a lens executes the FT of a 2D transparency
in its front focal plane in its rear focal plane. The lens computes the precise FT with
amplitude and phase in an analog method [11].
Figure 8.1 depicts the topology of a general-purpose optical processor for data pro-
cessing. The I/P plane, the processing plane, and the O/P plane make up the system.

FIGURE 8.1 An optical processor’s topology.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp08.indd 243 9/26/23 1:12 PM

 244 Photonics Elements for Sensing and Optical Conversions

The I/P plane displays the data to be processed; this plane will usually implement an
algorithm which helps in the conversion of signals from electricity to light. The con-
version is done via a Spatial Light Modulator (SLM). A 1D or 2D signal can be used
as an I/P. In the context of a 1D I/P signal, an acousto-optic cell is commonly utilized,
whereas 2D SLMs are used for 2D signals. Due to the lack of SLMs in the early
years, the I/P plane consisted of a stationary slide. As a result, the concepts and pos-
sibilities of optical processors could be proved, but no real-time applications could be
presented, rendering the processor worthless for most real-world applications. Lenses,
holograms (optically recorded or computer produced), and nonlinear components can
all be found in the processing plane. This is perhaps the most important component
of the processing, and it can be done at the speed of light in most optical processors.
The O/P plane, which detects the processing results, is made up of a photodetector, a
photodetector array, or a camera. Since the bulk of them are running at video rates,
Figure 8.1 indicates that the speed of the entire process is restricted by the speed of
its slowest component, which is usually the I/P plane SLM. The SLM is an essential
component in the creation of realistic optical processors, but it’s also one of their
weakest. Likewise, the poor performance and expensive price of SLMs have slowed
the development of a real-time optical processor.
Real-time pattern recognition was first thought to be one of the most potential tech-
nologies of optical computers, hence the following two optical correlator designs were
developed. Because the separation between the I/P plane and the O/P plane is four
times the focal length of the lenses, the basic correlator is termed 4-f in Figure 8.2(a).

FIGURE 8.2 (a) Optical setup of basic 4-f correlator [22], (b) autocorrelation peak for a
matched filter [22], (c) autocorrelation peak for a phase-only filter [22], (d) optical setup of
JTC [22], (e) O/P plane of the JTC [22].

BK-TandF-KAZANSKIY_9781032572949-230642-Chp08.indd 244 9/26/23 1:12 PM

 Optical Computing 245

This extremely simple design is established on Maréchal and Croce’s spatial filter-
ing work in 1953 [12] and was further refined by numerous researchers over the years
[13, 14]. The I/P scene is projected in the I/P plane, and Lens 1 performs the FT.
The reference’s actual conjugated FT is placed in the Fourier plane and hence mul-
tiplied by the I/P scene’s FT. Lens 2 uses a second FT to determine the relationship
between the I/P scene and the reference in the O/P plane. The fundamental issue of this
arrangement was developing a sophisticated filter using the reference’s FT, and A. V.
Lugt suggested in 1964 to utilize a Fourier hologram of the benchmark as a filter [15].
Figure 8.2(b and c) depicts the O/P correlation peak for autocorrelation when the corre-
lation filter is a matched filter and when the correlation filter is phase-only, respectively.
Weaver and Goodman [16] proposed a new optical correlator layout in 1966, the
joint transform correlator (JTC), which is shown in Figure 8.2(d). The two pictures,
the reference r(x, y) and the scene s(x, y), are arranged beside each other on the I/P
plane, which is the first lens FTs. After detecting the intensity of the combined spec-
trum, the FT is used.
The cross correlations between the scene and the reference are among the terms
that make up the second FT. This FT may be done optically using an SLM, as seen
in Figure 8.2(d). When the reference and scene are similar, the O/P plane of the JTC
is shown in Figure 8.2(e). The two cross-correlation peaks are the only ones worth
looking at. The CCD camera can be exchanged with an optical component such
as an optically addressed SLM or a photorefractive crystal to create a completely
optical processor. Because the JTC does not need the computation of a correlation
filter, it is the appropriate design for real-time applications, including target tracking
where the benchmark must be revised at a rapid rate. Coherent optical processors are
illustrated in Figure 8.2. Wave intensities, rather than complex wave amplitudes, are
used to carry information in incoherent optical processors. Incoherent processors
aren’t affected by phase changes in the I/P plane and don’t produce coherent noise.
The non-negative real value of the information, on the other hand, necessitates the
employment of a variety of methods to execute some image-processing applications
[17, 18]. Linear optical processing may be broken down into space-invariant activi-
ties like correlation and convolution, as well as space-variant operations like coordi-
nate transformations [19] and the Hough transform [20]. Logarithm transformation,
thresholding, and analog-to-digital conversion are examples of nonlinear processing
that may be done optically [21].

8.4 TYPES OF OPTICAL COMPONENTS USED TO

IMPLEMENT OPTICAL FUNCTIONS
Deep learning has an unquenchable thirst for processing resources. Classical deep
learning has silently developed a bottleneck due to interference of electrical impulses,
energy usage, and physical constraints [23], even though electronic parts established
on silicon can still sustain it presently. Academic and corporate circles are attempt-
ing to find new approaches for resolving electrical flaws that are less computationally
intensive. It has huge benefits in information transmission and OC due to its high
speed of 300,000 km per second, which is 300 times faster than that of an electron,
and its data-carrying capacity and variety, which is 2 × 104 times greater than that

BK-TandF-KAZANSKIY_9781032572949-230642-Chp08.indd 245 9/26/23 1:12 PM

 246 Photonics Elements for Sensing and Optical Conversions

of electric channels, as well as high parallelism and strong anti-interference [24].

Switching electricity with light has emerged as a viable and sustainable work style,
following the current trend.

8.4.1 Spatial Light Modulator (SLM)

Light wave processing has provided useful techniques for converting data into spa-
tially modulated coherent light waves using SLM apparatuses, allowing the creation
of digital holographic images [11]. The capacity to alter the phase and amplitude of
light in the far-field is one of the hologram’s most valuable qualities. The FT illus-
trates how a hologram (near field) interacts with its replay field (far-field). In free
space, the far-field might develop at the focal point of a positive lens or an indefi-
nite distance from the near field plane [25]. Waveforms from an existent item can
be reproduced using holograms. With developments in digital technology and light
wave processing, it is now feasible to compute interference patterns computationally
and construct synthetic wave fronts of any shape. Computer-generated holograms
(CGH), diffractive optical elements (DOE), phase/amplitude masks, diffractive grat-
ing, and other terms can be used to describe these interference patterns. Because
they all work on the concept of diffraction, the nomenclature is rather arbitrary.
CGHs and DOEs have historically been used to implement various optical opera-
tions in correlators as spatial filters [26–29]. Despite the advent of dynamically con-
trolled SLMs, the use of DOEs in Fourier correlators remains relevant in high-energy
applications [30] since they have a significantly higher damage threshold than SLM.
Research by V. A. Soifer’s team have designed multichannel spatial optical elements
that allow coherent light fields to be optically decomposed into a series of orthogonal
functions. Using angular harmonics to calculate the light field’s angular momentum
[31–35], detection and analysis of wave front aberrations using Zernike polynomi-
als [36–40], and using the optical Karhunen–Loeve decomposition, we may derive
decorrelated image characteristics [41, 42]. Relying on segmented spatial filters in
the sequence of diffraction gratings, an optical approach for producing a directions
field for fringed/contour pictures such as interferograms and fingerprints has been
established [43–46]. To deepen focus and adjust for defocusing and chromatic aber-
rations, imaging systems undergo phase apodization and optical wave front coding
[47–50]. The spiral phase plate as a phase rotor filter is suggested to be used to visu-
ally realize the m-th order Hankel transform as well as to optically discriminate
the light field with rotational symmetry [51]. In contrast to the commonly used 2D
Hilbert transforms in image processing, the vortex spatial phase filter is utilized to
execute the radial Hilbert transform, which is isotropic [52–54]. Figure 8.3 shows
the operation of the optical correlator with a spatial filter in the form of the spiral
phase plate for the implementation of the radial Hilbert transform to contrast the
image of the fundus vessels [52]. In this way, a correlation filter can be used not only
for the classical task of optical detection (recognition) (Figure 8.2) but also for the
optical implementation of a mathematical operation such as the Hilbert transform
(Figure 8.3).
The SLM is an electrically programmable device that may modify light in compli-
ance with a fixed spatial (pixel) pattern. It may typically be used to adjust the phase

BK-TandF-KAZANSKIY_9781032572949-230642-Chp08.indd 246 9/26/23 1:12 PM

 Optical Computing 247

FIGURE 8.3 Illustration of the optical implementation of the radial Hilbert transform using
the spiral phase plate as a spatial filter to contrast the image of the fundus vessels [52].

and/or amplitude of incoming light. As a result, SLM may easily realize phase-only,
amplitude-only, or a combination of phase-amplitude [55]. Commercially available
SLMs are quite “slow,” although faster ones are also known [56]. A digital micromir-
ror device (DMD) is considered a fast analog of SLM. DMDs show switching speed
ranging from several kHz to tens of kHz (settling time for full-scale angle change
is around 10 μs) [57], however, high speed comes at the expense of limited modula-
tion depth and diffraction efficiency [58]. Thus, in applications where high energy
efficiency is not required, it is possible to use DMDs instead of SLMs. There are a
variety of modulation techniques to choose from. EO-SLM (electro-optic spatial
light modulator) is one of the most appealing and commonly utilized. Liquid crystal
is used as the modulation material in EO-SLM. A micro-display is used for incident
light manipulation and collecting in a liquid crystal SLM, as well. This may be done
in two ways: as a transmissive display using LCD SLM technology or as a reflective
display using liquid crystal on silicon (LCoS) SLM technology. The liquid crystal
molecular alignment is one of the modulator’s most prominent characteristics. This
might take the form of a vertical, parallel, or twisted structure. As a result, with the
right polarizing optics, the incident beam of light parameters, such as amplitude,
phase, or their combination, may be efficiently adjusted.
The nematic LCoS technology is a form of SLM that allows for phase-only modu-
lation [59]. Furthermore, it belongs to the electrically addressed reflection modula-
tor class, in which the liquid crystal is controlled by a direct and precise voltage,
and the beam of the light wave front may be adjusted as well. The LCoS SLM may
be used to rebuild pictures from CGH as a diffractive modulator [60]. CGHs may
be used for a variety of communication applications, and it is increasingly being
used in indoor visible light communication networks [61]. Additionally, a range
of optimization approaches, including the iterative Fourier transform algorithm
(IFTA), linear Fourier transform (i.e. linear phase mask), simulated annealing, and

BK-TandF-KAZANSKIY_9781032572949-230642-Chp08.indd 247 9/26/23 1:12 PM

 248 Photonics Elements for Sensing and Optical Conversions

the Gerchberg–Saxton algorithm, can be used to quickly construct acceptable holo-

grams [62, 63]. The beam of the light wave front can be modified when the SLM is
used as a diffractive device for reconstructing graphics from CGH.
As previously stated, LCoS displays have gained a lot of traction as potential
micro-displays for a variety of SLM applications. Likewise, they have appealing and
important characteristics, such as excellent spatial resolution and light efficiency. As
a result, they’ve been employed in a wide range of optical utilizations, including com-
munication, reconfigurable interconnects [64], storage [65], diffractive optics [66],
metrology [67], and quantum computing [68]. They can also be used for light wave
processing and monitoring in wave shaper technology [69]. Another feature of the
LCoS is that it is extremely cost-efficient and can be configured in a variety of ways.
Several functionalities, such as group delay ripple correction, wavelength filtering,
and chromatic dispersion correction, are made possible by this. In addition, LCoS
technology may be used in flex grid, which has been regarded as a key component
for next-generation networks [65]. The classic fixed grid with 50 GHz spacing, as
regulated by the International Telecommunication Union (ITU) Telecommunication
Standardization Sector (ITU-T), has a variety of drawbacks. The fixed grid has been
reported to result in poor usage of the optical spectrum. Furthermore, it severely
limits the system’s transmission capacity. The flex grid implementation, on the other
hand, allows for the use of several modulation formats and their coexistence on a
common infrastructure. They can also be multiplexed densely and effectively, which
helps optical networks not only extend their reach but also increase the per-channel
data rate. It is also expected that the deployment of WSS and SDM would consider-
ably aid in the expansion of the network’s coverage and capacity [65].
Electrically addressed SLMs (EASLMs) and optically addressed SLMs (OASLMs)
are two types of SLMs that differ in how data is loaded into the apparatuses [70]. The
EA has been the most prominent approach in modern commercial SLMs because of
the growth of electronic information technology over the last few decades. However,
because data must be converted back and forth between the optical and electrical
domains, EASLMs are not the greatest solution for future all-optical information
processing systems. The OASLMs, on the other hand, enable light to be modulated
directly by light without having to go through an electronic-optical transformation
[71, 72]. Additionally, OASLMs are required for a variety of all-optical purposes that
EASLMs cannot handle, such as coherent to incoherent image conversion, real-time
optical correlation, and parallel all-optical processing [8, 73–76]. OASLMs may be
built-in theory using material nonlinearities, with modulation over the read light
achieved by spatially precisely modifying the characteristics of the materials through
nonlinear optical stimuli [77]. Natural materials, on the other hand, have insufficient
nonlinearities to permit efficient “light-control-by-light” inside nanoscale volumes.
This renders the apparatuses exceedingly bulky or necessitates a lot of pumping
power to collect big enough nonlinear modulations, making them unsuitable for the
nano-era.
The recent remarkable advancement in dynamic optical metasurface (MS) tech-
nology provides a chance to overcome challenges and proposes a unique framework
for nanoscaled SLM [78, 79]. The MS can also enhance optical engagements by
focusing light on nanoscale volumes, allowing for greater control of light fields in

BK-TandF-KAZANSKIY_9781032572949-230642-Chp08.indd 248 9/26/23 1:12 PM

 Optical Computing 249

response to external mechanical, chemical, and magnetic stimuli, for example. The
ultracompact EASLMs established on MS have recently been observed relying on
the active manipulation of beams of light by additional electrical fields [80]. An opti-
cally addressed spatial light modulator (OASLM) established on metasurface-based
OASLM (MS-OASLM) is proposed in [81], with a nonlinear polarization manipula-
tion of read light by another write light at the nanoscale as the operating mechanism.
It delivers 500-line pairs/millimeter (equivalent to a pixel size of just 1 μm), which
is more than ten times greater than a standard commercial SLM. The MS-OASLM
has exceptional compactness and a thickness of just 400 nm. MS-OASLMs like
this might pave the way for next-generation all-optical data processing and high-
resolution display technologies.

8.4.2 Plasmonic Switches
Moore’s law is expected to surpass its physical constraints, putting modern computer
systems centered on von Neumann architecture at a physical limit [82, 83]. Potential
alternate prospect next-generation computing approaches are necessary due to the
rapidly rising development of bandwidth needs with lower power utilization. As
a result of the usage of optical interconnects in high computing chips rather than
electrical interlocks to minimize power consumption and increase speeds, photonic
computing is expected to have a competent replacement for existing electronic com-
puting systems. Furthermore, Si photonics is considered a leading field owing to its
capacity to provide effective light modulation, light confinement, and compliance
with today’s CMOS production techniques [84, 85]. A decade ago, the shift from
electronic to photonic systems began. Related to current material discoveries, fabri-
cation technique advances, and more ongoing research efforts in this field, photonic
interconnects and circuits have witnessed significant advancements in recent years.
Photonic logic circuits rely heavily on optical switching and modulation. Optical
computations are established on photonic logic circuits. Utilizing the thermo-optic
effect, free carrier dispersion effect, and Pockels electro-optic effect, intensive
research in the direction of efficient optical switches, optical modulators, and optical
logic circuits has been presented [86–89]. All these tools are volatile, requiring con-
stant voltage to manage, and, as a result, they consume a lot of power. Furthermore,
due to their low electro-refractivity, they have long interaction durations. As a result,
these apparatuses have substantial footprints, making ultra-compact designs diffi-
cult to achieve [90, 91]. A novel type of material known as phase change materials
(PCMs) has found considerable application in both the electronics and photonics
domains in recent years [92, 93]. With high-speed switching between two steady
states, PCMs reveal changes in electrical and/or optical characteristics. Several pho-
tonic and plasmonic apparatuses have been proposed and investigated, including
on-chip optical modulators and optical switches established on PCMs [94, 95]. One
of the widely used PCMs in photonic apparatuses is Ge2Sb2Te5 (GST) [96].
Even though effective energy nonvolatile (NV) memory and OLGs based on
PCMs have been thoroughly researched and established over the last 20 years, PCM-
based photonic NV memories and photonic logic circuits have received little atten-
tion. Novel NV combinational and sequential logic circuit designs are investigated,

BK-TandF-KAZANSKIY_9781032572949-230642-Chp08.indd 249 9/26/23 1:12 PM

 250 Photonics Elements for Sensing and Optical Conversions

as well as NV hybrid electro-optic plasmonic circuits. The electro-optic devices are

made up of a plasmonic waveguide (WG) with a mono PCM layer. Changing the
phase of the PCM between amorphous to crystalline likewise changes the optical
losses in the WG. Electrical threshold flipping or thermal conduction heating via
externally applied radiators or the plasmonic WG metal itself as an integrated heater
can be used to generate phase shift in the PCM. All OLGs, a half adder circuit, and
sequential circuits may be built employing plasmonic switches as active components,
as illustrated. Furthermore, the plasmonic switches and logic functions have mini-
mal extinction ratios larger than 20 dB, are small, have little operational power, and
operate at high speeds. To develop an effective architecture for logic processes, pho-
tonics, plasmonics, and electronics are merged on the same platform.
A plasmonic slot WG with PCM as an active material coating the slot WG’s surface
is used in the NV hybrid electro-optic plasmonic switch [97]. Figure 8.4(a) depicts
the suggested plasmonic slot WG’s construction, which includes Si-Au tapered WGs,

FIGURE 8.4 (a) The broadband NV hybrid EO plasmonic switch is shown schematically.
The side view picture of the plasmonic slot WG with a thin coating of GST is shown in the
inset [97]. (b) The hybrid NV EO plasmonic switch serves as the central unit in the race-
track μ-RR-based EO switch [97]. (c) Normalized transmission spectra for the amorphous
and crystalline phases of the PCM layer of the racetrack μ-RR-based EO switch. (d, e) The
E-field mapping of the EO switch for the amorphous phase and the crystalline phase of the
PCM layer at the resonance wavelength of 1,558 nm, respectively [97]. (f) Asymmetric MZI-
based EO switch with hybrid NV EO plasmonic switch as an active component in one arm.
(g) Normalized transmission band for the crystalline and amorphous phases of the PCM layer
in the plasmonic switch for the asymmetric MZI based EO switch. The E-field distribution of
the EO switch for the (h) amorphous phase and (i) crystalline phase of the PCM layer at the
resonance wavelength of 1,548 nm [97].

