Tutorial Texts Series

• Analysis and Evaluation of Sampled Imaging Systems, Richard H. Vollmerhausen, Donald Reago, Ronald
Driggers, Vol. TT87
• Nanotechnology A Crash Course, Raúl J. Martin-Palma and Akhlesh Lakhtakia, Vol. TT86
• Direct Detection LADAR Systems, Richard Richmond, Stephen Cain, Vol. TT85
• Optical Design Applying the Fundamentals, Max J. Riedl, Vol. TT84
• Infrared Optics and Zoom Lenses, Second Edition, Allen Mann, Vol. TT83
• Optical Engineering Fundamentals, Second Edition, Bruce H. Walker, Vol. TT82
• Fundamentals of Polarimetric Remote Sensing, John Schott, Vol. TT81
• Fundamentals of Photonics, Chandra Roychoudhuri, Vol. TT79
• Radiation Thermometry Fundamentals and Applications in the Petrochemical Industry, Peter Saunders,
Vol. TT78
• Matrix Methods for Optical Layout, Gerhard Kloos, Vol. TT77
• Fundamentals of Infrared Detector Materials, Michael A. Kinch, Vol. TT76
• Practical Applications of Infrared Thermal Sensing and Imaging Equipment, Third Edition, Herbert
Kaplan, Vol. TT75
• Bioluminescence for Food and Environmental Microbiological Safety, Lubov Y. Brovko, Vol. TT74
• Introduction to Image Stabilization, Scott W. Teare, Sergio R. Restaino, Vol. TT73
• Logic-based Nonlinear Image Processing, Stephen Marshall, Vol. TT72
• The Physics and Engineering of Solid State Lasers, Yehoshua Kalisky, Vol. TT71
• Thermal Infrared Characterization of Ground Targets and Backgrounds, Second Edition, Pieter A. Jacobs,
Vol. TT70
• Introduction to Confocal Fluorescence Microscopy, Michiel Müller, Vol. TT69
• Artificial Neural Networks An Introduction, Kevin L. Priddy and Paul E. Keller, Vol. TT68
• Basics of Code Division Multiple Access (CDMA), Raghuveer Rao and Sohail Dianat, Vol. TT67
• Optical Imaging in Projection Microlithography, Alfred Kwok-Kit Wong, Vol. TT66
• Metrics for High-Quality Specular Surfaces, Lionel R. Baker, Vol. TT65
• Field Mathematics for Electromagnetics, Photonics, and Materials Science, Bernard Maxum, Vol. TT64
• High-Fidelity Medical Imaging Displays, Aldo Badano, Michael J. Flynn, and Jerzy Kanicki, Vol. TT63
• Diffractive Optics–Design, Fabrication, and Test, Donald C. O’Shea, Thomas J. Suleski, Alan D.
Kathman, and Dennis W. Prather, Vol. TT62
• Fourier-Transform Spectroscopy Instrumentation Engineering, Vidi Saptari, Vol. TT61
• The Power- and Energy-Handling Capability of Optical Materials, Components, and Systems, Roger M.
Wood, Vol. TT60
• Hands-on Morphological Image Processing, Edward R. Dougherty, Roberto A. Lotufo, Vol. TT59
• Integrated Optomechanical Analysis, Keith B. Doyle, Victor L. Genberg, Gregory J. Michels, Vol. TT58
• Thin-Film Design Modulated Thickness and Other Stopband Design Methods, Bruce Perilloux, Vol. TT57
• Optische Grundlagen für Infrarotsysteme, Max J. Riedl, Vol. TT56
• An Engineering Introduction to Biotechnology, J. Patrick Fitch, Vol. TT55
• Image Performance in CRT Displays, Kenneth Compton, Vol. TT54
• Introduction to Laser Diode-Pumped Solid State Lasers, Richard Scheps, Vol. TT53
• Modulation Transfer Function in Optical and Electro-Optical Systems, Glenn D. Boreman, Vol. TT52
• Uncooled Thermal Imaging Arrays, Systems, and Applications, Paul W. Kruse, Vol. TT51
• Fundamentals of Antennas, Christos G. Christodoulou and Parveen Wahid, Vol. TT50
• Basics of Spectroscopy, David W. Ball, Vol. TT49
• Optical Design Fundamentals for Infrared Systems, Second Edition, Max J. Riedl, Vol. TT48
• Resolution Enhancement Techniques in Optical Lithography, Alfred Kwok-Kit Wong, Vol. TT47
• Copper Interconnect Technology, Christoph Steinbrüchel and Barry L. Chin, Vol. TT46
• Optical Design for Visual Systems, Bruce H. Walker, Vol. TT45
 George H. Seward

Tutorial Texts in Optical Engineering

Volume TT88

Bellingham, Washington USA

Library of Congress Cataloging-in-Publication Data

Seward, George.
Optical design of microscopes / George Seward.
p. cm. -- (Tutorial texts in optical engineering ; v. TT88)
ISBN 978-0-8194-8095-8
1. Microscopes. 2. Optics. I. Title.

QH211.S478 2010
681'.413--dc22
2010000660

Published by

SPIE
P.O. Box 10
Bellingham, Washington 98227-0010 USA
Phone: +1 360 676 3290
Fax: +1 360 647 1445
Email: [email protected]
www.spie.org

Copyright © 2010 Society of Photo-Optical Instrumentation Engineers

All rights reserved. No part of this publication may be reproduced or distributed in any form or by
any means without written permission of the publisher.

The content of this book reflects the work and thought of the author(s). Every effort has been made
to publish reliable and accurate information herein, but the publisher is not responsible for the
validity of the information or for any outcomes resulting from reliance thereon.

Printed in the United States of America.

 Introduction to the Series
Since its inception in 1989, the Tutorial Texts (TT) series has grown to more than
80 titles covering many diverse fields of science and engineering. The initial idea
for the series was to make material presented in SPIE short courses available to
those who could not attend and to provide a reference text for those who could.
Thus, many of the texts in this series are generated by augmenting course notes
with descriptive text that further illuminates the subject. In this way, the TT
becomes an excellent stand-alone reference that finds a much wider audience
than only short course attendees.
Tutorial Texts have grown in popularity and in the scope of material covered
since 1989. They no longer necessarily stem from short courses; rather, they are
often generated by experts in the field. They are popular because they provide a
ready reference to those wishing to learn about emerging technologies or the
latest information within their field. The topics within the series have grown from
the initial areas of geometrical optics, optical detectors, and image processing to
include the emerging fields of nanotechnology, biomedical optics, fiber optics,
and laser technologies. Authors contributing to the TT series are instructed to
provide introductory material so that those new to the field may use the book as a
starting point to get a basic grasp of the material. It is hoped that some readers
may develop sufficient interest to take a short course by the author or pursue
further research in more advanced books to delve deeper into the subject.
The books in this series are distinguished from other technical monographs
and textbooks in the way in which the material is presented. In keeping with the
tutorial nature of the series, there is an emphasis on the use of graphical and
illustrative material to better elucidate basic and advanced concepts. There is also
heavy use of tabular reference data and numerous examples to further explain the
concepts presented. The publishing time for the books is kept to a minimum so
that the books will be as timely and up-to-date as possible. Furthermore, these
introductory books are competitively priced compared to more traditional books
on the same subject.
When a proposal for a text is received, each proposal is evaluated to
determine the relevance of the proposed topic. This initial reviewing process has
been very helpful to authors in identifying, early in the writing process, the need
for additional material or other changes in approach that would serve to
strengthen the text. Once a manuscript is completed, it is peer reviewed to ensure
that chapters communicate accurately the essential ingredients of the science and
technologies under discussion.
It is my goal to maintain the style and quality of books in the series and to
further expand the topic areas to include new emerging fields as they become of
interest to our reading audience.

James A. Harrington
Rutgers University
Contents
Preface .................................................................................................... xiii

Chapter 1 Optical Design Concepts / 1

1.1 A Value Proposition ............................................................................. 1
1.2 Specimen Model.................................................................................. 1
1.3 Detector Parameters ........................................................................... 1
1.4 Numerical Aperture ............................................................................. 1
1.5 Wave Propagation ............................................................................... 2
1.6 Geometric Aberrations ........................................................................ 3
1.7 Image Contrast .................................................................................... 4
1.8 Manufacturing...................................................................................... 5
1.9 Assembly ............................................................................................. 5

Chapter 2 Basic Microscope Concepts / 7

2.1 Magnification ....................................................................................... 7
2.2 Accommodation................................................................................... 7
2.3 Finite Tube Length .............................................................................. 8
2.4 Infinity-Corrected Objective ................................................................. 9
2.5 Tube Lens ......................................................................................... 10
2.6 Ocular Lens ....................................................................................... 11
2.7 Refractive Objects ............................................................................. 12
2.8 Diffractive Objects ............................................................................. 14
2.9 Dark Field .......................................................................................... 14

Chapter 3 Basic Geometric Optics / 17

3.1 Ray Tracing ....................................................................................... 17
3.2 Cardinal points .................................................................................. 17
3.3 Stops ................................................................................................. 18
3.4 Gaussian Lens Formula .................................................................... 19
3.5 Image Types...................................................................................... 21
3.6 Optical Power .................................................................................... 21
3.7 Paraxial Optics .................................................................................. 22
3.8 Relay Lens ........................................................................................ 23
3.9 Magnifier............................................................................................ 24

vii
viii Contents

Chapter 4 Aberrations / 27
4.1 Seidel Aberrations ............................................................................. 27
4.2 Chromatic Aberrations....................................................................... 29
4.3 Other Aberrations .............................................................................. 29
4.4 Aspheric Surfaces ............................................................................. 33

Chapter 5 Basic Physical Optics / 35

5.1 Importance of Physical Optics ........................................................... 35
5.2 Wave Equation .................................................................................. 35
5.3 Refractive Index ................................................................................ 36
5.4 Dispersion ......................................................................................... 40
5.5 Refraction and Reflection .................................................................. 41
5.6 Emission ............................................................................................ 43
5.7 Absorption ......................................................................................... 43
5.8 Evanescent Field ............................................................................... 44
5.9 Space-Angle Product ........................................................................ 45
5.10 Coherence ....................................................................................... 46
5.11 Airy Pattern...................................................................................... 46
5.12 Gaussian Beam Propagation .......................................................... 48
5.13 Transfer Functions .......................................................................... 49
5.14 Gaussian Estimate of Airy Pattern .................................................. 50
5.15 Scatter ............................................................................................. 51
5.16 Interference Filters .......................................................................... 51

Chapter 6 Fluorescence / 53
6.1 Absorption Parameters...................................................................... 53
6.2 Electron States .................................................................................. 54
6.3 Energy Diagrams............................................................................... 55
6.4 Fluorophores ..................................................................................... 57

Chapter 7 Optical Design Metrics / 63

7.1 CAD Tools ......................................................................................... 63
7.2 Wavefront Error ................................................................................. 63
7.3 Ray-Intercept Plot.............................................................................. 64
7.4 Spot Diagram .................................................................................... 66
7.5 Point-Spread Plot .............................................................................. 66
7.6 Encircled-Energy Plot ........................................................................ 68
7.7 Modulation Transfer Function............................................................ 68
7.8 Edge Spread ..................................................................................... 69
7.9 Lens Report ....................................................................................... 69
7.10 Relative Illumination ........................................................................ 70
7.11 Surface-Form Error ......................................................................... 70
7.12 Manufacturing Standards ................................................................ 72
Contents ix

Chapter 8 Image Contrast / 73

8.1 Radiometry ........................................................................................ 73
8.2 Expression of Contrast ...................................................................... 73
8.3 Shot Noise ......................................................................................... 75
8.4 Emittance Patterns ............................................................................ 76
8.5 Angular Collection Efficiency ............................................................. 77
8.6 Spatial Collection Efficiency .............................................................. 78
8.7 Full-Pixel Contrast ............................................................................. 79
8.8 Subpixel Contrast .............................................................................. 80
8.9 Point-Source Contrast ....................................................................... 81
8.10 Full-Pixel Airy Contrast .................................................................... 82

Chapter 9 Microlens Formats / 85

9.1 10XR Double Gauss.......................................................................... 85
9.2 10XR Microlens ................................................................................. 86
9.3 2XR Microlens ................................................................................... 87
9.4 1X Microlens...................................................................................... 88
9.5 2XR Telecentric Spectroscopy Lens ................................................. 89

Chapter 10 Illumination Systems / 93

10.1 Condenser ....................................................................................... 93
10.2 Abbe Illumination ............................................................................. 93
10.3 Nelson Illumination .......................................................................... 94
10.4 Diffusers .......................................................................................... 95
10.5 Köhler Illumination ........................................................................... 96
10.6 Matched Stops ................................................................................ 98
10.7 Light-Emitting Diodes ...................................................................... 98
10.8 Aspheric Plus Singlet Relay .......................................................... 100
10.9 Achromatic Aspheric Plus Doublet Relay ...................................... 101
10.10 Abbe Condenser ......................................................................... 101
10.11 Abbe Aspheric ............................................................................. 101
10.12 Total Internal Reflection Fluorescence Illumination ..................... 103

Chapter 11 Cover Strata / 105

11.1 Importance of Specimen Tolerance .............................................. 105
11.2 Perfect 10X for Air ......................................................................... 105
11.3 10X Objective with Cover Glass in Place of Air............................. 106
11.4 10X Objective with Microscope Slide in Place of Air ..................... 107
11.5 40X Objective with Silica Cover in Place of Glass ........................ 108
11.6 40X Objective with Tilted Cover Glass .......................................... 110
11.7 60X Objective with Silica Cover in Place of Glass ........................ 111
11.8 Strehl Ratio versus Optical Path Length ....................................... 111
x Contents

Chapter 12 Objective Lenses / 113

12.1 Formats ......................................................................................... 113
12.2 Aplanatic Surface .......................................................................... 113
12.3 10X Plan Achromat ....................................................................... 114
12.4 40X Fluor ....................................................................................... 115
12.5 60X Immersion TIRF ..................................................................... 117
12.6 100X Aplanat ................................................................................. 121
12.7 10X Schwarzschild ........................................................................ 122
12.8 20X Internal Parabola.................................................................... 123

Chapter 13 Tube Elements / 127

13.1 Doublet Tube Lens ........................................................................ 127
13.2 Doublet-Pair Tube Lens ................................................................ 127
13.3 Filter Types.................................................................................... 129
13.4 Filter within a Finite Conjugate Distance ....................................... 133
13.5 Warped Filter within an Infinity Correction..................................... 133

Chapter 14 Ocular Lenses / 139

14.1 Eyepiece........................................................................................ 139
14.2 Pupils............................................................................................. 139
14.3 Kellner Ocular................................................................................ 140
14.4 Plössl Ocular ................................................................................. 141
14.5 Erfle Ocular ................................................................................... 142

Chapter 15 Sensors / 147

15.1 CCD Sensors ................................................................................ 147
15.2 Active Pixel Sensors...................................................................... 151
15.3 Photomultiplier Tubes.................................................................... 152
15.4 Film................................................................................................ 154

Chapter 16 Human Vision / 157

16.1 Physiology ..................................................................................... 157
16.2 Contrast Sensitivity Function ......................................................... 158
16.3 Point Spread of a Lens .................................................................. 159
16.4 Lateral Inhibition of the Retina....................................................... 160
16.5 Temporal Feedback of Photoreceptors ......................................... 162
16.6 Saccation Point Spread ................................................................. 163
16.7 Vision Research ............................................................................ 165
16.8 Temporal Contrast Sensitivity Function......................................... 165
Contents xi

Chapter 17 Optical Materials / 169

17.1 Glass Types .................................................................................. 169
17.2 Glass Map ..................................................................................... 172
17.3 Fluorite .......................................................................................... 173
17.4 Short Flint ...................................................................................... 176
17.5 Anomalous Dispersion .................................................................. 176
17.6 Sellmeier Formula ......................................................................... 177
17.7 Environmentally Safe Glass .......................................................... 179
17.8 Glass Code.................................................................................... 180
17.9 Spectral Lines................................................................................ 180
17.10 Cost of Optics .............................................................................. 180
17.11 Structural Materials ..................................................................... 181

Chapter 18 Composition and Spectra of Materials / 183

18.1 Glass Structure.............................................................................. 183
18.2 Crown ............................................................................................ 183
18.3 Flint................................................................................................ 184
18.4 Long Crown ................................................................................... 186
18.5 Short Flint ...................................................................................... 186
18.6 Short-Flint Special ......................................................................... 187
18.7 Environmentally Safe Short Flint ................................................... 189
18.8 Dense Flint .................................................................................... 191

Chapter 19 Advanced Concepts / 195

19.1 Wave Equation .............................................................................. 195
19.2 Refractive Index ............................................................................ 196
19.3 Relative Partial Dispersion ............................................................ 198
19.4 Emission ........................................................................................ 199
19.5 Coherence ..................................................................................... 199
19.6 Gaussian Beam Power.................................................................. 201
19.7 Transfer Functions ........................................................................ 202
19.8 Scatter ........................................................................................... 204
19.9 Interference Filters ........................................................................ 205
19.10 Shot Noise ................................................................................... 206

Appendix: Prescriptions / 209

Works Consulted / 229
Recommended Reading / 231
References / 233
Index / 237
Preface
This book provides an introduction to optical design as it pertains to
microscopes. The large numerical aperture of a microscope creates issues that are
not present in systems such as telescopes and cameras. The importance of
microscope design is growing rapidly in 2010.
Microscopes are frequently employed in drug development, clinical tests, and
genomics; however, one system does not fit all applications. Considerable
expertise is required for the evaluation, design, and manufacture of these
instruments. An integrated relationship must be established between several
subsystems: the source, the illumination optics, the specimen, the objective lens,
the tube optics, and the sensor.
In this text, the optical prescriptions are specified with glass name, refractive
index, and Abbe number. Methods for color correction are described with
specific materials. The anomalous partial dispersion in the blue is described for a
long crown and a short flint. The origin of dispersion is related to glass
composition.
As this is a tutorial, several essential topics in optics are reviewed. Geometric
optics provides a simplistic description of ray tracing. Physical optics is
described in two chapters: the first chapter describes the basic concepts of wave
propagation with simple algebra, and the second chapter describes more
advanced concepts with vector calculus. Optical aberrations describe an optical
system’s departure from perfection. Expressions for image contrast are defined
for practical application to imaging systems. Expertise in these topics is essential
for proper design of an optical instrument.
Numerous chapters rely on simple plots and basic algebra, so even a novice
designer should easily comprehend the majority of chapters; however, a few
chapters require comprehension of calculus, vector operations, and Fourier
analysis. An expert in design should seek maximum comprehension of the
chapter on advanced concepts at the end of the book.
This book provides a foundation for the development of expertise in optical
design. Expertise is developed through education, practice, and exploration.

George H. Seward
March 2010

xiii
Chapter 1
Optical Design Concepts
1.1 A Value Proposition
A comprehensive optical design should yield both consistent manufacture and
maximum performance. Such features are critical to biological applications,
where the cost of consumables can be far greater than the cost of the optical
hardware. Maximum performance can reduce the magnitude of consumables on a
daily basis. Furthermore, a false negative can delay proper treatment of a critical
condition. Consistent manufacture builds credibility in a clinical test.
Sections 1.2 through 1.9 provide a review of essential optical design
concepts. Expertise within these concepts is developed through education and
practice. Perseverance is not a substitute for relevant expertise in optical design.

1.2 Specimen Model

The structure of a specimen determines the limits of optical performance. A
small deviation from the nominal thickness can introduce a significant wavefront
error that is not corrected by a lens. The light collected by the margin of the lens
is not focused within a diffraction-limited spot. The marginal light is spread over
many pixels, which lowers image contrast. Consequently, an optical design must
accommodate the tolerances of the specimen. In particular, the cover strata of
biological specimens present significant challenges in optical design.

1.3 Detector Parameters

Every detector has fundamental limitations, which may be expressed as
equivalent incident photons. A dark current is converted to photons per time per
area. Both full-well capacity and read noise are converted to photons per area.
The dependence on area is very important because image contrast is frequently
determined by image incidence (photons per second per area). Thus, an
equivalent incidence of the detector must be established.

1.4 Numerical Aperture

The margin of a lens is defined by the aperture of the lens. The angular extent of
the aperture varies with immersion medium. The numerical aperture (NA)
defines a constant throughout all media. The marginal NA is defined as

1
2 Chapter 1

ΝA M = n sin θ M , (1.1)

where n is the refractive index and θM is the angle of the marginal ray.i The
marginal ray is defined as a ray from the object point to the margin of the
aperture. The NA can decrease with field height, yielding less image brightness
at the field margin.
The marginal NA defines several important features of the lens system. The
spot diameter [Eq. (1.4)] depends on ΝA −M1 . The depth of focus depends on
ΝA −M2 [Eq. (1.8)]. The collection efficiency by solid angle roughly depends on
ΝA M2 [Eq. (8.11)]. The peak irradiance of a diffraction-limited spot depends on
ΝA M4 [Eq. (8.24)]. The NA is an essential metric for an optical instrument.
The paraxial NA is derived from the focal length f and entrance pupil
diameter φEnt:

φ Ent 1
ΝA P = = , (1.2)
2f 2F

where F is the f-number (f/#). The principal surface is a sphere with its center at
the focal point. The marginal NA can differ from the paraxial NA because
spherical aberration can bend marginal rays both too much or too little. The f/#
can inflate with field height, which yields less image brightness. The image-
space f/# may also inflate as the focal length grows with field position.

1.5 Wave Propagation

The space-angle product describes the convergence of a spherical wavefront as1

AΩ ≥ λ 2 , (1.3)

where A is the area of point spread, Ω is the solid angle of convergence, and λ is
the wavelength. The diffraction limit of the point spread is defined by the
minimum space-angle product.
A planar version of the space-angle product is

d ΝAG = 0.63λ , (1.4)

where d is the diameter of the Gaussian beam and NAG is the Gaussian NA. The
area of the Gaussian point spread doubles over the Rayleigh distance, which is

i
The term numerical aperture has the unique distinction of being a two-letter symbol that persists in
contrast to the more common practice of using a one-letter symbol in math and physics.
Optical Design Concepts 3

A0 λ
zR = = . (1.5)
λ Ω

A doubling of the point-spread area also indicates a halving of the point-spread

irradiance. Consequently, the Rayleigh distance indicates a depth of focus with a
defined range of contrast.
The space-angle product for a circular aperture is

d ΝA M = 1.22λ , (1.6)

where d is the diameter of the Airy disc.

A Gaussian estimate of an Airy point spread is

ΝAG = 0.71 ΝA M . (1.7)

The Gaussian depth of focus may be expressed by marginal NA as

λ
zR = . (1.8)
π ΝA 2M

This relationship provides an effective metric for depth of focus in relation to

wave propagation.

1.6 Geometric Aberrations

Geometric aberrations describe the departure of convergent rays from an image
point. As the geometric point spread grows beyond the diffraction limit, the
actual point spread becomes aberration limited. An aberration-limited spot
frequently contains a sharp central peak and a broad annular skirt. The central
peak may be defined by the central NA (NAC), which is smaller than the
marginal NA. Thus, there are four NAs for an aberration-limited lens: NAG, NAC,
NAM, and NAP . Their relationships by size are expressed as follows:

ΝAG < ΝAC ≤ ΝA M ≈ ΝA P . (1.9)

Geometric aberrations are described by polynomials within a ray-intercept plot: a

cubic plot indicates spherical aberration; a parabolic plot indicates coma;
dissimilar slopes indicate astigmatism or axial color; a vertical shift indicates
lateral color. Comprehension of ray-intercept plots is essential for evaluation of
optical designs.
4 Chapter 1

Production optics can be effectively modeled through patent prescriptions.

Specifications within patents should resemble the actual or a similar product. A
proper optical design tool can generate ray-intercept plots, wavefront errors,
point spreads, and line spreads. A practical field height may be derived from the
onset of coma and astigmatism. Chromatic aberrations are common in both
patent prescriptions and actual products. A central NA should be derived from
the metrics of the optical design tool. Collection efficiency is derived from an
encircled-energy plot or calculated from the central and marginal NAs.
A production lens can also be modeled from published modulation transfer
function (MTF) data. The onset of astigmatism is frequently displayed in MTF
data, and a central NA may be estimated from the shape of the MTF data.

1.7 Image Contrast

Ordinarily, increasing the paraxial NA improves image contrast by directing
more object light into a smaller point spread; however, this is not always true.
Increasing the paraxial NA can degrade image contrast by directing more object
light into the annular skirt of an aberration-limited point spread.
A contrast calculation is based on several features. The central NA
determines the collection efficiency of the object, whereas the marginal NA
determines the collection efficiency of the background. The size of an object
determines the appropriate calculation of collection efficiency by a pixel. A full-
pixel object has a 100% collection efficiency by pixel, while a subpixel object
does not. A fluorophore radiates from a point, while intrinsic fluorescence
radiates from a full pixel. These are important nuances in the detection of
fluorophores.
The resolution of an optical system may be defined in numerous formats. An
Airy disk is a frequently used metric in optical design. A geometric radius may
indicate a point spread far beyond the Airy radius. A root-mean-square (rms)
radius provides a practical metric for resolution. An ensquared energy plot
effectively defines collection efficiency of a pixel. A central NA defines a sharp
peak within the point spread. A wavefront error describes the limitations of plano
optics, such as filters and mirrors.
The specimen flux (quanta per time) must be considered. A fluorophore
might stop emitting after emission of 100 photons. Consequently, the collection
NA must be maximized. However, increasing the collection NA does not always
increase the image contrast.
The background radiance must also be considered. It can originate from the
foreground or background, or can even originate from the optical path. Impurities
in the cover strata can emit fluorescence, as can the intrinsic material of the lens.
Detector noise is yet another consideration. Dark current, read noise, and
dark noise are three common metrics of electronic detectors of light. The object
signal may overcome the read noise through integration; however, the object
signal cannot overcome the dark current by integration. The dark noise is the sum
of the dark current and read noise at a specific frame rate. The shot noise of the
object signal must also be considered.
Optical Design Concepts 5

There are numerous possibilities for the definition of contrast. An effective

model requires careful consideration of the application. It is important to note
that collection of more light does not always improve image contrast; an
optimum NA yields the maximum contrast.

1.8 Manufacturing
Surface quality is expressed in three common formats. Surface-form error
describes the deviation from the specified form; 50 nm is a typical specification
for surface-form error in optics for the visible spectrum. Surface imperfections
describe the size and number of digs, coating defects, and scratches (known as
“scratch/dig”). Surface imperfections must be extremely small when near the
object field. Surface texture defines the fine structure limits. The total integrated
scatter (TIS) of a surface is defined as

TIS = φ 2 , (1.10)

where φ is the rms phase error of the exiting wavefront.

The material quality is extremely important. Stress birefringence creates
wavefront aberration, whereas bubbles, inclusions, and striae cause scattering.
Impurities create extrinsic fluorescence and may discolor during ultraviolet (UV)
exposure. Crystalline inclusions may be dragged through the surface during
polish. Some vendors provide metrics for abrasion resistance. Material quality is
often overlooked during fabrication specifications. The cost of optical glass
increases rapidly as the purity increases; however, a thrifty choice of glass often
results in more expense due to rejections during system assembly.
The key to consistent manufacturing is specification of reasonable tolerances.
A reasonable tolerance is based on the current art of production. Such limitations
are not normally published data. They must be discovered through discussion
with the vendor. An acceptable lens design might fail in production due to
unreasonable tolerances.

1.9 Assembly
Assembly procedures should state specific steps toward a measurable result.
Otherwise, performance may vary between units. Tools may be required for
qualification of subsystems, and subsystems should be qualified for installation
into larger assemblies. A facile assembly procedure is essential for consistent
production. After assembly, the system must be qualified. A standard specimen
should be specified for measurement by the instrument and may require
significant expense.
Chapter 2
Basic Microscope Concepts
2.1 Magnification
Magnification specifies the ratio of an image dimension to its object dimension.
Magnification is typically cited in one of three formats: axial magnification,
radial magnification, or angular magnification. Axial magnification defines the
ratio of an axial image shift to an axial object shift. Radial magnification defines
the ratio of an image height to the object height. Axial magnification is
proportional to the square of the radial magnification. Angular magnification
defines the ratio of the image angle to the object height or angle.
Several types of dimensions are employed in defining these magnifications:
an axial dimension is a distance along the optical axis; a radial dimension is a
distance across the optical axis, and may be linear or angular. A radial dimension
may be called lateral or transverse, and an axial dimension may be called
longitudinal.
In a microscope, the “magnification” can be both radial and angular. The
objective lens indicates a radial magnification for a real image, while an ocular
lens specifies an angular magnification for a virtual image. A microscope does
not typically specify an axial magnification, although it is an important
parameter, since the depth of focus is scaled by the axial magnification.

2.2 Accommodation
The human eye comprises a static lens of the cornea and a dynamic lens within
the ciliary muscle. The standard far point of human vision is infinity ∞, which
may be expressed as zero diopters of magnification. A standard eye should focus
on an object at 250 mm. This represents an addition of four diopters by the
compressed dynamic lens. Therefore, the standard near point sNP of human vision
is 250 mm, or four diopters of magnification. The process of dynamic focus
within this range (250 mm–∞) is called accommodation.
The standard optical power of the near point is defined as

1
φ NP = = 40 diopters , (2.1)
s NP

7
8 Chapter 2

where the optical distance of the standard near point is 250 mm. A diopter is a
reciprocal meter.