BK-TandF-KAZANSKIY_9781032572949-230642-Chp08.indd 250 9/26/23 1:12 PM

 Optical Computing 251

a slot in an Au film to form a plasmonic slot WG, and a PCM coating for instance
GST on the slot WG’s surface. The arrangement is made up of three primary compo-
nents, the first of which is a dielectric to plasmonic mode converter in the I/P region
of a hybrid Si-Au tapered WG. The major optical modulation happens in the second
portion, which is a MIM plasmonic slot WG covered with a thin coating of GST, and
the third part is a plasmonic to dielectric mode converter in the O/P section, which
uses a hybrid Au-Si tapered WG. The cross-sectional image (in the y–z plane) of the
GST-coated plasmonic slot WG is shown in the left top inset of Figure 8.4(a) [97].
As illustrated in Figure 8.4(b), the EO switch is established on a racetrack micro-
ring resonator (μ-RR) with an NV hybrid EO plasmonic switch in the ring WG as an
active component. The optical mode is linked to the ring through a Si WG with the
same cross section as the hybrid plasmonic switch’s Si WG. The transmission spec-
tra for both phases are shown in Figure 8.4(c). The optical loss in the ring is mini-
mal when GST is in the amorphous phase. Consequently, light can pass through
the RR and propagate via the optical mode. For specific wavelengths, resonance
is obtained in the RR, and, after a roundtrip, light coupled back to the I/P WG’s
through-port undergoes a 180° out-phasing in comparison to light arriving from the
I/P WG’s I/P point.
As a result, destructive interference develops between light flowing from the I/P
point and light coupled back to the I/P WGs through the port. As a result, light depart-
ing the O/P port at resonance wavelengths is eliminated as shown in Figure 8.4(d).
Optical mode, on the other hand, cannot travel via the ring WG due to momentous
optical loss in the crystalline phase. Consequently, there is no intrusion, and optical
power is sent via the O/P port as shown in Figure 8.4(e) [97].
Established on the hybrid plasmonic switch anticipated in [97], an asymmetric
MZI is depicted in Figure 8.4(f). The asymmetric MZIs were designed using com-
pact and low-loss Y-junction-based splitters and combiners. The I/P segment’s split-
ter divides the I/P power evenly between the MZI’s two arms, while the O/P section’s
combiner merges the optical powers of the two arms. The asymmetric MZI’s hybrid
plasmonic switch is incorporated in the sorter arm. Figure 8.4(g–i) shows the trans-
mission spectra and field mappings of the MZI for GST amorphous and crystalline
phases, respectively.

8.4.3 Theoretical and Numerical Investigations

Artificial intelligence (AI), as among the most effective areas in computer science,
focuses on simulating the framework of the nervous system by constructing artifi-
cial neural networks (ANNs), which maintain connections between neurons in mul-
tiple layers of the neural networks (NN) and give it higher accuracy and robustness.
ANN’s research has advanced significantly since the 1980s. It has also effectively
solved many functional challenges for modern computers in the fields of pattern
recognition, intelligent robots, automatic control, prediction and estimation, biomed-
icine, economy, and other fields, all while retaining better intelligence attributes.
Advanced machine learning techniques, which include ANNs [98, 99], have gotten a
lot of interest because of their practical uses in important tasks like image identifica-
tion and speech processing [100]. NNs employ a lot of multiply-accumulate (MAC)

BK-TandF-KAZANSKIY_9781032572949-230642-Chp08.indd 251 9/26/23 1:12 PM

 252 Photonics Elements for Sensing and Optical Conversions

operations, which puts a huge amount of pressure on current electronic computing

technology (e.g. CPU, GPU, FPGA, ASIC). For MAC activities, application-specific
apparatuses are recommended. Most NNs depend solely on real-valued arithmetic,
even though complicated arithmetic might provide a large benefit. For example, a
single complex-valued neuron with orthogonal decision boundaries can help resolve
the symmetry challenge and the XOR issues, while a single real-valued neuron can-
not [101]. Research indicates that complex-valued arithmetic [102] might help NNs
function better by providing extensive representational capacity, quick convergence,
powerful applicability, and noise-resistant memory mechanisms. Since complex
numbers must be formed by two real numbers, which fuels the growth of MAC
operations – the most commonly frequently utilized computationally expensive
components of NN algorithms-conventional digital electronic computing platforms
experience considerable slowdown when implementing algorithms utilizing com-
plex-valued operations [103, 104]. To circumvent these difficulties, it has been sug-
gested that the computationally demanding process of building NNs be delegated to
OC, which is proficient in genuinely complex-valued arithmetic [75].
Low power consumption, fast processing time, huge data storage, and intrinsic
parallelism are all benefits of OC that cannot be matched by its electrical cousin.
Numerous optical NN algorithms have been suggested. Photonic chip-based optical
NNs, for example, have grown extremely popular due to their remarkable adapt-
ability, scalability, and durability. This system has previously shown neuromorphic
photonic weight banks [105], all-optical NNs [106], and optical reservoir computing
with tremendous results [107]. On an embedded silicon photonic device, a typical
fully connected NN was realized experimentally [108]. Although this optical device
is established on light interference, the NN algorithms used are real-valued, negat-
ing the merits of sophisticated NNs. The optical impulses were already transformed
to photocurrents before entering the accumulator, resulting in a highly parallelized
optical NN accelerator established on photoelectric multiplication, which was also
built for real arithmetic. On-chip training, optical nonlinear activations [109], and
different NN designs are further subjects linked to optical NNs [110].
Apart from OC platforms, analog electronic apparatuses have effectively exhib-
ited multilayer perceptrons [111, 112] and convolutional NNs, as compared to more
prevalent digital electronic apparatuses [113]. Some earlier research has looked at
complex-valued NNs on analog electrical apparatuses [114–116]. Complex-valued
reservoirs also result in enhanced system dynamics and enhanced efficiency in reser-
voir computation. Although optical NNs can handle information in multiple degrees
of freedom (e.g. magnitude and phase) using complex-valued numerical methods
and acquire more effective information processing and analysis, there have been few
investigations in OC platforms for integrating general-purpose and complex-valued
NNs [117]. Strongly dependent on traditional deep learning algorithms developed for
real-valued arithmetic on traditional electronic computers, existing optical solutions
have not walked into this prospective flatland. These real-valued optical NNs are
built exclusively on the intensity information of the light waves, ignoring the phase
information, which eliminates one of OC’s primary advantages.
These problems were answered by designing and demonstrating an opti-
cal neural chip (ONC) that performs complex-valued arithmetic, demonstrating

BK-TandF-KAZANSKIY_9781032572949-230642-Chp08.indd 252 9/26/23 1:12 PM

 Optical Computing 253

the benefits of chip-based complex-valued networks via OC. It has been proved
that an optical neural chip (ONC) can build fully complex-valued NNs [118]. The
complex-valued ONC’s system is analyzed in four contexts: basic Boolean tasks,
species classification of an Iris data set, nonlinear data set classification (Circle and
Spiral), and handwriting recognition. When referred to its real-valued equivalent,
the complex-valued ONC achieves strong skillsets (i.e. precision, quick resolution,
and the capacity to generate nonlinear decision boundaries). Figure 8.5(a) depicts

FIGURE 8.5 A complex-valued coherent optical NN’s composition. (a) An I/P layer, many
hidden layers, and an O/P layer make up an optical NN. During the initial I/P preparation and
network evolution, the light signals are encoded and modulated by both amplitude and phase
in our complex-valued architecture [118]. (b) The ONC’s diagram for implementing complex-
valued networks. On a single chip, I/P preprocessing, weight multiplication, and coherence
recognition are all combined. The MZIs in red are responsible for the division and modula-
tion of the light signals (i1–i6). One MZI distinguishes the reference light that will be utilized
for coherent detection later, and the other indicates the MZIs that were utilized to build the
6 × 6 complex-valued weight matrix. On-chip coherent tracking is established on the remain-
ing gray designated MZIs [118]. (c) The ONC system’s process. Signal and reference light is
generated using a 1,550-nm coherent laser. The amplitude and phase of the signal light on each
path are modified by the machine learning (ML) job. The light inference is used to perform
the weighted sum process passively. The measurement findings are transferred to the electri-
cal interface for processing, which includes activation function application and cost function
computation. The modified weight matrices are then used to reprogram the ONC chip [118].

BK-TandF-KAZANSKIY_9781032572949-230642-Chp08.indd 253 9/26/23 1:12 PM

 254 Photonics Elements for Sensing and Optical Conversions

the optical NN’s architecture, consisting of an I/P layer, many hidden layers, and
an O/P layer [118]. During the initial I/P signal preparations and network growth in
the complex-valued architecture, light signals are encoded and modulated both by
optical magnitude and phase. Figure 8.5(b) depicts the ONC design for implement-
ing complex-valued NNs [118].
A single-chip handles I/P preprocessing, weight multiplication, and coherence
detection. The I/P signals are generated using a coherent laser (λ = 1,550 nm). The
ONC is simply a multiport interferometer with a special arrangement of Mach–
Zehnder interferometers (MZIs). Each MZI is made up of two beam splitter–phase
shifter (BS–PS) pairs. The BS has a stable transmittance of 50:50, while the PS is
thermally regulated to modify the phase. Different MZIs presented in the figure have
distinct capabilities. The bottom point couples the light into the chip. The one chain
of MZIs is responsible for I/P light splitting and modulation. The other MZI label
distinguishes the baseline light that will be utilized for coherent detection. The on-
chip light division ensures that light signals flowing through multiple optical routes
are polarized similarly and have a constant relative phase. The machine learning
job determines the modulation of the I/P. For jobs involving real-valued I/Ps, the
amplitude of the light signals is modulated, and the relative phases between distinct
pathways are reduced to 0.
On-chip coherent tracking is based using the gray MZIs. The optical chip’s O/P
light signals contain both magnitude and phase information, whereas traditional
intensity monitoring systems merely retrieve magnitude data. The intensity and
coherent detection are both possible with the integrated chip. The purpose of coher-
ent tracking is to find the phase angles between both the reference and signal light.
The O/P current achieved by linking photodiodes at both O/Ps in a balanced man-
ner is I1α2AsA1cosфs, where As and A1 are the signal and reference light amplitudes,
respectively. Likewise, if the baseline light is phase shifted by π/2, the O/P cur-
rent is IQα2AsA1sinфs. The фs is then calculated using the ratio of I1 and IQ, which
also aids in the removal of physical noise from optical components. The activation
analysis determines which detecting approach is used. A transimpedance amplifier
(TIA) converts the recorded photocurrents into voltage signals, which are subse-
quently gathered and analyzed by a traditional processor with an analog-to-digital
converter (ADC). As demonstrated in Figure 8.5(c), feedback signals may be created
and routed back to the ONC to alter chip layouts [118].

8.4.4 Diffractive NN
Prior works to OLGs focused primarily on constructive/destructive interference
effects between the I/P light signals, encompassing linear [119–122] and nonlinear
interference [123]. The reported works are heavily reliant on accurate positioning of
the basic characteristics of two I/P light signals, the control light, and/or the pump
light, such as phase difference, polarization, and intensity; if the two nanowires are
close to each other, as in the plasmonic logic gate, there is also a mandatory rule on
the size of I/P beams of light to prevent a big false I/P. Consequently, greater tight
control of I/P light may more fully actualize constructive or destructive interfer-
ence, resulting in a bigger intensity contrast ratio between the two O/P optical logic

BK-TandF-KAZANSKIY_9781032572949-230642-Chp08.indd 254 9/26/23 1:12 PM

 Optical Computing 255

states “1” and “0,” which is a critical quality to evaluate an OLG’s performance.
The heavy dependency on precise I/P light management has two negative effects on
the development of miniaturized OLGs. First, the substantial optical components
required to perform these controls are considered, and downsizing becomes chal-
lenging. Second, because of the complexities of achieving perfect I/P light control,
their performance may be unstable, and the intensity contrast ratio between two O/P
logic states may become rather low in practical circumstances. It is therefore very
desired for compact OLGs to eliminate these important I/P light needs. Due to the
necessity of developing innovative designs for all-optical apparatuses and systems,
such a goal remains an open problem that has long been sought after.
All seven fundamental optical logic operations (OLOs) are realized in a small
system utilizing just plane waves as the I/P signal, thanks to a simple yet universal
design method called a diffractive NN [106]. A compound Huygens’ metasurface
(MS) implements the diffractive NN, which may somewhat imitate the functions
of an ANN [124]. After training, the compound MS can disperse or focus the I/P
encoded light in one of two tiny areas/points, one representing logic state “1” and
the other representing logic state “0.” Three basic OLGs – NOT, OR, and AND
– are experimentally confirmed at microwave frequency utilizing a two-layer high-
efficiency dielectric MS as a conceptual example. There are two significant advan-
tages to the design technique. First, the implementation of OLOs here eliminates the
need for sophisticated and exact control of I/P light characteristics, which sets this
technique apart from earlier work. Furthermore, the I/P layer’s architecture is quite
broad and strong, and it can be easily changed into numerous user-friendly and pro-
grammable formats. Second, if the transmittance state of the I/P layer is dynamically
tunable, for example, electrically tunable if the optical mask is generated by a spatial
light modulator (SLM) [125], the suggested technique can enable comprehensive
logic functionality in a single optical network.
The I/P layer is a common optical mask that is designed to generate numerous
zones, as shown in Figure 8.6(a) [126]. Each optical mask zone is set to have two
alternative states for optical transmission without sacrificing generality, and its high
(low) transmittance state signals whether it is (is not) selected for OC. Then, merely
by allocating each of the seven fundamental optical logic operators and the I/P logic
states to a specific region, it is simple and efficient to directly specify all seven basic
optical logic operators and the I/P logic states in the optical mask. The hidden lay-
ers are responsible for decoding the encoded I/P light and rendering the computed
result at the O/P layer. The pattern of the I/P layer is shown in Figure 8.6(b). For the
sake of simplicity, each region’s high (low) transmittance condition is considered to
have a transmittance of 100% (0%) [126]. A cascaded two-layer transmission MS
with an axial spacing of 170λo is used to create the concealed layers (one of the tun-
able parameters in the training process of diffractive NN). Each MS is made up of
30×42 meta-atoms [inset in Figure 8.6(c)] [126], each of which has a 0.570λo-width
square cross section. Taking use of its unique qualities such as high transmittance
and polarization insensitivity, a simple yet practical high-efficiency dielectric MS is
developed. The calculated field intensity after training is depicted in Figure 8.6(d–m)
[126]. Most of the fields are appropriately concentrated within one of two tiny, speci-
fied zones, as intended.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp08.indd 255 9/26/23 1:12 PM

 256 Photonics Elements for Sensing and Optical Conversions

FIGURE 8.6 (a) A diffractive NN is laid out for photon-based logic functions. A diffractive
NN is used to numerically demonstrate three basic logic operations: NOT, OR, and AND. Two
levels of MSs make up the hidden layers here [126]. (b) The I/P layer’s diagram. The white
(gray) region’s light transmittance is set to 100 percent (0 percent). (c) The MS’s transmittance
response, which is made up of a 2D array of subwavelength meta-atoms. Each meta-atom may
adjust the incoming light’s phase (gray line) and amplitude (black line) locally [126]. (d–m)
Intensity distribution for three logic operations with random I/P logic states at the O/P layer. The
O/P optical logic state is specified as “1” (“0”) if the field is concentrated on the tiny, prescribed
areas on the left (right). Two dashed circles in each panel highlight the selected regions [126].

8.4.5 Theoretical and Numerical Investigations

With the progress of science, there seems to be a significant change in computing in
many ways over the years. Mechanical methods were used to create the earliest comput-
ers (1623 to 1945) [127]. Researchers attempted to create apparatuses that could readily
answer mathematical problems in the early 17th century. Several scientists, including
Gottfried Leibniz, Wilhelm Schickard, and Blaise Pascal, attempted to develop a cal-
culator that could handle addition, subtraction, multiplication, and division. George
Schertz and Edward Schertz created a system that could handle 15-digit numbers using
a 4-bit difference engine. One of the institutions that employed the mechanical com-
puter for punch card technology for the enumeration was the U.S. Census Bureau, which
was designed by Herman Hollerith of the International Business Machines [127].
A wide range of components, including optical gates, optical switches, opti-
cal interconnects, and optical memory, are required to form an optical computer.
Because of its applicability in ultrafast information processing [128] and the ways to
carry out various logical operations in OC systems, all-optical logic gates (LGs) have
become popular recently [129]. As a result, building all-optical LGs is the first step
in achieving advanced digital functionality in optical computers. Electronic LGs
were previously employed, but the highest switching speed obtained was 50 ps with
an average power of 0.5 mW per switching [130]. The reduced capacitance of p-n
junctions in semiconductor-based LGs is the explanation for this.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp08.indd 256 9/26/23 1:12 PM

 Optical Computing 257

Despite today’s electronic LGs being tiny, switching is still restricted by interlink-
ing capacitance; on the other side, optical LGs have switching speeds in the femto-
second range and are only restricted by the speed of light traveling through them
[131]. The prototype of all-optical LGs can be done in a variety of methods. The first
technique employs a semiconductor optical amplifier (SOA), which has a high gain
owing to refractive index variations. The original way to make all-optical LGs was
to use one of three approaches to introduce nonlinearity: cross-gain modulation [132,
133], cross-phase modulation [132], or four-wave mixing [133, 134]. SOA was also
utilized to construct SOA-assisted interferometer-based gates, which are all-optical
LGs [135]. However, SOA-based gates have several drawbacks, including SOA-based
apparatuses being constrained by SOA’s slow carrier recovery time, unsteady gates
owing to polarization sensitivity, and the SOA Mach–Zehnder interferometer (MZI)
technique, which necessitates more than two SOAs and complicates the scheme by
needing the proper tuning of the filter for SOA with the help of fiber LGs, as detun-
ing of the filter. Nonlinear WGs, in which localized nonlinear media have been used
by adjusting the control power, is another way for developing all-optical LGs [136–
138]. Nonlinear WG-based gates have several drawbacks, including a substantial I/P
signal power need and polarization dependency, which pose production issues.
PCs are periodic structured dielectric or EM media with photonic bandgaps
(PBGs) that prevent light from propagating through them [139–142]. John was the
one who introduced PC [143]. Unlike semiconductor crystals, which alter the char-
acteristics of electrons, these crystals impact the characteristics of photons. Light
has various benefits over electrons, including the ability to move faster in dielectric
material than electrons in the conductive metal and a larger data capacity in the
dielectric than electrons. One of the most significant functions in the realm of high
data transmission that 2D PCs may achieve is all-optical LGs. Furthermore, there
are two main types of PC-based gates: PBG-based gates and non-PBG-based LGs.

8.4.5.1 Non-Photonic Bandgap-Based All-Optical Logic Gates

Instead of establishing the PBG of the apparatuses, the I/P beam of any wavelength is
injected at the I/P, and a logic function is conducted utilizing a self-collimated beam
(SCB) in non-PBG-based PC gates. In this case, incoming light transmits to a device
in a given direction without diffraction. The phenomena of total internal reflection
are utilized to construct all-optical LGs utilizing an SCB. Total internal reflection
refers to the angle of incidence being larger than the critical angle, according to the
relation θ>arcsin(nL/nH ), where nL represents the low refractive index and n H repre-
sents the high refractive index. In a 2D PC, a device for the photonic-integrated cir-
cuit (PIC) was formulated and established on an SCB [144]. The structure is useful
for making optical switches and LGs, both of which are important parts of a PIC.
The device with Si rods in the air was designed using square lattice geometry. To
achieve the out-phasing at the O/P, a low refractive index medium was created such
that one-half of the beam was allowed to propagate while the other was reflected,
suggesting the existence of transmitted and reflected beams for the I/P signals. By
altering the phase difference between the reflected and transmitted beams, the OR
and XOR gate structures were developed. When the phase difference between the
I/P beams was 2kπ + π/2, O/P O1 functioned as an OR gate and O/P O2 as an XOR

BK-TandF-KAZANSKIY_9781032572949-230642-Chp08.indd 257 9/26/23 1:12 PM

 258 Photonics Elements for Sensing and Optical Conversions

FIGURE 8.7 (a) The switch’s schematic representation. On the I/P faces I1 and I2, two beams
with different phases collide. The computed steady-state field pattern of the E-polarized
mode at 0.194(a/λ) when incoming beams propagate in the Г-M direction in (b) and (c). The
phase difference between the two incident beams is adjusted to π/2 and –π/2, respectively, by
the phase difference φ1 – φ 2 [144].

gate, as illustrated in Figure 8.7(a–c) [144]. O1 worked as an XOR gate and O2 acted
as an OR gate when the phase difference was 2kπ – π/2 owing to a phase difference
of –π/2 between the I/P beams. The device’s frequency range was 0.188 to 0.199,
with a 17-dB extinction ratio.
For the PIC implementation, an all-OLG architecture of NOT, OR, AND, and
XOR gates centered on the SCB was presented [145]. The structure was con-
structed using 2D square lattice geometry with air holes in Si as the foundation
material. The framework has been developed and established on the phase differ-
ence between the I/P signals, following the phenomena of the SCB. When light
is launched at both the I/P and reference points of the same intensities, there was
partial reflectance of the I/P light wave while the reference light wave was com-
pletely mirrored, and when these light waves interfered with each other, there was
an O/P that relied on the phases at the I/Ps and interfered in a constructive manner
or destructively. It was proposed to use SCB-based logic gates for AND, NAND,
XNOR, and NOR OLGs. The framework has been developed with triangular lat-
tice topology and rods made of Si material in an air background [146]. Table 8.1

TABLE 8.1
Evaluation of Different Kinds of Non-PBG-Based All-OLGs
Logic Operational
Operation Wavelength (nm) Lattice Type Polarization CR (dB) Ref.
OR, XOR – Square (rods in air) TE 17 [144]
WG for OLGs 1,550 Square (air holes in Si) TE/TM – [147]
NOT, OR, AND, 1,550 Square (rods in Si) TE 30 [145]
XOR
AND, NAND, 1,555.1 Square (rods in air) TE 6 [146]
XNOR, NOR
XOR, OR 1,550 Triangular (rods in air) – – [148]
AND 1,550 Triangular (rods in air) – – [149]

BK-TandF-KAZANSKIY_9781032572949-230642-Chp08.indd 258 9/26/23 1:12 PM

 Optical Computing 259

shows that different studies offered alternative architectures for creating vari-
ous gates. Most OLGs were created in [145], which implemented AND, NAND,
XNOR, and XOR OLGs. Si rods in an air environment are the structure utilized by
all researchers. However, in the case of contrast ratio (CR), the structure developed
in [146] for NOT gate produced the greatest results, i.e. 30 dB. Furthermore, OLGs
constructed with a self-collimated light wave have a few drawbacks, including a
low CR, a wide area, a high cost owing to the big size, and signal guiding in only
the vertical and horizontal directions.

8.4.5.2 PBG-Based All-OLGs

The PBG of the architecture is utilized to detect hidden frequency ranges that can-
not flow through the structure in PBG-based all-optical OLGs. By adding various
forms of defects, one of the suppressed frequencies can transmit across the structure.
Multimode interference (MMI), nonlinear Kerr effect, and interference are com-
monly used in the design of PBG-based OLGs. In [150], MMI-based AND and XOR
OLGs are introduced, as depicted in Figure 8.8(a), with A and B I/P points and X
and Y O/P points. It was constructed utilizing triangular lattice geometry, SiO2 as
the foundation material, and Si rods. An I/P signal with an out-phasing of π was
transmitted at point A. On point B, the signal was emitted with an out-phasing of
–π/2, which represented logic “1” and creates logic “1” at the O/P point. When there
was a phase of π at point A, which indicated logic “0,” and another signal with out-
phasing π/2 at point B, which expressed logic “0,” the O/P generated logic “1.” In
another scenario, when a signal with an out-phasing of π expressed logic “0” at I/P
point A while a signal with an out-phasing of π/2 stated logic “0” at point B, logic
“0” was identified at the O/P. When both I/Ps at point A and point B displayed logic

FIGURE 8.8 Schematic of OLGs established on MMI. Inspired by [150], (a) AND/XOR
OLG, (b) XOR/XNOR, (c) AND and NOR OLGs. [Inspired by [151].]

BK-TandF-KAZANSKIY_9781032572949-230642-Chp08.indd 259 9/26/23 1:12 PM

 260 Photonics Elements for Sensing and Optical Conversions

“1” with out-phasings of 0 and –π/2, the O/P point sensed logic “0.” The AND OLG
was constructed using the same concept by modifying the length of MMI and appro-
priately choosing I/P signal phases. Both AND and XOR OLGs obtained a CR of
roughly 6.79 dB.
The self-imaging mechanism underpins the operation of MMI apparatuses. The
tiny field at the I/P activates guided modes in the effective region, which consti-
tutes interference in that region, in this phenomenon. Binary phase-shift keyed
(BPSK) signals are utilized as I/P logic values because, in MMI, I/P values are often
expressed by the phase of the I/P signals, and O/P logic levels are expressed using
amplitude independent of phase. Silicon material was employed for the rods in the
suggested device, with air as the background oriented in a square lattice shape [151].
As indicated in Figure 8.8(b), A and B are the I/P points of the XOR/XNOR OLG
arrangement, whereas X and Y are the O/P points. When there is an out-phasing
of π at point A, logic “0” was stated as signal I/P, while logic “1” is explicit with
phase 0. When there was an out-phasing of 3π/2 at point B, logic “0” was expressed
as signal I/P, while logic “1” was explicit with phase 0. To realize the AND OLG
indicated in Figure 8.8(c), I/Ps A and B with phase π represented logic “0,” whereas
I/Ps A and B with phase 0 represented logic “1.” With an out-phasing of 3π/2, logic
“0” was established at O1 and O2. The sole change in designing the NOR OLG was
that O1 and O2 were locked at logic “1,” represented by an out-phasing of π/2. The
AND OLG realized a CR of 21 dB, whereas the NOR operation achieved a CR of
19 dB [151]. Table 8.2 demonstrates the effectiveness of PBG-based all-OLGs on

TABLE 8.2
Device Performance of Different Types of PBG-Based All-OLGs
Logic Operational
Operation Wavelength (nm) Lattice Type Polarization CR (dB) Ref.
XOR, XNOR, 1530–1565 Triangular TM XOR-28.6, XNOR-28.6, [152]
NAND, OR (rods in SiO2) NAND-25, OR-26.6
XNOR, XOR, 1550 Square (rods in TE 37.4–40.41 [19]
OR, NAND air)
NAND 1554 Square (rods in TM – [154]
air)
NOT, AND, OR, 1550 Triangular TE NOT-3.74, AND-11.47, [48]
XOR, XNOR, (holes in Si) OR-12.48, XOR-6.50,
NAND XNOR-6.50
OR, AND OR-1529, Triangular TM 6 [156]
AND-1538 (rods in air)
NAND, NOR, 1550 Square (rods in TE 17.59, 14.3, 10.52 [157]
XNOR air)
NOR, AND 1550 Triangular TE – [158]
(rods in air)
OR 1287.8 Triangular TE 7.27 [159]
(rods in air)

BK-TandF-KAZANSKIY_9781032572949-230642-Chp08.indd 260 9/26/23 1:12 PM

 Optical Computing 261

performance metrics like the CR. PBG-based OLGs established on the interference
phenomenon have a higher CR and are easier to construct than OLGs established on
other concepts [152–155].

8.4.6 Resonant Nanophotonic Constructions

Nanophotonic components are being suggested as a novel foundation for analog
optoelectronic computing [160–163]. It was shown that a layer of a well-designed
metamaterial can visually execute several essential mathematical operations (differ-
entiation and integration of light waves concerning a spatial coordinate, convolution
of light waves with a predefined core) [161]. It generated a lot of interest and aided
the creation of novel nanophotonic structures for AOC. Differentiation (integration)
of the pulse envelope is commonly understood in the context of temporal differen-
tiation (integration) of a light wave (optical pulse). The most important findings in
fiber-optic network design and production for the differentiation (integration) of opti-
cal pulses propagating in OFs were acquired [164–170]. Bragg resonant structures
and RRs have been proposed to perform time-domain differentiation (integration)
procedures. Because the spectra of reflection and transmission in the proximity of
resonances are characterized by the Fano profile, and in a certain frequency, inter-
vals may accurately estimate the transfer functions of differentiating and integrating
filters; resonant formations can be used to implement differentiation and integration
operations. It is worth noting that more complex systems with many resonators may
solve ordinary differential equations of various orders as well as solutions of differ-
ential equations in the temporal domain [171–174].
The employment of differentiation and integration of light pulses traveling in free
space is also of importance, in addition to the temporal changes of light signals
traveling through OFs. V.A. Soifer’s research team achieved several significant dis-
coveries in this field. The processes of differentiation (integration) of the pulse enve-
lope may be successfully conducted by employing a resonant diffraction grating, as
described in studies [175–180]. High-order derivatives may be calculated quickly
using a set of multiple stacked diffraction gratings [179]. The creation of the theory
of spatiotemporal transformations of light waves was established on the theoretical
explanation of diffraction of light pulses on resonant diffractive assemblies.
An optical correlator can conduct a wide range of spatial filtering functions on
light waves. The optical correlator, also known as the coherent optical Fourier pro-
cessor, is made up of two lenses that execute the FT optically and a spatial filter
that encapsulates the transmission function characterizing the incoming light beam’s
needed spatial transformation. As a spatial filter, a differentiating filter with a com-
plicated transmission function corresponding to the spatial frequency should be uti-
lized to accomplish the differentiation operation in such a scheme. Nevertheless, the
optical correlator’s significantly larger size limits its practical applicability. Using
nanophotonic configurations, spatial differentiators, and integrators with a thickness
similar to the wavelength of the modulated light wave may be created. In 2014, the
idea of implementing spatial changes of light waves utilizing nanophotonic appara-
tuses was introduced. The fundamental work in [160] was a theoretical investigation
of the execution of spatial differentiation, integration, and convolution operations

BK-TandF-KAZANSKIY_9781032572949-230642-Chp08.indd 261 9/26/23 1:12 PM

 262 Photonics Elements for Sensing and Optical Conversions

on light waves. There were two ways offered. The first was to employ tiny analogs
of optical correlators, with the traditional Fourier lenses being substituted by tiny
layers with a gradient refractive index and an MS serving as the spatial filter storing
the requisite transmission function. The second method was to utilize a multilayer
structure that was particularly built to accomplish the spatial transformation of the
I/P signal indicated by the convolution operator with a certain core.
Professor S.I. Bozhevolnyi’s team conducted the first study proving the capabil-
ity of differentiation and integration regarding a spatial factor using an MS in 2015
[163]. The first strategy was utilized for optical differentiation and integration, in
which the processes were carried out in an optical correlator comprised of a lens
and a reflecting spatial filter. A reflecting MS encrypting transmission function of a
differentiating (integrating) filter was employed as the filter. The MS in question was
made up of a series of the metal-insulator-metal resonant circuit [163, 181]. Fabry–
Perot resonances of plasmonic modes traveling in metal slots are supported by such
resonators. The experimental findings provided in [163] show that optical differen-
tiation (integration) may be implemented with the use of an MS. Simultaneously, it
should be highlighted that the MSs in [163] only conduct differentiation and integra-
tion tasks in conjunction with a lens, preventing the suggested system from being
compact. In subsequent work [182–184], dielectric MSs were used to encode the
essential transmission functions of the spatial filter, causing a reduction in losses due
to absorption in the metal claddings of nano-resonators and enhanced performance
in the deployment of differential and integral conversions.
Because the nanophotonic assembly directly executes the needed spatial change
of the I/P signal, the second option is more plausible. In this example, the opti-
cal correlator is replaced with a single assembly (with no extra lenses). It’s worth
noting that the structure’s reflection and transmission coefficients as functions of
spatial frequency (a tangential component of the incoming wave vector) match the
transfer functions defining the incident light beam’s conversion [163]. As a result,
nanophotonic assemblies with reflection (or transmission) coefficients that approxi-
mate the transfer functions of the differentiator or integrator would be required to
accomplish the elementary actions of spatial differentiation or integration of the
light beam. A resonant reflection or transmission band with a Lorentzian line form
can be used to mimic the integrator’s transfer function. About the zeros of reflection
and transmission happening near the resonance, the differentiator’s transfer func-
tion is very well modeled.
V.A. Soifer’s research team achieved the most important results in this sector. The
work of this group initially established that phase-shifted Bragg gratings (PSBGs)
[185–187], resonant diffraction gratings [188, 189], PC resonators [190–195], and
three-layer arrangements with W-shaped refractive index profiles may efficiently
execute the functions of differentiation and integration of the light wave profile
[196]. It’s worth noting that resonant nanostructures can also be used to compute the
Laplace operator optically [187]. In image processing, this method is utilized to rec-
ognize edges. The first investigation affirming the potential of differentiating a spa-
tial variable using the resonant diffraction grating was carried out in 2018 using the
apparatuses of the Collective Use Center “Nano-Photonics and Diffraction Optics”
[197], which was established by a cooperative endeavor of Samara University and the

BK-TandF-KAZANSKIY_9781032572949-230642-Chp08.indd 262 9/26/23 1:12 PM

 Optical Computing 263

Russian Academy of Sciences’ Image Processing Systems Institute [196]. It resulted

in a high level of distinction that was far superior to the quality of differentiation
attained using MSs [163]. It’s also important to note that differentiators and inte-
grators founded on resonant diffraction gratings and PSBGs are not only smaller
but also easier to fabricate than equivalent apparatuses that rely on correlators with
MSs. The emphasis of [198] is a differentiator comprising of a prism with a metal
film placed on one of its sides; specifically, the differentiator is composed of a prism
with a metal film-coated on one of its sides [199, 200]. The incident light wave’s
differentiation is carried out in this scenario in reflection due to the incident light
wave’s activation of a surface plasmon-polariton on the metal film’s surface. The
effectiveness of employing such a framework for optical edge detection was proved
in experiments.
The first-order differentiation of the transverse profile of an incoming light wave
regarding a spatial variable is demonstrated experimentally using a subwavelength
diffraction grating as shown in Figure 8.9(a) [189]. Figure 8.9(b) shows a typical
SEM picture of the manufactured grating. The experimental findings accord well
with the provided theoretical model, implying that the differentiation occurs in trans-
mission at oblique incidence and is linked to the grating’s guided-mode resonance.
As per this concept, the grating’s transfer function about the resonance is similar to
that of an exact differentiator. Figure 8.9(c and d) illustrates the incident Gaussian
light wave and transmitted light wave profiles, respectively. The incident light wave
has a Gaussian shape to it. The precise derivative agrees well with the form of the

FIGURE 8.9 (a) The diffraction of an optical light wave on a resonant diffraction assembly
made up of a grating on top of a slab waveguide layer put on a substrate [189]. (b) The dif-
fraction grating was created using ERP-40 electron resist on top of a TiO2 layer, as shown by
SEM [189]. (c) Measured profile of incident Gaussian light wave [189]. (d) Measured profile of
transmitted light wave [189]. (e) Analytically derived derivative of incident light wave [189].
(f) Profile of transmitted light wave estimated taking into account manufactured structural
flaws [189].