2.3 Finite Tube Length

Until the late 1900s, most microscopes employed a lone objective lens to create
an image at the end of a tube. The objective flange mates to the front end of a
tube, and an eyepiece mounts to the back end of the tube. The tube length defines
the nominal operation of the objective. Standard tube lengths vary by vendor
from 160 to 210 mm. An objective lens is designed specifically for a nominal
tube length, which is normally marked on the objective barrel. An objective lens
with finite tube length does not function properly at another tube length.
Figure 2.1 displays a thin-lens model of an objective lens with a finite tube
length. The lateral magnification is negative; therefore, the image is inverted. It is
upside down and backward with respect to the object. Consequently, the image
moves in the opposite direction of the object. The objective lens may be called
“the objective” without reference to a lens.
The absolute magnification M of an objective lens with a finite tube length is
fairly estimated by Eq. 3.7 as

LT
(1 + M ) ≈ , (2.2)
f Obj

where LT is the length of the tube, and fObj is the focal length of the objective lens.
An objective barrel is normally marked with both tube length and magnification.
The magnification M drops the negative polarity of the radial magnification. The
prescription of an objective is tailored to a particular magnification. Insertion of
plano optics within the convergent rays of the tube can create significant
aberration and defocus.

Figure 2.1 Objective lens with a finite tube length. The tube defines a finite distance to the
image conjugate. The chief ray (CR) travels from the center of the lens stop (LS) to the
margin of the image.
Basic Microscope Concepts 9

Figure 2.2 Objective lens with infinity correction. The cardinal points along the optical axis
(OA) are front focal point (FFP), front principal plane (FPP), back principal plane (BPP),
and back focal point (BFP). The lens stop defines the marginal ray (MR) and lens-stop
diameter (LSD). Body (B) defines working distance (WD), flange (F), and flange distance
(FD).

2.4 Infinity-Corrected Objective

An ∞-corrected microscope splits the objective lens into two parts: an objective
lens and a tube lens. The objective lens creates an infinitely distant image within
the tube. The tube lens converts the infinitely distant image into a real image at
the tube-lens focal point. Plano optics may be placed in the tube without creation
of aberration.
Figure 2.2 displays a modern objective lens with ∞ correction. The object is
located at the front focal point. The image conjugate is infinitely distant. The
objective prescription is specified for an infinitely distant image. The ∞
correction is frequently indicated on the objective barrel as ∞, or INF. The flange
distance specifies the nominal distance of the object from a flange on the barrel.
The working distance indicates a gap between the nominal object position and
the body of the microscope.
The paraxial NA of an objective lens defines the angular size of the lens stop
with respect to the field:

φ LS
ΝA P = n sin θ LS = , (2.3)
2f
10 Chapter 2

where θLS is the angle of the lens stop, φLS is the diameter of the lens stop, and f is
the focal length. The marginal ray is defined by the center of the object field and
the margin of the lens stop.

2.5 Tube Lens

A tube lens converts the infinitely distant conjugate of an objective into an image
at the back focal point of the tube lens. The magnification of an ∞-corrected
objective lens is specified as

fT
M Obj = , (2.4)
f Obj

where fT is the focal length of the tube lens and fObj is the focal length of the
objective lens. The magnification will certainly change if the focal length of the
tube lens is changed. The image at the back focal point of the tube-lens is both
real and inverted.
The ∞ correction of the tube lens permits insertion of plano optics without
incurring defocus or aberration. Multiple beamsplitters and filters can also be
inserted without issue. However, a warped beamsplitter is not a plano optic; the
optics must be flat within a specific tolerance.
The location of the lens stop determines the location of the chief ray within
the tube lens. Separation of rays by field within a tube lens enables selective
correction of off-axis aberrations. Typically, a tube lens may travel axially by
10% of its focal length without creating significant aberration. A smaller
allowance of 1% applies to lateral movement.
Figure 2.3 displays a tube lens in combination with an objective lens. The
lens stop is placed at the front focal point of the tube lens. If the lens stop is
considered an object, then the tube lens creates an infinitely distant image of the
lens stop. The distant lens stop defines a telecentric lens. A telecentric image
space is also indicated by an axial chief ray at the image. The object space is also
telecentric.

Figure 2.3 Objective lens (OL) with tube lens (TL). The objective lens has a back focal
point (Obj-BFP). The tube lens has a front focal point (TL-FFP), front principal plane (TL-
FPP), back principal plane (TL-BPP), and back focal point (TL-BFP). The chief ray travels
from the center of the lens stop to the margin of the image.
Basic Microscope Concepts 11

2.6 Ocular Lens

An ocular lens is a magnifier with a lens stop on the front side. It is frequently
called an eyepiece or ocular without reference to a lens. The ocular creates an
image of the lens stop at the back focal point of the ocular. The exit pupil of a
microscope is defined by a bright white disk floating above the ocular. This
floating disk is an image of the lens stop of the objective lens. It might even be an
image of a filament from the illumination system. The exit pupil of a microscope
is typically smaller than the 2.5-mm natural pupil of the eye. Consequently, the
eye operates within a diffraction limit when looking into a microscope. The exit
pupil of a microscope can significantly sharpen visual acuity.
The diameter of the exit pupil is derived from Eq. (3.8) as

f Oc
φ Ex = φ LS , (2.5)
f Oc − ( f Oc + f T )

where fOc is the focal length of the ocular. Simplification yields

f Oc
φ Ex = φ LS . (2.6)
fT

The eye relief specifies the distance from the back of the ocular to the exit
pupil. Eye relief allows room for spectacles. Furthermore, an eye cup may reduce
contributions from ambient light.
As displayed in Fig. 2.4, an ocular creates an image of the lens stop. This
image defines an exit pupil at the back focal point of the ocular lens (see Table
2.1). The ocular lens also magnifies an image at its front focal point. The
magnification of an ∞-corrected ocular lens is specified as

φ Oc 250 mm
M Oc = = , (2.7)
φ NP f Oc

where φOc is the power of the ocular, φNP is the power of the standard near point
at 250 mm, and fOc is the focal length of the ocular lens.

Table 2.1 Two common geometries for an exit pupil.

Objective Focal length Exit Brand

lens stop Tube Ocular pupil
10 200 25 1.25 Nikon
10 180 25 1.39 Olympus
12 Chapter 2

Figure 2.4 Tube lens with an ocular lens. The ocular lens has a front focal point (Oc-FFP),
principal plane (Oc-PP), and back focal point. The exit pupil is located at the ocular back
focal point. The exit pupil (EP) is an image of the lens stop. The eye relief (ER) defines the
distance from the last surface to the exit pupil. The chief ray travels from the center of the
lens stop through the margin of the image to the center of the exit pupil.

2.7 Refractive Objects

A refractive object, or refractile, deflects illumination beyond the vision lens
stop. Figure 2.5 displays a thin-lens model of a cell that acts on a single point of
illumination. In Fig. 2.5, the object is centered on the illumination, and the entire
illumination ray bundle passes through the vision lens stop. The center of the cell
appears bright white. In Fig. 2.6, the object is slightly decentered, while the entire
illumination ray bundle barely passes the vision lens stop. The cell remains
bright. In Fig. 2.7, the object is decentered enough for partial blocking by the
vision lens stop. The cell appears gray at this point. The edge of the cell
gradually grows darker in the image. In Fig. 2.8, the object is decentered enough
for complete blocking by the vision lens stop. The edge of the cell remains dark
over a short distance from the edge.

Figure 2.5 Bright field at the center of a cell. The cell image is bright in the center of the
cell. The illumination lens stop (ILS) defines the convergence of illumination. The vision
lens stop (VLS) defines the collection of illumination. The background appears bright. The
rays are bent at the cell principal plane (CPP).
Basic Microscope Concepts 13

Figure 2.6 Bright field near the edge of a cell. The cell image is bright near the edge of the
cell. The illumination lens stop defines the convergence of illumination. The vision lens
stop defines the collection of illumination. The background appears bright.

Figure 2.7 Bright field in transition at the edge of a cell. The cell image is gray during the
transition to dark at the edge of the cell. Refraction by the cell directs light into the opaque
portion of the lens stop.

The width of the gray portion of refraction is proportional to the illumination

NA (NAI):

wGR ∝ ΝA I , for ( ΝA I < ΝAV ) . (2.8)

As NAI deceases, the gray band decreases in width.

The width of the dark region is dependent on the vision NA (NAV) as below:

wDR ∝ (1 − ΝAV − ΝA I ) , for ( ΝA I < ΝAV ) . (2.9)

14 Chapter 2

Figure 2.8 Bright field at the dark edge of a cell. The cell image is dark at the edge of the
cell. Refraction by the cell directs light into the opaque portion of the lens stop.

As the vision NA grows, a steeper portion of the cell wall is required for
sufficient deflection. Both of these linear dependencies are dependent on a thin-
lens model. Actual dependencies can be nonlinear.
The first-order resolution of a refractive object is dependent on NAI and
NAV. Reduction of NAI should reduce the total width of an edge of refraction.
However, the width of the dark portion also displays important dependencies on
diffraction.

2.8 Diffractive Objects

A diffractive object implies some form of edge. An edge has numerous origins,
such as the phase delay of an index change or the absorption by a dye. A
refractive object has a diffractive component at the edge.
A circular object creates the well-known Airy pattern. The angular
components of the Airy pattern tend toward zero at a larger angle; however, they
are finite in magnitude up to 90 deg. Consequently, the vision lens stop defines
the angular extent of the circular object. The spatial resolution of diffraction is
dependent on wavelength and NAV:

λ
wD ∝ . (2.10)
ΝA V

2.9 Dark Field

Figures 2.9 and 2.10 display a dark field. In Fig. 2.9, the central portion of the
vision lens stop blocks the normal illumination, and the image background
appears dark. The center of the cell also looks dark. In Fig. 2.10, the edge of the
Basic Microscope Concepts 15

cell refracts light into the open annulus of the vision lens stop, and the edge of
the cell appears bright.
Diffraction may also create bright features in a dark field. An edge diffracts
light through the lens stop with little dependence on NAI. A subpixel object may
create significant diffraction through the vision lens stop. Moreover, a dark-field
contrast may reveal small features that are not visible in the bright field.

Figure 2.9 Dark field at the center of a cell. The cell image is dark at the center of the cell.
The center of the vision lens stop blocks illumination in the background. The background
appears dark.

Figure 2.10 Dark field at the edge of a cell. The cell image is bright at the edge of the cell.
Refraction by the cell directs light into the clear portion of the vision lens stop. The clear
portion may be a partial or full annulus. A partial annulus may rotate about the optical axis.
Chapter 3
Basic Geometric Optics
3.1 Ray Tracing
A ray is a convenient representation of wave propagation. A ray represents the
normal to a wavefront. Geometric optics defines the transport of light through
simple geometric constructions of ray transmission and reflection. Snell’s law is
the only consideration during transmission of a ray. Paraxial optics specifies
cardinal points without consideration of aberrations. Seidel aberrations indicate
defects through polynomials within ray-intercept plots. Comprehension of the
ensuing topics is essential for the effective design of an optical instrument.
It is important to remember that geometric optics is a simple model for more
complex wave propagation. Ignorance or negligence of physical optics frequently
yields a dysfunctional design. Ray tracing is a convenient and powerful tool;
however, it is not a complete description of an optical system.

3.2 Cardinal points

There are three cardinal points in a lens: the principal point, the focal point, and
the nodal point. These cardinal points are easily managed in systems with
identical refractive indices at the object and image. However, the immersion oil
of a high-power objective lens requires a more complex set of cardinal points.
The principal point of a lens defines an axial point of refraction by the lens.
The principal point of a lens may be extended into a principal plane of the lens or
even a principal surface of a spherical shape. All refraction occurs at the principal
plane. A lone principal plane defines a thin lens where all rays are bent only at
the principal plane. A principal plane may also be split into a front and back
principal plane, which is the definition of a thick lens. There is no radial
displacement of rays between the principal planes of a thick lens. A thick lens
provides a simple description of a complex lens through paraxial optics.
The focal point of a lens defines a point of convergence for incident axial
rays. A focal length is defined by the distance from a principal plane to a focal
point. If the refractive index of the image is different from the refractive index of
the object, then there are two different focal lengths: a front focal length for the
incident rays of the object and a back focal length for the exiting rays of the
image. The effective focal length (EFL) normally cites the focal length in air.
A nodal point of a lens defines a nodal ray that is not refracted by the lens.
The exiting nodal ray travels in the same direction as the incident nodal ray.
17
18 Chapter 3

Ordinarily, the nodal points are coincident with the principal points; however, if
the refractive index of the front is different from that of the back, then the nodal
points are separated from the principal points.

3.3 Stops
The aperture of a lens defines a stop, which restricts flow of light. A field stop
defines the spatial extent of an image field. A lens stop defines the angular extent
of the aperture at an image. Other stops might eliminate detrimental rays. The
word aperture can be ambiguously applied to a variety of thick-lens parameters;
therefore, it is important to reference an aperture by its proper name: field stop,
lens stop, or back focal plane. The marginal ray defines the angular extent of the
lens stop with respect to an image point. A chief ray travels through the margin
of the field stop and the center of the lens stop; it is frequently referenced in ray
calculations.
The marginal NA (NAM) is derived from the marginal ray of the lens stop as

ΝA M = n sin θM , (3.1)

where n is the refractive index and θM is the angle of the marginal ray. The NA
can be applied to several formats. The marginal ray can apply to the object, the
image, the focal point, or even the angular field of view. The NA must be
properly indentified by its location.
The f/# of a lens is another metric for stop size:

f
F= , (3.2)
φ EnP

where φEnP is the diameter of the entrance pupil as defined by an image of the
lens stop within object space. The f/# of a lens depends on an image at a focal
point. The f/# does not consistently correlate to NA, which is dependent on image
distance. The image-space f/# of a lens is

sI
FI = , (3.3)
φ ExP

where φExP is the diameter of the exit pupil, and sI is the image distance. As stated
earlier, the exit pupil is an image of the lens stop within image space.
Conversely, the entrance pupil is an image of the lens stop within object space.
As the image distance increases, the image-space f/# inflates, thus reducing the
brightness of subpixel objects.
The image-space f/# can inflate for several reasons. It doubles for a relay lens
with unity magnification and inflates at large field angles due to a reduction of
the solid angle with field position. A lens design might require f/# inflation for
Basic Geometric Optics 19

control of aberration. Spherical aberration might inflate the effective number for
subpixel objects. A low-f/# camera lens (such as f/1.4) should be carefully
examined before using it for low-light level applications. The image contrast will
not likely match calculations of the f/# stated by the vendor.

3.4 Gaussian Lens Formula

A thin-lens model describes a lens in terms of refraction at a principal plane.
Figure 3.1 displays the four cardinal points of a thin-lens diagram. By
convention, incident rays are governed by the front focal point, whereas exiting
rays are governed by the back focal point. The front focal length is longer than
the back focal length. However, the rules of the cardinal points are still observed.
An incident front ray becomes a back axial ray. A nodal ray maintains slope. A
front focal ray becomes a back axial ray.
The image conjugates are related by distance, focal length, and refractive
index:

nO n I nO
+ = , (3.4)
dO d I fO

where nO is the refractive index within object space, nI is the refractive index
within image space, dO is the object distance from the principal plane, dI is the
image distance from the principal plane, and fO is the focal length within object
space. This concept may also be written as the Gaussian lens formula:

Figure 3.1 Thin-lens model of a relay lens with refractive indices of 1.5 for incident rays on
the front side and 1.0 for exiting rays on the back side. Front focal point (FFP), nodal point
(NP), principal point (PP), back focal point (BFP), object distance (OD), and image
distance (ID) are indicated. The OD is positive to the left (from PP to the front space). The
ID is positive to the right (from PP to the back space).
20 Chapter 3

1 1 1
+ = , (3.5)
sO s I f

where sO is the optical distance of the object, sI is the optical distance of the
image, and f is the focal length within air. Optical distance is the equivalent
distance in air.
A traditional Gaussian lens formula employs a specific polarity convention.
A positive focal length or distance is measured from a principal plane toward a
focal point. Verbal descriptions of image conjugates can be difficult.
Consequently, a thin-lens diagram is warranted.
The lateral magnification ML can be defined in several formats:

hI −s I
ML = = , (3.6)
hO sO

where hI is the height of the image and hO is the height of the object. In terms of
focal length, the magnification is expressed as
s −M s
(1 − M L ) = I = L O . (3.7)
f f
f
ML = . (3.8)
f − sO

During positive lateral magnification, the image moves in the same direction
as the object. The axial magnification is

Δs I
MA = = M L2 . (3.9)
ΔsO

The tilt magnification is

MA
MT = = ML . (3.10)
ML

During positive lateral magnification, the image tilts in the same direction as the
object. The magnification of a lens is normally stated as the absolute value of the
lateral magnification:

M = ML . (3.11)
Basic Geometric Optics 21

The reduction of a lens is the reciprocal of the magnification:

1
R= . (3.12)
M

3.5 Image Types

Images may be considered real or virtual. A real image is formed by rays that
actually pass through the image point. Consequently, a finger may touch a real
image. A virtual image, on the other hand, is formed by rays that apparently
originate from the image point but do not exist at the image point. A virtual
image does not exist in real space. A finger cannot touch a virtual image.
Image orientation may be defined as the orientation of its upside and left
side. An erect image is “upside-up” and “left-side-left.” A flipped image is either
upside down or backward. An inverted image is flipped in both directions—it is
upside down and backward. A mirror image is flipped along just one direction.
Finally, an image conjugate is an image within a pair of images conjoined by a
lens system. The object may be considered an image conjugate.

3.6 Optical Power

The optical power (φ) of a lens is defined as the reciprocal of the focal length
within air:
n 1
φ= = , (3.13)
fn f

where n is the refractive index and fn is the focal length within the refractive
index. A diopter specifies a unit for optical power as reciprocal meters. Thus, a
focal length of 1000 mm indicates one diopter of power, while a focal length of
250 mm defines four diopters of power.
There are three formats of the optical power: transmission, external
reflection, and internal reflection. The polarity is positive when a collimated
incident beam becomes a convergent exiting beam. The optical power for
transmission is

1
φ T = Δn   , (3.14)
R

wherein Δn is the change in refractive index and R is the radius of the surface.
The optical power for an external reflection is

1
φ ER = 2   . (3.15)
R
22 Chapter 3

The optical power for an internal reflection is

1
φ IR = 2n   . (3.16)
R

At an air-to-glass interface (n = 1.0, 1.5), the optical power of transmission is

1
φ T = 0.5   . (3.17)
R

The optical power of an external reflection is

1
φ ER = 2   = 4 φ T . (3.18)
R

The optical power of an internal reflection is

1
φ IR = 3   = 6 φ T . (3.19)
R

Obviously, the effect of radius is much greater in reflection than transmission.

The effects of surface texture and surface form error display a similar
dependence.
Optical power may also define angular magnification. The angular height per
spatial height within air may be defined through optical power as

n
φ= , (3.20)
s

where s is spatial distance to the object. The angular height of an object is

defined as
h φ
α= =h , (3.21)
s n

where h is the spatial height of an object. The spatial height of an object or image
is independent of the local refractive index, while the angular height is inversely
proportional to the local refractive index.

3.7 Paraxial Optics

The term paraxial defines a regime where the rays are nearly equivalent to the
optical axis, and the following approximation is valid:

sin θ = θ . (3.22)
Basic Geometric Optics 23

The paraxial condition may also apply to an off-axis ray that is nearly equivalent
to the normal. The paraxial condition is a first-order approximation of Snell’s
law:

n 1θ1 = n 2 θ2 . (3.23)

This linear relationship enables a simple-matrix method for the construction of

images. The optical distance is the equivalent distance in air:

d
s= , (3.24)
n

where d is the spatial distance and n is the refractive index. The optical distance
is expressed in units of spatial length. The optical path length is a length of the
phase within a wave:
nd
Λ= , (3.25)
λ

where Λ is the optical path length and λ is the spatial wavelength. The optical
path length is expressed in units of phase angle, cycles, or wavelengths. The
optical path difference normally refers to the optical path length. It is important
to carefully define these parameters, as they are frequently confused.

3.8 Relay Lens

Figure 3.1 displays a thin-lens model of a relay lens. The cardinal points are
located in dissimilar refractive media. The object is immersed in oil at a
refractive index of 1.5 on the front side of the lens. The front focal length is 1.5
times the back focal length. An image of the object is formed on the back side of
the lens within air at a refractive index of 1.0. A focal ray is converted to an axial
ray at the principal plane, and vice versa. A nodal ray is transmitted at the
principal plane without refraction. The image location is established through two
of three rays: an incident axial ray, an incident focal ray, or an incident nodal ray.
Incident rays are governed by the front cardinal points, whereas exiting rays
are governed by the back cardinal points. The front focal point of a positive lens
is located on the incident side of the lens, whereas the front focal point of a
negative lens is located on the exiting side of the lens.
24 Chapter 3

Figure 3.2 Thick-lens model of relay lens with refractive indices of 1.5 for incident rays on
the front side and 1.0 for exiting rays on the back side. Front nodal point (FNP) and back
nodal point (BNP) are shown.

Figure 3.2 displays a thick-lens model of a relay lens. The principal plane is
split into front and a back principal planes. Rays translate from the front to back
principal planes without a change in height. There are now two principal points
and two nodal points. Refraction in a thick lens depends on cardinal points in the
same fashion as a thin lens. A thick-lens model provides accurate location of
cardinal points within a lens barrel.

3.9 Magnifier
Figure 3.3 displays a thin-lens model for a magnifier. The object is located inside
the front focal point. The image is located outside of the front focal point. The
real rays on the back side of the lens do not extend to the front side. The real rays
are extended through the front side as virtual rays that originate from a virtual
image. A virtual ray does not exist in real space.
In Fig. 3.4, the angular magnification of a magnifier is defined by an object
at its front focal point. The magnifier converts the object into an infinitely distant
image. The angular size of the image with respect to the principal point is

α n = hO φ M , (3.26)

where α is the angle in radians, n is the refractive index, hO is the object height, f
is the focal length, and φM is the power of the magnifier. Thus, the angular
magnification is the optical power of the magnifier. The lateral magnification of
an object at the standard near point is

φM s NP
ML = = . (3.27)
φ NP fM
Basic Geometric Optics 25

A typical magnifier displays 40 diopters of optical power, while the near-point

optical power is 4 diopters. Application of a 40-diopter magnifier to the standard
near point yields a lateral magnification of 10 diopters by angle.
The loupe was an early format of a magnifier. Its name is derived from the
French word for an imperfect gem; its modern translation is “magnifying glass.”
A modern loupe may contain one or more lens elements. Walker2 describes
several magnifiers in detail.

Figure 3.3 Thin-lens model of a magnifier. Phantom lines (dash, dash, long dash) indicate
virtual rays and a virtual object. Solid lines indicate real rays.

Figure 3.4 Magnification of a magnifier with the object at the front focal point. Object angle
(OA) is defined by object height (OH) times magnification in diopters of power.
Chapter 4
Aberrations
As a consultant, this author has reviewed the design and performance of
numerous instruments and three common problems were found: ignorance of
physical optics, negligence of aberrations, and absence of a tolerance budget. At
least one member of the design team should indentify the type of aberration
within the lens. Spherical aberration grows with NA, coma grows with field
angle and NA, and astigmatism grows with tilt of the optic. These are important
considerations for an optical system design.

4.1 Seidel Aberrations

A polynomial expression for the ray height at an image is expressed as follows:3

y=Mh
+ A1 p
(4.1)
+ B0 p 3 + B1h1 p 2 + B2 h 2 p1 + B3 h3
+  0 Ci hi p 5-i + ,
5

where h is the object height and p is the pupil position. The first line indicates the
paraxial image height, which is simply the product of the magnification M and
the object height h. The paraxial image height specifies an image system without
aberration. The second line indicates a first-order aberration: defocus A1. The
third line indicates third-order aberrations: spherical aberration B0, coma B1,
astigmatism and Petzval curvature B2, and distortion B3.
Defocus occurs in several formats: an axial error in the position of the sensor
creates defocus across the entire image; a curved image surface creates variable
defocus across the field; a tilted surface creates astigmatism; a variation in
refractive index creates axial color. Defocus is indicated by a dependency on p1.
Figure 4.1 displays a report for a spherical lens with spherical aberration.
There is a dependency on p3. Spherical aberration is created by an increased
power of the marginal focus with respect to the axial focus. The marginal rays
are bent too much due to the spherical shape of the surface.

27
28 Chapter 4

Figure 4.1 Spherical aberration ray fan and spot. The spherical lens has a 20-mm focal
length, BK7 glass, and 0.25 NA.
Aberrations 29

Figure 4.2 displays a report for an aspheric lens with coma. There is a
dependency on hp2. Coma creates the appearance of hairy stars which resemble
comets. Comet is derived from the Greek word komē, meaning hairy, and astēr,
meaning star. Coma is created by both an axial and a lateral shift of the marginal
focus.
Figure 4.3 displays a report for an aspheric lens with astigmatism. There is a
dependency on h2p, which occurs in the tangential plane. Astigmatism is derived
from the Greek a-, indicating an antonym, and stigma, meaning spot. Ergo,
“astigmatism” means “no spot,” as in a line. There is focus along the tangential
pupili EY, while the spot is spread within the sagittal planeii EX. Astigmatism is
created by an axial shift of the tangential focus with respect to the sagittal focus.
The tangential pupil covers a larger arc angle due to the tilt of the lens.
Consequently, the tangential power is greater than the sagittal power.
The Petzval radius describes the radius of the paraxial focal surface. A
defocus by field position depends on h2p. Distortion describes a growth in focal
length with field position. The increased image height depends on h3, which is
not normally revealed in a ray fan. Both Petzval curvature and distortion are
normally described through dedicated plots for those features. For example, a
grid distortion plot might reveal a barrel or pincushion distortion of a square.

4.2 Chromatic Aberrations

Figure 4.3 displays lateral color. The ray intercept of the F and C lines are
dependent on h. Lateral color is created by the normal dispersion of glass,
wherein the refractive index grows as the wavelength decreases. The F line (486
nm) is shifted more than the C line (656 nm).
Figure 4.4 displays a report for an aspheric lens at 0.25 NA. There is a large
slope in the ray-intercept plots for all wavelengths in the x pupil but not in the y
pupil. The ray intercept of the d line is essentially flat, which indicates no
defocus. The ray intercept of the F line displays a negative slope, which indicates
axial defocus. The F line is bent more strongly due to a higher refractive index.
As with lateral color, axial color is created by the normal dispersion of glass,
wherein the refractive index grows as the wavelength decreases.

4.3 Other Aberrations

Third-order aberrations indicate any sum of three for the powers of h and p.
There are also fifth-order aberrations that become more important at higher NAs.
Spherochromatism indicates a change in spherical aberration with wavelength.
Errors in manufacture or assembly can also create aberrations. A lateral shift
between elements of equal but opposite spherical aberration can create significant
coma. Surface form error can also add random error. Finally, thin optics could
warp during application of coatings.

i
The tangential plane contains the optical axis and the field point.
ii
The sagittal plane contains the chief ray and is normal to the tangential plane.
30 Chapter 4

Figure 4.2 Coma ray fan and spot. The aspheric lens has a 20-mm focal length, BK7
glass, and 0.25 NA.
Aberrations 31

Figure 4.3 Astigmatism and lateral color ray fan and spot. The spherical lens has a 20-
mm focal length, BK7 glass, and 0.02 NA.
32 Chapter 4

Figure 4.4 Axial color ray fan and spot. The aspheric lens has a 20-mm focal length, BK7
glass, 0.20 NA, 0-deg field, F line (486 nm), d line (588 nm), and C line (656 nm).
Aberrations 33

4.4 Aspheric Surfaces

An aspheric surface is frequently defined by its conic section plus a summation
of polynomials, as below:

( y R)
2

z=R +  An y n , (4.2)
1 − (1 + K )( y R )
2
1+

where R is the radius of curvature, y is the height above the z axis, and K is the
conic constant. A conic constant of 0.0 defines a sphere that is easily created by
polishing. A conic constant of –1.0 defines a parabola, whereby reflection
converges a collimated beam into a point. A conic constant between 0.0 and –1.0
defines an ellipse, whereby internal refraction effectively converges a collimated
beam into a point. Finally, a conic constant beyond –1.0 defines a hyperbola,
whereby external refraction effectively converges a collimated beam into a point.
The Brewster window defines a practical limit for refractive aspheric surfaces. A
reflective parabola is limited only by tolerances on the field and surface form and
texture.
The polynomial coefficients An are useful for correction of aberrations. An A2
coefficient is normally zero because it is prominent in the conic section. A conic
constant of –0.4334 at the lens stop may correct for spherical aberration at the
cost of coma, as in Fig. 4.2. An A4 near the lens stop may also correct third-order
spherical aberration. An A4 and A6 near the field stop may correct astigmatism,
field curvature, or distortion. A single aspheric surface may completely correct a
single aberration at a single field point. However, complete correction at one
field point normally requires aberration at another. An aspheric surface at the
lens stop works best over a small angular field, whereas an aspheric surface at the
field stop works best at a small NA.
Chapter 5
Basic Physical Optics
5.1 Importance of Physical Optics
Physical optics defines the transport of light by the physics of wave propagation.
Maxwell’s equations and Fourier transformations are applied toward accurate
descriptions of point spread and defocus. Comprehension of the ensuing topics is
essential for an effective design of an optical instrument. Development of
comprehension is a product of exploration and application of these principles.
Refraction, reflection, and scatter are all based on the wave nature of light.
The Airy pattern is dependent on the diffraction of a circular aperture, and the
depth of focus is dependent on physical optics.
Most sections of this chapter have a more detailed version in Chapter 19,
which focuses on advanced concepts. At least one member of the optical design
team should comprehend the advanced concepts.