BK-TandF-KAZANSKIY_9781032572949-230642-Chp08.indd 263 9/26/23 1:12 PM

 264 Photonics Elements for Sensing and Optical Conversions

transmitted light wave (Figure 8.9(e)). The profile of the transmitted light wave esti-
mated considering manufactured structural flaws is also shown in Figure 8.9(f). The
configuration under consideration might be used in the development of novel pho-
tonic apparatuses for light wave shaping, optical data processing, and AOC.

8.5 GROWTH IDEAS, CONSTRAINTS, AND MISCONCEPTIONS

For the many methods of OC, there are certain common obstacles. To begin with, large-
scale integration of optical-electrical chips must be manufactured to boost the paral-
lelism of OC systems at the physical level. In addition, optical-electrical co-package
technology is required to decrease the price of data transport between the electrical
and optical domains. Second, current optical transmitters and modulators are pri-
marily intended for optical transmission rather than processing. Because most imple-
mentations demand high bit depth I/P data, OC methods allow substantially greater
extinction ratio (ER) and linearity of optical apparatuses than optical communication.
Furthermore, because optical apparatuses with a greater ER and linearity may
allow high-efficiency optical coding for data I/P, the overall performance will be
enhanced. Eventually, a new architectural design is necessary. The optical-electrical
transformation might severely limit the energy efficiency of the hybrid computing
device, making it impossible to take use of the benefits of OC in a traditional com-
puter design. The new architectural design might have a significant speed-up fac-
tor while retaining as much customizability as feasible in the meantime. Finally,
there are a few investigations on algorithms that are appropriate for analog OC.
Algorithms are now built using Boole logics, which are ideal for digital computer
systems. They are, nonetheless, impossible to match when it comes to the function-
ality given by OC. If algorithms for OC are created, their operation overhead and
execution time will be significantly reduced compared to present methods.
Despite the numerous hurdles, OC’s potential has been growing. To begin with,
various manufacturing methods have been involved in the development of larger-
scale optical-electrical chip integration. For instance, Light-matter delivered the
world’s first 4096 MZI integrated chip, called “Mars,” demonstrating the capa-
bility of large-scale integration and giving researchers in OC more confidence.
Furthermore, the previously described wavelength division multiplexing (WDM)
and mode division multiplexing (MDM), as well as the spatial optical system, are all
sustainable with increasing parallelism. Moreover, by directly employing a faster-
speed optical device with low-bit depth optical coding, the poor ER and linearity
of optical apparatuses may be adjusted. In terms of data I/P performance, a 2-GHz
optical modulator with OOK and a 1-GHz optical modulator with PAM4 are equal.
This type of compensation, on the other hand, is only possible in computer processes
that can be transformed into a linear combination of low-bit depth operations in the
time domain. Using low-bit depth quantization for application I/P data, on the other
hand, is a common method for making current optical instruments practical in OC.
Light wave looping should be extensively used to maintain data in the optical
domain for as prolonged as feasible in hybrid computing systems to decrease the over-
head from optical-electrical conversion. The temporal delay induced by light wave
looping might be minor due to the fast propagation speed of light. New designs can

BK-TandF-KAZANSKIY_9781032572949-230642-Chp08.indd 264 9/26/23 1:12 PM

 Optical Computing 265

TABLE 8.3
Some Important Applications of OC
Applications References
1 Parallel processing [202–205]
2 Optical switches [4, 206, 207]
3 Optical data storage [201, 208, 209]
4 Data communication [210, 211]
5 All-optical logic operation [128, 212, 213]

be inspired by dataflow approaches. Finally, OC techniques might consider the com-

plicated operators available in the optical domain. To minimize implementation risk
and processing time, several sets of Boole logic operators in present algorithms can
be substituted with a single complicated operator. As a result, integrating complicated
operators with Boole logic operators in an algorithm might be a promising technique
to construct OC algorithms. OC’s possibilities have been expanding. The ever-increas-
ing need for ANNs, as well as their processing requirements, will continue to push
research into OC patterns. Optical sensing and communication may provide another
opportunity for OC to be used. Furthermore, high-complexity computing algorithms
in the optical domain, including FT, convolution, and equation solving, might signifi-
cantly improve systematic efficiency. Some of the important applications of OC and
current state-of-the-art research works are summarized in Table 8.3.
Investigators from the universities of Oxford, Exeter, and Münster have developed
a novel technology that allows them to retain more optical data in a smaller space
on-chip than formerly conceivable. This method achieves on the phase-change opti-
cal memory cell, which utilizes light to record and read data, and might result in a
quicker, more energy-efficient computer memory. Unlike today’s computers, which
utilize electrical impulses to store data in one of two states – zero or one – the opti-
cal memory cell stores data via light. The researchers achieved optical memory with
over 32 states, or levels, equating to 5 bits. This is a significant step toward the devel-
opment of an all-optical computer, which is a long-term aim for many researchers
in this field [201].
In the coming years, we can expect the creation of all-optical computers in the
form of research samples, however industrial prototypes that have a market niche
require the implementation of effective technological solutions in electronics.
Nevertheless, the advantages of photonic solutions over electronic ones listed in the
review paper make it necessary for scientists to continue research in this direction
and systematize the accumulated results. From the point of view of the authors, at
present, solutions in the optical implementation of certain types of special proces-
sors may turn out to be effective, for example, replacing the used electronic analog
machines with similar optical ones, providing the solution of certain types of dif-
ferential equations, contour detection, etc. The proposed solutions can be promis-
ing where initially there is an optical signal at the input of a special processor, for
example, a video stream. In this case, the optical neural network allows processing

BK-TandF-KAZANSKIY_9781032572949-230642-Chp08.indd 265 9/26/23 1:12 PM

 266 Photonics Elements for Sensing and Optical Conversions

and recognizing some images faster than converting them into electronic form and
processing it on a digital computer using graphics accelerators.
The fact that processing is a nonlinear process in which several signals must
interplay is a key problem for optical computing. Light, which is an EM wave, can
only engage with another EM wave in the possession of electrons in a material, and
the intensity of this interaction is significantly less for EM waves than for electrical
signals in a traditional computer. As a response, processing elements for an optical
computer may require more power and have bigger size than processing elements for
a traditional electronic computer employing transistors. Another myth is that optical
transistors should be susceptible to incredibly high frequencies since light travels far
faster than electrons’ drift velocity and at frequencies measured in THz. The pace
at which an optical transistor may react to a signal is still restricted by its spectral
bandwidth, since every EM wave must respect the transform limit. Practical con-
straints such as dispersion frequently restrict fiber-optic communications channels to
bandwidths of tens of GHz, just marginally better than many silicon transistors. To
achieve far quicker operation than electronic transistors, effective means of sending
ultrashort pulses along extremely dispersive waveguides would be required.

8.6 CONCLUSION
Artificial neural networks (ANNs) have been used to build artificial intelligence
in recent times, resulting in unimaginable demands for computing resources.
Nevertheless, due to Moore’s Law’s insufficiency and the failure of Dennard’s
scaling laws, traditional computer hardware related to electronic transistors and
von Neumann architecture would be unable to meet such an incomprehensible
demand. Conversely, analog optical computing (AOC) provides an alternate method
for unleashing enormous processing power to speed up a wide range of compute-
intensive operations. To perform specific computing operations, AOC makes use
of physical features of light, including amplitude and phase, as well as the interplay
between light and optical devices. Because of the unique mathematical portrayal
of computational processes in one specific analog optical computing system, it is a
specialized computing system. AOC can achieve higher data processing acceleration
in specialized applications, including pattern recognition and numerical calculation,
as opposed to traditional digital computing. The popularity of optically implemented
neural networks has grown in recent years, attributable to the growing quantity of
databases that must be handled, placing a strain on the effectiveness of traditional
digital, electronic computers. The key consideration in implementing a viable optical
computer (including a neural one) is to integrate the linear part of the system, from
which optics derives its competitive advantage, with nonlinear components and I/P–
O/P interfaces while preserving the optical interconnections’ speed and power effi-
ciency. In this chapter, several optical components such as spatial light modulators,
plasmonic switches, neural networks, diffractive neural networks, photonic crystal
all-optical logic gates, and resonant nanophotonic structures are used to implement
optical computing have been reviewed and their advancements have been discussed.
We believe that this chapter will be beneficial to the scientific community working
on the topic of optical computing.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp08.indd 266 9/26/23 1:12 PM

 Optical Computing 267

REFERENCES
1. de Lima, T. F., Tait, A. N., Mehrabian, A., Nahmias, M. A., Huang, C., Peng, H.-T.,
Marquez, B. A., Miscuglio, M., El-Ghazawi, T., Sorger, V. J., Shastri, B. J., and Prucnal,
P. R. 2020. Primer on silicon neuromorphic photonic processors: Architecture and
compiler. Nanophotonics. 9(13):4055–73. https://doi.org/10.1515/nanoph-2020-0172.
2. Minzioni, P., Lacava, C., and Tanabe, T. et al. 2019. Roadmap on all-optical processing.
J. Opt. 21(6):063001. https://doi.org/10.1088/2040-8986/ab0e66.
3. Boyd, G. D., Chirovsky, L. M. F., and Morgan, R. A. 1991. Dynamic optical switching
of symmetric self-electro-optic effect devices. Appl. Phys. Lett. 59(21):2631. https://doi.
org/10.1063/1.105920.
4. Zasedatelev, A. V., and Baranikov, A. V. et al. 2021. Single-photon nonlinearity at room
temperature. Nature. 597:493–7. https://doi.org/10.1038/s41586-021-03866-9.
5. Brunner, D., Marandi, A., Bogaerts, W., and Ozcan, A. 2020. Photonics for computing
and computing for photonics. Nanophotonics. 9(13):4053–4. https://doi.org/10.1515/
nanoph-2020-0470.
6. Goswami, D. 2003. Optical computing. Resonance. 8:8–21. https://doi.org/10.1007/
BF02834399.
7. Optalysys. 2022. https://optalysys.com/.
8. Smith, S. 1985. Lasers, nonlinear optics and optical computers. Nature. 316:319–24.
https://doi.org/10.1038/316319a0.
9. Yeh, P., Chiou, A., Hong, J., Backwith, P., Chang, T., and Khoshnevisan, M. 1989.
Photorefractive nonlinear optics and optical computing. Opt. Eng. 28(4):328–43.
https://doi.org/10.1142/9789812832047_0075.
10. Gillette, J. 1989. CD-ROM data storage technology: Benefits and limitations in docu-
ment publication. Book Res. Q 5:37–43. https://doi.org/10.1007/BF02683798.
11. Goodman, J. 1968. Introduction to Fourier optics. San Francisco, CA: McGraw-Hill.
ISBN: 978-0-07-023776-6.
12. Maréchal, A., and Croce, P. 1953. Un filtre de frequences spatiales pour l’amelioration
du contraste des images optiques. Comptes Rendus de l’Académie des Sciences
237(12):607–9.
13. Cutrona, L. J., Leith, E. N., and Palermo, C. J. et al. 1960. Optical data processing
and filtering systems. IEEE Trans. Inf. Theory. 6(3):386–400. https://doi.org/10.1109/
TIT.1960.1057566.
14. Neill, E. 1956. Spatial filtering in optics. IRE Trans. Inf. Theory. 2(2):56–65. https://doi.
org/10.1109/TIT.1956.1056785.
15. Lugt, A. V. 1964. Signal detection by complex spatial filtering. IEEE Trans. Inf. Theory.
10(2):139–45.
16. Weaver, C., and Goodman, J. 1966. A technique for optically convolving two functions.
Appl. Opt. 5(7):1248–9. https://doi.org/10.1364/AO.5.001248.
17. Rhodes, W., and Sawchuk, A. 1981. Incoherent optical processing. In Optical informa-
tion processing: Fundamentals, ed. S. H. Lee, 69–110. Berlin, Heidelberg: Springer-
Verlag. https://doi.org/10.1007/3540105220_10.
18. Leith, E. 1989. Incoherent optical processing and holography. In Optical processing
and computing, ed. H. Arsenault, 421–440. SanDiego, CA: Academic Press. https://doi.
org/10.1016/B978-0-12-064470-4.50017-X.
19. Goodman, J. 1981. Linear space-variant optical data processing. In Optical informa-
tion processing: Fundamentals, ed. S. H. Lee, 235–260. Berlin, Heidelberg: Springer-
Verlag. https://doi.org/10.1007/3540105220_13.
20. Ambs, P., Lee, S. H., Tian, Q., and Fainman, Y. 1986. Optical implementation of the
Hough transform by a matrix of holograms. Appl. Opt. 25(22):4039–45. https://doi.
org/10.1364/AO.25.004039.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp08.indd 267 9/26/23 1:12 PM

 268 Photonics Elements for Sensing and Optical Conversions

21. Lee, S. 1981. Nonlinear optical processing. In Optical information processing:

Fundamentals, ed. S. H. Lee, 261–303. Berlin, Heidelberg: Springer-Verlag. https://doi.
org/10.1007/3540105220_14.
22. Ambs, P. 2010. Optical computing: A 60-year adventure. Adv. Opt. Technol.
2010:372652. https://doi.org/10.1155/2010/372652.
23. Schwabe, R., Zelinger, S., Key, T., and Phipps, K. 1998. Electronic lighting interfer-
ence. IEEE Ind. Appl. Mag. 4:46–8. https://doi.org/10.1109/2943.692532.
24. Hu, W., Li, X., Yang, J., and Kong, D. 2010. Crosstalk analysis of aligned and mis-
aligned free-space optical interconnect systems. J. Opt. Soc. Am. A. 27(2):200–5. https://
doi.org/10.1364/JOSAA.27.000200.
25. Carpenter, J. 2012.Holographic mode division multiplexing in optical fibres. PhD diss.
Cambridge, UK: University of Cambridge.
26. O’Neill, E. L. 1956. Spatial filtering in optics. IRE Trans. Inform. Theory. IT-2:56–65.
https://doi.org/10.1109/TIT.1956.1056785.
27. Stroke, G. W. 1966. An introduction to coherent optics and holography. New York:
Academic Press.
28. Preston, K. 1972. Coherent optical computers. New York: McGraw-Hill. ISBN:
978-0-07-050785-2.
29. Lugt, A. V. 1974. Coherent optical processing. Proc. IEEE. l62(10):1300–19. https://doi.
org/10.1109/PROC.1974.9624.
30. Perrin, M., and Metzger, G. 1975. Principles and feasibility of an optical prepro-
cessor in high energy physics. Nucl. Instrum. Methods. 126:509–18. https://doi.
org/10.1016/0029-554X(75)90801-0.
31. Kotlyar, V. V., Khonina, S. N., and Soifer, V. A. 1998. Light field decomposition in
angular harmonics by means of diffractive optics. J. Mod. Opt. 45(7):1495–506. https://
doi.org/10.1080/09500349808230644.
32. Khonina, S. N., Kotlyar, V. V., Soifer, V. A., Paakkonen, P., and Turunen, J. 2001.
Measuring the light field orbital angular momentum using DOE. Opt. Mem. Neural
Netw. 10(4):241–55.
33. Kotlyar, V. V., Khonina, S. N., Soifer, V. A., and Wang, Y. 2002. Light field orbital
angular moment measurement with the help of diffractive optical element. Avtometriya.
38(3):33–44.
34. Kotlyar, V. V., Kovalev, A. A., and Volyar, A. V. 2020. Topological charge of optical
vortices and their superpositions. Comput. Opt. 44(2):145–54. https://doi.org/10.18287/
2412-6179-CO-685.
35. Reddy, A. N. K., Anand, V., Khonina, S. N., Podlipnov, V. V., and Juodkazisk, S. 2021.
Robust demultiplexing of distinct orbital angular momentum infrared vortex beams
into different spatial geometry over a broad spectral range. IEEE Access. 9:143341–8.
https://doi.org/10.1109/ACCESS.2021.3120836.
36. Porfirev, A. P., and Khonina, S. N. 2016. Experimental investigation of multi-order
diffractive optical elements matched with two types of Zernike functions. Proc. SPIE.
9807:98070E. https://doi.org/10.1117/12.2231378.
37. Degtyarev, S. A., Porfirev, A. P., and Khonina, S. N. 2017. Zernike basis-matched
multi-order diffractive optical elements for wavefront weak aberrations analysis. Proc.
SPIE. 10337:103370Q. https://doi.org/10.1117/12.2269218
38. Khorin, P. A., Volotovskiy, S. G., and Khonina, S. N. 2021. Optical detection of values
of separate aberrations using a multi-channel filter matched with phase Zernike func-
tions. Comput. Opt. 45(4):525–33. https://doi.org/10.18287/2412-6179-CO-906.
39. Khonina, S. N., Karpeev, S. V., and Porfirev, A. P. 2020. Wavefront aberration sensor
based on a multichannel diffractive optical element. Sensors. 20(14):3850. https://doi.
org/10.3390/s20143850.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp08.indd 268 9/26/23 1:12 PM