5.2 Wave Equation

The wave equation in Chapter 19 [Eq. (19.1)] may be reduced to a simple
expression for the spatial frequency k and the temporal frequency ω:

k 2 = εμω2 , (5.1)

where ε is electric permittivity and μ is magnetic permeability. The electric

permittivity is positive for a dielectric and negative for a conductor.
The spatial frequency in radians per distance is

2π
k= , (5.2)
λ

where λ is the spatial period. The temporal frequency ω in radians per time is
expressed as

2π
ω= , (5.3)
T
where T is the temporal period.

35
36 Chapter 5

The velocity of the wave is

ω 1
= = nc , (5.4)
k εμ

where n is the refractive index, and c is the speed of light.

The original language of electromagnetism employed the words “conductor”
and “dielectric,” as in, “a conductor displays electric current at low frequencies,
whereas a dielectric does not.” By persistence of this nomenclature, a glass is still
considered a dielectric even although it displays significant bound electron
current. The bound electron current of a glass defines the refractive index and
dispersion of a glass.
Chapter 19 provides more information on the wave equation, electron
current, and refractive index.

5.3 Refractive Index

The refractive index n is derived from

n 2 = εμc 2 . (5.5)

The spatial frequency within a refractive medium can be expressed as

ω 2π
k =n =n , (5.6)
c λ

where λ is the wavelength in vacuum.

The complex refractive index N is expressed as

N = n + iκ , (5.7)

where the real part n is the refractive index and the imaginary part κ is the
extinction coefficient. The refractive index is largely dependent on bound charge
motion, while the extinction coefficient is dependent on free carrier motion.
Application of the complex refractive index to a plane wave yields the
electric field as follows:

E = E 0 exp i ( N k ⋅ R − ωt )  , (5.8)

where the real part indicates a cosine function and the imaginary part indicates an
exponential decay. The complex format facilitates differential operations.

The refractive index n defines an oscillation of the electric field as

Basic Physical Optics 37

 2π 
E = E0 cos  n z − ωt  . (5.9)
 λ 

The extinction coefficient κ defines a decay of the electric field as

 −2π 
E = E0 exp  κz. (5.10)
 λ 

The extinction coefficient of a glass is normally zero, which implies an

infinite depth of the electric field. Conversely, the extinction coefficient of a
metal is large, which implies a shallow depth.
The motions of free and bound carriers are quite different. Free electrons
move quickly toward cancellation of an external electric field. Consequently, the
electric field of a metal decays exponentially toward zero within a fraction of the
spatial wavelength. Bound electrons also move toward cancellation of the
external field; however, a restoring force of the bond greatly limits cancellation
of the external field. Consequently, the electric field of glass returns to zero faster
than air, but not within a spatial wavelength. This effect of partial cancellation is
expressed by a shortened wavelength.
Chapter 19 describes mathematical models for the refractive index of free
and bound electrons. A Drude model applies to free carriers through two
parameters: plasma frequency and lifetime. A Lorentz model applies to bound
carriers through three parameters: plasma frequency, lifetime, and resonant
frequency. In both models, a lifetime indicates the average duration of oscillation
before absorption. A metal is described by free carriers, and a glass is described
by bound carriers.
Figure 5.1 displays a Drude model for the free electrons of silver. The plasma
frequency occurs at 150 nm, and the lifetime is 10 plasma cycles. The refractive
index is nearly zero throughout the visible regime, while the extinction
coefficient is > 3 from 400 nm and beyond. Most of the reflection is due to free
carrier motion without scatter. Consequently, the absorption is nearly zero and
the reflectance is nearly 100% [Eq. (5.6)].
Figure 5.2 displays a Drude model for the free electrons of nickel. The
plasma frequency occurs at 120 nm, and the lifetime is five plasma cycles. The
refractive index grows throughout the visible regime, and the extinction
coefficient is > 3 from 400 nm and beyond. The short lifetime of five plasma
cycles indicates significant absorption. Consequently, the absorption is large and
the reflectance is near 70% [Eq. (5.6)]. Nickel appears darker than silver due to
electron scattering, even though the higher plasma frequency of nickel indicates a
higher electron density. Silver has superior electron mobility due to less frequent
scattering, as is indicated by a longer lifetime. Electron mobility has a profound
effect on reflectance.
38 Chapter 5

Figure 5.1 Drude model of silver. Refractive index is n, extinction coefficient is κ, and
reflectance is R at the d line (589 nm).

Figure 5.2 Drude model of nickel.

Figures 5.3 and 5.4 display the Lorentz spectra of silica. Both Lorentz
models employ a bound resonance at 100 nm. The refractive index resembles the
empirical data for silica,4 while the extinction coefficient varies greatly. The
lifetimes are varied for the purpose of exercise of the Lorentz model. In Fig. 5.3,
a lifetime at 800 bound cycles creates a reasonable estimate of the refractive
index; it grows more rapidly with a shorter wavelength, which defines normal
dispersion. In Fig. 5.4, the lifetime is much shorter at 100 bound cycles. The
refractive index does not grow more rapidly at shorter wavelengths, which
defines anomalous dispersion. Actually, in the example of Fig. 5.4, the refractive
index profile is flat from 0.2–1.0 μm, while the extinction profile is broadened.
Basic Physical Optics 39

Figure 5.3 Lorentz model of silica.

Figure 5.4 Lorentz model of silica with a shortened lifetime. Absorption reduces the
magnitude of dipole current and dispersion. A shorter lifetime indicates low dispersion of
short flint.

Figure 5.5 displays the transmission of a Lorentz model of silica at different

lifetimes, as derived from Eq. (5.9). The internal transmittance is calculated from
Eq. (5.21). The lifetime at 800 cycles creates a narrow absorption band, which is
largely confined to the ultraviolet regime (< 400 nm). The length of the spectrum
is defined by the separation of two absorption bands: a first at 0.1 μm, and a
second at 9 μm. The lifetime at 800 cycles defines a long spectrum in the blue,
which extends far into the ultraviolet. The lifetime at 100 cycles creates a broad
absorption band, which extends into the blue regime (440–490 nm). This defines
a short spectrum in the blue, which is a feature of short flint glass. Chapters 17
and 18 provide more information on glass types and refractive indices.
40 Chapter 5

Figure 5.5 Transmittance of a Lorentz model of silica. Absorption increases the width of
the absorption band. A shorter lifetime creates absorption of blue wavelengths. A short
lifetime displays a short spectrum in the blue.

Aluminum represents a combination of Drude and Lorentz spectra. Its

refractive index and extinction coefficient are 1.15 and 7.15 at the d line,
respectively. Aluminum behaves like silver with a bound oscillator at 800 nm.
Consequently, there is small dip in reflectance at 800 nm. The reflectance of
aluminum is near 92% at the d line.

5.4 Dispersion
During normal dispersion, the refractive index grows more rapidly with a shorter
wavelength. The Abbe number provides a metric for dispersion:

nd − 1
νd = , (5.11)
nF − nC

where nd is the d line of helium at 587.6 nm, nF is the F line of hydrogen at 486.1
nm, and nC is the C line of hydrogen at 656.3 nm.
The partial dispersion describes the change in refractive index as a fraction of
the change between the F and C lines. For example, the partial dispersion of the g
and F lines is defined as

n g − nF
PgF = . (5.12)
nF − nC
Basic Physical Optics 41

The g and F lines are both active in the blue region of color vision. Other
spectral lines are listed later in Table 17.6. The SCHOTT glass company defined
two glass types as normal in partial dispersion: a crown K7 and a flint F2.
There are two methods for decreasing dispersion νd: increasing resonant
frequency or decreasing resonant lifetime. A higher resonant frequency places
the resonant frequency deeper into the UV; this creates a flat index profile in the
visible regime. It also defines a long-spectrum crown, with a large distance
between resonant wavelengths. On the other hand, a shorter resonant lifetime
represents less dipole current; this also flattens the refractive index profile. A
shorter resonant lifetime also indicates absorption in the blue wavelength regime;
this defines a short-spectrum flint. Additional information on materials can be
found in Chapters 17 through 19.

5.5 Refraction and Reflection

During refraction, the spatial frequencies along the surface must be the same for
both materials. Consequently, only the normal component of the propagation
vector k changes as expressed in Fig. 5.6. The condition of tangential spatial
frequencies yields a form of Snell’s law as follows:

k1 sin θ1 = k2 sin θ2 . (5.13)

During reflection, the tangential components must remain constant while the
magnitude of the propagation vector k remains constant. Consequently, the
reflection angle must equal the incident angle.

Figure 5.6 Propagation vectors of refraction and reflection.

42 Chapter 5

The Fresnel reflection and refraction coefficients are based on the continuity
conditions along the surface. There are two polarizations. A senkrecht
polarization specifies an upright electric field to the plane of incidence. A parallel
polarization specifies a parallel electric field to the plane of incidence. The
reflection coefficients for the electric field of a wave at normal incidence are

N1 − N 2
rSN = − rPN = , (5.14)
N1 + N 2

where rSN is the reflection coefficient for senkrecht polarization at normal

incidence, and rPN is the reflection coefficient at parallel polarization at normal
incidence. A shown in Fig. 5.7, the reflection coefficient flips polarity when the
order of materials is reversed. Consequently, a film with a thickness much less
than a wavelength displays zero reflection.
The reflectance specifies the reflection of irradiance, which is proportional to
the squared magnitude of the electric field. Consequently, the reflectance at
normal incidence is
2
N − N2
2
R= r = 1 . (5.15)
N1 + N 2

The reflectance at normal is a fair approximation for the sum of both

polarizations within the Brewster angle:

N2
tan θ B = . (5.16)
N1

The reflectance of a typical glass-to-air interface is 4%, where the refractive

index for the glass is 1.5.

Figure 5.7 Reflection coefficients at normal incidence.

Basic Physical Optics 43

5.6 Emission
Derivation of the irradiance of a dipole emission is

 
p
2
  sin 2 θ 
ID =   , (5.17)
 32 π2 ε 0 c3   r 2 
 

where p  is the temporal curvature of the dipole moment. This pattern resembles a
toroid with an inner diameter of zero. The irradiance is zero along the axis of the
toroid.
The total radiant flux of a dipole is

2

p
P= . (5.18)
12πε0 c 2

If dipole rotation creates an isotropic emission, then the radiant power is evenly
distributed over 4 π steradians. The irradiance of a spherical emission is

2

p 1
IS = . (5.19)
48 π2 ε0 c 3 r 2

The spherical irradiance is two-thirds the maximum for that of dipole emission.
A dipole can also absorb photon energy where electric-field polarization
might be critical. A constrained molecule can display anisotropic absorption with
polarization of the excitation field.

5.7 Absorption
Attenuation of the electric field may be expressed through the extinction
coefficient as

 2π   2π 
E = E0 exp  − κz  cos  nz − ωt  . (5.20)
 λ   λ 

The resulting irradiance profile is

1
I = cε0 E0 exp ( −αz ) ,
2
(5.21)
2

where c is the speed of light, ε0 is the electric permittivity of free space, and α is
the absorption coefficient:
44 Chapter 5

2π 4π
α=2 κ= κ. (5.22)
λ λ

The absorption coefficient applies to the electric irradiance, whereas the

extinction coefficient applies to the electric field.

5.8 Evanescent Field

The spatial and temporal curvatures display a polarity. The curvature of a
sinusoid is negative, which defines an inward curvature toward zero. The
curvature of an exponential is positive, which defines an outward curvature from
zero. An inward curvature defines an oscillation. An outward curvature defines
growth or decay. An evanescent field vanishes over a short distance and is
mathematically expressed as positive spatial curvature.
Figure 5.8 displays the crests of the internal and external waves during total
internal reflection. There is an internal wave and an external wave. The internal
wave obeys

k x2 + k y2 + k z2 = k I2 , (5.23)

where kI is the ordinary spatial frequency of the internal medium. The external
wave obeys

k x2 + k y2 − α 2z = k E2 , (5.24)

Figure 5.8 Wave crests of total internal reflection: (a) top view of the xz plane, and (b)
side view of the xy plane at the midsection of the top view. A solid line shows a positive
peak, and a dashed line shows a negative peak.
Basic Physical Optics 45

where kE is the ordinary spatial frequency of the external medium. The ordinary
spatial frequency is smaller in the external medium. The external wave requires
an outward curvature α 2z , because the sum of the tangential inward curvatures
k x2 + k y2 exceeds the inward curvature of the external medium k E2 . In terms of
spatial frequency, the tangential spatial frequency is faster than the external
spatial frequency. In terms of wavelength, the tangential wavelength is shorter
than the external wavelength. These extraordinary conditions define the
exponential decay of an evanescent field.

5.9 Space-Angle Product

An electromagnetic wave does not focus to a point as indicated by a thick-lens
model. A ray represents a normal vector to a wavefront. Consequently, an
electromagnetic wavefront converges to a point spread. A Gaussian laser beam
provides a simple expression for this convergence through the space-angle
product:1

AΩ = λ 2 , (5.25)

where A is the area of point spread, Ω is the solid angle of convergence, and λ is
the spatial wavelength. A Gaussian version of the space-angle product is

π
dG ( 2 ΝAG ) = λ = 1.27λ , (5.26)
4

where dG is the diameter of the Gaussian beam and NAG is the Gaussian NA.
The Gaussian space-angle product is based on a fundamental principle of
Fourier summation:
1
σ x σk ≥ , (5.27)
2

where σx is the standard deviation of spatial position, and σk is the standard

deviation of spatial frequency.
An extraordinary spatial frequency is defined as

ω
k>N . (5.28)
c

The existence of an extraordinary spatial frequency requires an evanescent field.

Consequently, the NA may exceed a hemisphere in the presence of an evanescent
field:
46 Chapter 5

π
ΝA M > n sin   . (5.29)
2

5.10 Coherence
The axial version of the space-angle product is

4 2
Δz Δλ = λP , (5.30)
π

where Δλ is the wavelength range, and λP is the peak wavelength. The coherence
length within a refractive medium is

λ 2P
ΛC ≈ . (5.31)
n Δλ

Chapter 19 provides additional information on coherence.

5.11 Airy Pattern

An Airy pattern is created by diffraction of waves at a circular aperture. A plot of
an Airy pattern is displayed in Fig. 5.9 along with a Gaussian estimate.
The Airy pattern is frequently defined for an arbitrary aperture and distance.5
Conversion to marginal NA yields the subsequent relations for optical design.
The electric field of an Airy pattern is

2 J1 ( ρ k ΝA M )
E0 , (5.32)
ρ k ΝA M

where ρ is the radius about the optical axis, k is the spatial frequency, NAM is the
marginal NA, E0 is the electric field at the origin, and J1 is a Bessel function of
the first kind of the first order. A Bessel function is a solution to a second-order
differential equation with cylindrical boundary conditions.6 The irradiance of an
Airy pattern is expressed as

2
2 J1 ( ρ k ΝA M )
I Airy ( ρ ) = I 0 , (5.33)
ρ k ΝA M

where I0 is the peak irradiance (power per area). The first zero crossing of the
first-order Bessel function is
Basic Physical Optics 47

Figure 5.9 Point-spread functions of an Airy pattern and a Gaussian estimate. Standard
deviation of a Gaussian estimate to an Airy point spread is < 0.01 in magnitude.

γ J1 = 3.8317 . (5.34)

The diameter of the first dark ring of an Airy pattern is

λ  γ J  λ 
φ Airy = 1.22 =  1  . (5.35)
ΝA M  π   ΝA M 

The peak irradiance of the Airy pattern is related to the total power P0 (quanta per
time) by

π ΝA 2M
I0 = P0 . (5.36)
λ2

The space-angle product of the Airy disk is derived from Eq. (5.35) as

2γ J1
φ Airy ( 2 ΝA M ) = 2.44λ = λ, (5.37)
π

where NAM is the marginal NA of the circular aperture responsible for the Airy
pattern.
48 Chapter 5

The space-angle product of a linear slit is

wsinc ( 2 ΝA M ) = 2.00λ , (5.38)

where wsinc is the width of a sinc function defined as

sin ( x )
sinc ( x ) = . (5.39)
x

The irradiance of a slit is expressed as

I slit ( x ) = I 0 sinc2( xk ΝA M ) , (5.40)

where I 0 is the peak irradiance (power per area).

5.12 Gaussian Beam Propagation

A Gaussian profile is the natural mode of a propagating beam of laser light. A
plot of a Gaussian beam profile was displayed in Fig. 5.7 as an estimate of an
Airy pattern.
The irradiance profile (power per area) is

 r2 
I = I 0 exp  −2 2  , (5.41)
 rG 

where rG is the Gaussian radius.

The space-angle product of the Gaussian beam is

4
dG ( 2 ΝAG ) = 1.27λ = λ , (5.42)
π

where NAG is the Gaussian NA. All dimensions for space and angle refer to the
Gaussian radius rG.
A Gaussian beam propagates as follows:

φ2 = φG2 + β2 z 2 , (5.43)

where φ is the beam diameter, φG is waist diameter, β is the full angle of

divergence, and z is the distance of propagation. The waist of the beam is located
at z = 0. The radius of Gaussian profile rG defines both the beam diameter and the
full angle of divergence.
Basic Physical Optics 49

The radius of the Gaussian wavefront is expressed as

1 φ2
R= . (5.44)
z β2

At zero distance, an infinite radius indicates a flat wavefront. At a large distance,

the radius R approaches the propagation distance z. The Rayleigh distance is
defined as the distance from the beam waist to the point-of-minimum wavefront
radius of curvature. As mentioned in Chapter 1, the Rayleigh distance is

A0 λ
zR = = , (5.45)
λ Ω

where A0 is the area of the waist. The area of the beam diameter doubles over the
Rayleigh distance:

AR = 2 AG . (5.46)

The distribution of radiant flux power in a Gaussian profile will be described in

Chapter 19.

5.13 Transfer Functions

The MTF of an optical system can be either coherent or incoherent. A laser beam
is an example of a coherent system, whereas white light is an example of an
incoherent system. A narrowband filter may define a partially coherent system.
The MTF of a coherent system is the amplitude of the aperture function.
Thus, a circle creates an MTF in the shape of a cylinder. The cutoff frequency is
the spatial frequency of the margin

ΝA M 1.22
kM = = . (5.47)
λ φ Airy

The MTF of a coherent system applies to the electric field.

The MTF of an incoherent system is the convolution of two cylinders, which
resembles a cone. The cutoff frequency of an incoherent system is twice that of a
coherent system 2kM.
An incoherent MTF can resemble a triangle, where the cutoff frequency is
the conical frequency kC. The conical estimate of an Airy transformation [Eq.
(19.34)] defines an effective NAM from the conical frequency as

kC λ
ΝA M = . (5.48)
1.6
50 Chapter 5

Equation (5.48) is extremely useful in application to MTF data for an off-the-

shelf product. Quite often the functional NA is far smaller than the state NA.
Chapter 19 provides more information on the triangle transfer function.

5.14 Gaussian Estimate of Airy Pattern

The peak irradiance of a Gaussian estimate of an Airy pattern is equal to the peak
irradiance of an Airy pattern when

2 ΝAG = ΝA M . (5.49)

A Gaussian estimate of an Airy pattern is an effective tool for calculation of

important parameters, such as resolution, contrast, and depth of focus. Figure 5.9
displays the point spreads of an Airy pattern and its Gaussian estimate. The
standard deviation of a Gaussian estimate from the Airy point spread is < 0.01.
Figure 5.10 displays the MTFs of an Airy pattern and its Gaussian estimate. The
standard deviation of a Gaussian estimate from the Airy MTF is < 0.05.

Figure 5.10 Modulation transfer functions of an Airy pattern and a Gaussian estimate.
Standard deviation of a Gaussian estimate from Airy MTF is < 0.05 in magnitude.
Basic Physical Optics 51

5.15 Scatter
Surface texture may create a random phase delay during reflection or
transmission. The scatter (or scatterance) applies to the irradiance as

2
S= s =φ 2 , (5.50)

where φ is the rms of the phase delay. The scatterance is frequently called the
total integrated scatter (TIS).
At an air-to-glass interface with refractive indices of 1.0 and 1.5, the
scatterance of transmission is

2
 2πσ 
S T = 0.25   , (5.51)
 λ 

where σ is the rms of surface texture and λ is the wavelength in air. The
scatterance of external reflection is

2
 2πσ 
S ER = 4   = 16 S T . (5.52)
 λ 

The scatterance of internal transmission is

2
 2πσ 
S IR = 9   = 36 S T . (5.53)
 λ 

Obviously, the effect of surface texture is much greater in reflection than in

transmission. At 588 nm in wavelength, a 1% scatterance is created by different
surface textures: 18 nm for transmission, 5 nm for external reflection, and 3 nm
for internal reflection.
Chapter 19 provides more information on surface texture and scatter.

5.16 Interference Filters

An interference filter comprises a dielectric stack of high- and low-index
materials. The spectral transmittance or reflectance is transformed by tilt as

λ′ = λ cos θ . (5.54)

At normal incidence, the spatial frequency of reflection is half that of the filter
lattice.
Chapter 6
Fluorescence
6.1 Absorption Parameters
The absorption coefficient α applies to irradiance (power per area), while the
extinction coefficient κ applies to electric field (force per charge). The
transmittance T applies to the irradiance:

IT
T= = exp ( −αz ) , (6.1)
I0

where I0 is the incident irradiance and IT is the transmitted irradiance.

The absorption coefficient α is derived from the extinction coefficient κ as

4π
α= κ. (6.2)
λ

The absorption coefficient α is derived from the molar extinction coefficient αρ

as

α = ρ αρ , (6.3)

where ρ is the molar concentration. The units of molar concentration are moles
per liter (mol/L). The units of molar extinction coefficient are concentration per
distance [moles per liter per centimeter (mol·L–1cm–1)]. It is important to
correctly define these similar metrics: absorption coefficient α, extinction
coefficient κ, and molar extinction coefficient αρ.
The absorbance for a small α is defined as

A = 1 − T = αz . (6.4)

53
54 Chapter 6

During an actual measurement, reflections of the vessel must be considered:

T = (1 − R ) exp ( −αz ) . (6.5)

The effect of reflections is eliminated by comparison to a known standard:

T2 exp ( −α 2 z )
= = exp ( α1 − α 2 ) z . (6.6)
T1 exp ( −α1 z )

 T2 
1 −  = ( α 2 − α1 ) z . (6.7)
 T1 

6.2 Electron States

The quantum mechanics of absorption and emission are effectively displayed by
the electron states of the hydrogen atom.7 An emission or absorption of a photon
requires a change in symmetry as an electron transfers from one state to another.
The orbital quantum number j must change by ±1 for this requirement. The
orbital quantum number also describes the angular momentum of electron orbit
about the nucleus.
All particles have an intrinsic angular momentum known as spin. The spin
quantum number may be ±½ . The spin of two electrons with identical orbitals
must be antiparallel. One electron has “spin up” while the other has “spin down.”
The antiparallel configuration is called a singlet state because there is only one
possible combination of spin components: spin up and spin down. The
antiparallel configuration creates a total spin quantum number of 0. During
transition to another state, the electron spins remain antiparallel. However, after
promotion to another state, an electron spin can be flipped by an external action
(EA). The electrons now display a total spin quantum number of 1. The electron
spin may now add to the orbital momentum in three different orientations:
parallel, perpendicular, and antiparallel. This configuration is a triplet state. The
triplet state has a lower energy than the singlet state because the electrons are
spinning in the same direction. The triplet state cannot return to the ground
singlet state without a spin flip by an EA. Entering a triplet state may
permanently stop fluorescence in a process known as “photobleaching.”
The angular momentum of an electron state is often described in
spectroscopic notation as follows:

N 2 s +1 L j , (6.8)

where N is the principal quantum number, s is the total spin quantum number, L
indicates the orbital angular momentum by letter (S, P, D, or F), and j is the total
angular momentum quantum number l + s. The multiplicity of the state is
Fluorescence 55

expressed by 2s + 1. Thus, a singlet state of s = 0 has a multiplicity of 1, where j

equals the orbital quantum number l, and a triplet state of s = 1 has a multiplicity
of 3, wherein j equals l + 1 (parallel), l (perpendicular) and l – 1 (antiparallel).

6.3 Energy Diagrams

A Jablonski energy diagram of a typical fluorophore is displayed in Fig. 6.1. The
states are named singlet or triplet with a subscript for excitation level. The
ground state is a singlet state S0. An electron may be promoted to one of two
excited singlet states S1 or S2. Vibrational states are represented by discrete
energy levels of each state. The lifetime of a fluorescent process is typically
nanoseconds.
An electron energy is frequently described in units of electron volts (eV).
The energy of a photon in electron volts is derived from Planck’s law as

1240
E= eV nm . (6.9)
λ

An EA may convert the excited singlet state S1 into a similar triplet state T1 by
spin flip of an electron. The parallel spins of the triplet state have lower energy
than the antiparallel spins of the singlet state. Consequently, the electron cannot
return to S1 without another EA. The triplet state can return to the singlet ground
state S0 by phosphorescence (P), however, an EA is required for a spin flip
during phosphorescence. Consequently, the lifetime of a phosphorescent process
is typically seconds.

Figure 6.1 Jablonski diagram. S indicates a singlet state with opposing electron spins. T
indicates a triplet state with parallel electron spins. Absorption (A) promotes an electron
between singlet states. Emission (E) occurs at a similar energy to absorption.
Fluorescence (F) indicates an emission at a lower energy than A. An external action (EA)
converts a singlet state into a triplet state. Phosphorescence (P) indicates a slow process,
which is forbidden without another EA.
56 Chapter 6

Figure 6.2 Franck-Condon diagram. Electron energy is plotted versus configuration

coordinate q. S0 indicates a singlet ground state with antiparallel electron spins. S1
indicates the first excited singlet state. Numbers 1, 3, and 8 indicate vibration mode.
Absorption and emission spectra resemble mirror images.

A triplet state might trigger permanent transformation of the system into a

nonfluorescent state. Photobleaching indicates this typical fate of an organic
fluorophore. Photobleaching can limit the total emission of a fluorophore to
<200 photons.
A Franck-Condon diagram displays the effects of configuration on
absorption and emission. In Fig. 6.2, the configuration coordinate q expresses the
shape of the molecule in different vibration modes 1 through 8. A Maxwell-
Boltzmann distribution describes the population of vibration states as

exp ( −εi / kT ) , (6.10)

where εi is the energy of a vibration, k is the Boltzmann constant, and T is the

temperature. The vibrational state spreads the configuration beyond the electronic
state. The spatial overlap of the vibration modes determines the probability of
absorption and emission. During absorption, the overlap of configurations grows
rapidly with energy, while the opposite occurs during emission. Consequently,
the absorption and emission spectra are mirror images of each other.
Figure 6.3 displays the spatial overlaps of configurations in a Franck-Condon
diagram. During absorption in Fig. 6.3(a), the overlap grows rapidly with
increasing absorption energy until the final vibration state 6 matches the spatial
Fluorescence 57

Figure 6.3 Franck-Condon spatial overlap of states. (a) Absorption is dependent on

spatial overlap of a single ground state and numerous excited states. Heavy lines indicate
overlap of excited states with the lowest ground state. (b) Emission is dependent on
spatial overlap of a single excited state and multiple ground states. Heavy lines indicate
overlap of ground states with the lowest excited state.

limit of the initial vibration state 1. This indicates an absorption plot that rises
rapidly from zero with increasing energy. At greater absorption energy than A16,
the overlap decreases by exponential decay. During emission in Fig. 6.3(b), the
overlap grows exponentially with increasing emission energy until the final
vibration state 6 matches the limit of the initial vibration state 1. At greater
emission energy than E16, the overlap of the final state with the initial state
decreases rapidly toward zero. If the vibration modes of the ground and excited
states are similar in amplitude but different in location, then the emission
spectrum is a mirror image of the absorption spectrum. This trend toward mirror
image is common in fluorophores.

6.4 Fluorophores
Carbon is the most common element of a fluorophore. A benzene ring C6H6
displays a hexagonal structure with strong covalent bonds along the hexagonal
plane and weak dipole attractions across the hexagonal plane. Benzene is a
building block for organic molecules. Fluorophores are also called
chromophores.
Indole is a common organic compound within live organisms. The intrinsic
fluorescence of a live organism is frequently due to indole groups. Intrinsic
fluorescence is also called autofluorescence. The structure of indole in Fig. 6.4
comprises a hexagonal benzene ring and a pentagonal pyrrole ring. A pyrrole
ring contains five elements: four CH groups and one NH group. Neither benzene
nor pyrrole is fluorescent;8 however, their combination as indole is fluorescent.
58 Chapter 6

Figure 6.4 Indole structure. Vertices without labels are the carbon sites. Each carbon
requires four bonds. (a) HC bonds are displayed. (b) A carbon vertex with three bonds
must have a phantom hydrogen (H).

There is a permanent dipole moment from hexagon to pentagon due to the

additional proton of the nitrogen element. A polar solvent can align to the
molecule with significant effect on the electron states, and also shift or distort
emission spectra.
Intrinsic DNA base elements are adenine, thymine, guanine, and cytosine.
They display essentially no fluorescence.9 In Fig. 6.5, adenine is bound to
thymine. The weak connections are an attractive force between dipoles. Although
this is a “bound state,” it is not considered a “bond” due to lack of electron
sharing. Adenine absorbs in the UV, but displays insignificant fluorescence.
Adenine cannot be used to track its complementary base thymine.

Figure 6.5 Adenine and thymine structure. Adenine and thymine are not fluorescent.
Dotted lines indicate weak connections between base pairs of RNA. The H* represent
sites that may also be carbon elements of the RNA backbone.
Fluorescence 59

Figure 6.6 2-AP and thymine structure. 2-AP is a fluorescent analog of adenine. Dotted
lines indicate weak connections between base pairs of RNA. H* represent sites that may
also be carbon elements of the RNA backbone.

An analog of adenine is required as a probe for thymine. As shown in Fig.

6.6, 2-aminopurine (2-AP) is very similar to adenine. An NH2 group switched
location with a CH group. The adenine analog 2-AP absorbs in the UV spectrum
and emits fluorescence in the visible spectrum.
Tryptophan is a common protein that contains a fluorescent indole group.
The molecular structure of tryptophan is displayed Fig. 6.7, while its absorption
and emission spectra are displayed in Fig. 6.8. Tryptophan is a common source of
background fluorescence. Fortunately, its emission is near zero at 500 nm and
beyond.

Figure 6.7 Tryptophan structure. An indole group provides fluorescence.

60 Chapter 6

Figure 6.8 Tryptophan spectra. Absorption (solid) and emission (dotted) are nearly mirror
images.

Fluorescein is an extrinsic compound to a living organism. The structure of

fluorescein is displayed in Fig. 6.9, where the phantom hydrogen atoms are not
displayed. The absorption and emission spectra are displayed in Fig. 6.10.
Fluorescein may be “functionalized” by modification of the benzene ring. As
shown in Fig. 6.11, a single phantom hydrogen at the bottom benzene ring is
replaced with an isothiocyanate group (–N=C=S). The resulting fluorophore is
fluorescein isothiocyanate (FITC). The isothiocyanate group is an electrophile
that easily bonds to exposed electron pairs. This enables FTIC to bond to other
molecules as a fluorescent probe or marker. Other functional groups may be
attached, such as antibodies, which may target attached specific proteins.
Green fluorescent protein (GFP) occurs both naturally and synthetically.9 A
natural form of GFP Class 1 exhibits extremely high quantum efficiency: 79% of
absorbed photons are converted to emitted photons at 504 nm in peak
wavelength. As seen in Fig. 6.12, GFP contains a hexagonal and pentagonal
group. Photon absorption ejects a proton from an OH group. This leaves a
complex negative anion with an emission in the visible. Synthetic variations of
GFP display other colors: yellow 518–529 nm, cyan 490 nm, and blue 440–448
nm. The chromophore of GFP is surrounded by a barrel that is formed by ribbons
of protein.
Fluorescence 61

Figure 6.9 Fluorescein structure. A synthetic green fluorescent molecule. A carbon vertex
with three bonds must have a phantom H.

Figure 6.10 Fluorescein spectra. Absorption (solid) and emission (dotted) display
significant overlap. Spectra are mirror images.
62 Chapter 6

Figure 6.11 Fluorescein isothiocyanate (FTIC) structure. FTIC is a functionalized version

of fluorescein. The isothiocyanate group (–N=C=S) bonds easily to exposed electron
pairs.

Figure 6.12 Green fluorescent protein (GFP). Excitation of a molecule triggers ejection of
a proton. Variable node V determines the spatial extent of an exited state and the
wavelength of fluorescence. The excited state of green emission ends at node V. The
excited state of red emission ends at ribbon node R. The chromophore sits within a barrel
formed by a chain of linear proteins in the form of a ribbon.
Chapter 7
Optical Design Metrics
7.1 CAD Tools
There are numerous genres of computer-aided design (CAD) tools. They offer a
variety of metrics in a range from simple illumination calculations to complex
wave propagation. It is important to remember that every tool has its limitations.
Mastery of these tools requires education and practice.
Various CAD tools are demonstrated in Secs. 7.2–7.9 through application to
a doublet with a 40-mm focal length. The doublet functions as a 5X objective in
combination with a 200-mm tube lens. The object field height at 1.2 mm
translates into a 6.0-mm image height, which accommodates the 11.3-μm
diagonal of a 1000 × 1000 array of 8-μm pixels. The projected pixel at the object
is 0.8 μm.
The ensuing metrics employ the object side of the lens as the image. The
effective object distance (or tube length) is infinite. The entrance pupil is 10 mm.
The image NA is 0.0125. The Airy radii are 2.37 μm for the F line (486 nm),
2.86 μm for the d line (588 nm), and 3.20 μm for the C line (656 nm).
Some metrics require sampling of the image. Sampling normally occurs at
2N × 2N, where N is an integer. Errors due to insufficient sampling must be
considered. The shape of the ray-intercept plot should predict any calculations
based on sampling.

7.2 Wavefront Error

The wavefront error is probably the most physically significant metric of a CAD
tool. A convergent wavefront should be spherical. A collimated wavefront should
be flat. A reflecting telescope with 0.05 NA can easily maintain a wavefront error
at λ/10. Unfortunately, such quality is not easily achieved in a refractor at 0.12
NA or higher.
Figure 7.1 displays the wavefront of the 5X doublet. The optical path
difference indicates the optical path length from a perfect sphere around the
image. The optical path length is specified in the wavelengths. The abscissa
indicates the position within the pupil or lens stop. The wavefront error is
independent of the direction of travel through the lens. On reversal of the
direction of travel, the optical path difference indicates the optical path length in
wavelengths from a perfectly flat wave in collimation. The wavefront error is an
effective measure of collimated wavefronts.

63
64 Chapter 7

Figure 7.1 Wavefront error of 5X doublet. A solid line indicates the F line (486 nm). A
short dash indicates the d line (589 nm). A long dash indicates the C line (656 nm).

The wavefront error of Fig. 7.1 reveals several important features. The d line
(588 nm) indicates a wavefront error of 0.1λ, which is considered perfect within
the limits of an Airy pattern. The p4 shape of the F line (486 nm) indicates a
fourth-order spherical aberration in wavefront error. The positive wavefront error
of the margin indicates a marginal convergence greater than the perfect optimum.
The negative wavefront error of the C line (656 nm) indicates axial color, where
the convergence is less than the perfect optimum.
A collimating optic, such as a beam expander or a laser, is normally qualified
by a wavefront error, which may be calculated and/or measured. A CAD tool
may define fabrication and assembly tolerances for an acceptable wavefront
error. A wavefront analyzer may provide empirical data as confirmation of
wavefront quality. The wavefront error is an essential parameter for application
of laser beams.

7.3 Ray-Intercept Plot

The ray-intercept plot (also known as a ray fan) is an essential tool for optical
design. It quantifies the geometric point-spread error as a function of ray position
within the aperture or lens stop. The geometric point-spread error may be
expressed in units of space or angle. The ray intercept is typically specified in
microns (μm).
Optical Design Metrics 65

Originally, the terms sagittal and tangential referred to cylindrical

coordinates. The sagittal vector represented the radius, as in an archer shooting
arrows various directions from the same point. The tangential vector represents
the tangent of the circle to the head of the sagittal vector. The tail of the
tangential vector is located at the head of the sagittal vector, and the tangent
vector is parallel to the circle. In medicine, a sagittal plane splits the body into
left and right portions. in fact, a sagittal plane normally splits a system into two
symmetrical parts; however, this is not so in the geometric optics of 2009.
The orientation of a sagittal plane of geometric optics is opposite from other
conventions. The tangential plane (yz) of geometric optics contains the chief ray
and the optical axis. The tangential plane may also be called the y plane. The
sagittal plane (xz) of geometric optics contains the chief ray and is normal to the
tangential plane. The sagittal plane may also be called the x plane. The sagittal
plane of geometric optics does not split an optical system into two symmetric
parts.
Figure 7.2 displays the ray intercept of the 5X objective. The error is plotted
for two orthogonal planes. The object field height is normally specified within
the tangential plane yz. The tangential error EY is plotted versus the tangential
pupil position PY. The sagittal error EX is plotted versus the sagittal pupil
position PX.

Figure 7.2 Ray-intercept plot of a 5X doublet. The tangential error EY is plotted versus the
tangential pupil PY. The sagittal error EX is plotted versus the sagittal pupil PX. A solid
line indicates the F line (486 nm). A short dash indicates the d line (588 nm). A long dash
indicates the C line (656 nm).
66 Chapter 7

The ray intercept of Fig. 7.2 reveals several important features. The fans are
normally defined for three different positions: the on-axis position at zero field
height, the margin of the field, and 0.7 times the margin of the field, which
represents half the field area.
The on-axis F line (486 nm) indicates a paraxial focus through a flat slope in
the central pupil zone. However, the marginal portion of the F line indicates both
third- and fifth-order spherical aberration. The on-axis d and F lines both display
third-order spherical aberration. The slopes of the F-line plots decrease steadily
with field height. This indicates the defocus of Petzval curvature. The dissimilar
slopes of F-line plots between EY and EX indicate astigmatism. A small
downward curvature in the tangential F-line plots indicates coma. The
astigmatism is larger than the coma.

7.4 Spot Diagram

A spot diagram is an effective representation of optical point spread on a large
screen with color display. It quickly displays the point spread with reference to a
pixel or an Airy disk. A spot diagram is an effective graphic to display to
personnel without expertise in optical design. Ordinarily, the spots of three
wavelengths are displayed as blue, green, and red; however, spot diagrams are
largely absent from this book due to the limitations of monochrome printing.
Figure 7.3 displays a spot diagram of the 5X doublet. A circle displays the
Airy diameter at each field point. The on-axis rays are barely contained within
the Airy disk. The elliptical shape of the off-axis points indicates astigmatism. A
pointed end of an ellipse indicates coma. The bottom header specifies three
important parameters: the Airy radius, the rms radius, and the geometric radius.
The Airy radius indicates a diffraction-limited resolution, the rms radius indicates
an aberration-limited resolution, and the geometric radius indicates the entire
spread of the rays.

7.5 Point-Spread Plot

A CAD tool can assign a uniform or Gaussian profile to the pupil. A point-spread
plot includes the effects of wave propagation. Planar waveforms are summed to
create the point spread. The optical path difference of each surface is
incorporated as phase information in the summary. Sampling arrays determine
the accuracy of the point-spread plot. The accuracy degrades rapidly with phase
delays beyond a wavelength.
Figure 7.4 displays a point spread for the 5X objective. The point spread is
monochromatic, and it has large aberration. It is not a perfect Airy pattern.
Consequently, the central peak is displayed as less than unity. This central peak
specifies the Strehl ratio, which is the ratio of the peak irradiance of point spread
to the corresponding peak of the Airy pattern. The Strehl ratio for the off-axis
Optical Design Metrics 67

Figure 7.3 Spot diagram of a 5X doublet. Wavelength is at the d line. Ray locations are
indicted by ×. The Airy disc is defined by a circle within the field at 1.2 mm.

Figure 7.4 Point spread of a 5X doublet that is monochromatic at d line. The field is at 1.2
mm. The Strehl ratio is 21%. Sampling is at 128 × 128.
68 Chapter 7

Figure 7.5 Encircled-energy plot of a 5X doublet that is monochromatic at the d line.

Sampling is at 128 × 128.

monochromatic point spread is 21%. A point spread is frequently normalized to

display the shape. The scale of a point spread must be scrutinized.

7.6 Encircled-Energy Plot

Encircled energy is derived from the point-spread plot. It is an essential metric
for low-light applications, where the collection of light should be maximized
within a block of four pixels. An encircled-energy plot is far more useful than a
point-spread plot.
Figure 7.5 displays an encircled-energy plot for a 5X doublet. The
monochromatic encircled energy at 4 μm varies greatly by field: 85% on axis,
75% at 0.85 mm, and 54% at 1.20 mm. The diffraction limit indicates 88% at 4
μm. The flat spot on the diffraction-limit plot line indicates an Airy radius at 2.8
μm.

7.7 Modulation Transfer Function

Figure 7.6 displays a modulation transfer function (MTF) for the 5X doublet. A
single field is shown because multiple fields can overlap in a confusing manner.
The conical frequency [further discussed in Eq. (19.34)] at 100 cycles/mm
indicates a central NA of 0.037 at the d line (588 nm) [Eq. (5.29)]. The paraxial
NA is 30% of the central NA.
Optical Design Metrics 69

Figure 7.6 MTF of a 5X doublet, where T is the tangential plane and S is the sagittal
plane. The modulus is the modulation index. Sampling is at 64 × 64 along the plot scale. A
conical approximation (dashed line) indicates a conical frequency at approximately 100
cycles per millimeter.

Some vendors provide MTF data for lenses, while many do not. A conical
approximation may be derived from published data. Ronchi rulings may provide
empirical measurement at spatial frequencies up to 500 line pairs/mm.

7.8 Edge Spread

An edge spread is an effective measurement for optics with large NAs. Spherical
aberration can spread the base of point spread beyond recognition. Observation
of a high-contrast edge can reveal important aberrations that are not apparent
from a point spread.
Figure 7.7 displays the edge spread of the sagittal edge along the tangential
plane for the d line at 1.20 mm. The edge spread indicates 10% at –8.0 μm and
90% at 4.5 μm. The leading edge is much larger than the Airy radius, while the
trailing edge is approximately twice the Airy radius.

7.9 Lens Report

A lens report displays several metrics in one graphic. A report can be easily
pasted into a design log, such as a Microsoft Word file. Figure 7.8 displays a
70 Chapter 7

Figure 7.7 Edge spread of a 5X doublet. The edge is along the sagittal plane (xz), and
edge spread is along the tangential plane (yz). Sampling is at 128 × 128.

report for the 5X doublet. The line spread and encircled energy are
monochromatic at the d line (588 nm). The right end of the flat spot of the
monochromatic encircled energy is a fair estimate of the Airy radius. Due to the
space constraints of this book, only one field point is active in each quadrant.
The current Fig. 7.8 is described by Prescription 7.8 in the Appendix. A
boxed cell indicates a variable during operation. (A corresponding prescription to
every lens report is located in the Appendix. The prescription numbers correlate
to the figure numbers.)

7.10 Relative Illumination

A relative illumination plot displays the effects of reduction of NA with field
angle. A dependency on cos3θ is derived from the solid angle of collection. The
addition of a Lambertian radiance creates a dependency on cos4θ. Numerous lens
types (such as a double Gauss) reduce NA at the field margin to control
astigmatism. Consequently, the effective f/# of a double Gauss inflates rapidly
with field angle, and f/# inflation reduces brightness.

7.11 Surface-Form Error

An interferometer may qualify a surface as round or flat within a fraction of a
wavelength. A surface-form error indicates a deviation from a reference surface
Optical Design Metrics 71

Figure 7.8 Lens report of a 5X doublet. Top left: layout. Top right: line spread. Bottom left:
ray intercept. Bottom right: encircled energy. A single point at the field margin (1.2 mm) is
employed for each quadrant.

by wavelengths or fringes. Optical elements are fabricated within tolerances for

surface-form error at a test wavelength.
The fringes must be converted to optical path lengths to calculate image
quality. An optical path length of 0.1λ is considered near perfect in practice. In
transmission, the number of test fringes NF converts to optical path length in
cycles as

Δn N F  λT 
ΛT =   cyc , (7.1)
2  λA 

where Δn is the change in refractive index, λT is the test wavelength, and λA is the
application wavelength. In reflection, the test fringes convert optical path length
as
 λT 
ΛR = n NF   cyc , (7.2)
 λA 

where n is the refractive index. Obviously, the effective surface-form error is

much greater in reflection than in transmission.
72 Chapter 7

7.12 Manufacturing Standards

The International Standards Organization (ISO) provides standards for the
specification of optical systems. The ISO standards are mainly comprised of the
old standards of two organizations, the American National Standards Institute
(ANSI) and Deutsche Industrie Normen (DIN). The title of the ISO standard is
ISO 10110–Preparation of drawings for optical elements and systems.
Volume defects of a material of an optical element are specified within three
categories: 0, birefringence; 1, bubbles and inclusions; and 2, inhomogeneity and
striae. Birefringence indicates the anisotropy of the refractive index. Bubbles and
inclusions specify the acceptable limits for voids and extrinsic solids. Inclusions
may be crystalline blocks that may gouge the surface during polishing.
Inhomogeneity expresses the variation of the refractive index. Striae frequently
occur in the shape of filaments or cords.
Surface qualities are specified in several categories: λ, surface coating; 3,
surface form; 4, centering; 5, surface imperfections; and 6, laser irradiation
damage threshold. The surface coating is typically a thin-film coating. The
surface form specifies the deviation of the surface from perfection in fringes. The
centering of the surface specifies either a tilt or a lateral displacement of the
surface from the optical axis. The surface imperfections specify the acceptable
limits of scratches, digs, and chips.
Surface texture defines the fine structure of the material. It may be expressed
as an rms of the texture width, which is similar to the depth, or as a power
spectrum density, wherein the rms amplitude per spectral range is cited.
Chapter 8
Image Contrast
8.1 Radiometry
Radiometry generally applies to the distribution of quanta over space, angle, and
time. Specifically, radiometry applies to energy, whereas photometry applies to
luminous energy, measured in Talbots. In the current discussion, radiometry
applies to both. The terminology is defined in Table 8.1.
Within the field of optics, flux is generally accepted as flow. However, in
physics, flux indicates the strength of a field per area. Sensors for commercial
applications are frequently defined by luminance in lux, whereas sensors for
scientific applications are normally specified in photons or electrons. Quantum
efficiency indicates the conversion of photons to electrons.

8.2 Expression of Contrast

The contrast of an object is defined as

Φ OP − Φ BP
C= , (8.1)
Φ BP + Φ D

where ΦOP is the flux of the object pixel, ΦBP is the flux of the background pixel,
and ΦD is the flux of the detector. The range of contrast is from –1 to +∞. A
negative contrast indicates a dark object within a bright background. A positive
contrast indicates an object that is brighter than the background. Quite often both
objects are rather dark.
The flux of the detector has numerous components, as described in Chapter
15. The two most common are dark current and read noise. They are often stated
as electron currents. The flux of the detector may be expressed as

Φ DC + Φ RN
ΦD = , (8.2)
ηQ

where ΦDC is the mean flux of the dark current in electrons per pixel, ΦRN is the
mean flux of the read noise in electrons per pixel, and ηQ is the quantum

73
74 Chapter 8

Table 8.1 Terms and symbols of radiometry.

Name Symbol Description Units
Quanta Q Quanta photon
Luminous energy joule (J)
Radiant energy Talbot (T)

Fluence F Quanta per area quanta per area

joules per area
Talbot per pixel

Flux Φ Quanta per second quanta per second

Optical power flow
Radiant flux watts (W)
Luminous flux lumens (lm)

Directance D Quanta per angle quanta per steradian

Radiant intensity watts per steradian (W·sr–1)
Luminous intensity lumens per steradian (lm·sr–1)
candela (cd)

Exitance E Exiting flow per area flow per area

flow per pixel
watts per area (W·cm–2)
lumens per area (lm·m–2)
lux (lx)

Incidence I Incident flow per area flow per area, flow per pixel
Irradiance watts per area (W·cm–2)
Illuminance lumens per area (lm·m–2), lux (lx)

Radiance L Flow per area per angle flow per pixel per hemisphere
Radiance watts per area per angle (W·cm–2sr–1)
Luminance candelas per area (cd·m–2)

efficiency of conversion of photons into electrons by the detector. The dark

current is normally thermal promotion of carriers into the conduction band. The
read noise is created during conversion of electrons into voltage. There are
numerous other sources of noise.
Visibility is defined as
Φ max − Φ min
V= . (8.3)
Φ max + Φ min

Its range is from 0 to 1. It is also the modulation index of the MTF.

Background radiance must be carefully considered. It may originate from the
foreground or background of the object, or even from the optical path. The cover
strata might emit fluorescence.
Image Contrast 75

Detector noise must also be considered. Dark current, read noise, and dark
noise are three common metrics of electronic detectors of light. The object signal
may overcome the read noise through integration. However, the object signal
cannot overcome the dark current by integration. The dark current and read noise
should be extracted from the dark noise of a detector specification. The shot
noise of the dark current and the object signal should also be considered.

8.3 Shot Noise

A shot is a fixed quanta, such as a lead sphere, a photon, or an electron. It is not
possible to detect a portion of a shot. If the expected mean photon count is 2
photons, then the actual count can be 0, 1, 2, 3, or more. The probability of a
count is described as a Poisson distribution:

μ n exp ( −μ )
P ( n) = , (8.4)
n!

where μ is the mean and n is the count. The Poisson distribution for a mean dark
current of 10 electrons is displayed in Fig. 8.1, where the count ranges from 5 to
15. The Poisson distribution for a mean dark current of 100 electrons is displayed
in Fig. 8.2, where the count ranges from 80 to 120.
As the mean count increases, the standard deviation of the shot count
approaches the square root of the count. The flux of the shot noise may be
expressed in arbitrary units of flux as follows:

Φ SN = qS Φ as Q > 1000 , (8.5)

where qS is the discrete quantum of the shot, Φ is the mean flux, and Q is the
expected quantum. The contrast with shot noise grows by the square root of the
expected quantum Q:

Figure 8.1 Poisson distribution for a mean quanta of 10.

76 Chapter 8

Figure 8.2 Poisson distribution for a mean quanta of 100.

Φ
= Φ SN = qS Q as Q > 1000 . (8.6)
Φ SN

The total flux of the shot noise of the system is fairly estimated by a sum of the
squares:

Φ 2S = qS Φ O + qS Φ B + qS Φ D . (8.7)

Ideally, the shot noise of the object should be the limiting noise of the system.
However, this is not always possible.
Chapter 19 provides more detail on shot noise.

8.4 Emittance Patterns

There are two common emittance patterns in microscopy: isotropic and
Lambertian. An isotropic emittance pattern is uniformly distributed over a
hemisphere. The radiance of an isotropic emittance is defined as

E
LI = , (8.8)
2π

where E is the emittance. A radiance of a Lambertian object is dependent on

cosθ:

E
LL = cos θ . (8.9)
π

A Lambertian emittance is created by either a diffuse reflectance10 or a specular

transmission of isotropic emittance.11
Image Contrast 77

8.5 Angular Collection Efficiency

The collection efficiency for a sphere of isotropic emission is

ΩC θ
ηSIE = = sin 2 n , (8.10)
4π 2

where θn is an angle within the refractive medium of emission. In small-angle

format, the marginal NA simplifies this expression:

ΝA 2M
ηSIE = . (8.11)
4n 2

The collection efficiency for a hemisphere of isotropic emission is

ΩC θ
ηHIE = = 2 sin 2 n . (8.12)
2π 2

In small-angle format, this becomes

ΝA 2M
ηHIE = = 2ηSIE . (8.13)
2n 2

The collection efficiency for a hemisphere of a Lambertian emission is derived

by integration of a normalized Lambertian radiance as follows:

θn  cos θ 
ηHLE =    ( 2π sin θ d θ ) = sin θn ,
2
(8.14)
0
 π 

where the first group in parentheses is the normalized Lambertian radiance, and
the second group in parentheses is the differential solid angle.
At small angles, the collection efficiency for a hemisphere of a Lambertian
emission is

ΝA 2M
ηHLE = = 2ηHIE . (8.15)
n2

The Lambertian emittance is more concentrated at small angles than an isotropic

emission.
78 Chapter 8

Figure 8.3 Spatial collection efficiency of point spreads on a pixel array.

8.6 Spatial Collection Efficiency

A point spread does not normally fit within a single pixel. A collection efficiency
of the pixel ηP is defined by the portion of a point spread that overlaps a pixel.
The maximum reliable collection by a pixel is 25%.
Figure 8.3 displays several point spreads within a pixel array. A point spread
may fit within a pixel, as indicated by the 100% point spread. However, it is
likely to split by 2 to 4 pixels. Consequently, the minimum reliable collection by
a pixel is 25% for point spreads of a radius at pixel width or less. Reduction of
the point spread does not increase the minimum reliable collection of pixel.
An optimum point spread fills four pixels. This yields a maximum signal
strength of a point source object. They may also be combined into a super pixel
of 2 × 2 pixels.
Image-processing algorithms benefit from contiguous full pixels. A point
spread may fit within a 3 × 3 super pixel. This yields a consistent pixel collection
efficiency of 14%, wherein at least one pixel is filled completely by the point
spread. More pixels are definitely better for image processing; however, more
pixels can yield smaller object flux per pixel while detector flux per pixel
remains the same. The following expressions for contrast indicate the detriments
of oversampling. The developer of an algorithm must consider the decreased
contrast with background and noise that results from increased sampling of the
object.
Image Contrast 79

8.7 Full-Pixel Contrast

Étendue describes the space-angle product for rays. The term is derived from the
French word for occupancy. In a geometric optical system, the etendue of a pixel
is conserved from sensor to projected pixel:10

A0 Ω0 = A1Ω1 =  = An Ωn , (8.16)

where A is the area of an image and Ω is the solid angle of the lens at the center
of the image. The contrast of a full-pixel object10 is expressed in large-angle
format as follows:
LO − LB
CFPO = −1
. (8.17)
LB + Φ D APP ΩO−1

The terms of pixel-based radiometry are defined Table 8.2. They are largely
based on the projection of the pixel on the object.

Table 8.2 Terminology of pixel-based radiometry.

Symbol Name
AO Area of the object
APP Area of the projected pixel at the object
dO Dimension of the object
dP Dimension of the pixel at the detector
dPP Dimension of the projected pixel at the object
EB Emittance of the background
EO Emittance of the object
ΦD Flux of the detector
ΦO Flux of the object
ηC Spherical collection efficiency of the central aperture
ηD Quantum efficiency of the detector
ηM Hemispherical collection efficiency of the marginal aperture
ηP Efficiency of a pixel
ηSR Efficiency of the Strehl ratio
LB Radiance of the background
LO Radiance of the object
NAM Marginal numerical aperture at the object
NAC Central numerical aperture at the object
n Refractive index at the object
λ Spatial wavelength within air
θM Marginal angle of the lens stop at the object
θC Central angle of the lens stop at the object
ΩO Angle of collection at the object
80 Chapter 8

The small-angle format of contrast for a full-pixel object is

LO − LB
CFPO = . (8.18)
LB + Φ D ( M 2 d P−2 )( π−1ΝA −M2 n 2 )

The full-pixel contrast displays several dependences of interest. The effect of the
detector is enhanced by the magnification and refractive index of the object. The
effect of the detector is reduced by the dimension of the pixel and the marginal
NA.

8.8 Subpixel Contrast

The contrast of a distant subpixel object is proportional to the square of the object
dimension and the square of the resolution.10 Figure 8.4 displays the
redistribution of subpixel radiance within a full pixel. The large-angle format of
subpixel contrast is adapted to a microscope as follows:

CSPO =
( LO − LB ) AO APP−1 =
AO
CFPO .
−1 −1
(8.19)
LB + Φ D A Ω PP O APP

The small-angle format is

CSPO =
( LO − LB ) dO2 . (8.20)
LB ( d P2 M −2 ) + Φ D ( π−1 ΝA −M2 n 2 )

Figure 8.4 Redistribution of the radiance of a subpixel object. Contributing signals include
the object (O), background (back), path, and equivalent noise of detector (END).
Image Contrast 81

The subpixel contrast displays several dependences of interest. The effect of

the background radiance is scaled by the area of the projected pixel d P2 M −2 .
Consequently, the effect of the background radiance is enhanced by the
dimension of the pixel and reduced by the magnification. The effect of the
detector is reduced by the marginal NA and enhanced by the refractive index of
the object.
The pixel collection efficiency in the above relationships apply to the
geometric boundaries of a pixel. However, an optical system is rarely described
by the geometric conditions. A geometric point expands into a spot. The radius of
point spread normally equals a pixel width or greater. The diffraction-limited
spot is specified by an Airy radius or a Gaussian radius, whereas the aberration-
limited spot is specified by an rms radius or a geometric radius.

8.9 Point-Source Contrast

The contrast of a point-source object is a special case for a subpixel object. The
geometric point source is transformed into a spot. An encircled-energy plot may
calculate the pixel collection from the optical prescription. An MTF
measurement may yield a central NA and a Gaussian NA.
The geometric point-source contrast is

Φ O ηP ηC
CGPS = . (8.21)
LB APP ΩO + Φ D

The small-angle format is

Φ O ηP ( π ΝA C2 n −2 )
CGPS = . (8.22)
LB ( d P2 M −2 )( π ΝA 2M n −2 ) + Φ D

The geometric point-source contrast displays several dependences of interest.