 Optical Computing 269

40. Khorin, P. A., Porfirev, A. P., and Khonina, S. N. 2022. Adaptive detection of wave aber-
rations based on the multichannel filter. Photonics. 9(3):204. https://doi.org/10.3390/
photonics9030204.
41. Soifer, V. A., Golub, M. A., and Khonina, S. N. 1993. Decorrelated features of images
extracted with the aid of optical Karhunen-Loeve expansion. Pattern Recognit. Image
Anal. 3(3):289–95.
42. Soifer, V. A., and Khonina, S. N. 1994. Stability of the Karhunen-Loeve expansion in
the problem of pattern recognition. Pattern Recognit. Image Anal. 4(2):137–48.
43. Soifer, V. A., Kotlyar, V. V., and Khonina, S. N. 1996. An optical method of directions
field construction. Avtometriya. 1:31–6.
44. Soifer, V. A., Kotlyar, V. V., Khonina, S. N., and anad Skidanov, R. V. 1996. Optical
methods of fingerprints identification. Comput. Opt. 16:78–89.
45. Soifer, V., Kotlyar, V., Khonina, S., and Skidanov, R. 1998. Optical-digital methods
of fingerprint identification. Opt. Lasers Eng. 29(4–5):351–9. https://doi.org/10.1016/
S0143-8166(97)00122-X.
46. Khonina, S. N., Kotlyar, V. V., Skidanov, R. V., and Soifer, V. A. 2003. Optodigital
system for identifying fingerprints in real time. J. Opt. Technol. 70(8):586–9. https://
doi.org/10.1364/JOT.70.000586.
47. Khonina, S. N., and Demidov, A. S. 2014. Extended depth of focus through imaging
system’s phase apodization in coherent and incoherent cases. Opt. Mem. Neural Netw.
23(3):130–9. https://doi.org/10.3103/S1060992X14030035.
48. Khonina, S. N., and Ustinov, A. V. 2015. Generalized apodization of an incoherent
imaging system aimed for extending the depth of focus. Pattern Recognit. Image Anal.
25(4):626–31. https://doi.org/10.1134/S1054661815040100.
49. Khonina, S. N., Ustinov, A. V., and Porfirev, A. P. 2019. Dynamic focal shift and extend-
ing depth of focus based on the masking of the illuminating beam and using an adjustable
axicon. J. Opt. Soc. Am. A. 36(6):1039–47. https://doi.org/10.1364/JOSAA.36.001039.
50. Khonina, S. N., Volotovskiy, S. G., Dzyuba, A. P., Serafimovich, P. G., Popov, S. B.,
and Butt, M. A. 2021. Power phase apodization study on compensation defocusing
and chromatic aberration in the imaging system. Electronics. 10(11):1327. https://doi.
org/10.3390/electronics10111327.
51. Khonina, S. N., Kotlyar, V. V., Skinkaryev, M. V., Soifer, V. A., and Uspleniev,
G. V. 1992. The phase rotor filter. J. Mod. Opt. 39(5):1147–54. https://doi.org/10.1080/
09500349214551151.
52. Ananin, M. A., and Khonina, S. N. 2009. Modelling of optical processing of images
with use of the vortical spatial filter. Comput. Opt. 33(4):466–72.
53. Davis, J. A., McNamara, D. E., Cottrell, D. M., and Campos, J. 2000. Image processing
with the radial Hilbert transform: Theory and experiments. Opt. Lett. 25(2):99–101.
https://doi.org/10.1364/OL.25.000099.
54. Guo, C.-S., Han, Y.-J., Xu, J.-B., and Ding, J. 2006. Radial Hilbert transform with
Laguerre-Gaussian spatial filters. Opt. Lett. 31(10):1394–6. https://doi.org/10.1364/
OL.31.001394.
55. Lazarev, G., and Hermerschmidt, A. 2012. LCOS spatial light modulators: Trends
and applications. In Optical imaging and metrology: Advanced technologies, eds. W.
Osten, and N. Reingand, 1–29. Weinheim: Wiley-VCH.
56. Pivnenko, M., Li, K., and Chu, D. Sub-millisecond switching of multi-level liquid
crystal on silicon spatial light modulators for increased information bandwidth. Opt.
Express. 2021;29(16):24614–28. https://doi.org/10.1364/OE.429992.
57. Jin, D., Zhou, R., Yaqoob, Z., and So, P. T. C. 2018. Dynamic spatial filtering using
a digital micromirror device for high-speed optical diffraction tomography. Opt.
Express. 26(1):428–37. https://doi.org/10.1364/OE.26.000428.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp08.indd 269 9/26/23 1:12 PM

 270 Photonics Elements for Sensing and Optical Conversions

58. Turtaev, S., Leite, I. T., Mitchell, K. J., Padgett, M. J., Phillips, D. B., and Cizmar,
T. 2017. Comparison of nematic liquid-crystal and DMD based spatial light modu-
lation in complex photonics. Opt. Express. 25(24):29874–84. https://doi.org/10.1364/
OE.25.029874.
59. Phase spatial light modulator LCOS-SLM. 2012. Opto-semiconductor handbook, 1–
14. Tokyo: Hamamstsu Photonics.
60. Kovachev, M. et al. 2006. Reconstruction of computer generated holograms by spatial
light modulators. In Multimedia content representation, classification and security, ed.
B. Gunsel, A. K. Jain, A. M. Tekalp, B. Sankur, 706–713. Berlin, Heidelberg: Springer.
https://doi.org/10.1007/11848035_93.
61. Younus, S., Hussein, A., Alresheedi, M., and Elmirghani, J. 2017. CGH for indoor vis-
ible light communication system. IEEE Access. 5:24988–5004. https://doi.org/10.1109/
ACCESS.2017.2765378.
62. Torii, Y., Balladares-Ocana, L., and Martinez-Castro, J. 2013. An iterative Fourier
transform algorithm for digital hologram generation using phase-only information and
its implementation in a fixed-point digital signal processor. Optik. 124(22):5416–21.
https://doi.org/10.1016/j.ijleo.2013.03.112.
63. Ripoll, O., Kettunen, V., and Herzig, H. P. 2004. Review of iterative Fourier-transform
algorithms for beam shaping applications. Opt. Eng. 43(11):2549–56. https://doi.
org/10.1117/1.1804543.
64. Roelens, M. A. F. et al. 2008. Dispersion trimming in a reconfigurable wavelength selec-
tive switch. J. Lightw. Technol. 26(1):73–8. https://doi.org/10.1109/JLT.2007.912148.
65. Wang, M. et al. 2017. LCoS SLM study and its application in wavelength selective
switch. Photonics. 4(2):22. https://doi.org/10.3390/photonics4020022.
66. Turunen, J., and Wyrowski, F. 1998. Diffractive optics for industrial and commercial
applications. Berlin: Wiley-VCH Verlag GmbH. ISBN: 978-3-05-501733-9.
67. Osten, W., Kohler, C., and Liesener, J. 2005. Evaluation and application of spatial light
modulators for optical metrology. Optica Pura y Applicada. 38(3):71–81.
68. Varga, J. J. M. et al. 2015. Preparing arbitrary pure states of spatial qudits with a
single phase-only spatial light modulator. J. Phys. Conf. Ser. 605:012035. https://doi.
org/10.1088/1742-6596/605/l/012035.
69. Schröder, J., Roelens, M., Du, L., Lowery, A., and Eggleton, B. 2012. LCOS based
waveshaper technology for optical signal processing and performance monitoring.
17th Opto-Electronics and Communications Conf: 859–60. https://doi.org/10.1109/
OECC.2012.6276666.
70. Barbier, P., and Moddel, G. 1997. Spatial light modulators: Processing light in real time.
Opt. Photon. News. 8(3):16–21. https://doi.org/10.1364/OPN.8.3.000016.
71. Shrestha, P., Chun, Y., and Chu, D. 2015. A high-resolution optically addressed spa-
tial light modulator based on ZnO nanoparticles. Light. Sci. Appl. 4:e259. https://doi.
org/10.1038/lsa.2015.32.
72. Chen, P., Ma, L.-L., Hu, W., Shen, Z.-X., Bisoyi, H., Wu, S.-B., Ge, S.-J., Li, Q., and
Lu, Y.-Q. 2019. Chirality invertible superstructure mediated active planar optics. Nat.
Commun. 10(1):2518. https://doi.org/10.1038/s41467-019-10538-w.
73. Shih, M., Shishido, A., and Khoo, I. 2001. All-optical image processing by means of
a photosensitive nonlinear liquid-crystal film: Edge enhancement and image addition–
subtraction. Opt. Lett. 26(15):1140–2. https://doi.org/10.1364/OL.26.001140.
74. Zhang, J., Wang, H., Yoshikado, S., and Aruga, T. 1997. Incoherent-to-coherent conver-
sion by use of the photorefractive fanning effect. Opt. Lett. 22(21):1612–4. https://doi.
org/10.1364/OL.22.001612.
75. Woods, D., and Naughton, T. 2012. Photonic neural networks. Nat. Phys. 8:257–9.
https://doi.org/10.1038/nphys2283.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp08.indd 270 9/26/23 1:12 PM

 Optical Computing 271

76. Solodar, A., Kumar, T., Sarusi, G., and Abdulhalim, I. 2016. Infrared to visible
image up-conversion using optically addressed spatial light modulator utilizing liq-
uid crystal and InGaAs photodiodes. Appl. Phys. Lett. 108(2):021103. https://doi.
org/10.1063/1.4939903.
77. Gelbaor, K., Matvey, K., Victor, L., Neil, C., and Abdulhalim, I. 2014. Liquid crys-
tal high-resolution optically addressed spatial light modulator using a nanodimen-
sional chalcogenide photosensor. Opt. Lett. 39(7):2048–51. https://doi.org/10.1364/
OL.39.002048.
78. Li, S.-Q., Xu, X., Veetil, R., Valuckas, V., Paniagua-Domnguez, R., and Kuznetsov,
A. 2019. Phase-only transmissive spatial light modulator based on tunable dielectric
metasurface. Science 364(6445):1087–90. https://doi.org/10.1126/science.aaw6747.
79. Li, J., Yu, P., Zhang, S., and Liu, N. 2020; Electrically-controlled digital metasurface
device for light projection displays. Nat. Commun. 11(1):3574. https://doi.org/10.1038/
s41467-020-17390-3.
80. Park, J., and Jeong, B. G. et al. 2020. All-solid-state spatial light modulator with inde-
pendent phase and amplitude control for three-dimensional lidar applications. Nat.
Nanotechnol. 16:69–76. https://doi.org/10.1038/s41565-020-00787-y.
81. Gong, S., Ren, M., Wu, W., Cai, W., and Xu, J. 2021. Optically addressed spatial light
modulator based on nonlinear metasurface. Photonics Res. 9(4):610–4. https://doi.
org/10.1364/PRJ.416189.
82. Heck, M. 2018. Optical computers light up the horizon. https://phys.org/news/2018-
03-optical-horizon.html.
83. Waldrop, M. 2016. The chips are down for Moore’s law. Nat. News. 530:144. https://doi.
org/10.1038/530144a.
84. Atabaki, A. H. et al. 2018. Integrating photonics with silicon nanoelectronics for the
next generation of systems on a chip. Nature. 556:349–54. https://doi.org/10.1038/
s41586-018-0028-z.
85. Paniccia, M. 2011. A perfect marriage: Optics and silicon. Integrated silicon-based
photonics now running at 50 Ggps, with Terabit speeds on the horizon. Optik Photon.
6(2):34–8. https://doi.org/10.1002/opph.201190327.
86. Liu, K., Ye, C., Khan, S., and Sorger, V. 2015. Review and perspective on ultrafast
wavelength-size electro-optic modulators. Laser Photon. Rev. 9(2):172–94. https://doi.
org/10.1002/lpor.201400219.
87. Chung, S., Nakai, M., and Hashemi, H. 2019. Low-power thermo-optic silicon modula-
tor for large-scale photonic integrated systems. Opt. Express. 27(9):13430–59. https://
doi.org/10.1364/OE.27.013430.
88. Lee, B., Biberman, A., Chan, J., and Bergman, K. 2009. High-performance modula-
tors and switches for silicon photonic networks-on-chip. IEEE J. Sel. Top. Quantum
Electron. 16(1):6–22. https://doi.org/10.1109/JSTQE.2009.2028437.
89. Ying, Z. et al. 2019. Integrated multi-operand electro-optic logic gates for optical com-
puting. Appl. Phys. Lett. 115(17):171104. https://doi.org/10.1063/1.5126517.
90. Xiao, X. et al. 2013. High-speed, low-loss silicon Mach-Zehnder modulators with dop-
ing optimization. Opt. Express. 21(4):4116–25. https://doi.org/10.1364/OE.21.004116.
91. Shu, H., Jin, M., Tao, Y., and Wang, X. 2019. Graphene-based silicon modulators.
Front. Inf. Technol. Electron. Eng. 20:458–71. https://doi.org/10.1631/FITEE.1800407.
92. Zhang, Q. et al. 2018. Broadband non-volatile photonic switching based on optical
phase change materials: Beyond the classical figure-of-merit. Opt. Lett. 43(1):94–7.
https://doi.org/10.1364/OL.43.000094.
93. Yang, Z., and Ramanathan, S. 2015. Breakthroughs in photonics 2014: Phase change
materials for photonics. IEEE Photon. J. 7(3):0700305. https://doi.org/10.1109/JPHOT.
2015.2413594.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp08.indd 271 9/26/23 1:12 PM

 272 Photonics Elements for Sensing and Optical Conversions

94. Badri, S., and Farkoush, S. 2021. Subwavelength grating waveguide filter based on
cladding modulation with a phase-change material grating. Appl. Opt. 60(10):2803–10.
https://doi.org/10.1364/AO.419587.
95. Wang, J., Wang, L., and Liu, J. 2020. Overview of phase-change materials based
photonic devices. IEEE Access. 8:121211–45. https://doi.org/10.1109/ACCESS.2020.
3006899.
96. Badri, S., Gilarlue, M., Farkoush, S., and Rhee, S. 2021. Reconfigurable bandpass opti-
cal filters based on subwavelength grating waveguides with a Ge2Sb2Te5 cavity. J. Opt.
Soc. Am. B. 38(4):1283–9. https://doi.org/10.1364/JOSAB.419475.
97. Ghosh, R., and Dhawan, A. 2021. Integrated non-volatile plasmonic switches based
on phase-change-materials and their application to plasmonic logic circuits. Sci. Rep.
11:18811. https://doi.org/10.1038/s41598-021-98418-6.
98. Prieto, A. et al. 2016. Neural networks: An overview of early research, current frame-
works and new challenges. Neurocomputing. 214:242–68. https://doi.org/10.1016/j.
neucom.2016.06.014.
99. LeCun, Y., Bengio, Y., and Hinton, G. 2015. Deep learning. Nature. 521(7553):436–44.
https://doi.org/10.1038/nature14539.
100. Krizhevsky, A., Sutskever, I., and Hinton, G. 2017. Imagenet classification with
deep convolutional neural networks. Commun. ACM. 60(6):84–90. https://doi.
org/10.1145/3065386.
101. Nitta, T. 2004. Orthogonality of decision boundaries in complex-valued neural net-
works. Neural Comput. 16(1):73–97. https://doi.org/10.1162/08997660460734001.
102. Aizenberg, I. 2011. Complex-valued neural networks with multi-valued neurons.
Berlin, Heidelberg: Springer-Verlag. ISBN: 978-3-642-20352-7.
103. Peng, H.-T., Nahmias, M., De Lima, T., Tait, A., and Shastri, B. 2018. Neuromorphic
photonic integrated circuits. IEEE J. Sel. Top. Quantum Electron. 24(6):6101715.
https://doi.org/10.1109/JSTQE.2018.2840448.
104. Sze, V., Chen, Y.-H., Yang, T.-J., and Emer, J. 2017. Efficient processing of deep neural
networks: A tutorial and survey. Proc. IEEE. 105:2295–329. https://doi.org/10.1109/
JPROC.2017.2761740.
105. de Lima, T. F. et al. 2019. Machine learning with neuromorphic photonics. J. Lightw.
Technol. 37(5):1515–34. https://doi.org/10.1109/JLT.2019.2903474.
106. Lin, X. et al. 2018. All-optical machine learning using diffractive deep neural net-
works. Science. 361(6406):1004–8. https://doi.org/10.1126/science.aat8084.
107. Vandoorne, K. et al. 2014. Experimental demonstration of reservoir computing on a
silicon photonics chip. Nat. Commun. 5:3541. https://doi.org/10.1038/ncomms4541.
108. Harris, N. C. et al. 2018. Linear programmable nanophotonic processors. Optica.
5(12):1623–31. https://doi.org/10.1364/OPTICA.5.001623.
109. Williamson, I. A. D. et al. 2019. Reprogrammable electro-optic nonlinear activa-
tion functions for optical neural networks. IEEE J. Sel. Top. Quantum Electron.
26(1):7700412. https://doi.org/10.1109/JSTQE.2019.2930455.
110. Bueno, J. et al. 2018. Reinforcement learning in a large-scale photonic recurrent neural
network. Optica. 5(6):756–60. https://doi.org/10.1364/OPTICA.5.000756.
111. Ambrogio, S. et al. 2018. Equivalent-accuracy accelerated neural-network training
using analogue memory. Nature. 558:60–7. https://doi.org/10.1038/s41586-018-0180-5.
112. Li, C. et al. 2018. Efficient and self-adaptive in-situ learning in multilayer memristor
neural networks. Nat. Commun. 9:2385. https://doi.org/10.1038/s41467-018-04484-2.
113. Yao, P. et al. 2020. Fully hardware-implemented memristor convolutional neural net-
work. Nature 577:641–6. https://doi.org/10.1038/s41586-020-1942-4.
114. Rakkiyappan, R., Velmurugan, G., and Li, X. 2015. Complete stability analysis of
complex-valued neural networks with time delays and impulses. Neural Process. Lett.
41:435–68. https://doi.org/10.1007/s11063-014-9349-6