The effect of the background radiance is scaled by the area of the projected pixel.
Consequently, the effect of the background radiance is enhanced by the
dimension of the pixel and reduced by the magnification. The effect of the
detector is reduced by the marginal NA and enhanced by the refractive index of
the object.
An increased sampling of the point spread decreases the pixel collection
efficiency. However, the above relationship indicates steady contributions from
the background and the detector while the object signal is reduced by increased
sampling. The developer of an algorithm must consider the reduction of object
contrast when increasing the sampling of point spread.
82 Chapter 8

8.10 Full-Pixel Airy Contrast

A special case is defined by the diffraction-limited emission of a fluorophore.
The Airy pattern at the object is defined by the central NA. The spatial Airy
pattern does not change between air and the immersion medium. However, the
angle of collection does change between air and the immersion medium. The
spherical collection efficiency (SCE) of a fluorophore is defined by the central
NA within the refractive medium. The background fluorescence is isotropic
within the immersion medium. The hemispherical collection efficiency (HCE) of
the background is based on the marginal NA. These conditions define a large-
angle format of the contrast for an Airy pattern as follows, with the help of Eqs.
(5.20), (8.10), and (8.12):

CFPA =
EO ηC
=
( Φ O π ΝA C2 λ −2 )  n −2 sin 2 (θC / 2) 
. (8.23)
−1
EB ηM + Φ D APP EB  2n −2 sin 2 (θ M / 2)  + Φ D d P−2 M 2

The small-angle format is

CFPA =
( Φ π ΝA λ )( ΝA / 4n ) =
O
2
C
−2 2
C Φ ( π ΝA λ n )
2
O
4
C
−2 −2

. (8.24)
E ( ΝA / 2n ) + Φ d M
B
2
M
2
E ( 2ΝA n ) + Φ ( 4d
D
−2
P
2
B
2
M
−2
D
−2
P M2)

The full-pixel Airy contrast displays several dependences of interest. The effect
of the detector increases with magnification. The object signal grows faster than
the background until NAM exceeds NAC. However, as NAM exceeds NAC, the
object signal remains constant while the background signal grows rapidly with
NAM. Ideally, NAM should not exceed NAC. Also, a higher sampling of the Airy
pattern through a larger magnification degrades contrast of the object with the
detector.
The marginal and central NAs are related by the efficiency of the Strehl ratio
as follows:

ηSR ΝA 2M = ΝA C2 . (8.25)

The full-pixel Airy contrast is expressed in terms of the Strehl ratio as follows:

Φ O ( η2SR π ΝA 4M λ −2 n −2 ) Φ O ( η2SR πλ −2 ΝA 2M 2−1 )

CFPA = = . (8.26)
EB ( 2 ΝA 2M n −2 ) + Φ D ( 4d P−2 M 2 ) EB + Φ D ( 2d P−2 M 2 ΝA −M2 n 2 )

The full-pixel Airy contrast increases rapidly with marginal NA until the
Strehl ratio starts dropping below unity.
A developer of an algorithm must consider the detrimental effects of
increased sampling on the Airy pattern. The required magnification increases the
Image Contrast 83

effect of detector noise. Also, the increased marginal NA of the larger

magnification does not collect more object signal, whereas it does collect more
background. Thus, an increased sampling reduces contrast when the Strehl ratio
is reduced.
Chapter 9
Microlens Formats
Most camera lenses are designed for a range of 500 mm to ∞. This range
accommodates both a head shot and a landscape. A typical camera lens for a 35-
mm-film format is 52 mm in focal length. According to Eq. (3.7), a
magnification of 0.1 requires an object distance of 11f, or 550 mm. This may also
be considered a 10X reduction. The performance of a typical camera lens
degrades rapidly beyond a 0 to 10X reduction range.
The terms macrolens and microlens are interchangeable. In photography, a
macrolens converts a small object into a large image within a print. In
microscopy, a microlens looks at a small object. They both operate within a
reduction range of 10XR to 1X. A magnification range of 1X to 10X is achieved
by simply reversing the lens. Nikon frequently employs microlens as the name
for a lens with near-unity magnification.

9.1 10XR Double Gauss

Figure 9.1 displays a 10XR double-Gauss lens for application to a film format
with a 35-mm diagonal. The lens, as described in Smith,12 indicates a large
spherical aberration and a decreasing relative illumination with field height. The
decreased relative illumination indicates f/# inflation with field angle. Rays of the
off-axis field must be clipped to manage astigmatism. Inflation of f/# with field
angle is a characteristic of a double-Gauss lens. The actual field is adapted to a
sensor with a 12-mm diagonal, or 6.0-mm field height, or 6.9-deg field angle.
The f/# inflation at 6.9 deg is < 1%.
The ray-intercept plot indicates enormous aberration at a 6.0-mm field
height. The tangential error indicates 154 μm of spherical aberration. The sagittal
error indicates 51 μm of spherical aberration. The edge spread is nearly 40 μm.
The encircled energy at 10 μm is much less than 30%. The flat spot of the
ensquared plot occurs at much less than 5 μm. The diffraction limit might be
suitable for a 10-μm pixel, while the aberration limit is certainly not.
The low f/# indicates a high brightness; however, the aberrations indicate
poor contrast. A full-pixel object at > 40 μm in width should display a brightness
in accord with the f/#. However, edges will spread to > 20 μm wide. A 40-μm-
wide stripe transforms into a round peak at 80 μm in width. A subpixel object
will spread into a mound of half height at 40 μm in diameter.

85
86 Chapter 9

Figure 9.1 10XR double-Gauss F/1.4. Focal length is 52 mm. Lens f/# is 1.4. Image space
f/# is 1.5. Image NA is 0.31. Edge spread: 512 × 512 sampling. Encircled energy:
512 × 512 sampling. Prescription by G. H. Smith (2009), Lens 23.3.12

9.2 10XR Microlens

Figure 9.2 displays a microlens at 10XR13 for application to an image sensor with
a 12-mm diagonal of 6.0-mm image field height. The sagittal plane indicates very
little spherical aberration, as the plots are nearly linear. The C, d, and F lines are
not well corrected to each other. The tangential plane indicates a small amount of
coma in the negative portion of the pupil. Overall, this lens functions quite well
at 10XR. The image-space f/# is 3.2.
The microlens at 10XR functions effectively for charge-coupled device
(CCD) sensors. The edge spread is approximately 10 μm, which is on the order
of most CCD pixels but not CMOS pixels. The flat of the encircled plot indicates
a diffraction limit of approximately 2.2 μm in radius. The aberration limit is
approximately 7 μm, which is again appropriate for most CCD pixels. At 10XR,
this is a terrific lens for CCD sensors with 8- to 10-μm pixels. However, there is
significant axial color as indicted by the broad line spread.
Microlens Formats 87

Figure 9.2 10XR microlens F/2.9. Focal length is 100 mm. Image space f/# is 3.2. Image
NA is 0.15. Edge spread: 64 × 64 sampling. Encircled energy: edge: 64 × 64 sampling.
Prescription by Matsui.13 Flat plate added as a filter in object space.

9.3 2XR Microlens

Figure 9.3 displays a microlens at 2XR for application to an image sensor with a
12-mm diagonal of 6.0-mm image field height. The sagittal plane indicates both
third- and fifth-order spherical aberration. The C, d, and F lines are not well
corrected to each other. The tangential plane indicates a small amount of lateral
color for the F line (486 nm), as it shifts upward by 5 μm. The tangential plane
indicates a small amount of astigmatism for the F line, as it slopes upward. The
image-space f/# is 4.3.
The microlens at 2XR functions adequately for CCD sensors. The edge
spread is approximately 16 μm, which is nearly 2 pixels of a CCD pixel. The flat
of the encircled plot indicates a diffraction limit of approximately 3.0 μm in
radius. The aberration limit is approximately 14 μm, which is nearly 2 pixels of a
CCD pixel. At 2XR, this lens is limited by aberration beyond a CCD pixel size. It
is marginally optimized for 10 μm without manufacturing considerations. A
tolerance budget during manufacture would indicate larger errors. There is
significant axial color as indicted by the broad line spread.
88 Chapter 9

Figure 9.3 2XR microlens F/2.9. Focal length is 100 mm. Image space f/# is 4.3. Object
NA is 0.058. Image NA is 0.116. Edge spread: d line, image field 6.0 mm, 64 × 64
sampling. Encircled energy: edge: d line, image field 6.0 mm, 64 × 64 sampling.
Prescription by Matsui.13 Flat plate added as a filter in object space.

9.4 1X Microlens
Figure 9.4 displays a microlens at 1X for application to an image sensor with a
12-mm diagonal of 6.0-mm image field height. The sagittal plane indicates both
third- and fifth-order spherical aberrations, where the fifth-order aberration is
larger than the third. The C, d, and F lines are not well corrected to each other.
The tangential plane indicates a small amount of lateral color for the F line and a
small amount of coma for all wavelengths, as the plots all curve downward. The
image-space f/# is 5.5.
The microlens at 1X functions poorly for CCD sensors. It has similar
problems as the microlens at 2XR but with larger magnitude. The edge spread is
> 20 μm. The flat of the encircled plot indicates a diffraction limit of
approximately 4.0 μm in radius. The aberration limit is > 20 μm. The microlens
at 1X aberration beyond a CCD pixel size is not optimized for a 10-μm pixel. A
tolerance budget in manufacture would indicate larger errors.
Microlens Formats 89

Figure 9.4 1X microlens F/2.9. Focal length is 100 mm. Image space f/# is 5.5. Object NA
is 0.089. Image NA is 0.089. Edge spread: 64 × 64 sampling. Encircled energy: edge: 64 ×
64 sampling. Prescription by Matsui.13 Flat plate was added as a filter in object space.

9.5 2XR Telecentric Spectroscopy Lens

Figure 9.5 displays a telecentric spectroscopy lens (TSL)14 at 2XR for application
to an image sensor with a 12-mm diagonal of 6.0-mm image field height. The
sagittal plane indicates no spherical aberration at the d line. The reduced
spectrum ends at the Argon line (514 nm), which displays a small third-order
spherical aberration. The C line (656 nm) displays an 8-μm third-order spherical
aberration. The argon, d, and C lines are corrected within a small spherical
aberration. The spherochromatism is well balanced for the three wavelengths.
The TSL at 2XR is not an achromat, but it might be called a semiapochromat
because it is nearly corrected for three wavelengths. The tangential plane
indicates a small amount of coma at 8 μm in the positive portion of the pupil. The
image-space f/# is 4.0.
The TSL at 2XR functions effectively for CCD sensors. The edge spread
rises sharply over 4 μm for ease of edge of detection by the image-processing
algorithm. The total width of the edge spread is about 10 μm, which is nearly 1
pixel. The flat of the encircled plot indicates a diffraction limit of approximately
3.0 μm in radius. The aberration limit is < 10 μm for most of the rays. At 2XR
this lens is limited by aberration at a typical CCD pixel size of 10 μm. It is
optimized for 10 μm with consideration of manufacture. A reasonable tolerance
budget of 0.1-mm decenter in manufacture does not indicate larger errors.
90 Chapter 9

Figure 9.5 2XR telecentric spectroscopy lens. Wavelengths: Ar (argon 514 nm), d line
(588 nm), and C line (656 nm). Focal length is 193 mm. Image space f/# is 4.0. Object NA
is 0.063. Image NA is 0.126. Edge spread: d line, image field 6.0 mm, 64 × 64 sampling.
Encircled energy: edge: d line, image field 6.0 mm, 64 × 64 sampling. Prescription patent
pending by Seward.14

Figure 9.6 displays the two groups of the TSL. Group 1 corrects the primary
aberrations, which are largely dependent on NA. The aspheric surface of group 1
manages the third- and fifth-order spherical aberration. The doublet manages the
axial color and coma. Group 2 corrects the secondary aberrations of field angle.
The aspheric surface of group 2 manages astigmatism and field curvature. The
concave surface also manages astigmatism and field curvature. A lanthanum
krone (LaK) of the aspheric lens of group 2 manages lateral color. The flange of
the asphere of group 2 provides the precise axial location of group 2. The chief
rays are telecentric, which is beneficial for sensors with a microlens over each
pixel. This lens is highly optimized for spectroscopy. A filter is placed in the
infinity correction zone of group 1. Consequently, filters can be swapped without
disruption of focus. The reduced spectrum of 514–656 nm eliminates chromatic
aberrations of the blue spectrum, which are frequently not active in spectroscopy.
Microlens Formats 91

Figure 9.6 Lens groups of 2XR telecentric spectroscopy lens. Aspheric surface (ASP),
filter of group 1 (F), flange of group 2 (F), chief ray (CR), cover glass (C), and sensor (S).

The object space defines a 0.063 NA. Application of Eq. (1.8) to the d line
yields a Gaussian depth of 94 μm. A 0.19-mm depth of focus is reasonable for
consistent operation. A reasonable tolerance is essential for automated loading as
well as accommodation of tilt of the specimen plate. Effective management of
tolerances is a critical part of instrument design.
The hemispherical collection efficiency (HCE) of the TSL is displayed in
Fig. 9.7. The object space of the marginal NAs are calculated from the marginal
ray angles. The 2XR TSL (0.063) displays a total HCE as 0.000 55. The 2XR
ML (0.058) displays a total HCE as 0.000 46. The 1X ML (0.089) displays a total
HCE as 0.001 07.
The ML at 1X is a fine lens for imaging of fluorescent arrays. The 1X ML
collects nearly twice as much light as the TSL and ML at 2XR. However, the
maximum HCE of the 1XR requires a 20-μm pixel. The higher object NA defines
a smaller depth of focus, which is difficult to manage by tolerance of placement.
The TSL offers superior color correction to both versions of the ML. A filter
may be swapped without a disruption of focus. The fewer pieces define a simpler
tolerance budget and better transmittance. However, a disadvantage of the TSL is
cost.
An aspheric lens in small quantities is expensive to manufacture. A
coordinated process of interferometry, grinding, and polishing is required. An
aspheric lens with a 60-mm diameter and a wavefront error of 0.1λ costs about
$3000 per unit in the year 2009. Fortunately, this cost may be recovered through
a reduction in consumption of chemicals during operation.
The TSL is suited for applications wherein photons are precious.
Fluorophores can photobleach after just 200 photons of emission, reagents can be
expensive, and potential drug compounds from extinct plants may be used only
in small amounts. There are many applications for the telecentric spectroscopy,
even with a 10X cost over a mass-production lens.
92 Chapter 9

Figure 9.7 Hemispherical collection efficiency (HCE) of a telecentric spectroscopy lens

(TSL) and a microlens (ML). The HCE is plotted versus radial position within the point
spread. The HCE plot is an encircled-energy plot from ZEMAX that is scaled by the HCE.
The encircled-energy plot occurs in air (n = 1.00). HCE occurs in water (n = 1.33).
Marginal NAs of object space are indicated within parentheses.
Chapter 10
Illumination Systems
10.1 Condenser
As shown in Fig. 10.1, a condenser is a relay lens for the field stop of an
illumination system. The illumination field stop is a real object. The condenser
creates a virtual image of the illumination field stop at the specimen. The
illumination field should be slightly larger than the vision field. Assembly
tolerances dictate the amount of additional size of the illumination field beyond
the vision field. The condenser also creates a virtual image of the illumination
lens stop at the vision lens stop.

10.2 Abbe Illumination

Abbe illumination locates a source at a conjugate of the illumination field with a
small NA. Ernst Abbe promoted a small NA of illumination, which facilitates the
Abbe sine condition.15 The small illumination NA of Abbe may be defined as
< 0.75 times the vision NA.
Ernst Abbe worked for Carl Zeiss. According to the Zeiss archives, Abbe
eventually became the chief executive at Zeiss Optics, through which position he
stimulated growth of the company from 25 to 1,400 employees. Abbe promoted
numerous social reforms at Zeiss Optics, such as paid vacation, profit sharing,
pensions, sick time, and an eight-hour workday.

Figure 10.1 Condenser lens (CL), illumination field (IF), and stops: illumination field stop
(IFS), illumination lens stop (ILS), and vision lens stop (VLS). Cardinal points: Front focal
point (FFP), principal point (PP), and back focal point (BFP).

93
94 Chapter 10

Abbe is frequently credited with many advancements of optics in the late

1800s. He defined the term “numerical aperture” in 1878,16 which replaced terms
such as “water angle” and “balsam angle.” However, he was not responsible for
“critical illumination,” which is properly credited to Nelson in the next section.

10.3 Nelson Illumination

Nelson illumination locates a source at a conjugate of the illumination field with
a large NA. The critical illumination NA of Nelson may be defined as ≥ 0.75
times the vision NA. Nelson illumination originally applied to the flame of an oil
lamp as the light source. The isotropic radiance of the flame could be placed at a
field conjugate without a diffuser. The extent of the source defines the
illumination field stop. A large angular size of the lens stop creates a small radius
of partial coherence at the illumination field.
In Fig. 10.2, the image of a source is relayed onto the illumination lens stop.
A source collector lens gathers as much light as is practical. A source field lens is
located near the illumination field stop. The source field lens directs light of the
field stop into the lens stop.
In 1875, Edward Nelson discovered “the advantage of a large axial cone.”15
However, his contradiction of small-cone illumination created trouble for him. In
1889, Abbe openly condemned the large-angle cone.15 Abbe also denounced a
large annulus by Smith.15 Eventually, after repeated exhibitions by the Royal
Society of Microscopy and others, the large-angle cone of Nelson was adopted in
high-class microphotography. Smith’s annular illumination was also proven
through photographic evidence. In 1910 Nelson documented his critical
illumination, 35 years after his discovery and 21 years after its public
condemnation. Nelson defined critical illumination as “the aperture [of
illumination]…is not less than three-quarters of the NA of the observing
objective.”15
Critical illumination is beneficial for maximum separation of two
overlapping point spreads. However, it is not required for detection of incoherent
fluorescent emissions. Critical illumination minimizes the coherent interaction
between point spreads.

Figure 10.2 Source relay of critical illumination. Source (S), source collector lens (SCL),
source field lens (SFL), source image (S′), illumination field stop (IFS), and illumination
lens stop (ILS). Illumination NA should be 0.75 times vision NA.
Illumination Systems 95

The Abbe resolutioni of Nelson illumination may be defined as

λ
d Abbe = , where ΝA S ≤ ΝAV , (10.1)
ΝAV + ΝA S

where dAbbe is the minimum separation of the features, NAV is the NA of the
vision system, and NAS is the NA of the source. A point source defines a
completely coherent illumination field. The Abbe resolution of a coherent field is
λ/NAV. A broad source NAS ≥ 0.75 NAV defines an incoherent system. The Abbe
resolution of a broad source is limited by the vision system. The effective NA of
the illumination field cannot exceed that of the vision field. Consequently, the
Abbe resolution of an incoherent system cannot exceed 0.5λ/NAV. The partial
coherence of the illumination resembles the point spread of an Airy pattern as
follows:

2 J1 ( ρ k NA S )
Γ (ρ) = . (10.2)
ρ k NA S

10.4 Diffusers
The coil structure of a filament is extremely nonuniform. Consequently,
placement of a filament near an image conjugate creates an image of the filament
at the sensor. A diffuser is required for reduction of the filament structure of the
image.
A diffuser spreads light over both space and angle. Consequently, the
etendue of the illumination increases and the radiance decreases. Increased
uniformity by diffusion requires significant reduction of radiance. Consequently,
the image brightness decreases. An improved uniformity over space and angle
normally requires a reduction of image brightness.
Diffusers are notorious for inconsistent performance. The effects of an acid
etch are dependent on time of exposure and composition of materials. Sand
blasting employs variable grain size. Opal glass employs diffuse scattering
throughout its volume. An opal diffuser creates an extremely isotropic radiance
in transmission. However, much incident light is backscattered. The
inconsistency of a diffuser in production might create enormous variations
between instruments.
The emission of a filament is uniform over angle. Consequently, conversion
of the angular extent of a filament into a spatial extent is highly beneficial.
Ideally, the filament should be located at an image conjugate of the lens stop.
This is a foundation of Köhler illumination.

i
The Abbe resolution applies to the spacing of a grating.
96 Chapter 10

10.5 Köhler Illumination

In 1893, Köhler conceived an illumination system where the flame of an oil lamp
was focused on the lens stop of a relay lens.17 Köhler stated two important
principles, as translated in the English reprint on the centennial anniversary of his
original article in German.18 The first principle of Köhler illumination is: “the
light source is almost in the back focal plane of the condenser lens.”18 This
ensures sufficient defocus of the light source at the vision field. The image of the
source is distant from the condenser lens. The second principle of Köhler is: “the
location of the object…is beyond the focal length of the condenser.”18 This
places the illumination field stop at a different location from the image of the
source. Köhler also indicated an image of the illumination lens stop at the vision
lens stop.
A modern interpretation of Köhler is based on three principles: the
illumination lens stop is located at the front focal point of the condenser; an
image of the source is located at the illumination lens stop; and the vision lens
stop is located at an image of the illumination lens stop. The first condition
ensures maximum defocus of the filament. The second and third conditions
minimize reduction of image brightness at the field margin. These conditions can
be mutually exclusive because most objective lenses do not have the vision lens
stop located precisely at a focal point of the objective. The telecentric
configuration of the first condition is most common; however, it can create an
illumination profile with a significant dependence on field height.
In a Köhler illumination scheme, the field stop collects light from a filament
at sufficient distance for excellent uniformity in the illumination field. The
uniformity within the field stop is determined by the angular distribution of the
filament’s emission. The spatial distribution of the filament is located at the lens
stop.
In Fig. 10.3, the filament relay of a Köhler system is displayed. The filament
is relayed by the source collector lens on the illumination lens stop. An image of
the filament is located on the illumination lens stop and the vision lens stop. The
angular distribution of the source is converted by the source collector lens into a
spatial distribution at the illumination field stop. Consequently, the spatial
uniformity is excellent without any use of a diffuser. The illumination field stop
is relayed onto the illumination field by the condenser lens.
The position of the filament is critical to Köhler illumination. Fabrication
tolerances cannot reliably place an image of the filament within the lens stop.
Consequently, the filament must be actively aligned, or the filament emission
must be diffused. A typical filament has an operational lifetime of 1000 h.
Illumination Systems 97

Figure 10.3 Principles of Köhler illumination. The image of a source (S′) is distant from the
illumination field (IF). The image of an illumination lens stop (ILS′) is distant from the
illumination field (IF).

The stops of Köhler illumination are displayed in Fig. 10.4. A first variable
iris is placed at the illumination field stop. The illumination field stop defines an
illumination field that should overfill the vision field. Illumination far beyond the
vision field is detrimental to contrast because the specimen might scatter light in
the vision field. A second variable iris is placed at the illumination lens stop. The
illumination lens stop defines an illumination NA which should underfill the
vision lens stop. Illumination beyond the vision lens stop is highly detrimental to
contrast because the objective barrel is normally stainless steel, which reflects
extraneous light into the vision system.

(a)

(b)

(c)

Figure 10.4 Stops of Köhler illumination: (a) maximum extent of rays, (b) relay of the
illumination field stop (IFS), and (c) relay of the illumination lens stop (ILS).
98 Chapter 10

An iris may also be called a diaphragm, which a is flexible membrane that

covers on opening; however, a diaphragm does not normally have a hole in the
middle as an iris does. A diaphragm of the lung controls breathing, a diaphragm
of the ear converts air pressure into nerve signals, and a diaphragm of a speaker
converts an electrical signal into air pressure. The word iris is derived from the
Greek iris, meaning rainbow. Interestingly, a rainbow is a naturally occurring
annulus that does resemble the iris of an eye, which is an optical stop.
At the limit of a point source, NAS = 0 in Eq. (10.2), a Köhler illumination
field becomes completely coherent, which is detrimental to image quality.
Constructive interference between image point spreads normally creates
detrimental features such as fringes. A laser beam resembles a collimated point
source at NAS = 0. The axial coherence of a monochromatic laser [Eq. (5.31)]
creates both circular and linear fringes within an image. The small angular size of
a Köhler source can create similar fringes over a short range of lateral coherence
[Eq. (10.2)].

10.6 Matched Stops

In both Nelson and Köhler illumination, the space-angle product of the
illumination is maintained by two variable irises. A first iris is placed at the
illumination field stop, and the second is placed at the illumination lens stop. The
space-angle product of the illumination should fill the objective lens stop and the
vision field without overfilling either.
Overfilling the space-angle product of the vision system creates extraneous
light in the background of the image. Illumination beyond the vision field may be
scattered by the specimen into the vision system. Illumination beyond the vision
lens stop may be scattered by the lens barrel into the vision system. A major
benefit of Köhler illumination is maximum image brightness without reduction of
contrast. The stops are adjusted to maintain the space-angle product without
overfilling either the vision field stop or vision lens stop.

10.7 Light-Emitting Diodes

A light-emitting diode (LED) offers several advantages over an incandescent
filament. The lifetime of an LED can reach 100,000 h, while an incandescent
halogen bulb can only reach 2,000 h. LEDs can offer a spatially uniform emitter
for critical illumination. Their angular emittance pattern is typically Lambertian.
An isotropic internal emission is converted to an external Lambertian profile by
Snell’s law.11 The peak of a Lambertian radiance profile is twice that of an
isotropic profile; thus, more of the light is directed into a lens stop.
LEDs do, however, have disadvantages. An LED requires a binary
semiconductor. Silicon and germanium are elementary semiconductors with
indirect bandgaps, which cannot offer direct transition to an open state of the
valance band. Gallium arsenide and other binary semiconductors have direct
bandgaps, which offer transition to the open states of the valence band.
Consequently, an LED requires a binary compound.
Illumination Systems 99

The radiance of the sun is eye safe at 0.5 deg but is not eye safe after
magnification. The irradiance of the sun’s image at the retina does not change
after magnification; however, the cooling geometry of the image is much
different. During normal vision, the 0.5-deg image of the sun is absorbed over a
small sphere that creates a spherical cooling geometry, which permits safe
dissipation of heat. During magnification, however, the image of the sun is
absorbed over a disk that creates a planar cooling geometry, which cannot
dissipate heat quickly enough to prevent damage to the eye. Consequently, any
magnification of the sun might create a retinal burn.
This warning about the sun also applies to high-brightness LEDs. The
American National Standards Institute (ANSI) has specific information on
diffuse emitters, such as LEDs, as part of its standard for safe use of lasers,
Z136.1.19
The eye has two defense mechanisms against bright light: a blink and a large
saccade. A blink is the closure of the lid. The typical reaction time for a blink is
0.18 s. The ANSI standards19 define an aversion response of 0.25 s for visible
light; this defines a 5-mW limit for a class III(a) laser, which is safe within a
blink of an eye. On the other hand, a class III(b) laser is not safe within a blink of
an eye. A large saccade is a gaze shift that normally occurs within 10 s. The
ANSI standards19 define an aversion response of 10 s for invisible light.
Figure 12.14 in Schubert11 indicates emission by material type. Gallium
nitride with indium (GaInN) emits in the blue (450–490 nm) and green (515–570
nm) wavelength ranges. GaInN can be grown on a sapphire (Al2O3) substrate;
however, there is significant strain due to lattice mismatch. Gallium phosphide
(GaP) can emit weakly in the yellow (570–580 nm), which is near the peak of
human sensitivity at 555 nm. Gallium arsenide with aluminum and indium
(AlGaInAs) emits strongly in the orange (585–620 nm). Aluminum gallium
arsenide (AlGaAs) emits strongly in the red (625–740 nm). There is a significant
dead spot in LED emission in the green side of the yellow wavelength range
(530–580 nm). LEDs offer much longer lifetimes than incandescent filaments.
Schubert11 provides a practical review of photopic vision, which is based on a
red, green, and blue cones.
The peak wavelength of emission is strongly dependent on concentrations of
the components during vacuum deposition. Typically, the peak wavelength varies
by as much as the full width at half maximum of emission. Many dyes and
fluorophores are dependent on a narrow band of excitation. A tolerance for peak
wavelength of excitation should be established. The peak wavelength varies
greatly by production run; a typical variation is 3–5% of the nominal peak
wavelength.
A white LED is generated by a blue LED and white phosphor. The phosphor
creates a uniform spectrum distribution along with a variable blue peak. A white
LED has an artificial appearance in comparison to an incandescent bulb.
The packaging of an LED is dependent on its power. A low-power package
encapsulates the die and leads in epoxy, which also provides structural support. A
high-power LED package requires a heat sink in direct contact with the LED
100 Chapter 10

chip. The LED chip of a high-power package can be exposed to the air or
covered by a lens. A high-power LED package typically employs silicone as the
encapsulate.
Epoxy resin provides excellent stability over time. However, an epoxy resin
cannot tolerate high temperatures (> 120° C). Epoxy might also exhibit reduced
transmission of red, violet, and UV rays. The refractive index of epoxy is
typically 1.6, while the refractive index of GaP is approximately 3.5. A higher
refractive index of the encapsulate promotes extraction of photons from the LED
chip. A low-power package frequently employs an epoxy lens at a 5-mm-
diameter T1-3/4 package. The placement error of the chip within a T1-3/4
package is large, as the decenter can exceed the emitter width.
A silicone encapsulate can tolerate the higher temperatures (190° C) of high-
brightness LEDs. Polysiloxane is the correct chemical name, which indicates a
chain of siloxane groups R2SiO, with a variety of side groups R. The refractive
index can range from 1.4 to 1.6 with dependency on the composition of the side
groups R. The stiffness is also dependent on the composition of the side groups.
A high-brightness LED typically employs a 5.6-mm-diameter hemispherical
dome as the encapsulate.
There are numerous lenses that fit over the standard 5.6-mm dome. An LED
lens typically comprises a central refractor and an annular reflector.20 The
combination of refractive and reflective elements defines a catadioptric LED
lens. The material normally used is polymethylmethacrylate (PMMA), with a
refractive index of 1.49. The uniformity of a catadioptric LED lens is not suitable
for application to microscopes.

10.8 Aspheric Plus Singlet Relay

Figure 10.5 displays a lens report for a condenser comprising an aspheric lens
and a singlet. They are both stock items as indicated in the prescription. A small
amount of coma limits the d line to a sharp focus of 10 μm. The axial color limits
the polychromatic focus to 160 μm.
A magenta ring is present at the perimeter of the illumination field. The
polychromatic edge spread indicates two distinct bends. The first bend at near
zero indicates the sharp edge of the green light at the d line. The second bend at
20 μm indicates the sharp edge of the red light at the C line. Beyond the second
bend, there is only blue light at the F line. The region between the bends
comprises only red and blue, which appear as a magenta ring of 20-μm width.
The magenta ring is an important metric during focus of the condenser. First,
an operator closes the illumination field stop until the edge of the illumination
field fits within the vision field. Second, the operator drives the condenser focus
toward maximum brightness of the magenta ring. This procedure ensures the best
focus of the green d line toward consistent operation of the instrument. A smaller
illumination NA might increase visibility of the magenta ring.
Illumination Systems 101

Figure 10.5 Aspheric relay 4XR 0.25. The magenta ring is defined by two bends in the
edge spread. Polychromatic edge spread employs 1024 × 1024 sampling. Both lenses are
stock production.

10.9 Achromatic Aspheric Plus Doublet Relay

Figure 10.6 displays a lens report for a condenser comprising an aspheric lens
and a doublet. The doublet is a custom item, whereas the aspheric lens is a stock
product. A small amount of third-order spherical aberration limits the d line to a
sharp focus of 10 μm. The axial color limits the polychromatic focus to 40 μm.

10.10 Abbe Condenser

Figure 10.7 displays a lens report for a condenser comprising two plano-convex
singlets that are stock items. This combination defines an Abbe condenser, which
has numerous formats. A large spherical aberration limits the green d line to a
broad focus of 50 μm. The axial color limits the polychromatic focus to 80 μm.
Observation of a magenta ring requires an NA of approximately 0.10.
Observation of the Abbe sine condition may eliminate the spherical aberration.

10.11 Abbe Aspheric

Figure 10.8 displays a lens report for an aspheric condenser. The axial color
limits the polychromatic focus to 50 μm.
102 Chapter 10

Figure 10.6 Achromatic aspheric relay 4XR 0.25. A custom doublet in combination with a
stock aspheric lens.

Figure 10.7 Abbe illumination relay 4XR 0.25. Stock spherical elements.
Illumination Systems 103

Figure 10.8 Aspheric Abbe relay 4XR 0.25. Custom aspheric elements utlizing B270 glass
for molding.

10.12 Total Internal Reflection Fluorescence Illumination

A total internal reflection fluorescence (TIRF) illumination scheme employs total
internal reflection for application to fluorescence. The evanescent field of TIR
provides a shallow excitation of a liquid on the back side of the cover glass. The
depth of the evanescent field is normally less than 100 nm. A small molecule is
bound to the back side of the cover. The small molecule is efficiently excited
without excitation of the entire depth of the liquid. A single fluorophore can be
detected without competition from the intrinsic fluorescence of the bulk of the
liquid.
Figure 10.9 displays a TIRF illumination scheme. A laser beam enters the
“back aperture” of the lens. A virtual focus is located at the tube focal plane of
the TIRF lens. The virtual focus is converted into a collimated beam at the
specimen. The real focus occurs inside a calcium fluoride lens element. Purity of
calcium fluoride is essential for preventing background radiance from the real
focus with large irradiance. Background radiance may be caused by fluorescence
and/or nonlinear dipole currents.
The angle of TIR within the cover is
n 
θTIR = arcsin  L  , (10.3)
 nC 
104 Chapter 10

where nL is the refractive index of the liquid and nC is the refractive index of the
cover. The angle of the TIRF beam is described by a thick-lens model, where the
principal plane is a sphere within the cover slip. The angle of the TIRF beam in
the cover is defined by the height of the beam hB as follows:

 h 
sin θBC =  B , (10.4)
 nC f 

where nC f is the focal length within the refractive index of the cover. Conversion
from glass to silica shortens the immersion focal length. The angle of TIR at
water grows from 61 deg in borosilicate to 66 deg in silica.

Figure 10.9 TIRF illumination 60X. The TIR of the laser beam provides shallow excitation
on the back side of the cover. The angle of TIR is 60.4 deg for D-263 and water. The
prescription is the same as the 60X immersion lens of Fig. 12.5.
Chapter 11
Cover Strata
11.1 Importance of Specimen Tolerance
A specimen tolerance budget is essential for consistent performance of an optical
instrument. The cover strata should have a specification for nominal thickness
and tolerance. Exceeding the specimen tolerance might create enormous
spherical aberration. Spherical aberration can reduce contrast without an apparent
growth in spot size. Other important issues for the specimen are intrinsic
fluorescence, surface-form error, surface defects, and surface texture.

11.2 Perfect 10X for Air

A perfect 10X objective within air is defined by Prescription 11.1 in the
Appendix. Equation (2.4) defined the focal length of a 10X objective as 20 mm
in combination with a 200-mm tube lens. A typical sensor field is defined by a
12-mm diagonal sensor. Thus, the object field of a 10X objective is 1.2 mm in
diameter or 0.6 mm in height.
A virtual object is located at the back focal point. A spherical principal plane
defines the lens stop. The object distance from the lens stop is –20 mm in the
prescription. The radius of the lens stop equals the back focal length at 20 mm.
Rays traveling leftward from the object define paths of perfection without
aberration. The rightward-traveling rays after the lens stop should overlay the
paths of perfection, unless there is deviation from the cover strata of the virtual
object.
The object-space NA is defined within the perfect part of the lens
prescription. The lens-stop diameter at 10 mm is defined by the object-space NA.
The image-space NA may vary due to aberrations in the cover strata of the image
side. A planar lens stop is an acceptable alternative to a spherical lens stop.
However, a planar lens-stop diameter requires an inflated diameter for the same
object NA. The lens stop and focal length define the f/# as 2.0. A spherical lens
stop provides a more accurate definition of f/#. The distance from the lens stop is
variable, as indicated by a box around the thickness of the lens stop.

105
106 Chapter 11

Figure 11.1 Perfect 10X objective in air. Focal length is 20 mm, object NA is 0.25, and the
lens stop is 10 mm in diameter. Line-spread is polychromatic, and encircled energy is at
the d line. Field height is at 0.6 mm. Axial defocus of 1 nm is intentionally included for
visualization of ray-intercept plots.

The corresponding merit function for a perfect 10X objective employs three
basic metrics: the ray intercept of a top marginal ray, the ray intercept of a
bottom marginal ray, and a Strehl ratio for the d line at the field margin. A
process of iteration drives a single variable (lens-stop thickness) toward the
maximum Strehl ratio at the field height.
Figure 11.1 displays a lens report for perfect 10X objective within air. A
nanometer of axial length beyond perfect focus creates a small amount of
defocus. The ray-intercept plots for all three wavelengths are colinear. The
amplitude of defocus error is far below the Airy radius at 1.5 μm for the d line.
The edge spread is 0.8 μm wide. The flat of the encircled-energy plot ends near
1.5 μm.

11.3 10X Objective with Cover Glass in Place of Air

In Fig. 11.2, a typical cover glass is added to the perfect 10X objective in air. A
0.17-mm thickness of D263M glass represents a No. 1.5 cover slip, which has a
manufacturing tolerance of 0.16–0.19 mm. Other standard types of cover glass
are cited in Table 11.1. There is a hint of spherical aberration and coma.
Cover Strata 107

Figure 11.2 Glass cover in image space for a perfect 10X objective in air. Glass cover:
D263M at 0.17 mm thick. Focal length is 20 mm, object NA is 0.25, and the lens stop is 10
mm in diameter. Line-spread is polychromatic, and encircled energy is at the d line. Field
height is at 0.6 mm.

However, the image is considered diffraction limited. There is a slight departure

from the diffraction limit in the encircled-energy curve. A 10X objective at 0.25
NA can operate properly with or without a 0.17-mm cover glass.
A cover glass is manufactured by pulling a ribbon of glass off a molten tin
bath. The surface is largely free of digs and scratches. There are no scratches due
to pulling of impurities during polish. However, there are significant tolerances
of thickness, which become more important as the objective NA becomes larger.
Table 11.1 displays a list of typical cover-glass tolerances.

11.4 10X Objective with Microscope Slide in Place of Air

In Fig. 11.3, a typical microscope slide is added to the perfect 10X objective in
air. A 1.0-mm thickness of D263M represents a typical microscope slide. There
is enough spherical aberration for a small departure from the diffraction limit of
the d line. There is an even smaller amount of axial color as indicated by the
dissimilar slopes at the origin of the ray intercepts. The Strehl ratio of the margin
is 96% for the d line and 91% for the polychromatic combination of the F, d, and
C lines. A 10X objective at 0.25 NA can operate properly with a 1.0-mm glass
slide.
108 Chapter 11

Table 11.1 Cover glass tolerances.

Dimensions (mm)
Number Nominal Minimum Maximum Range Typical area
0 0.10 0.085 0.115 0.030 15 × 15 to 24 × 60
1 0.15 0.130 0.160 0.030 15 × 15 to 24 × 60
1.5 0.17 0.160 0.190 0.030 15Ø, 15 × 15, 24 × 50
2 0.21 0.190 0.230 0.040
3 0.30 0.280 0.320 0.040
4 0.37 0.320 0.420 0.040 24 × 50
5 0.55 0.500 0.600 0.100
Slide 1.00 0.960 1.060 0.100 25 × 75
Slide 1.20 1.100 1.200 0.100 25 × 75

Figure 11.3 Glass slide in image space for a perfect 10X objective. Glass slide: D263M at
1.0 mm thick. Focal length is 20 mm, object NA is 0.25, and the lens stop is 10 mm in
diameter. Line spread is polychromatic, and encircled energy is at the d-line. Field height
is at 0.6 mm.

11.5 40X Objective with Silica Cover in Place of Glass

Prescription 11.4 (see Appendix) defines a cover glass within a perfect object
space. The correct cover glass is 0.17 mm thick. The correct cover material
within object space is D263M. The effective focal length of the 40X objective is
5.0. The optical distance [Eq. (3.24)] from the lens stop to the object is also 5.0
Cover Strata 109

mm. The spatial distance is > 5.0 mm due to immersion with the cover glass. On
the image side, silica replaces D263M as the cover material. The material in
image space may be glass, silica, water, or air. Silica is used in this example for
simplicity. At higher NAs with immersion oil, the choice of image material
becomes important as total internal reflection becomes possible.
Figure 11.4 displays a perfect 40X objective with silica in place of D263M
on the image side. The object-space NA is 0.75. The image-space NA is 0.74.
There are small amounts of axial color and third-order spherical aberration. There
is a large fifth-order spherical aberration at the margin. However, the encircled-
energy plot indicates diffraction-limited performance. The Strehl ratio of the
margin is 100% for the d line and 98% for the polychromatic combination of the
F, d, and C lines. Thus, a 40X objective at 0.75 NA can operate properly with a
silica cover in place of a glass cover. However, the NA cannot increase with
inclusion of a larger spherical aberration.

Figure 11.4 Silica specimen cover replacing a glass cover in image space for a perfect
40X objective. Silica is used in place of D263M for the cover material at 0.17 mm thick.
Focal length is 5 mm, object NA is 0.75, and the lens stop is 8.6 mm in diameter. Line-
spread is polychromatic, and encircled energy is at the d line. Field height is at 0.15 mm.
110 Chapter 11

Figure 11.5 Tilt of the glass specimen cover in image space for a perfect 40X objective.
Cover is at a tilt of 0.5 deg (9 mrad). Cover material is D263M at 0.17 mm thick. Focal
length is 5 mm, object NA is 0.75, and the lens stop is 8.6 mm in diameter. Line-spread is
polychromatic, and encircled energy is at the d line. Field height is at 0.15 mm.

11.6 40X Objective with Tilted Cover Glass

Figure 11.5 displays a perfect 40X objective with a specimen tilt of 0.5 deg (9
mrad). Three object points are required for optimization: 0.0, +0.15, and –0.15
mm. An image tilt of –1.0 deg in silica is required as compensation for the +0.5-
deg tilt of the cover within image space. The object-space NA is still 0.75. There
is a large amount of coma, which blurs the left edge of the line spread. There are
much smaller amounts of astigmatism and fifth-order spherical aberration. The
encircled-energy plot indicates a large departure from the diffraction limit. The
Strehl ratios for the d line are 78% at 0.15 mm, 64% at +0.15 mm, and 85% at
–0.15 mm. A 40X objective at 0.75 NA is significantly compromised by coma
due to a 0.5-deg tilt of the cover glass.
There is also significant tilt of the specimen plane at the sensor. The tilt of
the specimen plane within the cover glass is –0.5 deg with respect to the optical
axis. The equivalent tilt of the specimen plane within air is approximately 0.3
deg. After 40X magnification, the image of the specimen plane at the sensor is
tilted by 12 deg. The amplified effects of specimen tilt demand careful
consideration at any magnification.
Cover Strata 111

Figure 11.6 Silica specimen cover replacing a glass cover in image space for a perfect
60X objective. Silica is used in place of D263M for the cover material at 0.17 mm thick.
There is immersion oil between the lens stop and cover. Focal length is 3.3 mm, object NA
is 1.40, and the lens stop is 9.3 mm in diameter. Line-spread is polychromatic, and
encircled energy is at the d line. Field height is at 0.10 mm.

11.7 60X Objective with Silica Cover in Place of Glass

Figure 11.6 displays a perfect 60X objective lens with silica in place of D263M
as the cover material. Immersion oil fills the gap between the lens stop and the
cover. The focal length is 3.3 mm. The immersion focal length is 5.0 mm. The
radius of curvature of the lens stop is 5.0 mm. The object-space NA is 1.40. The
image-space NA is 1.35. The diameter of the lens stop is 9.3 mm. There is an
enormous amount of third-order spherical aberration. The Strehl ratio of the
margin is 2% for the d line. A 60X objective at 1.40 NA cannot operate properly
with a silica cover in place of a glass cover.

11.8 Strehl Ratio versus Optical Path Length

Due to additional cover strata, the Strehl ratio is very dependent on marginal NA.
Figure 11.7 displays the Strehl ratio versus marginal NA for several common
errors. The plots are primarily identified by optical path lengths as 50λ, 200λ, and
2000λ. These path lengths represent three common thickness errors: a cover glass
at 30 μm of thickness tolerance, a microscope slide at 120 μm of tolerance, and a
microscope slide at 1.2 mm in thickness. The 50λ path indicates a maximum NA
112 Chapter 11

of 0.50 before a cover tolerance degrades the spot; the 200λ path indicates a
maximum NA of 0.35 before a slide tolerance degrades the spot; the 1200λ path
indicates a maximum NA of 0.20 before a slide degrades the spot.
Consideration of specimen tolerances is critical for consistent performance of
an instrument. Application of a binder adds significant amounts of unexpected
optical path length when using a glass cover. The tolerance on the binder
thickness might exceed the nominal thickness of the cover. A large amount of
binder might enable tilt of the cover. A credible tolerance for a specimen should
be defined before an optical instrument is designed. Consistent performance is a
mark of quality in optical design and manufacture.

Figure 11.7 Strehl ratio versus optical path length of an additional cover. Optical path
lengths are cited in wavelengths. Spatial thickness of D263M is cited in parentheses. 50λ
(29 μm) indicates a tolerance for a cover glass. 200λ (120 μm) indicates a tolerance for a
microscope slide. 2000λ (1200 μm) indicates a thickness for a microscope slide.
Chapter 12
Objective Lenses
12.1 Formats
There are numerous types of objectives; several are listed below:
1. An aplanati is free from spherical aberration and coma.
2. A plan objective has a flat field: there is no Petzval curvature.
3. An achromat is corrected at two wavelengths.
4. An apochromat is corrected at three wavelengths.
5. A semiapochromat is nearly corrected at three wavelengths.

The original fluorite objectives corrected for color with fluorite glass, but
lanthanum glass eventually replaced fluorite in most applications. A modern
fluor objective contains LaK while excluding many schwer flints (SF). A fluor
objective transmits well in the UV; consequently, a fluor objective may also
imply application to fluorescence.
The magnifications of the following objectives are based on a 200-mm tube
lens, which is a standard tube lens for Nikon microscopes. An Olympus scope
employs a 180-mm focal length as its tube lens.

12.2 Aplanatic Surface

An aplanatic surface is free of both spherical aberration and coma. There are
three configurations of an aplanatic surface. An aplanatic surface of the flat kind
has little practical value: the object and image are located at the surface. An
aplanatic surface of the concentric kind is very intuitive: the object and image are
concentric with the surface. There is no refraction in the second configuration.
An aplanatic surface of the scaled-divergence kind is not obvious: it is defined by
the following mathematical relation, where the divergence is scaled by the
refractive index:

ΝA I ΝA O
= , (12.1)
nI nO

i
Aplanat is derived from the Greek a–plan-etes. a-, meaning not; plan, meaning moving; and etes,
meaning star. However, an aplanat does not transform a star into a planet during an exposure.
113
114 Chapter 12

Figure 12.1 Aplanatic surface of the scaled-divergence kind. The NA is scaled by the
refractive index between the object and the image. The lens stop is proximal to the image.

where NAI is the NA of the image, nI is the refractive index of the image, NAO is
the NA of the object, and nO is the refractive index of the object.
Figure 12.1 displays a single aplanatic surface of the scaled-divergence kind.
This eliminates the spherical aberration. The divergence within the glass
borosilicate crown (BK7) is roughly 1.5 times that of air. Proper placement of the
lens stop near the image conjugate also eliminates coma. The image field is
curved. The aplanatic surface is free of both spherical aberration and coma.
In Fig. 12.2, the addition of a meniscus defines two more aplanatic surfaces.
The first and third surfaces S1 and S3 are aplanatic due to scaling of divergence
by the refractive index. The second surface S2 is aplanatic due to concentric
image conjugates of the surface. Once again, proper location of the lens stop is
required for elimination of coma, and the object field is curved.

12.3 10X Plan Achromat

Figure 12.3 displays a 10X plan achromat.21 The focal length is 20 mm. The
paraxial NA is 0.25. In the ray-intercept plot, correction at two wavelengths (588
and 656 nm) indicates an achromat, and lack of linear slope indicates plan. The
encircled-energy plot indicates nearly diffraction-limited performance at 0.6 mm
and slightly aberration-limited performance at 0.6 mm. The flat of the encircled-
energy plot indicates an Airy radius of 1.2–1.4 μm. A 0.25 NA indicates an Airy
radius of 1.4 μm at 588 nm. The edge spread displays a high-contrast transition
over 1.0 μm. The SCE is 1.9% in air and 0.9% in water.
Objective Lenses 115

Figure 12.2 Aplanatic front end. First and third surfaces S1 and S3 define aplanatic
surfaces of the scaled-divergence kind. The second surface S2 defines an aplanatic
surface of the concentric kind.

There are two distinct lens groups. The first group resembles a single Gauss:
a biconvex singlet and meniscus doublet. The meniscus lens is shaped for
management of spherical aberration, coma, lateral color, and Petzval curvature.
The second group is a doublet: it manages the flange focal distance and the
residual aberrations of the first group.
The lens employs super-schwer crown (SSK), lanthanum schwer flint
(LaSF), and schwer flint (SF). The SSK glass has a refractive index of
approximately 1.6; the LaSF and SF have a large refractive index at 1.8.
Performance in the blue is compromised for two reasons: the partial dispersion
and absorption in the blue spectrum. Both limitations originate in the SFs.
This lens functions extremely well for human vision, although its
performance in the blue compromises electronic vision.

12.4 40X Fluor

Figure 12.4 displays a 40X fluor.22 The focal length is 20 mm. The paraxial NA
is 0.75. The original lens had a focal length of 1 mm and a 0.17-mm-thick cover
glass. The lens was scaled to 5 mm in focal length while maintaining the cover
glass at 0.17 mm thick. The entrance pupil at 7.5 mm was derived from a 0.75
NA. The flat of the encircled-energy plot indicates an Airy radius of 0.4–0.5 μm.
116 Chapter 12

Figure 12.3 10X plan achromat at 0.25 NA. NAP = 0.25 is derived from a 5.0-mm entrance
pupil and a 20-mm EFL. Prescription by Fukutake.21

An NA of 0.75 indicates an Airy radius of 0.5 μm at 588 nm. The Gaussian depth
is 1.9 μm. The edge spread displays a high-contrast transition over 0.3 μm. The
SCE is 8.7% in water.
The ray-intercept plot indicates several key features. The lens is focused for
the d line. There is a small amount of spherical aberration at the margin, which is
likely clipped in application. The d and F lines are effectively corrected with each
other. The lens is an achromat, even though the patent specifies a
semiapochromat. The lens is fairly corrected for the C, d, and F lines, because
they all lie within an axial range of 3 μm. However, the g line (436 nm) displays
significant axial color at 9 μm. Perhaps in the current lens, semiapochromat
indicates correction for three wavelengths within a tolerance, but not a fourth.
There are three distinct groups of the lens, and they each serve a primary
purpose. Optimization of the prescription may require a small deviation from the
group’s primary role.
The first group resembles an aplanatic front end. The first surface is an
aplanatic surface of the concentric kind. The second surface is an aplanatic
surface of the scaled-divergence kind. This nearly hemispheric lens employs a
high-index lanthanum flint. The third surface is an aplanatic surface of the
concentric kind, and the fifth surface is an aplanatic surface of the scaled-
divergence kind. The third through fifth surfaces define an aplanatic doublet.
Objective Lenses 117

Figure 12.4 40X fluor at 0.75 NA. NAP = 0.75 is derived from a 7.5-mm entrance pupil and
5-mm EFL. Prescription by Misawa.22

The second group resembles a single Gauss, where the doublet is a triplet.
The positive elements are all long-spectrum fluorite krone glass: FK5 and FK56
provide a low index and low dispersion. The low index limits the NA of the
objective, while the low dispersion reduces lateral color.
The third group defines a negative element for extension of track length
beyond the focal length. The distance from lens stop to object should be 60 mm,
while the effective focal length is 5 mm. This configuration requires a strong
negative element at the lens stop. It is somewhat aplanatic with respect to the lens
stop.

12.5 60X Immersion TIRF

Figure 12.5 displays a 60X immersion lens.23 It can be employed for total internal
reflection fluorescence (TIRF). See Sec. 10.12 for the illumination path of TIRF.
The paraxial NA is 1.4, the marginal NA is 1.2, and the central NA is
approximately 1.0. A hemispheric surface defines the last air-to-glass interface of
the lens. The next concave surface holds a drop of oil. The oil has a refractive
index of 1.518 and an Abbe number of 58.9. The cover material is likely D-263,
which has a refractive index of 1.5255 and Abbe number 55. The cover is 0.17
118 Chapter 12

Figure 12.5 60X immersion lens. Rays indicate an NAP as 1.4. NAM is 1.2 and NAC is
approximately 1.0. Part (a) displays 8 lens groups and 15 lens elements. Part (b) displays
group 8 comprising: a nearly hemispheric surface (HS), an oil immersion, a cover glass,
and water as the object medium. Prescription by Yamaguchi.23

mm thick. The working distance between the lens and cover is 0.15 mm. The
object is located in water on the back side of the cover. The object may also be
located within a binder.
The 60X immersion lens is suitable for TIRF. The marginal ray (MR) at 70.2
deg is beyond the critical angle of 60.7 deg at the cover-to-water interface. This
enables an evanescent field with a < 100-nm depth into the water. Small
molecules may bond to the back surface of the cover. The evanescent field
excites the surface-bound molecules without excitation of the bulk liquid.
The 60X TIRF lens employs a sequence of four aplanats at the object. The
first aplanat comprises a doublet of oil and lanthanum schwer flint (LaSF35). The
convex surface of oil is an aplanatic surface of the concentric kind. The large
index of LaSF35 at 2.02 reduces the exiting NA and reduces the spherical
aberration. The convex surface of the LaSF35 is an aplanatic surface of the
scaled-divergence kind. Three of the aplanats employ a combination of a long-
spectrum krone CaF2 and a short flint KzFH1. The combination of calcium
fluoride and lead borate is a common method for color correction in the blue. A
short flint from 2009 may be comprised of neodymium and silicate. These four
aplanats gradually bend the marginal rays inward towards the optical axis.
At the back of the lens, there are two meniscus doublets. The opposing
concave surfaces create a powerful negative lens. Their proper alignment is very
difficult to attain. However, their negative power is essential for maintenance of
flange focal length and compensation for aberrations of positive aplanats.
Figure 12.6 displays a report for the 60X TIRF lens at 1.4 NA. The ray-
intercept plot displays a large spherical aberration of 18 μm at the margin for the
d line. The marginal ray angle at 70.2 deg within the cover specifies a marginal
NA at 1.23, which is far below the NA. The encircled-energy plot indicates a
central peak within 0.4 μm and a broad base beyond 5 μm. Also within the
encircled-energy plot, the flat of the 0.1-mm field occurs at 0.4 μm, while the flat
of the diffraction limit is at 0.25 μm. An Airy diameter of 0.8 μm at 588 nm
Objective Lenses 119

Figure 12.6 60X immersion lens at 1.4 NA. NAP = 1.4 is derived from a 9.6-mm entrance
pupil and a 3.37-mm EFL. NAM = 1.23. NAC = 0.90. Edge spread is at the d line and 0.1-
mm field. Prescription by Yamaguchi.23

indicates a central NA of 0.90. The edge spread displays a sharp edge with poor
contrast. The edge rises quickly over 0.2 μm, while the tails extend beyond 5 μm.
Figure 12.7 displays a report for the 60X TIRF lens at 1.0 NA. The ray-
intercept plot displays a small spherical aberration of 2 μm at the margin for the d
line. The axial color indicates < 3 μm of axial shift. The marginal ray angle at
40.7 deg within the cover specifies a marginal NA at 0.93, which is close to the
paraxial NA. The Gaussian depth of focus is 0.43 μm at 588 nm. The encircled-
energy plot indicates a central peak within 0.3 μm without a broad base. Also
within the encircled-energy plot, the flat of the 0.1-mm field is nonexistant, while
the flat of the diffraction limit is also at 0.35 mm. An Airy diameter of 0.8 μm at
588 nm indicates a central NA of 0.9. The edge spread displays a sharp edge with
excellent contrast. The edge rises quickly over 0.2 μm with very little tail. Figure
12.8 displays the SCE by spot radius of the 60X immersion lens. The SCE plot is
an encircled plot, which is scaled by spherical collection efficiency.
The SCE for 0.7 NA is quite different from the SCEs of 1.0 NA and 1.4 NA.
The SCE at 0.7 NA indicates a different central peak and a much different
magnitude. The flat indicates an Airy radius at 0.45 μm. The maximum
magnitude is 0.07. The area of the central peak is nearly twice that of 1.0 NA.
120 Chapter 12

Figure 12.7 60X immersion lens at 1.0 NA. NAP = 1.0 is derived from a 6.7-mm entrance
pupil and a 3.37-mm EFL at the d line. NAM = 0.93. NAC = 0.9. Prescription by
Yamaguchi.23

Figure 12.8 SCE of a 60X immersion lens. The SCE is plotted versus radial position
within the point spread. The SCE plot is an encircled-energy plot from ZEMAX which is
scaled by the SCE. The encircled-energy plot occurs in cover glass (n = 1.52). Spherical
collection occurs in water (n = 1.33). Paraxial NAs at 1.4, 1.0, and 0.7, respectively.
Marginal NAs at 1.2, 0.9, and 0.7, respectively.
Objective Lenses 121

The spherical collection is nearly half that of 1.0 NA. Consequently, the
irradiance of the central peak at 0.7 NA is nearly one-quarter that of 1.0 NA.
Increasing the NA greatly increases the irradiance of the point spread. However,
this trend is not always true. At some point the increased NA might direct more
light into a broad skirt without increasing the peak irradiance.
The SCEs at NAs of 1.0 and 1.4 are similar at small radii, but much different
at large radii. The SCE for 1.4 NA is slightly larger that of 1.0 NA from 0–0.3
μm. This indicates a common central peak in the point spread. However, the SCE
for 1.0 NA remains flat beyond 1 μm, while the SCE for 1.4 NA grows steadily
until 4 μm. This indicates a broad skirt in the point spread for 1.4 NA, which is
absent in the point spread for 1.0 NA. The maximum SCE for 1.4 NA is 0.31,
and the maximum SCE for 1.0 NA is 0.14. The similarity at small radii is due to
shared central NA where the lens is diffraction limited. The additional annulus of
the 1.4 NA does not tighten the spot because the wavefront of the annulus has
spherical aberration.
In combination with a tube lens at 200-mm focal length, the 60X objective
creates a 21-μm radius for the central peak at both 1.4 and 1.0 NA. The skirt of
1.0 NA extends to a 60-μm radius, while the skirt of the 1.4 NA extends to a 240-
μm radius. At 8 μm per pixel, the skirt of the 1.4 NA extends to a radius of 30
pixels. The central peak of both extends to a radius of 3 pixels. The central peak
is 5 pixels wide, the skirt for the 1.0 NA is 15 pixels wide, and the skirt for the
1.4 NA is 60 pixels wide.
Half of the light collected by the 1.4 NA is directed into the background of
the image. The shot noise of the skirt is nearly equal to that of the central peak.
The skirt noise of a single point spread is spread over many pixels. It might not
be above the read noise of the sensor. It might not be detected in a single point
spread. However, the skirt noise of multiple point spreads can be significant. The
contrast of a central peak with background can be greatly reduced by the
summation of multiple skirts. Increasing the NA does not always improve
contrast.

12.6 100X Aplanat

In Figs. 12.9 and 12.10, a 100X aplanat is displayed from Smith.24 The lens
employs a finite tube length of 180 mm. Consequently, its focal length is near 1.8
mm. The front end comprises three aplanatic surfaces of the same configuration
in Fig. 12.2. The magnification of the front end is approximately 3.5X. The two
additional doublets provide another 30X. The doublets are comprised of CaF2,
F2, and BaF2. There is a small curvature toward the object field. Consequently, it
is not a plan objective. Its lack of spherical aberration and coma define an
aplanat.
122 Chapter 12

Figure 12.9 100X aplanat at 1.0 NA. NAP = 1.3 is derived from a 3.5-mm entrance pupil
and a 1.8-mm EFL at the d line. Adapted from Smith.24

In Fig. 12.9, the central NA is 1.0. The on-axis point displays a small fifth-
order spherical aberration. This is caused by small differences in refractive index
between the sphere BK7, the oil TYPE A, and the cover K7. These small changes
in index at plano interfaces create spherical aberration. In Fig. 12.10, the paraxial
NA is 1.3. The fifth-order spherical aberration is large.