BK-TandF-KAZANSKIY_9781032572949-230642-Chp08.indd 272 9/26/23 1:12 PM

 Optical Computing 273

115. Wang, H., Duan, S., Huang, T., Wang, L., and Li, C. 2016. Exponential stability of com-
plex-valued memristive recurrent neural networks. IEEE Trans. Neural Netw. Learn.
Syst. 28(3):766–71. https://doi.org/10.1109/TNNLS.2015.2513001.
116. Velmurugan, G., Rakkiyappan, R., Vembarasan, V., Cao, J., and Alsaedi, A. 2017.
Dissipativity and stability analysis of fractional-order complex-valued neural networks
with time delay. Neural Netw. 86:42–53. https://doi.org/10.1016/j.neunet.2016.10.010.
117. Hirose, A. 1994. Applications of complex-valued neural networks to coherent optical
computing using phase-sensitive detection scheme. Inf. Sci. Appl. 2(2):103–17. https://
doi.org/10.1016/1069-0115(94)90014-0.
118. Zhang, H., Gu, M., and Jiang, X. D. et al. 2021. An optical neural chip for implement-
ing complex-valued neural network. Nat. Commun. 12:457. https://doi.org/10.1038/
s41467-020-20719-7.
119. Wei, H. et al. 2011. Cascaded logic gates in nanophotonic plasmon networks. Nat.
Commun. 2:387. https://doi.org/10.1038/ncomms1388.
120. Wei, H. et al. 2011. Quantum dot-based local field imaging reveals plasmon-based
interferometric logic in silver nanowire networks. Nano Lett. 11(2):471–5. https://doi.
org/10.1021/nl103228b.
121. Fu, Y. L. et al. 2012. All-optical logic gates based on nanoscale plasmonic slot wave-
guides. Nano Lett. 12(11):5784–90. https://doi.org/10.1021/nl303095s.
122. Sang, Y. G. et al. 2018. Broadband multifunctional plasmonic logic gates. Adv. Opt.
Mater. 6(13):1701368. https://doi.org/10.1002/adom.201701368.
123. Xu, Q., and Lipson, M. 2007. All-optical logic based on silicon micro-ring resonators.
Opt. Express. 15(3):924–9. https://doi.org/10.1364/OE.15.000924.
124. Raeker, B., and Grbic, A. 2019. Compound metaoptics for amplitude and phase. Phys.
Rev. Lett. 122(11):113901. https://doi.org/10.1103/PhysRevLett.122.113901.
125. Khonina, S. N., Kazanskiy, N. L., Karpeev, S. V., and Butt, M. A. 2020. Bessel beam:
Significance and applications–A progressive review. Micromachines. 11(11):997.
https://doi.org/10.3390/mi11110997.
126. Qian, C., Lin, X., Lin, X., Xu, J., Sun, Y., Li, E., Zhang, B., and Chen, H. 2020.
Performing optical logic operations by a diffractive neural network. Light Sci. Appl.
9:59. https://doi.org/10.1038/s41377-020-0303-2.
127. Onifade, A. 2004. History of the computer. IEEE Conf. on the History of Electronics: 21.
128. Cotter, D. et al. 1999. Nonlinear optics for high-speed digital information processing.
Science. 286(5444):1523–8. https://doi.org/10.1126/science.286.5444.1523.
129. Suzuki, Y., Shimada, J., and Yamashita, H. 1985. High-speed optical-optical logic gate
for optical computers. Electron. Lett. 21(4):161–2. https://doi.org/10.1049/el:19850114.
130. Meindl, J. 1995. Low power microelectronics: Retrospect and prospect. Proc. IEEE.
83(4):619–35. https://doi.org/10.1109/5.371970.
131. Yavuz, D. 2006. All-optical femtosecond switch using two-photon absorption. Phys.
Rev. A. 74(5):053804. https://doi.org/10.1103/PhysRevA.74.053804.
132. Kim, J. H. et al. 2002. All-optical XOR gate using semiconductor optical amplifiers
without additional input beam. IEEE Photonics Technol. Lett. 14(10):1436–8. https://
doi.org/10.1109/LPT.2002.801841.
133. Sharaiha, A., Topomondzo, J., and Morel, P. 2006. All-optical logic AND-NOR gate
with three inputs based on cross-gain modulation in a semiconductor optical amplifier.
Opt. Commun. 265(1):322–5. https://doi.org/10.1016/j.optcom.2006.03.036.
134. Nesset, D., Tatham, M., and Cotter, D. 1995. All-optical AND gate operating on 10
Gbit/s signal-sat the same wavelength using four-wave mixing in a semiconductor laser
amplifier. Electron. Lett. 31(11):896–7. https://doi.org/10.1049/el:19950600.
135. Webb, R. P. et al. 2003. 40 Gbit/s all-optical XOR gate based on hybrid-integrated
Mach-Zehnder interferometer. Electron. Lett. 39(1):79–81. https://doi.org/10.1049/
el:20030010.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp08.indd 273 9/26/23 1:12 PM

 274 Photonics Elements for Sensing and Optical Conversions

136. Wu, Y.-D., Shih, T.-T., and Chen, M.-H. 2008. New all-optical logic gates based on the
local non-linear Mach-Zehnder interferometer. Opt. Express. 16(1):248–57. https://doi.
org/10.1364/OE.16.000248.
137. McGeehan, J., Giltrelli, M., and Willner, A. 2007. All-optical digital 3-input AND gate
using sum- and difference-frequency generation in PPLN waveguide. Electron. Lett.
43(7):409–10. https://doi.org/10.1049/el:20073430.
138. Wang, J., Sun, J., and Sun, Q. 2007. Single-PPLN-based simultaneous half-adder, half-
subtracter, and OR logic gate: Proposal and simulation. Opt. Express. 15(4):1690–9.
https://doi.org/10.1364/OE.15.001690.
139. Butt, M. A., and Kazanskiy, N. L. 2021. Two-dimensional photonic crystal hetero-
structure for light steering and TM-polarization maintaining applications. Laser Phys.
31(3):036201. https://doi.org/10.1088/1555-6611/abd8ca.
140. Butt, M. A., Khonina, S. N., and Kazanskiy, N. L. 2021. 2D-photonic crystal het-
erostructures for the realization of compact photonic devices. Photonics Nanostruct.
44:100903. https://doi.org/10.1016/j.photonics.2021.100903.
141. Kazanskiy, N. L., Butt, M. A., and Khonina, S. N. 2021. 2D-heterostructure photonic
crystal formation for on-chip polarization division multiplexing. Photonics. 8(8):313.
https://doi.org/10.3390/photonics8080313.
142. Kazanskiy, N. L., and Butt, M. A. 2020. One-dimensional photonic crystal waveguide
based on the SOI platform for transverse magnetic polarization-maintaining devices.
Photonics Lett. Poland. 12(3):85–7. https://doi.org/10.4302/plp.v12i3.1044.
143. John, S. 1987. Strong localization of photons in certain disordered dielectric superlat-
tices. Phys. Rev. Lett. 58(23):2486. https://doi.org/10.1103/PhysRevLett.58.2486.
144. Zhang, Y., Zhang, Y., and Li, B. 2007. Optical switches and logic gates based on self-
collimated beams in two-dimensional photonic crystals. Opt. Express. 15(15):9287–
92. https://doi.org/10.1364/OE.15.009287.
145. Fan, R. et al. 2016. 2D photonic crystal logic gates based on self-collimated effect. J.
Phys. D. 49(32):325104. https://doi.org/10.1088/0022-3727/49/32/325104.
146. Christina, X. S., and Kabilan, A. P. 2012. Design of optical logic gates using self-colli-
mated beams in 2D photonic crystal. Photonic Sens. 2(2):173–9. https://doi.org/10.1007/
s13320-012-0054-7.
147. Hou, J. et al. 2009. Polarization insensitive self-collimation waveguide in square lat-
tice annular photonic crystals. Opt. Commun. 282(15):3172–6. https://doi.org/10.1016/j.
optcom.2009.04.051.
148. Kabilan, A. P., Christina, X. S., and Caroline, P. E. 2010. Photonic crystal based all
optical OR and XO logic gates. 2010 Second Int. Conf. on Computing, Communication
and Networking Technologies: 1–4. https://doi.org/10.1109/ICCCNT.2010.5591766.
149. Xavier, S. C. et al. 2016. Compact photonic crystal integrated circuit for all-optical logic
operation. IET Optoelectron. 10(4):142–7. https://doi.org/10.1049/iet-opt.2015.0072.
150. Ishizaka, Y. et al. 2011. Design of ultra compact all-optical XOR and AND logic gates
with low power consumption. Opt. Commun. 284(14):3528–33. https://doi.org/10.1016/j.
optcom.2011.03.069.
151. Lin, Y. et al. 2014. Design and optimization of all-optical AND and NOR logic gates
in a two-dimensional photonic crystal for binary-phase-shift-keyed signals. 7th Int.
Conf. on Biomedical Engineering and Informatics: 965–969. https://doi.org/10.1109/
BMEI.2014.7002912.
152. Liu, W. et al. 2013. Design of ultra compact all-optical XOR, XNOR, NAND and OR
gates using photonic crystal multi-mode interference waveguides. Opt. Laser Technol.
50:55–64. https://doi.org/10.1016/j.optlastec.2012.12.030.
153. Shaik, E., and Rangaswamy, N. 2017. Multi-mode interference-based photonic crystal
logic gates with simple structure and improved contrast ratio. Photonic Netw. Commun.
34(1):140–8. https://doi.org/10.1007/s11107-016-0683-7.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp08.indd 274 9/26/23 1:12 PM

 Optical Computing 275

154. Serajmohammadi, S., and Absalan, H. 2016. All optical NAND gate based on non-
linear photonic crystal ring resonator. Inf. Process. Agric. 3(2):119–23. https://doi.
org/10.1016/j.inpa.2016.04.002.
155. Rani, P. et al. 2017. Realization of all optical logic gates using universal NAND gates
on photonic crystal platform. Superlattices Microstruct. 109:619–25. https://doi.
org/10.1016/j.spmi.2017.05.046.
156. Younis, R. M., Areed, N. F. F., and Obayya, S. S. A. 2014. Fully integrated AND
and OR optical logic gates. IEEE Photonics Technol. Lett. 26(19):1900–3. https://doi.
org/10.1109/LPT.2014.2340435.
157. Rao, D. G. S., Swarnakar, S., and Kumar, S. 2020. Performance analysis of all-optical
NAND, NOR, and XNOR logic gates using photonic crystal waveguide for optical com-
puting applications. Opt. Eng. 59(5):057101. https://doi.org/10.1117/1.OE.59.5.057101.
158. Parandin, F., and Karhanehchi, M. 2017. Terahertz all-optical NOR and AND logic
gates based on 2D photonic crystals. Superlattices Microstruct. 101:253–60. https://doi.
org/10.1016/j.spmi.2016.11.038.
159. Pirzadi, M., Mir, A., and Bodaghi, D. 2016. Realization of ultra-accurate and compact
all-optical photonic crystal or logic gate. IEEE Photonics Technol. Lett. 28(21):2387–
90. https://doi.org/10.1109/LPT.2016.2596580.
160. Sun, C., Wade, M., and Lee, Y. et al. 2015. Single-chip microprocessor that communi-
cates directly using light. Nature. 528:534–8. https://doi.org/10.1038/nature16454.
161. Silva, A., Monticone, F., Castaldi, G., Galdi, V., Alu, A., and Engheta, N. 2014.
Performing mathematical operations with metamaterials. Science. 343(6167):161–3.
https://doi.org/10.1126/science.1242818.
162. Solli, D., and Jalali, B. 2015. Analog optical computing. Nat. Photonics. 9:704–6.
https://doi.org/10.1038/nphoton.2015.208.
163. Pors, A., Nielsen, M., and Bozhevolnyi, S. 2015. Analog computing using reflective
plasmonic metasurfaces. Nano Lett. 15(1):791–7. https://doi.org/10.1021/nl5047297.
164. Ferrera, M., Park, Y., and Razzari, L. et al. 2010. On-chip CMOS-compatible all-
optical integrator. Nat. Commun. 1(3):29. https://doi.org/10.1038/ncomms1028.
165. Pasquazi, A., and Peccianti, M. et al. 2011. Sub-picosecond phase-sensitive optical
pulse characterization on a chip. Nat. Photonics. 5(10):618–23. https://doi.org/10.1038/
nphoton.2011.199.
166. Fernández-Ruiz, M., Carballar, A., and Azaña, J. 2013. Design of ultrafast all-optical
signal processing devices based on fiber Bragg gratings in transmission. J. Lightw.
Technol. 31(10):1593–600. https://doi.org/10.1109/JLT.2013.2254467.
167. Asghari, M., and Azana, J. 2011. Photonic integrator-based optical memory unit. IEEE
Photon. Technol. Lett. 23(4):209–11. https://doi.org/10.1109/LPT.2010.2096806.
168. Ashrafi, R., and Azaña, J. 2012. Figure of merit for photonic differentiators. Opt.
Express. 20(3):2626–39. https://doi.org/10.1364/OE.20.002626.
169. Rutkowska, K. A., Duchesne, D., Strain, M. J., Morandotti, R., Sorel, M., and Azana,
J. 2011. Ultrafast all-optical temporal differentiators based on CMOS-compatible
integrated-waveguide Bragg gratings. Opt. Express 19(20):19514–22. https://doi.
org/10.1364/OE.19.019514.
170. Li, M., Janner, D., Yao, J., and Pruneri, V. 2009. Arbitrary-order all-fiber temporal
differentiator based on a fiber Bragg grating: Design and experimental demonstration.
Opt. Express 17(22):19798–807. https://doi.org/10.1364/OE.17.019798.
171. Tan, S., Wu, Z., Lei, L., Hu, S., Dong, J., and Zhang, X. 2013. All-optical computation
system for solving differential equations based on optical intensity differentiator. Opt.
Express. 21(6):7008–13. https://doi.org/10.1364/OE.21.007008.
172. Tan, S., Xiang, L., Zou, J., Zhang, Q., Wu, Z., Yu, Y., Dong, J., and Zhang, X. 2013.
High-order all-optical differential equation solver based on microring resonators. Opt.
Lett. 38(19):3735–8. https://doi.org/10.1364/OL.38.003735.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp08.indd 275 9/26/23 1:12 PM