12.7 10X Schwarzschild

Figure 12.11 displays a 10X Schwarzschild based on geometric relations in
Riedl.25 The focal length is 20 mm. The marginal NA is 0.25, and the central NA
is 0.117. In the ray-intercept plot, correction at two wavelengths (588 and 656
nm) indicates an achromat; lack of linear slope indicates focus on the curved
object surface. The encircled-energy plot indicates nearly diffraction-limited
performance at 0.6 mm and slightly aberration-limited performance at 0.6 mm.
The flat of the encircled-energy plot indicates an Airy radius of 1.2–1.4 μm. The
paraxial NA 0.25 indicates an Airy radius of 1.4 μm at 588 nm. The edge spread
displays a high-contrast transition over 1.0 μm.
The efficiency of annular collection is based on the marginal and central
NAs:
Objective Lenses 123

Figure 12.10 100X aplanat at 1.3 NA. NAP = 1.3 is derived from a 4.5-mm entrance pupil
and a 1.8-mm EFL at the d line. Adapted from Smith.24

 arcsin ( ΝA M n )  2  arcsin ( ΝA C n ) 
η AC = sin 2   − sin  . (12.2)
 2   2 

The efficiency of annular collection in the current Schwarzschild is 1.2% in air

and 0.7% in water.

12.8 20X Internal Parabola

Figure 12.12 displays a 20X internal parabola after Larson et al.26 and Krogmeier
et al.27 The focal length is 10 mm. The marginal NA is 1.00. In the ray-intercept
plot, a large amount of coma exists. The encircled-energy plot indicates
diffraction-limited performance at 0.3 mm and aberration-limited performance at
the 0.3-mm field. The geometric radius of the encircled-energy plot is 250 μm.
At 20X magnification, this spot radius becomes a 5.0-mm-diameter spot, which
provides a generous tolerance for placement of a ¾-in. photomultiplier tube. The
SCE is 17% in water. An illuminator with a central NA of 0.25 defines an
annular collection efficiency of 16.1% in water.
An issued patent for the internal parabola26 provides a short tutorial on valid
claims. The patent claims a chip comprising a mircofluidic channel, an
illuminator, and a concave collector. Individually, any one of these items does
124 Chapter 12

Figure 12.11 10X Schwarzschild at 0.25 NA. NAM = 0.25 is derived from a 10.0-mm
entrance pupil and a 20-mm EFL. NAC = 0.117 is derived from a 10-mm entrance pupil
and a 44.7-mm distance from convex mirror to field.

not comprise an invention; however, their combination does constitute an

invention. A parabolic reflector by itself is certainly prior art from nearly a
century ago. However, there are numerous patents on combinations of a
parabolic reflector and a refractor. A signal lamp for railroads28 employs a central
refracting lens and an annular internal parabolic reflector as an improvement over
metallic parabolic reflectors. The internal reflection of the annular collector is
impervious to corrosion. Addition of several annular refractors greatly reduces
the diameter of the parabolic reflector.29 An LED mounted to the focal plane of a
concave mirror comprises a novel combination.30 An annular parabolic cap31
collimates exiting light from the cylindrical surface of the LED lens. A
cylindrical void defines a collection angle beyond a hemisphere within an
annular reflector and a central refractor.32 A cylindrical void contains a convex
refractor, while the central refractor and the annular reflector share a planar exit
face.20 There are numerous combinations of parabolic reflectors and refractors.
The surface texture of the internal reflector is critical. At an index of 1.50, an
internal reflector requires a surface texture 36 times smaller during refraction. An
internal reflector at visible wavelengths has challenging tolerances for surface
texture.
Objective Lenses 125

Figure 12.12 20X solid parabola at 1.00 NAP. NAM = 1.00 is derived from a 30-mm
entrance pupil and a 10-mm EFL.
Chapter 13
Tube Elements
13.1 Doublet Tube Lens
A stock 200-mm doublet may serve effectively as a tube lens as shown in Fig.
13.1. The performance on axis is diffraction limited. At a 6.0-mm image field,
there are several aberrations. The ray intercepts of the x plane indicate defocus in
the form of Petzval curvature. The ray intercepts of the y plane indicate
astigmatism and coma. The bulk of the aberration is field curvature.
Consequently, the production doublet is effective within a small field with a
CCD sensor.
A custom 200-mm doublet may provide a tube lens with flat field (Fig. 13.2).
The performance on axis is diffraction limited. At a 6.0-mm image field, there is
no field curvature. However, the ray intercepts of the y plane indicate a small
coma. The encircled-energy plot indicates nearly diffraction-limited
performance. A telecentric lens stop is essential to this design. The custom
doublet is an effective design for a CCD sensor with a 12-mm diagonal.

13.2 Doublet-Pair Tube Lens

A typical infinity-corrected microscope employs a doublet pair as shown in Fig.
13.3. The first doublet is plano convex. It provides three features: optical power,
correction of spherical aberration, and correction of axial color. The second
doublet is a meniscus lens with little optical power. It provides correction of
coma and lateral color. The “bending” of the second lens from flat to curved
corrects the coma of the first lens.
The doublet pair based on the Abbe number provides excellent correction of
geometric aberration (Fig. 13.4). The performance on axis is diffraction limited at
the d line. There is only axial color. At a 6.0-mm image field, there is no field
curvature. However, the ray intercepts of the y plane indicate a significant lateral
color for the g line (426 nm). The encircled-energy plot indicates nearly
diffraction-limited performance at the d line at a 6.0-mm field height. A
telecentric lens stop is essential to this design. The doublet pair is an effective
solution for a CCD sensor with a 12-mm diagonal.

127
128 Chapter 13

Figure 13.1 Stock doublet acts as a 200-mm tube lens. Lens stop is 10 mm in diameter. A
telecentric lens stop is created by placement of a lens stop at the front focal point.

Figure 13.2 Custom-doublet tube lens.

Tube Elements 129

Figure 13.3 Doublet-pair tube lens layout. A custom doublet acts as a 200-mm tube lens.
The lens stop is 10 mm in diameter. A telecentric lens stop is created by placement of a
lens stop at the front focal point.

The doublet pair based on the Abbe number (shown in Fig. 13.4) may be
considered an achromat for the d and g lines. Figure 13.5 displays the axial color
of the doublet pair. The “U shape” indicates complete correction for two
wavelengths. However, the plot does not display the “S shape” of an apochromat,
where a vertical line indicates three corrected wavelengths. The lens is certainly
an achromat; however, it is not apochromatic. The doublet pair might be called a
semiapochromat because it is substantially corrected for three wavelengths.
Implementation of real glass from the 1990s yields improved performance of
the doublet pair as shown in Fig. 13.6. The chromatic correction is greatly
improved by application of known glass types to the prescription. The g line
displays less lateral color. However, there is a cost to correct lateral color. The
lanthanum flint LaFN7 and the short-flint special KzSFN4 are highly reactive
with the atmosphere and cleaning agents. The schwer krone SK10, the barium
schwer flint BaSF4, and the short-flint special KzSFN4 are obsolete as of 2010.
Implementation of this design requires new glass types along with new values for
radius and thickness. See Chapters 17 through 19 for additional information on
glass types and correction of lateral color.
The doublet pair with real glass in Fig. 13.6 should be considered an
achromat. Figure 13.7 displays the axial color of the doublet pair with real glass.
The lens with real glass is barely an achromat. However, the doublet pair may be
called a semiapochromat because it is substantially corrected for three
wavelengths.

13.3 Filter Types

There are numerous types of filters: a long-pass filter transmits long
wavelengths; a short-pass filter transmits short wavelengths; a bandpass filter
transmits a narrow band; a dichroic mirror typically reflects shorter wavelengths;
a notch filter reflects a narrow wavelength range.
Most filters comprise a dielectric stack on a 3-mm-thick substrate of fused
silica. According to Eq. (5.35), the filter wavelength displays a dependency on
the angle of incidence:

λF
λ′F = , (13.1)
cos θ

where λF is the filter wavelength at normal incidence and λ ′F is the equivalent

wavelength at an incident angle θ.
130 Chapter 13

Figure 13.4 Doublet-pair tube lens based on the Abbe number. The g line indicates lateral
color. Glass is based on numeric values of the refractive index and the Abbe number at
the d line.

Figure 13.5 Axial color of a doublet-pair tube lens based on the Abbe number. A custom
doublet acts as a 200-mm tube lens. The “U shape” indicates an achromat.
Tube Elements 131

Figure 13.6 Doublet-pair tube lens based on real glass. Application of real glass yields
better color correction. However, all but one of the glass types is obsolete as of 2010.

Figure 13.7 Axial color of a doublet-pair tube lens based on real glass. A custom doublet
acts as a 200-mm tube lens. The “U shape” indicates an achromat.
132 Chapter 13

Filter cubes employ thin 1.1-mm thick beamsplitters that reduce the
displacement of the beam to < 1 mm. However, they are notorious for warping.
The surface tension of the dielectric stack may deform a round filter into a
spherical surface. However, surface tension frequently deforms a rectangular
filter far beyond a sphere. The effects of deformation are small in transmission,
but huge in reflection. A flatness specification is warranted for a filter, especially
when the reflection of a filter is part of the imaging system.
A colored-glass filter employs absorption for transmission of long
wavelengths. A colored-glass filter is independent from angle of incidence.
However, colored-glass filters do emit fluorescence: absorption of a green laser
beam by a colored-glass filter may create an orange emission from the filter.
A hard filter is comprised exclusively of metal oxides on a silica substrate.
Numerous layers define the shape of the transmission spectrum. They are
extremely durable during exposure to the atmosphere. Hard filters are a
development from wavelength multiplexing in fiber-optic communications. The
dielectric films are deposited under vacuum on a large substrate and then cut into
smaller sizes.
A soft filter comprises two filters. The first filter is a short dielectric stack on
the first substrate. The first filter provides multiple peaks in transmission. The
second filter is a colored glass that absorbs the unwanted peaks in transmission.
Absorption by the second filter frequently converts to fluorescence. The first and
second filters are bonded to each other inside a ring, which provides a barrier
against attack by the atmosphere. A soft filter may have a significant wedge that
shifts the image location. There is a preferred direction of travel in a soft filter,
because the dielectric filter should be the incident filter. Otherwise, the
absorption filter encounters excess light, which promotes damage to the
absorption filter.
Neutral density (ND) filters come in two types. An adsorptive ND filter
absorbs light throughout the volume of the filter. It appears gray or black. A
reflective ND filter employs a thin film of inconel for both reflection and
absorption. Inconel is an alloy that is extremely resistant to oxidation and
corrosion. Its primary composition is nickel (72%), chrome (15%), and iron
(10%).
Filters are normally specified by optical density (OD), as follows:

OD = − log10T , (13.2)

where T is the transmittance. Thus, an optical density of 3 indicates a

transmission of 0.001. A small optical density of 0.05 adds linearly to another
small optical density, such as 0.10. However, large optical densities do not
necessarily add together due to large reflections.
Tube Elements 133

Figure 13.8 Filter doublet. An absorbance indicates an absorption filter without reflection
(A). R indicates a reflection filter without absorption. Light within a high-reflectance cavity
has equal probability of exiting through either filter.

Figure 13.8 displays a filter doublet wherein light may circulate in the cavity
between the filters. An absorption filter is described by absorbance A, while a
reflection filter is described by reflectance R. A sequence of two absorption
filters yields the square of the transmittance values as the total transmittance.
Consequently, their optical densities add to the total optical density. A sequence
of two reflection filters yields far more than the square of the transmittance
values. Addition of the second filter cuts the transmittance in half. Consequently,
the optical densities of two reflective filters do not add to the total optical
density. They might even disrupt each other if the separation is less than the
coherence length of the transmitted spectrum.

13.4 Filter within a Finite Conjugate Distance

A perfect tube conjugate is defined by a 10-mm stop and a 200-mm focal
distance. This defines a tube NA at 0.025, which is rather small. It can tolerate a
large thickness of cover strata without creation of spherical aberration. However,
it cannot tolerate much thickness for a tilted plate.
Figure 13.9 displays a 50-mm beamsplitting cube in perfect tube conjugate.
The ray intercept for the x plane indicates axial color. The ray intercept for the y
plane indicates both axial and lateral color. A 50-mm beamsplitting cube is a
reasonable limit for a 0.025 NA conjugate. This represents a wavefront error of
0.26λ.
Figure 13.10 displays a 1.4-mm beamsplitting plate in perfect tube conjugate.
The ray intercept for the x plane indicates proper focus. The back focal length is
driven to proper focus with the x plane. The ray intercept for the x plane indicates
astigmatism and lateral color. A 1.4-mm beamsplitting plate is acceptable within
a 0.025 NA conjugate. This represents a wavefront error of 0.34λ.
134 Chapter 13

Figure 13.9 Beamsplitting cube within a finite conjugate distance. The cube consists of 50
mm of silica. The polychromatic Strehl ratio is 90%. A perfect lens is defined by a 10-mm
lens stop and a 200-mm EFL, and NAP = 0.025. Exception from other figures: encircled
energy applies to a polychromatic combination of F, d, and C lines.

13.5 Warped Filter within an Infinity Correction

A warped plate may also create astigmatism in a collimated beam of the infinity-
corrected tube. Two common filter thicknesses are evaluated: 30 and 1.1 mm.
A tilted plate creates a lateral displacement of the beam:

 1 − sin 2 θ 
ΔyP = d P sin θ  1 − . (13.3)
 n 2 − sin 2 θ 


In fused silica at 45 deg, the lateral displacement is 0.32 times the plate
thickness. A 3.0-mm thickness creates a 1.0-mm beam shift, while a 1.1-mm
thickness creates a 0.4-mm beam shift.
In Fig. 13.11, a tilted filter at 3.0 mm in thickness is located within a tube of
an infinitely distant conjugate. A warp radius of 580 mm creates an astigmatism
at a Strehl ratio of 0.9. This represents a peak wavefront error of 0.40λ within a
10-mm lens stop.
Tube Elements 135

Figure 13.10 Plate tilted within a finite conjugate distance. A parabolic reflector defines
the tube lens as a 10-mm lens stop and a 200-mm EFL. The plate at a 45-deg tilt consists
of 1.4 mm of silica. The Strehl ratio of the d line is 90%. A perfect lens is defined by a 10-
mm lens stop and a 200-mm EFL, and the NAP = 0.025.

In Fig. 13.12, a tilted filter at 1.1 mm in thickness is located within a tube of

an infinitely distant conjugate. A warp radius of 350 mm creates an astigmatism
at a Strehl ratio of 0.9. This represents a peak wavefront error of 0.35λ within a
10-mm lens stop.
A 1.1-mm-thick beamsplitter is commonly used within a filter block of a
microscope. At 1.1 mm in thickness, a 25 × 36-mm filter plate can warp
significantly. The corners are susceptible to greater bending than the center. This
creates additional astigmatism and defocus at the corners. A tolerance on radius
of curvature or transmitted wavefront error is warranted.
The above examples employ a parabolic reflector as the tube lens. Each
system is driven to the focus of the xz plane. The parabolic reflector is perfect for
an on-axis object, while it is not suitable for an off-axis object. Consequently, an
on-axis object is employed in the examples for an infinity-corrected tube. Ideally,
a tolerance for a filter curvature should be calculated in combination with an
actual tube-lens design.
136 Chapter 13

Figure 13.11 Warped plate at 3.0-mm thickness with an infinity correction. The warped
plate at 45-deg tilt consists of 3.0 mm of silica. A 580-mm radius of curvature is derived
from a Strehl ratio of the d line at 90%.
Tube Elements 137

Figure 13.12 Plate at 1.1-mm thickness within an infinity correction. The warped plate at
45-deg tilt consists of 1.4 mm of silica. A 350-mm radius of curvature is derived from a
Strehl ratio of the d line at 90%. A tube lens is a parabolic reflector with a 10-mm lens
stop, 200-mm EFL, and NAP = 0.025.
Chapter 14
Ocular Lenses
14.1 Eyepiece
An eyepiece is a magnifier with its entrance pupil defined by the lens stop of an
objective lens. It is also known as an ocular lens. A typical ocular has a focal
length of 25 mm, which is one-tenth of the near point of human vision at 250
mm. An image conjugate of the tube lens serves as the object of an ocular. The
object is located at the front focal point of the eyepiece. The angular size of the
image is 10 times greater than the angular size of the object at the standard near
point. Consequently, the angular magnification of a typical eyepiece is 10X. The
angular magnification of an arbitrary eyepiece is defined by Eqs. (2.7) and (3.27).

14.2 Pupils
The entrance pupil of an eyepiece is defined by the lens stop of the objective
lens. In a system with finite tube length, the object NA is defined by the diameter
of the objective lens stop and the length of the tube. In a system with infinite tube
length, a tube lens creates a distant image of the objective lens stop, and the
object NA is defined by the diameter of the lens stop and the focal length of the
tube lens.
The exit pupil of an eyepiece is an image of the entrance pupil, which is the
lens stop. During proper illumination, the exit pupil appears as a bright white
disk floating above the eyepiece. The exit pupil should fit within the pupil of a
human eye. The distance from the last surface of the ocular to the exit pupil is the
eye relief.
The eye relief is an important functional parameter. At short distances of 10
to 15 mm, the eye relief prevents contact with eyelashes. At longer distances of
25 to 30 mm, the eye relief provides space for eyeglasses. A flexible eye cup on
the ocular acts as a shield from room light for an operator without eyeglasses.
The eye cup is folded backward along the ocular for operators with eyeglasses.
The diameter of the exit pupil defines an artificial pupil for the operator’s
eye. At small pupil values of 1.25 mm, the human eye is diffraction limited: this
is true even for people with minor astigmatism. This artificial pupil provides a
significant advantage over the natural pupil of 2.5 mm at which the eye is

139
140 Chapter 14

aberration limited. Proper room light is required for dilation to a natural size of
2.5 mm.
The field stop of the ocular is placed at an image field of the tube. The field
stop of the ocular defines the vision field at the specimen. An image of the field
stop may be projected onto the vision field. The size of the field stop is
dependent on the type of ocular lens. The diameter of the field stop can vary from
15 mm in a Kellner ocular to 25 mm in an Erfle ocular.
In a microscope of finite tube length in 1980, an image is located at 160 mm
from the objective flange—this defines a tube length of 160 mm with an image
located at its end. The field stop of the ocular is located at the end of the tube. A
field stop of 22 mm defines a chief ray angle of 3.9 deg in an eyepiece of
25-mm EFL. Beyond this angle, the image quality becomes more difficult to
maintain. Consequently, the lens stop of a microscope of finite tube length is
typically 22 mm or less.
In a microscope with infinity correction, a 10-mm lens stop and a 200-mm
tube lens create a tube NA of 0.025. The field stop of the ocular is located at the
back focal point of the tube lens.

14.3 Kellner Ocular

In Fig. 14.1, a Kellner eyepiece employs a doublet and a plano-convex singlet.
The doublet acts as a magnifier on the image, while the singlet acts as a relay
lens for the entrance pupil (which is the lens stop). The singlet is called a field
lens because it is located near the image field, where it has little effect on the
magnification of the field. Defects of the singlet can represent a significant
portion of a ray bundle. Consequently, the singlet is placed near the image field
but not at the image field.
In Fig. 14.2, a Kellner ocular lens is formed by two stock items, a plano-
convex lens of 25-mm EFL and doublet of 30-mm EFL. The maximum field is
limited to 7.5 mm in the field stop and 17 deg in the image. The collinear ray-
intercept plots of the x plane indicate a nearly perfect spot on axis with a small
Petzval curvature. The ray-intercept plots of the y plane indicate axial color,
lateral color, and astigmatism. The field-stop diameter is limited to 15 mm by the
outer diameter of the doublet.

Figure 14.1 Kellner ocular at a finite distance to the lens stop. The lens stop (LS) is 8 mm
in diameter. The tube length is 160 mm. The object NA is 0.025. The field stop (FS) is 15
mm in diameter. The exit pupil (EP) is 1.6 mm in diameter at an eye relief of 15.5 mm. the
field of the image is 17 deg. The field lens (FL) is a standard plano-convex lens at 50 mm
in focal length. The magnifier (M) is a stock doublet at 30 mm in focal length. The EFL is
25.0 mm at the d line. The angular magnification is 10X the near-point view.
Ocular Lenses 141

Figure 14.2 Kellner lens report. Wavelengths are at the F, d, and C lines. Object height is
7.5 mm in the lens report. Maximum field diameter is 15.0 mm.

A Kellner lens is an excellent low-cost option with perfect image quality on

axis. The angular field at 17 deg and the field stop at 15 mm are both small by
today’s standards. Replacement of the doublet with a plano-convex singlet
creates a Ramsden eyepiece, which is a predecessor to the Kellner.

14.4 Plössl Ocular

A Plössl eyepiece is formed by two modifications to the Kellner. The first
modification involves splitting the doublets of a Kellner eyepiece into two
doublets, and the second modification involves removing the field lens. Splitting
the magnifier of the Kellner into two doublets greatly increases the diameter of
the magnifier while maintaining its focal length. The increased diameter of the
magnifier permits elimination of the field lens. A Plössl benefits from a
telecentric object space because less bending of rays is required. A Plössl
eyepiece is common in telescopes.
In Fig. 14.3, a Plössl ocular is formed by a pair of standard-stock doublets at
50 mm in focal length. The collinear ray-intercept plots of the x plane indicate
near perfection on axis. The sloped ray-intercept plots of the x plane indicate
field curvature. The ray-intercept plots of the y plane indicate astigmatism and
lateral color. The field-stop diameter is limited to 25 mm by the outer diameter of
the doublet.
142 Chapter 14

Figure 14.3 Plössl lens report. The lens stop is 50 mm in diameter at 1000 mm from the
field stop. The object NA is 0.025. The exit pupil diameter is 2.0 mm at an eye relief of
19.0 mm. Doublets are identical stock products at 50 mm in focal length. The EFL of the
doublet pair is 26.6 mm at the d line. The angular magnification is 9.4 times the near-point
view. Object height is 9.0 mm in the lens report. Maximum field diameter is 25.0 mm.

A Plössl lens is an excellent low-cost option for a wide-field eyepiece. The

angular field at 28 deg and the field stop at 25 mm are both large by today’s
standards. Glass optimization can reduce the lateral color.

14.5 Erfle Ocular

A historical development of the Erfle eyepiece is described by two prescriptions.
The original patent by Erfle in 1923 is displayed in Fig. 14.4. It was based on the
limited selection of available glass at the time. An Erfle eyepiece of 1997 is
displayed in Fig. 14.5 with higher-index glass and low-dispersion lanthanum
glass. The glass types of these Erfle lenses are displayed in the glass map of Fig.
14.6.
The Erfle design of 1923 employs glass from the lower region of the glass
map: borosilicate krone (BK) and flint (F). The refractive index does not exceed
1.62. Consequently, the element curvatures are steeply curved, which enhances
lateral color and astigmatism. The dispersion of these glass types does not exceed
36. Consequently, there is limited color correction at large field angles. The field-
stop diameter is limited to 25 mm by the outer diameter of the elements.
Ocular Lenses 143

Figure 14.4 Erfle lens of 1923. The lens stop is 10 mm in diameter at 200 mm from the
field stop. The object NA is 0.025. The exit pupil diameter is 2.3 mm at an eye relief of
18.7 mm. The elements are custom to the design. The EFL of the doublet pair is 25.2 mm
at the d line. The angular magnification is 9.9X the near-point view. Object height is 9.0
mm in the lens report. Maximum field diameter is 25.0 mm.

The Erfle lens of 1923 is aberration limited. The collinear ray-intercept plots
of the x plane indicate near perfection on axis. The flat ray-intercept plots of the x
plane indicate no field curvature. The ray-intercept plots of the y plane indicate
astigmatism. The encircled-energy plot indicates a spot radius of 3 mrad. The
line-spread plot indicates a line width of 4 mrad.
The Erfle lens of 1997 employs glass from the upper region of the glass map,
such as schwer flint (SF), schwer krone (SK), and lanthanum krone (LaK). The
refractive indices greatly exceed 1.62. Consequently, the element curvatures are
less steep, thus reducing astigmatism. The dispersion exceeds 36. Consequently,
there is improved color correction at the field margin. A negative relative partial
dispersion ΔPgF of the lanthanum krone LaK8 provides correction of color in the
blue. Additional information on glass types is found in chapters 17 through 19.
The Erfle lens of 1997 is nearly diffraction limited. The collinear ray-
intercept plots of the x plane indicate near perfection on axis. The sloped ray-
intercept plots of the x plane indicate a small field curvature. The ray intercept
plots of the y plane indicate small lateral color. The difference in slopes between
the x and y planes indicates a small astigmatism. The astigmatism and lateral
144 Chapter 14

Figure 14.5 Erfle lens of 1997. The lens stop is 35 mm in diameter at 200 mm from the
field stop. The object NA is 0.25 telecentric. The exit pupil (EP) is 2.1 mm in diameter at
an eye relief of 18.8 mm. The field doublet (FD) and the pupil doublet (PD) are custom
doublets. The middle singlet (MS) is a custom biconvex singlet. The EFL of the lens
system is 29.0 mm at the d line. The angular magnification is 10.0 times the near-point
view. Object height is 9.0 mm in the lens report. Maximum field diameter is 25.0 mm.

color are balanced in magnitude. The lens is substantially optimized. The flat of
the encircled-energy plot indicates a spot radius of 1.0 mrad. The line-spread plot
indicates a line width of 0.4 mrad.
The pupil doublet and the middle singlet contribute most of the optical
power. The field doublet has little effect on the angle of the chief ray. At 9.0 mm
in the object, the first doublet bends the chief ray by < 1 deg. The pupil doublet
and the middle singlet contribute negative astigmatism due to their positive
powers. The pupil doublet is substantially corrected for lateral color, while the
middle singlet adds negative lateral color.
The field doublet effectively counters the aberrations of the pupil doublet and
middle singlet. The biconcave flint of the field doublet creates a large positive
lateral color. The steep convex surface of the field doublet creates a large angle
of incidence for a chief ray. Consequently, the steep convex surface of the field
doublet creates a positive astigmatism.
The distortion of the Erfle 1997 eyepiece is displayed in Fig. 14.7. A square
box in the object field appears as a pincushion in the image field. This represents
positive distortion that is caused by a negative growth in focal length with field
Ocular Lenses 145

height. The shorter focal length of the margin creates a greater magnification at
the margin. A linear feature of the object is curved in the image field. An
addition of a negative field singlet may reduce positive distortion.
Most of today’s eyepieces are derivatives of the Erfle eyepiece. A Skidmore
eyepiece splits the middle single into two singlets. An aspheric field lens may
completely eliminate field curvature or distortion.

Figure 14.6 Glass map for Erfle oculars.

Figure 14.7 Grid distortion for the Erfle lens of 1997. The pincushion distortion is a
positive distortion of 6.3% between the object and image. Wavelength is at the d line.
Object height is 9.0 mm.
Chapter 15
Sensors
15.1 CCD Sensors
A CCD sensor is a charge-coupled device. Potential wells for electrons are
created by extrinsic doping of a semiconductor. The metal-oxide gate of a field-
effect transistor determines the depth of a well. An external voltage may raise or
lower a gate. A specific gate sequence transfers electrons between wells. The
charge-coupling process converts photoelectrons into a video signal. Holst and
Lomheim,33 provide a broad review of CCD and complementary metal-oxide
semiconductor (CMOS) sensors. Janesick34 provides a more rigorous review of
CCD technology.
Figure 15.1 displays the structure of a potential well of a CCD sensor. An n-
type material donates negatively charged electrons. A p-type material donates
positively charged holes. The bulk material is p-type silicon. The Fermi level
defines an equal probability for a hole and an electron. n-implantation raises the
Fermi level by donation of electrons. Consequently, the conduction-band edge is
lower in the n-doped region. This defines the base of the well. The full-well

Figure 15.1 Potential well of a CCD. Photon absorption promotes electrons (filled circles)
from valence band to conduction band. N-doping of bulk p-substrate creates a potential
well for photoelectrons. Shallow p-doping isolates the potential well from surface states. A
buried channel is defined by deep n-implantation and shallow p-implantation.
147
148 Chapter 15

capacity is proportional to the depth of n-implantation. A deep well is defined by

deep implantation of n-type atoms. The extinction depth of silicon is
approximately 100 nm at 500 nm in wavelength. The well should exceed this
depth for maximum quantum efficiency of conversion of photons into electrons.
A buried channel is formed in two steps: deep implantation of n-type atoms
and shallow implantation of p-type atoms. Deep n-implantation requires a small-
mass ion with a high energy. A small-mass ion loses a small fraction of its
momentum during a collision with the base network of large-mass elements.
Nitrogen and phosphorous are low-mass n-type ions. A high-energy ion is
created by a particle accelerator with a high voltage. Shallow p-implantation
requires a large-mass ion with a small energy. A large-mass ion element quickly
transfers momentum to other large-mass elements of the base network. Gallium
and indium are p-type ions with a large mass.
In summary, a buried channel requires two specific fabrication techniques: a
high-energy low-mass n-implantation and a low-energy high-mass p-
implantation. These features are becoming less common in silicon foundries,
which produce CMOS devices such as flash memory and cell-phone cameras.
The metal gates occupy a significant portion of the pixel area. Consequently,
the quantum efficiency of a front-illuminated CCD may be only 10%. A back-
thinned CCD permits back illumination without obscuration by electrical
contacts. The quantum efficiency of a back-illuminated CCD may exceed 90%;
however, removal of the back surface is an expensive process. The back surface
must be thinned to much less than the extinction depth.
Figure 15.2 displays a CCD circuit. Each field-effect transistor (FET) defines
a well. The voltage of the metal-oxide gate may raise the bottom of the well. The
raised gate drives photoelectrons into an adjacent well. A column is defined by a
vertical array of pixels, where each pixel holds a packet of electrons. A clock
sequence drives a packet along a column into a shift register. Another clock
sequence transfers a row of packets to a sense node. The sense node converts the
packets of charge into an analog or digital video signal.
There are two important types of noise from a CCD: dark current and read
noise.34 The dark current is created by thermal promotion of carriers into the
conduction band. Surface states can randomly eject electrons as spurious charge
or popcorn noise. A buried channel is essential for reduction of spurious charge
from surface states. The read noise is created during amplification at the sense
node. The thermal noise of the load resistor is typically the main source of read
noise. Numerous types of noise exist within a CCD sensor.34
A Kodak KAI-1020 provides an example of noise in a CCD sensor at 40° C.
Its pixel dimension is 7.4 μm. The full-well capacity of the 7.4-μm pixel in the
KAI-1020 is 40,000 electrons. The dark current is typically 0.2 nA/cm2, which
translates to 600 electrons per pixel per second. The typical read noise is 10
electrons per pixel per second. At a frame rate of 30 Hz, the dark current
contributes 20 electrons toward a total dark noise at 30 electrons. These
specifications are readily available from the Kodak device specifications for the
KAI-1020. 35
Sensors 149

Figure 15.2 Circuitry of a CCD. A pixel is defined by a metal-oxide semiconductor (MOS)

gate on a field-effect transistor (FET). A gate sequence delivers an electron packet to a
capacitor at the sense node. The sense FET converts the packet to a video signal. Drive
voltage (VD) sends current through the sense FET and load resistor (RL). A reset gate
restores the sense node to the reference voltage (VR).