 276 Photonics Elements for Sensing and Optical Conversions

173. Lu, L., Wu, J., Wang, T., and Su, Y. 2012. Compact all-optical differential-equation
solver based on silicon microring resonator. Front. Optoelectron. 5:99–106. https://doi.
org/10.1007/s12200-012-0186-9.
174. Yang, T., Dong, J., Lu, L., Zhou, L., Zheng, A., Zhang, X., and Chen, J. 2014. All-
optical differential equation solver with constant-coefficient tunable based on a single
microring resonator. Sci. Rep. 4:5581. https://doi.org/10.1038/srep05581.
175. Bykov, D. A., Doskolovich, L. L., and Soifer, V. A. 2012. Integration of optical
pulses by resonant diffraction gratings. JETP Lett. 95(1):6–9. https://doi.org/10.1134/
S0021364012010031.
176. Soifer, V. A., ed. 2017. Diffractive optics and nanophotonics. Boca Raton, FL: CRC
Press. ISBN: 978-1-4978-5447-7.
177. Bykov, D. A., Doskolovich, L. L., Golovastikov, N. V., and Soifer, V. A. 2013. Time-
domain differentiation of optical pulses in reflection and in transmission using the same
resonant grating. J. Opt. 15(10):105703. https://doi.org/10.1088/2040-8978/15/10/105703.
178. Bykov, D. A., Doskolovich, L. L., and Soifer, V. A. 2011. Temporal differentia-
tion of optical signals using resonant gratings. Opt. Lett. 36(11):3509–11. https://doi.
org/10.1364/OL.36.003509.
179. Bykov, D. A., Doskolovich, L. L., and Soifer, V. A. 2012. On the ability of resonant
diffraction gratings to differentiate a pulsed optical signal. J. Exp. Theor. Phys.
114(5):724–30. https://doi.org/10.1134/S1063776112030028.
180. Bugaev, A. 2020. Resonant nanophotonic structures for analog optical computing.
2020 Int. Conf. on Information Technology and Nanotechnology (ITNT): 1–5. https://
doi.org/10.1109/ITNT49337.2020.9253358.
181. Pors, A., and Bozhevolnyi, S. I. 2016. Plasmonic metasurfaces for efficient phase
control in reflection. Opt. Express. 21(22):27438–51. https://doi.org/10.1364/OE.21.
027438
182. Chizari, A., Abdollahramezani, S., Jamali, M. V., and Salehi, J. A. 2016. Analog optical
computing based on a dielectric meta-reflect array. Opt. Lett. 41(15):3451–4. https://doi.
org/10.1364/OL.41.003451.
183. Abdollahramezani, S., Chizari, A., Dorche, A. E., Jamali, M. V., and Salehi, J. A. 2017.
Dielectric metasurfaces solve differential and integro-differential equations. Opt. Lett.
42:1197–200. https://doi.org/10.1364/OL.42.001197.
184. Babashah, H., Kavehvash, Z., Koohi, S., and Khavasi, A. 2017. Integration in analog
optical computing using metasurfaces revisited: Toward ideal optical integration. J.
Opt. Soc. Am. B. 34(6):1270–9. https://doi.org/10.1364/JOSAB.34.001270.
185. Doskolovich, L. L., Bykov, D. A., Bezus, E. A., and Soifer, V. A. 2014. Spatial differ-
entiation of optical beams using phase-shifted Bragg grating. Opt. Lett. 39(5):1278–81.
https://doi.org/10.1364/OL.39.001278.
186. Golovastikov, N., Bykov, D., Doskolovich, L., and Bezus, E. 2015. Spatial optical inte-
grator based on phase-shifted Bragg gratings. Opt. Commun. 338:457–60. https://doi.
org/10.1016/j.optcom.2014.11.007.
187. Bykov, D. A., Doskolovich, L. L., Bezus, E. A., and Soifer, V. A. 2014. Optical com-
putation of the Laplace operator using phase-shifted Bragg grating. Opt. Express.
22(21):25084–92. https://doi.org/10.1364/OE.22.025084.
188. Golovastikov, N. V., Bykov, D. A., and Doskolovich, L. L. 2014. Resonant diffraction
gratings for spatial differentiation of optical beams. Quantum Electron. 44(10):984–8.
https://doi.org/10.1070/QE2014v044n10ABEH015477.
189. Bykov, D. A., Doskolovich, L. L., Morozov, A. A., Podlipnov, V. V., Bezus, E. A.,
Verma, P., and Soifer, V. A. 2018. First-order optical spatial differentiator based on a
guided-mode resonant grating. Opt. Express. 26(8):10997–1006. https://doi.org/10.1364/
OE.26.010997.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp08.indd 276 9/26/23 1:12 PM

 Optical Computing 277

190. Kazanskiy, N. L., Serafimovich, P. G., and Khonina, S. N. 2013. Use of photonic crystal
cavities for temporal differentiation of optical signals. Opt. Lett. 38(7):1149–51. https://
doi.org/10.1364/OL.38.001149.
191. Kazanskiy, N. L., Serafimovich, P. G., and Khonina, S. N. 2012. Use of photonic crystal
resonators for differentiation of optical impulses in time. Comput. Opt. 36(4):474–8.
192. Kazanskiy, N. L., and Serafimovich, P. G. 2014. Coupled-resonator optical waveguides
for temporal integration of optical signals. Opt. Express. 22(11):14004–13. https://doi.
org/10.1364/OE.22.014004.
193. Serafimovich, P. G., and Kazanskiy, N. L. 2015. Active photonic crystal cavities for
optical signal integration. Opt. Mem. Neural Networks. 24(4):260–71. https://doi.
org/10.3103/S1060992X15040050.
194. Serafimovich, P. G., and Kazanskiy, N. L. 2016. Optical modulator based on coupled
photonic crystal cavities. J. Mod. Opt. 63(13):1233–8. https://doi.org/10.1080/0950034
0.2015.1135258.
195. Serafimovich, P. G., Stepikhova, M. V., Kazanskiy, N. L., Gusev, S. A., Egorov, A.
V., Skorokhodov, E. V., and Krasilnik, Z. F. 2016. On a silicon-based photonic-crys-
tal cavity for the near-IR region: Numerical simulation and formation technology.
Semiconductors. 50(8):1112–6. https://doi.org/10.1134/S1063782616080212.
196. Golovastikov, N. V., Doskolovich, L. L., Bezus, E. A., Bykov, D. A., and Soifer, V.
A. 2018. An optical differentiator based on a three-layer structure with a W-shaped
refractive index profile. J. Exp. Theor. Phys. 127(2):202–9. https://doi.org/10.1134/
S1063776118080174.
197. Kazanskiy, N. L., and Skidanov, R. V. 2019. Technological line for creation and
research of diffractive optical elements. Proc. SPIE. 11146:111460W. https://doi.
org/10.1117/12.2527274.
198. Kazanskiy, N. L. 2017. Efficiency of deep integration between a research univer-
sity and an academic institute. Procedia Eng. 201:817–31. https://doi.org/10.1016/j.
proeng.2017.09.604.
199. Zhu, T., Zhou, Y., Lou, Y., Ye, H., Qiu, M., Ruan, Z., and Fan, S. 2017. Plasmonic
computing of spatial differentiation. Nat. Commun. 8:15391. https://doi.org/10.1038/
ncomms15391.
200. Wesemann, L., Panchenko, E., Singh, K., Gaspera, E. D., Gomez, D. E., Davis, T. J., and
Roberts, A. 2019. Selective near-perfect absorbing mirror as a spatial frequency filter
for optical image processing. APL Photon. 4:100801. https://doi.org/10.1063/1.5113650.
201. Li, X., Youngblood, N., Rios, C., Cheng, Z., Wright, C. D., Pernice, W. H. P., and
Bhaskaran, H. 2019. Fast and reliable storage using a 5 bit, nonvolatile photonic mem-
ory cell. Optica. 6(1):1–6. https://doi.org/10.1364/OPTICA.6.000001.
202. Stucke, G. 1989. Parallel architecture for a digital optical computer. Appl. Opt.
28(2):363–70. https://doi.org/10.1364/AO.28.000363.
203. Jenkins, B. K., and Giles, C. L. 1986. Parallel processing paradigms and optical com-
puting. Proc. SPIE. 625:22–9. https://doi.org/10.1117/12.963476.
204. Feldmann, J. et al. 2021. Parallel convolutional processing using an integrated photonic
tensor core. Nature. 589:52–8. https://doi.org/10.1038/s41586-020-03070-1.
205. Stark, H. 1975. An optical-digital computer for parallel processing of images. IEEE
Trans. Comput. C-24(4):340–7. https://doi.org/10.1109/T-C.1975.224227.
206. White, I., Aw, E. T., Williams, K., Wang, H., Wonfor, A., and Penty, R. 2009. Scalable
optical switches for computing applications. J. Opt. Netw. 8(2):215–24. https://doi.
org/10.1364/JON.8.000215.
207. Rashed, A. N. Z., Mohammed, A. E. N., Zaky, W. F., Amiri, I. S., and Yupapin, P. 2019.
The switching of optoelectronics to full optical computing operations based on nonlin-
ear metamaterials. Results Phys. 13:102152. https://doi.org/10.1016/j.rinp.2019.02.088.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp08.indd 277 9/26/23 1:12 PM

 278 Photonics Elements for Sensing and Optical Conversions

208. Gu, M., Li, X., and Cao, Y. 2014. Optical storage arrays: A perspective for future big
data storage. Light Sci. Appl. 3:e177. https://doi.org/10.1038/lsa.2014.58.
209. Alexoudi, T., Kanellos, G. T., and Pleros, N. 2020. Optical RAM and integrated optical
memories: A survey. Light Sci. Appl. 9:91. https://doi.org/10.1038/s41377-020-0325-9.
210. Zhu, H. H., and Zou, J. et al. 2022. Space-efficient optical computing with an inte-
grated chip diffractive neural network. Nat. Commun. 13:1044. https://doi.org/10.1038/
s41467-022-28702-0.
211. Salmani, M., Eshaghi, A., Luan, E., and Saha, S. 2021. Photonic computing to accel-
erate data processing in wireless communications. Opt. Express. 29(14):22299–314.
https://doi.org/10.1364/OE.423747.
212. Qiu, C., Xiao, H., Wang, L., and Tian, Y. 2022. Recent advances in integrated optical
directed logic operations for high performance optical computing: A review. Front.
Optoelectron. 15:1. https://doi.org/10.1007/s12200-022-00001-y.
213. Lee, K.-Y. et al. 2011. Optical logic operation for AND Gate based on planar photonic
crystal circuit. Proc. SPIE. 8308:83081S. https://doi.org/10.1117/12.905637.

BK-TandF-KAZANSKIY_9781032572949-230642-Chp08.indd 278 9/26/23 1:12 PM

 General Conclusion
Nikolay Lvovich Kazanskiy1,2
1 Samara National Research University, Samara, Russia
2 Image Processing Systems Institute – Branch of the Federal
Scientific Research Centre “Crystallography and Photonics”
of Russian Academy of Sciences, Samara, Russia

Notwithstanding considerable advances made in designing nanophotonic struc-

tures, devices, and components, their practical uses have been limited so far. It is
the shared opinion of this book’s authors that before long, the situation will start
changing rapidly. In support of this view, one may mention numerous experiments
that have corroborated theoretical and numerical estimations and first successful
attempts of replicating devices comprising nanophotonic components. Because of
this, the contributors to the monograph have chosen not to discuss methods for solv-
ing Maxwell’s equations, which are described in a number of well-known books
[1–3], and other general approaches employed in diffractive nanophotonics [4–5],
focusing instead on a number of promising research directions relating to the devel-
opment of sensorics and optical transformations.
Silicon photonic waveguides discussed in Chapter 1 are attractive thanks to the
mature and affordable manufacturing technology. The authors have analyzed sev-
eral types of silicon-based optical fibers (OF), including conventional ridge, comb,
slit, hybrid, and suspended waveguides. Conventional OF designs, such as comb and
ridge, can offer better confinement of propagating modes, making them convenient
for using as optical interconnects in a crystal chip. Directly structuring the silicon
material enables optical interconnects to be implemented in laser signal transmission
systems that can carry enormous data volumes, while featuring low energy con-
sumption and not suffering from heating and signal deterioration. Silicon-based slit
and hybrid plasmonic OFs are extremely sensitive to the environment, which makes
them ideal candidates for sensor applications. Advantages of silicon OFs include a
low cost of the material, mature and well-established processing and manufacturing
techniques in the microelectronics industry, as well as the possibility of easy integra-
tion into electric devices and components in a single substrate.
In Chapter 2, we numerically and experimentally studied potentialities that comb
photonic-crystal waveguides have for solving a wide range of problems. Among such
problems, special mention should be made of temporal differentiation and integra-
tion (including the multiple one) of optical signals, development of an active cavity
enabling vertical electric or plane optical pumping, enhancing the sensitivity of an
optical sensor system, designing an optical modulator, and so forth. A technique for
manufacturing a photonic crystal resonator by ion etching of a hole array in a silicon
strip waveguide was also described.
In Chapter 3, we focused on the analysis of plasmonic refractive index sensors. Cutting-
edge developments in special-type plasmonic structures based on metal-insulator-metal

DOI: 10.1201/9781003439165-9 279

BK-TandF-KAZANSKIY_9781032572949-230642-Conclusion.indd 279 9/26/23 1:11 PM

 280 Photonics Elements for Sensing and Optical Conversions

waveguides and their application in refractive index sensors were discussed. Various
types of plasmonic waveguides, their geometry, materials, and manufacturing proce-
dures, as well as potential energy losses were analyzed. Design and numerical simu-
lation of two sensor configurations based on a metal-insulator-metal waveguide that
can be used both for refractive index measurements and as a plasmonic filter were dis-
cussed. An important part of the chapter was concerned with the discussion of sensorics
potentialities based on the combined use of surface plasmon resonances and nanopar-
ticles. In the concluding part of the chapter, we made an attempt to predict the future of
plasmonic sensors, coming to a conclusion that the future looks bright.
In Chapter 4, the authors conducted an analysis of plasmonic nanolasers, cover-
ing various modern aspects. In the recent decade, great advances were made both
in the theory and experimental demonstrations of nanolasers, attracting significant
interest from researchers and industry. The authors described surface plasmons,
typical amplifying materials, and an operation principle of plasmonic nanolasers.
They analyzed experimental demonstrations of the devices, classified according to
a confinement technique used for light enhancement, including a solitary particle, a
Fabry–Perot resonator, a whispering gallery mode resonator, and a metal particle/
hole array, before giving a concise description of major parameters of nanolasers.
Next, the authors discussed the use of plasmonic nanolasers in integrated photonic
circuits, sensors, and biology. In conclusion, the authors outlined directions of fur-
ther development and assessed prospects of plasmonic nanolasers.
In Chapter 5, we presented several important practical uses of metasurfaces (MS).
An MS is a two-dimensional or planar version of a metamaterial of subwavelength
thickness. Thanks to their small size and light weight, MSs have been extensively
studied and utilized in electromagnetic devices. MSs possess a remarkable ability
to block, absorb, concentrate, scatter, or direct waves ranging from a millimeter to
the visible spectrum of wavelengths at grazing, normal, and tilted incidence. We set
forth methods for MS design, analyzed currently used types of metasurfaces (anten-
nae, metalenses, absorbers, sensors, camouflage MS, modulators, and polarizers),
their limitations, and their advantages.
In Chapter 6, the authors analyzed several types of fiber-optic sensors. Major ben-
efits of these sensors include the compatibility with the existing fiber-optic data trans-
mission networks, high resistance to the external chemical and radiation factors, the
immunity to electromagnetic interference, high sensitivity to the design parameters,
and easy integration into measurement systems. Depending on a particular type of
fiber-optic sensor, the output signal can be in the form of either an output amplitude
or a phase/frequency shift. For the amplitude modulation of light, multimode optical
fibers can be utilized, thus making it possible to enhance the sensitivity and adjust
measurement ranges by using diffractive optical elements incorporated in the mea-
surement system. Under illumination by coherent light, the amplitude sensors are
sensitive to noise and perform well only in short-length fiber-optic lines, especially
when filtering individual modes. Fiber-optic sensors that rely on a frequency shift
of transmitted radiation show essentially higher immunity to disturbances thanks to
operating on a principal wavelength. They also can easily be combined into mea-
surement circuits in which each individual sensor is identifiable, enabling measure-
ments at a considerable distance from the object of interest. In these sensors, fibers

BK-TandF-KAZANSKIY_9781032572949-230642-Conclusion.indd 280 9/26/23 1:11 PM

 General Conclusion 281

are microstructured to create fiber Bragg gratings. These days, microstructuring tech-
niques have been rapidly developing and were also dealt with in this chapter.
In Chapter 7, building on the expertise accumulated studying wavefront aberra-
tions in the human eye, the authors demonstrated prospects of developing diffractive
aberrometers and wavefront sensors for professional ophthalmology. The approaches
and devices proposed enable the optical system of the human eye to be better char-
acterized, leading to a more accurate computer-aided diagnosis and thus making
possible a transition to the personified high-tech medicine.
Chapter 8 is based on our recent review [6]. I felt it relevant to include the review
material in this monograph for a possibility of real-time processing of data coming
from optical sensors, while omitting a stage of optical-digital transformation followed
by computer-aided numerical analysis, by directly using a specialized optical chip
seems enticing. To date, an optical computer has not yet been developed and mass-
produced, with good solutions still waiting to be offered for optical transistors, optical
memory, and many more, which would allow overcoming the enormous inertia of
the many well-established manufacture procedures in electronics. Nonetheless, the
urgent demand for optical technology relies on the fact that the time response of elec-
trical circuits limits today’s computers [5, 7]. A solid transmission medium restricts
signal speed and volume while also generating heat that destroys components. The
extreme downsizing of microscopic electronic apparatuses also causes “cross talk”
or signal mistakes that compromise a system’s dependability. These and other chal-
lenges have prompted researchers to look for photonic solutions. The advantages of
the photonic solutions over the electronic ones set forth in this chapter give research-
ers an impetus to pursue their efforts in this direction, systematizing the accumulated
bulk of knowledge. In this chapter, we reviewed several optical components, such as
spatial light modulators, plasmonic switches, neural networks, diffractive neural net-
works, photonic crystal all-optical logic gates, and resonant nanophotonic structures
that are used to implement optical computing, and discussed their advantages. From
our viewpoint, at the time being, efficient solutions can be found through developing
separate types of specialized optical processors capable of replacing currently used
electronic-analogous computers for solving some types of differential equations, con-
tour detection and extraction, and others. The proposed solutions may show promise
in situations where an optical signal (e.g. a video stream) is directly fed to the input of
a specialized processor. In this case, an optical neural network enables some images
to be processed and recognized in a faster and more economic way than by previously
digitizing and digitally processing them using graphics processing units.
It gave me great pleasure to collaborate on the book project with a wonderful
team of contributing authors. I hope that reading this book will also be a pleasant
process for the readers.

REFERENCES
1. Soifer, V. A., ed. 2014. Diffractive nanophotonics. Boca Raton, FL: CRC Press. ISBN:
978-1-4665-9070-0.
2. Soifer, V. A., ed. 2017. Diffractive optics and nanophotonics. Boca Raton, FL: CRC
Press. ISBN: 978-1-4987-5447-7.

BK-TandF-KAZANSKIY_9781032572949-230642-Conclusion.indd 281 9/26/23 1:11 PM

 282 Photonics Elements for Sensing and Optical Conversions

3. Soifer, V. A., ed. 2013. Computer design of diffractive optics. Cambridge, UK:
Woodhead Publishing. ISBN: 978-1-84569-635-1.
4. Soifer, V. A., Kotlyar, V. V., and Doskolovich, L. L. 2009. Diffractive optical elements
in nanophotonics devices. Comput. Opt. 33(4):352–68.
5. Soifer, V. A. 2014. Diffractive nanophotonics and advanced information technologies.
Her. Russ. Acad. Sci. 84(1):9–20. https://doi.org/10.1134/S1019331614010067.
6. Kazanskiy, N. L., Butt, M. A., and Khonina, S. N. 2022. Optical computing: Status and
perspectives. Nanomaterials. 12(13):2171. https://doi.org/10.3390/nano12132171.
7. Brunner, D., Marandi, A., Bogaerts, W., and Ozcan, A. 2020. Photonics for computing
and computing for photonics. Nanophotonics. 9(13):4053–4. https://doi.org/10.1515/
nanoph-2020-0470.

BK-TandF-KAZANSKIY_9781032572949-230642-Conclusion.indd 282 9/26/23 1:11 PM

 Appendix A: Listing for
Calculating the Sensor
Matched with the Zernike
Basis Functions
// radial part of the Zernike basis functions
double Z (int n, int m, double r, double r0){
double Zer = 0;
m = abs(m);
for(int p=0;p<=(n-m)/2;p++){
Zer += (pow(-1,p) * fact(n-p) * pow((fact(p)*fact((n+m)/
2-p)*fact((n-m)/2-p)),-1) * pow(r/r0,n-2*p));
}
return Zer;
}

// the Zernike basis functions in trigonometric

representation
complex <double> Z_trigon_cortes (int n, int m, int N, int j,
int k, double a){
double hx = 2*a/N;
double xj=-a+j*hx;
double hy = 2*a/N;
double yk=-a+k*hy;
complex <double> coef (m*atan2(yk,xj),0);
if (m>=0){
complex <double> expp = cos(coef);
double r = sqrt(xj*xj+yk*yk);
if (r>a) return 0;
complex <double> Zer = Z(n,m,r,a)*expp;
return Zer;
}
if (m<0){
complex <double> expp = sin(coef);
double r = sqrt(xj*xj+yk*yk);
if (r>a) return 0;
complex <double> Zer = Z(n,m,r,a)*expp;
return Zer;
}
}

283

BK-TandF-KAZANSKIY_9781032572949-230642-AppA.indd 283 9/26/23 5:10 PM

 284 Appendix A: Listing for Calculating the Sensor Matched

// array of values of Zernike basis functions in Cartesian

representation
void Z_exp_cortes_Arr (complex <double> **arr, int n, int m,
int N, double a){
for(int j=0;j<=N*2;j++){
for(int k=0;k<=N*2;k++){
arr[j][k] = Z_exp_cortes(n,m,N,j,k,1);
}
}
}

// array of values of Zernike basis functions in trigonometric

representation
void Z_trigon_cortes_Arr (complex <double> **arr, int n, int
m, int N, double a){
for(int j=0;j<=N*2;j++){
for(int k=0;k<=N*2;k++){
arr[j][k] = Z_trigon_cortes(n,m,N,j,k,1);
}
}
}

// optical Zernike analyzer (multichannel)

void Zernik_Optic_Analizer_Multy_Channel(complex <double> **E,
complex <double> **Filtr, int N, int n, int m, double a,
Bitmap *img1, Bitmap *img2){

complex <double> **Z1 = new complex <double> *[2*N];

for (int i=0;i<=2*N;i++)
Z1[i] = new complex <double>[2*N];

complex <double> **Z16 = new complex <double> *[2*N];

for (int i=0;i<=2*N;i++)
Z16[i] = new complex <double>[2*N];

complex <double> Psi1;

complex <double> Psi2;

double hx = 2*a/N;
double hy = 2*a/N;

for(int j=0;j<=N;j++){
for(int k=0;k<=N;k++){
double xj=-a+j*hx;
double yk=-a+k*hy;

Psi1 = Z_exp_cortes(2,0,N,j,k,1);
Psi2 = Z_exp_cortes(2,2,N,j,k,1);

complex <double> coef (0,3*atan2(yk,xj));

}
}

BK-TandF-KAZANSKIY_9781032572949-230642-AppA.indd 284 9/26/23 5:10 PM

 Appendix A: Listing for Calculating the Sensor Matched 285

//Z_exp_cortes_Ampl_Phase(E,img1,img2,2,2,N,a);
Z_exp_cortes_Ampl_Phase(Z1,img1,img2,1,1,N,a);
Z_exp_cortes_Ampl_Phase(Z16,img1,img2,7,7,N,a);

for(int j=0;j<2*N;j++){
for(int k=0;k<2*N;k++){

double xj=-a+j*hx;

double yk=-a+k*hy;

complex <double> e0 (0,0);

complex <double> e1 (0,2*100*xj);

complex <double> e2 (0,2*70*yk);
complex <double> e3 (0,2*70*xj+2*70*yk);

complex <double> e4 (0,-2*70*xj);

complex <double> e5 (0,-2*70*yk);
complex <double> e6 (0,-2*70*xj+2*70*yk);

complex <double> e7 (0,2*70*xj-2*70*yk);

complex <double> e8 (0,-2*70*xj-2*70*yk);

complex <double> e9 (0,4*70*xj);

complex <double> e10 (0,4*70*yk);
complex <double> e11 (0,4*70*xj+4*70*yk);

complex <double> e12 (0,-4*70*xj);

complex <double> e13 (0,-4*70*yk);
complex <double> e14 (0,-4*70*xj+4*70*yk);

complex <double> e15 (0,4*70*xj-4*70*yk);

complex <double> e16 (0,-4*70*xj-4*70*yk);

Filtr[j][k]=E[j][k]*(conj(Z1[j][k])*exp(e1)+
conj(Z2[j][k])*exp(e2)+conj(Z3[j][k])*exp(e3)+conj(Z4[j][k])*
exp(e4)+conj(Z5[j][k])*exp(e5)+conj(Z6[j][k])*exp(e6)+
conj(Z7[j][k])*exp(e7)+conj(Z8[j][k])*exp(e8)+conj(Z9[j][k])*
exp(e9)+conj(Z10[j][k])*exp(e10)+conj(Z11[j][k])*exp(e11)+
conj(Z12[j][k])*exp(e12)+conj(Z13[j][k])*exp(e13)+conj(Z14[j]
[k])*exp(e14)+conj(Z15[j][k])*exp(e15)+conj(Z16[j][k])*
exp(e16));
}
}

double max = 0;
for(int i=0;i<N;i++){
for(int j=0;j<N;j++){
if (max < abs(Filtr[i][j])) max = abs(Filtr[i][j]);
}
}

BK-TandF-KAZANSKIY_9781032572949-230642-AppA.indd 285 9/26/23 5:10 PM

 286 Appendix A: Listing for Calculating the Sensor Matched

int a;
for(int i=0;i<=N;i++){
for(int j=0;j<=N;j++){
a = (int) (abs(Filtr[i][j])*255/max);
img1->SetPixel(i, j, Color(a,a,a));
}
printf("%d \n", a);
}

fft_draw(Filtr,N,N,img1,1,2);

// Zernike math analyzer

complex <double> *Zernik_Math_Analize___txt_to_excel(FILE *m,
FILE *m1, complex <double> **E, int N, double a){

complex <double> **Z1 = new complex <double> *[2*N];

for (int i=0;i<=2*N;i++)
Z1[i] = new complex <double>[2*N];
complex <double> **Z16 = new complex <double> *[2*N];
for (int i=0;i<=2*N;i++)
Z16[i] = new complex <double>[2*N];

complex <double> *_C = new complex <double> [2*N];

complex <double> *norm = new complex <double> [2*N];

Z_trigon_cortes_Arr(Z1,0,0,N,a);
Z_trigon_cortes_Arr(Z15,4,4,N,a);

for(int j=0;j<=N;j++){
for(int k=0;k<=N;k++){
norm[0] += abs(Z1[j][k]*Z1[j][k]);
norm[14] += abs(Z15[j][k]*Z15[j][k]);

}
}

for(int j=0;j<=N;j++){
for(int k=0;k<=N;k++){
_C[0] += (E[j][k]*conj(Z1[j][k]) / norm[0]);
_C[14] += (E[j][k]*conj(Z15[j][k]) / norm[14]);
}
}

printf("Math_Analizer_Zernike \n");

BK-TandF-KAZANSKIY_9781032572949-230642-AppA.indd 286 9/26/23 5:10 PM

 Appendix A: Listing for Calculating the Sensor Matched 287

fprintf(m,"C[%d][%d] = %f", 0, 0,abs(_C[0]),"w","w");

fprintf(m,"C[%d][%d] = %f", 4, 4,abs(_C[14]),"w","w");

fprintf(m, "\n", "w", "w");

fprintf(m1,"%f",abs(_C[0]), "w", "w");

fprintf(m1,"%f",abs(_C[14]), "w", "w");

fprintf(m1, "\n", "w", "w");

return _C;
}

BK-TandF-KAZANSKIY_9781032572949-230642-AppA.indd 287 9/26/23 5:10 PM

BK-TandF-KAZANSKIY_9781032572949-230642-AppA.indd 288 9/26/23 5:10 PM
 Appendix B: Listing for
Calculating a Distorted Image
// simulation of imaging in the focal plane of an optical
system with introduced aberrations
complex <double>** Wavefront_FRONT_BACK__PSF(complex <double>
C00, complex <double> C11,complex <double> C1_1,
complex <double> C22, complex <double> C20,
complex <double> C2_2,
complex <double> C33, complex <double> C31,
complex <double> C3_1, complex <double> C3_3,
complex <double> C44, complex <double> C42,
complex <double> C40, complex <double> C4_2, complex <double>
C4_4,
complex <double> _C00, complex <double> _
C11,complex <double> _C1_1,
complex <double> _C22, complex <double> _C20,
complex <double> _C2_2,
complex <double> _C33, complex <double> _C31,
complex <double> _C3_1, complex <double> _C3_3,
complex <double> _C44, complex <double> _C42,
complex <double> _C40, complex <double> _C4_2, complex
<double> _C4_4,
int N, Bitmap *ampl1, Bitmap *phase1, Bitmap
*_ampl1, Bitmap *_phase1, Bitmap *ampl2, Bitmap *phase2,
Bitmap *_ampl2, Bitmap *_phase2, Bitmap *fft0, Bitmap *fft1,
Bitmap *_fft0, Bitmap *_fft1, Bitmap *ampl3, Bitmap *phase3,
Bitmap *fft2, Bitmap *fft3, double a, float zeros ){

complex <double> **arr1 = new complex <double> *[2*N];

for (int i=0;i<=2*N;i++)
arr1[i] = new complex <double>[2*N];

complex <double> **_arr1 = new complex <double> *[2*N];

for (int i=0;i<=2*N;i++)
_arr1[i] = new complex <double>[2*N];

complex <double> **arr2 = new complex <double> *[2*N];

for (int i=0;i<=2*N;i++)
arr2[i] = new complex <double>[2*N];

complex <double> **_arr2 = new complex <double> *[2*N];

for (int i=0;i<=2*N;i++)
_arr2[i] = new complex <double>[2*N];

289

BK-TandF-KAZANSKIY_9781032572949-230642-AppB.indd 289 9/26/23 1:10 PM

 290 Appendix B: Listing for Calculating a Distorted Image

complex <double> **arr3 = new complex <double> *[2*N];

for (int i=0;i<=2*N;i++)
arr3[i] = new complex <double>[2*N];

C00 *= 1E3;
C4_4 *= 1E3;

_C00 *= 1E3;
_C4_4 *= 1E3;

//sum(Cij*Zij)
for(int i = 0; i<N; i++){
for(int j = 0; j<N; j++){
arr1[i][j] = C00*Z_trigon_cortes(0,0,N,i,j,1)+
C11*Z_trigon_cortes(1,1,N,i,j,1)+
C4_4*Z_trigon_cortes(4,-4,N,i,j,1);
arr1[i][j] *= 2.*M_PI;

}
}

for(int i = 0; i<N; i++){

for(int j = 0; j<N; j++){
_arr1[i][j] = _C00*Z_trigon_cortes(0,0,N,i,j,1)+
_C11*Z_trigon_cortes(1,1,N,i,j,1)+
_C4_4*Z_trigon_cortes(4,-4,N,i,j,1);

_arr1[i][j] *= 2.*M_PI;

}
}
//amplitude and phase sum(Cij*Zij)
Ampl_Phase(arr1,N,ampl1,phase1,0);
Ampl_Phase(_arr1,N,_ampl1,_phase1,0);

//reduce circle
//exp(i*sum(Cij*Zij))
complex <double> pokasatel = 0;
complex <double> _pokasatel = 0;
complex <double> I (0,1);
double a = 1;
double hx = 2*a/N;
double hy = 2*a/N;
for(int i = 0; i<N; i++){
for(int j = 0; j<N; j++){
double xi =-a+i*hx;
double yj =-a+j*hy;
double r = sqrt(xi*xi+yj*yj);
if (r>a) arr2[i][j] = 0;

BK-TandF-KAZANSKIY_9781032572949-230642-AppB.indd 290 9/26/23 1:10 PM

 Appendix B: Listing for Calculating a Distorted Image 291

else {
pokasatel = I*arr1[i][j];
_pokasatel = I*_arr1[i][j];
arr2[i][j] = exp(pokasatel);
_arr2[i][j] = exp(_pokasatel);

arr3[i][j] = exp(pokasatel + _pokasatel);

}
}
}

//amplitude and phase exp(i*sum(Cij*Zij))

Ampl_Phase(arr2,N,ampl2,phase2,0);
Ampl_Phase(_arr2,N,_ampl2,_phase2,0);
Ampl_Phase(arr3,N,ampl3,phase3,0);

fft(arr2,N,N,1,zeros);
fft(_arr2,N,N,1,zeros);
fft(arr3,N,N,1,zeros);
Ampl_Phase(arr2,N,fft0,fft1,0);
Ampl_Phase(_arr2,N,_fft0,_fft1,0);
Ampl_Phase(arr3,N,fft2,fft3,0);

return arr2;
}

// modeling an image distorted by aberrations

void Aber_Image (complex <double> **arr1, complex <double>
**arr, Bitmap *bm, Bitmap *bm1, Bitmap *bm2, Bitmap *bm3,
Bitmap *bm4, Bitmap *bm5, Bitmap *bm6, int n, int m, double
Cnm, int N, double a){

complex <double> I (0,1);

int jj=30;
int kk=30;

/*n=4;
m=-4;
Cnm=1;*/

int background = 0;
int figure = 255;

for(int j=0;j<=N*2;j++){
for(int k=0;k<=N*2;k++){
arr1[j][k] = background;
}
}

BK-TandF-KAZANSKIY_9781032572949-230642-AppB.indd 291 9/26/23 1:10 PM

 292 Appendix B: Listing for Calculating a Distorted Image

for(int j=-jj;j<=jj;j++){
for(int k=-kk;k<=kk;k++){
arr1[(int)N/2][(int)N/2+k] = figure;
arr1[(int)N/2+j][(int)N/2] = figure;

arr1[(int)N/2-1][(int)N/2+k] = figure;
arr1[(int)N/2+j][(int)N/2+1] = figure;

arr1[(int)N/2+1][(int)N/2+k] = figure;
arr1[(int)N/2+j][(int)N/2-1] = figure;
}
}

Ampl_Phase(arr1,N,bm,bm1,1);

fft_draw(arr1,N,N,bm1,-1,1);

for(int j=0;j<=N*2;j++){
for(int k=0;k<=N*2;k++){
arr[j][k] = Z_trigon_cortes(n,m,N,j,k,1);
}
}

Ampl_Phase(arr,N,bm2,bm3,1);

for(int j=0;j<=N*2;j++){
for(int k=0;k<=N*2;k++){
arr[j][k] = exp(Cnm*2.*M_PI*I*Z_trigon_cortes(n,m,N,
j,k,1));
}
}

Ampl_Phase(arr,N,bm4,bm5,1);

for(int j=0;j<=N*2;j++){
for(int k=0;k<=N*2;k++){
arr[j][k] *= arr1[j][k];
}
}

fft_draw(arr,N,N,bm6,-1,1);

BK-TandF-KAZANSKIY_9781032572949-230642-AppB.indd 292 9/26/23 1:10 PM