The integration of dark current can be reduced by cooling or electron

multiplication. Thermoelectric (TE) cooling may greatly reduce both read noise
and dark current; however, TE cooling requires a dry enclosure for the sensor.
Electron multiplication by a gain register before the sense node may quickly
amplify the electron count without integration of dark current. An electron
multiplication CCD (EMCCD) eliminates dark current at a cost of spurious
charge. An EMCCD is an attractive alternative to TE cooling, which is more
common in optical instruments.
A third type of noise is the shot noise of the photoelectrons. The full-well
capacity of an 8-μm pixel is typically 40,000 electrons as the saturation signal.35
This requires a 16-bit digital number (65536) for counting every photoelectron.
However, the shot noise is the square root of the expected count. The shot noise
for a full-well capacity of 40,000 electrons is 200 electrons. The maximum
signal-to-noise (SNR) ratio is 200. The shot noise occupies 8 bits (256). Thus, a
16-bit full-well capacity has an 8-bit range above the shot noise. An 8-bit digital
video signal should be sufficient for a maximum contrast at 200 SNR with shot
noise. The maximum contrast with shot noise occurs at the full-well capacity.
There are two common modes of frame transfer in CCDs for microscopes:
full-frame transfer and interline transfer. A frame is defined by the pixel format,
such as 1000 × 1000. A typical data rate is 40 MHz for an 8-μm pixel. The frame
rate for this example is approximately 40 Hz. Additional operations per frame
may reduce the frame rate to 30 Hz. The read noise increases with data rate due
to a larger temporal bandwidth. The dark-current noise decreases with frame rate
150 Chapter 15

due to a longer temporal integration. The data and frame rates should be chosen
wisely with regard to the application.
A full-frame transfer device requires a CCD array with two adjacent frames
on the same chip. A bright frame is exposed to the image. A dark frame is buried
beneath an opaque shield. A full frame is transferred from the bright frame to the
dark frame in approximately 1000 data cycles. A data cycle is 0.025 ms at 40
MHz. A full-frame transfer to the video signal requires one million data cycles,
which is 25 ms. Additional photoelectrons are collected in the bright frame
during frame transfer to the dark frame. A bright spot during frame transfer can
smear into a dark background.
An interline-frame transfer device requires an array of bright and dark
columns. A bright column is exposed to the image during integration. A dark
column is buried beneath an opaque shield. A bright column transfers charge to a
dark column in one data cycle. The dark column is transferred to the video signal
while the bright column collects more light. A progressive scan delivers an entire
frame to the video signal in a sequence of contiguous rows during one million
data cycles, or 25 ms. A progressive scan may employ dual outputs toward a
double frame rate. An interlaced scan delivers half of the frame during 0.5
million data cycles, or 13 ms. However, there is a delay of 13 ms between the
odd and even rows. Consequently, a rapidly moving feature may not correlate
properly between odd and even rows.
An interline-transfer CDD might employ a microlens as shown in Fig. 15.3.
The microlens is spherical or cylindrical. Its shape is determined by the surface
tension of a polymer during solidification. A microlens collects light over the
entire pixel. A spherical lens directs light from a lens stop into a spot on the full-
frame sensor. A cylindrical microlens directs light from a lens stop into a bright
column of an interline sensor. The microlens of Fig. 15.3 may increase the
quantum efficiency of an interline sensor from 10 to 40%. The projected lens
stop of the microlens is defined by a distant image of the bright column.
Consequently, the microlens is telecentric. An F/2.8 microlens defines a half
angle of collection at 10 deg. Many wide field lenses cannot accommodate this
telecentric acceptance angle.

Figure 15.3 Microlens (ML) of a CCD. The microlens is centered over the bright column
(BC). Dark columns (DC) are located beneath dark shields (DS). Marginal rays (MR) from
the lens stop (not shown) are directed onto the bright column.
Sensors 151

Figure 15.4 Circuitry of an active pixel sensor (APS). An APS contains a photodiode (PD)
and three transistors. The signal is delivered directly to the video signal. Drive voltage VD
sends current through the sense FET and load resistor RL. The reset gate restores the
sense node to the reference voltage VR.

15.2 Active Pixel Sensors

An active pixel sensor (APS) employs amplification of a photodiode output
within each pixel. Figure 15.4 displays a three-transistor circuit for an APS. Each
pixel contains a photodiode, sense transistor, select transistor, and reset transistor.
An APS frequently employs the same fabrication methods as CMOS technology
for flash drives, which can tolerate large amounts of noise. A CMOS APS
normally displays much more noise than a similar CCD sensor.
A Kodak CMOS sensor KAC-00401 provides an example of CMOS noise at
20° C. Its pixel dimension is 6.7 μm. The typical dark current is 0.9 nA/cm2,
which translates to 2100 electrons per pixel per second. At 30 frames per second
(fps), the dark current contributes 70 photoelectrons. The typical dark current is
39 mV per pixel per second. At 30 fps, the dark current contributes 1.3 mV.
There is ~ 1 V per 50,000 electrons at the sense node. The typical read noise is
4.3 bits within a 12-bit analog-to-digital converter. If 41,000 photons are scaled
to a digital count of 4096 (12 bits), then 1 bit equals 10 photons, and a 4.3-bit
read noise equals 200 photons. The combination of dark current and read noise
yields a total dark noise of 270 photoelectrons at 30 fps. This is nine times
greater than the CCD sensor. After consideration of the smaller area of the
CMOS pixel, the noise per area of the CMOS device is 10 times greater than that
of the CCD device. These specifications are readily available from the Kodak
device specifications for the KAC-00401.36
152 Chapter 15

Figure 15.5 Potential well of a CMOS detector. Photon absorption promotes electrons
(filled circles) from valence band to conduction band. N-implantation of the bulk p-
substrate creates a potential well for photoelectrons. The shallow well is exposed to
surface states. CMOS foundries do not typically provide deep n-implantation or shallow p-
implantation.

Figure 15.5 displays a typical structure of a CMOS detector. The n-

implantation forms a shallow well with exposure to surface states. Many CMOS
foundries do not offer deep n-implantation or shallow p-implantation. These
processes are not required for flash memory. Most CMOS devices operate at
3.3 V, which offers lower power consumption than a CCD at 5 volts.

15.3 Photomultiplier Tubes

A photomultiplier tube (PMT) defines the vision field as a single pixel. It is
reasonable to assign a pixel collection efficiency of 100% to a PMT, because the
spot will likely fall completely within the single pixel. The dark current is
extremely low. A high voltage of nearly 1000 V is required for operation. Its
temporal response is limited by a resistor-capacitor (RC) circuit.
Figure 15.6 displays the electrical circuit of a PMT. The photocathode of the
PMT has a low work function for electrons. Consequently, UV and visible
photons may eject electrons into the vacuum of the PMT. The ejected electron is
driven by an electric field toward a first dynode. The impact of the electron on
the dynode ejects multiple electrons, which are driven toward a second dynode.
This process repeats at numerous dynodes until termination on the anode
capacitor and load resistor.
A PMT may operate in reflection or transmission mode. A reflection-mode
PMT employs a reflective photocathode, as illustrated in Fig. 15.6. The electrons
are ejected from the front of the photocathode as if they were reflected. A
transmission-mode PMT employs a transmissive photocathode. The electrons are
ejected from the photocathode as if they were transmitted through the
photocathode.
Sensors 153

Figure 15.6 PMT circuit with components: photocathode (PC), dynode (D), dynode
resistor (RD), anode (A), anode capacitor (CA), load resistor (RL), and high voltage (HV)
(500–800 V).

The charge accumulation by the anode capacitor is expressed as

Q = Φ I ηQ g D τ I , (15.1)

where Φ is the incident flux on the detector, ηQ is the quantum efficiency of

conversion of photons into electrons, gDG is the efficiency of the gain in the
dynode structure, and τI is the time of integration.
The temporal decay of the output capacitor is described as

Q  −t 
V= exp  , (15.2)
CA  RL C A 

where CA is the anode capacitance and RL is the load resistance across the anode
capacitor. A second load resistor may be added in parallel to the first for
increased speed by addition of a second discharge path.
A reflection-mode PMT requires a transparent widow for passage of photons
into the vacuum and on the reflection-mode photocathode. The electrons are
excited near the surface where a strong electric field drives them away from the
photocathode toward the first dynode. There is little chance for an electron to
relax to a bound state within a reflection-mode PMT. The spatial response of a
reflection-mode photocathode is very nonuniform. There is a strip of maximum
quantum efficiency at the side of the cathode in proximity to the first dynode.
A transmission photocathode is a thin film on the inside of silica window.
The photoelectrons must travel through the film before entering the vacuum.
Consequently, there is significant opportunity for relaxation to bound states
within the photocathode. The quantum efficiency is compromised by the
154 Chapter 15

transmission mode. However, the quantum efficiency is spatially uniform across

the transmission-mode photocathode. Such uniformity promotes consistency
between instruments.

15.4 Film
Photographic film employs grains of silver halide (AgBr, AgCl, AgI), which
convert photons into metallic silver. This process has three main steps. First a
photon is converted into a valence electron with mobility in the grain. Second, an
excited electron falls into a trap as an invisible record of photon absorption.
Third, a developer converts the excited silver halide into colloidal metallic silver.
Silver displays one of the largest intrinsic reflectance of any metal within the
visible wavelength regime. An interpolation between 2.0 and 2.5 eV estimates
the refractive index as 0.263 + i3.94 at the d line (2.11 eV). The magnitude of the
complex vector is 3.95. Within air at index of 1.00, this yields a reflectance of
94% at a maximum absorbance of 6%. Within an emulsion at an index of 1.50,
this yields a reflectance of 92% at a maximum absorbance of 8%. Most of the
optical density is due to reflectance. Consequently, the generation of heat is
minimized.
Figure 15.7 displays an exposure curve for a monochrome film for visible
photography. Table 15.1 displays specific parameters. The 400-speed film has a
3.7-μm grain, while the 100-speed film has a 7.5-μm grain. The 400-speed film
responds to a lower exposure as measured in fluence (Talbots per area).

Figure 15.7 Optical density of photographic film, exhibiting exposure curves of

monochrome film at speeds of 400 and 100. Exposure is measured in fluence [Talbots per
area (lx·s)]. Optical density was defined in Eq.(13.2). Parameters of T-400: 60 cyc/mm as
a high-frequency limit, 7.5-μm Gaussian pixel, 0.004 lx·s. Parameters of T-100: 120
cyc/mm as high-frequency limit, 3.7-μm Gaussian pixel, 0.001 lx·s. The threshold count is
230 photons for both T-100 and T-400. Both require development in D-76 for 10 min at
20° C. Data points were measured from plots within the Kodak specification F-4016.
Sensors 155

Table 15.1 Monochrome film parameters.

Frequency (cyc/mm) Lens stop Pixel size Read noise

Film 0.5 MTF 0.0 MTF NAG NAM (μm) (lux·s) (p/µm2) (photons)

T-400 60 120 0.10 0.14 7.5 1 × 10–3 4 230

T-100 120 240 0.20 0.28 3.7 4 × 10–3 16 230

However, the 400-speed film has a larger dark threshold. Both films exhibit the
same dark threshold at 230 photons per crystal. The 100-speed film exhibits
better contrast with shot noise because 4X as many photons are required for the
same optical density. At a 4X photon count, the contrast with shot noise is 2X
greater [Eq. (8.6)]. In terms of decibels, at a specific OD, the T-100 film requires
6 dB more photons toward a 3-dB greater contrast with shot noise.
The marginal NA of the 400-speed film nearly matches that of a typical tube
lens, 0.025. Consequently, the 400-speed film is optimized for microscopy.
However, as mentioned earlier, the larger pixel of the 100-speed films creates
better contrast with shot noise.
Frequency data are measured from modulation transfer curves of the Kodak
specifications for T-MAX film.37 The marginal NA is derived from the cutoff
frequency in Eq. (19.37), and the Gaussian NA is derived from Eq. (5.49).
Chapter 16
Human Vision
16.1 Physiology
There are three basic components of human vision: the point spread of the lens,
the lateral inhibition of the retina, and the saccation of the eye. Each component
is described by a MTF, which acts on the irradiance. The physiology of the retina
includes rods, cones, horizontal cells, bipolar cells, amacrine cells, and ganglion
cells.38
The human eye evolved in numerous steps.39 The eye began as a simple
photoreceptor that detected light and shadow. Then, an eye cup created a camerai
with a finite angular field of view. A retina created an image. A fixed lens
sharpened the image. A variable iris controlled brightness. And lastly, a dynamic
lens controlled focus. Even Charles Darwin confessed to such a progression by
natural selection as “absurd.”ii However, Darwin did believe that incremental
improvements could define a progression by natural selection. The intermediate
steps are present within vertebrates of the current day.39
The foveaiii comprises tightly packed cones with very few rods. The absence
of blood vessels and other cells creates a pit that defines the spatial limit of the
fovea. The foveola (the center of the fovea) comprises red, green, and blue cones
without any rods. The annular portion beyond the foveola contains a few rods but
is still largely clear of blood vessels, bipolar cells, amacrine cells, and ganglion.
The current model employs an average cone diameter at 3.8 um. At this diameter,
in a hexagonal packing scheme, the foveola contains 5000 cones within a 200-μm
diameter, while the fovea contains 20,000 photoreceptors within a 400-μm
diameter.40 Within the fovea, there is a single incremental ganglion per
photoreceptor. Beyond the fovea, rods are more common.
The photoreceptors connect to bipolar cells that feed into ganglia. An
incremental bipolar-to-ganglion path responds to an increment of light. Sjostrand
et al.40 calculated three ganglia per foveal cone from experimental quantification
of both. Thus, a single cone within the fovea connects to three ganglia: one
incremental ganglion, one decremental ganglion, and a third type, such as a
parasol ganglion. An incremental bipolar-to-ganglion path responds to an
increment of light. A decremental bipolar-to-ganglion path responds to a
i
A camera is a closed chamber with a hole. From Greek kamera, meaning vault.
ii
See full quotation in Lamb et al.39
iii
Fovea, Latin, meaning pit.
157
158 Chapter 16

decrement of light. The bipolar and ganglion cells of the fovea are laterally
displaced beyond the fovea. The displacement of these cell bodies creates a pit.
Horizontal cells create lateral inhibition from adjacent photoreceptors. The
horizontal cells secrete gamma amino butyric acid (GABA), which inhibits
transduction from the photoreceptors to the bipolar cells.
The periphery is defined by the region beyond the fovea, where blood vessels
and other cells are prevalent. These features of the peripheral retina may scatter
incident light, which degrades the optical point spread. Both cones and horizontal
cells grow larger with distance from the fovea. A receptor pool comprises
multiple receptors with connection to a single ganglion. The effective size of a
peripheral photoreceptor is much larger than a foveal receptor, due to receptor
pooling. The morphology of a peripheral horizontal cell is much different from a
foveal horizontal cell. Numerous features of the periphery degrade spatial
resolution as defined by a point spread of the system.
This model is a practical implementation of transfer functions of the lens,
retina, and saccation. The variety of cone diameters is represented by a Gaussian
point spread. A foveal cone is connected to a single incremental bipolar-to-
ganglion path.

16.2 Contrast Sensitivity Function

Contrast sensitivity is defined by the reciprocal of the threshold visibility. The
peak contrast sensitivity of human vision is approximately 160 at 275 cyc/rad for
the natural pupil at a 2.5-mm diameter. The high-frequency limit occurs at about
2500 cyc/rad at a contrast sensitivity of 1.
The contrast sensitivity function (CSF) is modeled by the product of transfer
functions of the lens, retina, and saccation:

GL ( k ) GR ( k ) GS ( k )
CSF ( k ) = 160 , (16.1)
GL ( k P ) GR ( k P ) GS ( k P )

where k is the spatial frequency, GL(k) is the transform of the lens, GR(k) is the
transform of the retina, GS(k) is the transform of saccation, and kP is the spatial
frequency of the peak contrast sensitivity. Each transform is described in detail in
the subsequent sections.
Figure 16.1 displays the CSFs of three pupil diameters. Each pupil requires a
different radius of saccation for maximum contrast with the lens-point point
spread. Obviously, the artificial pupil displays higher peak frequency than the
natural pupil. Consequently, the artificial pupil creates a sharper image. The exit
pupil of a microscope is an extremely important part of the optical system.
Human Vision 159

Figure 16.1 Contrast sensitivity plot of human vision. The angular diameter of a
photoreceptor (PR) is 0.23 mrad. The angular diameter of a horizontal cell is 13 times that
of the PR at 3.0 mrad. Each plot is identified by a pupil diameter and a saccation radius.
The natural pupil diameter at 2.5 mm employs a saccation radius at 10.1 PR diameters
toward a peak sensitivity at 280 cyc/rad. The artificial pupil diameter at 1.4 mm employs a
saccation radius at 2.7 PR diameters toward a peak sensitivity at 400 cyc/rad. The
artificial pupil diameter at 1.25 mm employs a saccation radius at 3.0 PR diameters toward
a peak sensitivity at 385 cyc/rad.

16.3 Point Spread of a Lens

There are two important formats of the point spread of a lens: a diffraction-
limited format of an artificial pupil and an aberration-limited format of the
natural pupil. The exit pupil of a microscope frequently defines an artificial pupil
at a 1.4 or 1.25 mm diameter. The natural pupil of the eye is defined as 2.5 mm,
which is typical for an office or a laboratory environment. The effective focal
length of the eye is 16.7 mm. Figure 16.2 displays the MTFs of the lens at three
pupil diameters.
The transform of an artificial pupil at 1.4 mm is defined by the diffraction
limit of the pupil [Eq. (19.32)]. After Eq. (19.33), the cutoff frequency for an
artificial pupil is

 2 ΝA AP 
k AP =   f E ( cycles per radian ) , (16.2)
 λ 

where NAAP is the NA of the artificial pupil and fE is the focal length of the eye.
The artificial NA of a 1.4-mm pupil is 0.042, which indicates a cutoff frequency
of 2444 cyc/rad for the d line. The angular extent of an Airy disk of this lens is
derived from Eqs. (5.19) and (1.2) as

φ Airy λ
β Airy = = 2.44 , (16.3)
fE φ AP
160 Chapter 16

Figure 16.2 Transformations by a lens of human vision. The 2.5-mm pupil is based on the
model of Artal and Navarro.41 The 1.4-mm pupil is based on the Airy pattern of an exit
pupil of a 180-mm tube lens. The 1.25-mm pupil is based on the Airy pattern of an exit
pupil of a 200-mm tube lens.

where φAP is the diameter of the artificial pupil. The angular extent of the Airy
disk is 1.0 mrad (3.5 arcmin) at 1.4 mm and 587 nm. The equivalent Gaussian
diameter is 0.7 mrad.
The transform of an artificial pupil at 1.25 mm defines an 0.037 artificial NA
and a cutoff frequency of 2200 cyc/rad for the d line. The angular extent of the
Airy disk is 1.15 mrad (3.9 arcmin) at 1.25 mm and 587 nm. The equivalent
Gaussian diameter is 0.8 mrad.
The transform of the lens at the natural 2.5-mm pupil is accurately described
by Artal and Navarro.41 Their empirical transform of the natural pupil GNP(k) acts
on flux per angle:

GNP ( k ) = (1 − CNP ) exp ( − A NP k ) + CNP exp ( − B NP k ) , (16.4)

where k is an angular frequency, and the constants of the natural pupil are
defined as follows:

ANP = 2.8 mrad/cyc, B NP = 1.0 mrad/cyc, CNP = 0.36 . (16.5)

This relation is valid up to 50 cycles per degree, or 2900 cyc/rad.

16.4 Lateral Inhibition of the Retina

The lateral inhibition of the retina is created by the photoreceptors, the horizontal
cells, and the bipolar cells. Figure 16.3 displays a simple model of transduction in
Human Vision 161

Figure 16.3 Transduction circuit of the retina. Each node represents an electric potential
of a cell. There are two feed-forward paths. The incremental path comprises: a
photoreceptor (PR), an incremental bipolar cell (BInc) with an inverted input, and an
incremental ganglion (GInc). The decremental path comprises: a photoreceptor, a
decremental bipolar cell (BDec), and a decremental ganglion (GDec). The inverted feedback
of the horizontal cell (H) on the center photoreceptor provides lateral inhibition from the
surrounding photoreceptors.

the retina. Upon exposure to light, a photoreceptor acquires more negative

charge; thus, its electric potential decreases in response to an increment of light.
There are two feed-forward paths: the incremental path creates a positive
response to an increment of light, and the decremental path creates a positive
response to a decrement of light.iv Tessier-Lavigne42 provides an effective
introduction to the transduction of the retina with simple figures. Sterling43
provides a more rigorous discussion through numerous citations of empirical
data.
The negative feedback of the horizontal cells creates lateral inhibition of on-
center bipolar cells from off-center photoreceptors. This defines a condition of
center-surround antagonism. The input signals to the photoreceptors may be
expressed by a Taylor series:6

g ′( x ) g ′′( x )
g( x + a) = g( x) − a+ a2
1! 2!
(16.6)
g ′( x ) g ′′( x )
g( x − a) = g( x) + a+ 2
a .
1! 2!

iv
In the field of neuroscience, the incremental path is the “on-center” path, while the decremental
path is the “off-center” path. A negative electric potential of a cell is called “polarization.”
Hyperpolarization increases the magnitude of the negative electric potential across the cell
membrane. Hypopolarization decreases the magnitude of the negative electric potential across the
cell membrane.
162 Chapter 16

The summation of signals at the incremental ganglion is expressed as

 g( x − a) + g ( x) + g( x + a) 
−g( x) +  . (16.7)
 3 

Combining Eqs. (16.6) and (16.7) reveals a line spread of the retina as

1
g R ( x ) = g ′′( x ) a 2 . (16.8)
3

The dependency on the second derivative indicates the retina as a detector of

spatial curvature as opposed to magnitude or slope. This dependency on spatial
curvature is very close to its actual function. Human vision is based on curvature
and not on magnitude.
The line spread of the fovea may be also described by a difference of
Gaussians:

 −4 x 2   −4 x 2 
g F ( x ) = exp  2  − exp  2  . (16.9)
 β PR   βH 

where βPR is the full angle of the photoreceptor and βH is the full angle of the
horizontal cell. This equation effectively represents the receptor fields within the
fovea, but not the periphery. The transfer function of the fovea is

 −π2β2PR k 2   −π2β2H k 2 
GF ( k ) = exp   − exp  , (16.10)
 8   8 

Figure 16.4 displays the transform of the fovea.

16.5 Temporal Feedback of Photoreceptors

There is also temporal feedback within a photoreceptor 44,45 that drives the gain of
the photoreceptor to zero in response to a steady input of light. The peak of
temporal contrast sensitivity occurs at approximately 15 Hz for an illuminance of
200 lux, which is typical for office lighting. The cutoff frequency for office
lighting is approximately 80 Hz. According to the data of Kelly,44,45 both of these
frequencies increase as illuminance does.
The temporal feedback is countered by saccation.46 A saccade defines a
motion within the limits of a small volume or “sack.” A large saccade of > 10
photoreceptors occurs on average every 200 ms. A microsaccade of < 10
photoreceptors occurs every 12 ms, which specifies a rate of 80 Hz. The
microsaccades are essential for continuous regeneration of the image in the
presence of temporal feedback.
Human Vision 163

Figure 16.4 Transformation of the fovea of human vision. A photoreceptor is

approximately 0.23 mrad (0.8 arcmin, 3.8 μm) in Gaussian diameter. A horizontal cell is 13
times larger at approximately 3.0 mrad (10 arcmin, 50 μm).

16.6 Saccation Point Spread

The probability of saccation within a radius is an exponential function as derived
by Steinmann:46

 −r 
PS ( r ) = 1 − exp  , (16.11)
 αS 

where αS is the mean of saccation. The point spread of saccation is

 −r 
g S ( r ) = exp  . (16.12)
 αS 

The line spread of saccation is described as

 −x 
g S ( x ) = exp  . (16.13)
 0.6 α S 

The transfer function of saccation is derived from the line spread as

1
GS ( k ) = . (16.14)
1 + i 2πk ( 0.6 α S )
164 Chapter 16

Figure 16.5 Correlation of saccation transform to other transforms. Product of transforms

of the lens and photoreceptor [Lens (2.5 mm)·PR] establishes a target for the saccation
transform with a radius of 10.1 PR diameters.

As shown in Fig. 16.5, the magnitude of a saccation transform is matched to

the product of transforms of the lens and photoreceptor. The lens point spread is
typically much larger than the photoreceptor. If the output signal of the retina
controls the magnitude of saccation, then the retina should drive the saccation
magnitude toward the point spread of the lens and photoreceptor. Consequently,
the saccation angle is adjusted for maximum correlation between the saccation
transform and the combined transform of the lens and photoreceptor.
Figure 16.6 displays the saccation transforms in optimum correlation to the
lens transforms at three common pupil diameters. The smaller pupils are artificial
pupils. The exit pupil of a microscope or telescope may significantly sharpen the
point spread of the eye as indicated by the larger MTF at higher frequencies.

Figure 16.6 Transformation by saccation of human vision. Each saccation radius has a
specific pupil diameter.
Human Vision 165

Figure 16.7 Contrast-sensitivity plot of human vision in log-log format. The parameters of
the plot are the same as the natural pupil of Fig. 17.1. Peak sensitivity is 160 at 275
cyc/rad for the natural pupil at 2.5 mm. The experimental data is an average of five
subjects. The low-frequency slope of 2 indicates a frequency-squared dependence, which
is a departure from reality. At low frequencies, the experimental slope approaches 1. Data
points were derived from application of a CAD program to a scanned image of De
Valois.47 Theoretical data is plotted at increments of 5 cyc/rad (5, 10, 15… cyc/rad).

16.7 Vision Research

Vision research indicates the study of the retina and the brain. The shape of a
CSF plot in log-log format reveals important information about transduction. In
the low-frequency portion, a frequency-squared dependency indicates a Gaussian
field of a horizontal cell. Conversely, a linear frequency dependency indicates an
exponential decay as the field of a horizontal cell. These low-frequency
dependencies are certainly relevant to vision research; however, they are not as
important for instrument design.
Figure 16.7 displays a log-log plot of the CSF with experiment data from
DeValois.47 The peak spatial frequency is near 250 cyc/rad. Thus, a 2 mrad
object matches the central lobe of the point spread of the natural pupil. The high-
frequency limit is near 2500 cyc/rad. The low-frequency limit is < 10 cyc/rad.
Peripheral spatial vision requires numerous modifications to the model. The
photoreceptors become larger with distance from the fovea.48 They are also
directed into pools by lateral connections. A horizontal cell in the periphery has
fewer and longer protrusions.49 A peripheral horizontal cell should be modeled as
an exponential decay. Such dependencies are important to neural science.

16.8 Temporal Contrast Sensitivity Function

The temporal CSF displays two components: a feed-forward and a feedback. A
feed-forward transduction may resemble the charging of a resistor and capacitor
(RC). The transform of an RC circuit is
166 Chapter 16

1
GRC ( ω) = , (16.15)
1 + iωτ RC

where ω is the temporal frequency and τRC is the time constant of the RC circuit.
A cascade of four similar reactions creates a feed-forward transform

2
1
4
 1 
GFF ( ω) = =  , (16.16)
1 + ( ωτ FF ) 
2
1 + iωτ FF

where τFF is the feed-forward time constant. The exponent of 4 represents a

cascade of four similar feed-forward reactions. The feed-forward process is
governed by electric transduction of ions.
The low-frequency limit of the temporal CSF is estimated by the feedback of
the horizontal cells. A cascade of two similar reactions creates a feedback
transform

2
GFB ( ω ) = 1 − exp ( −iωτ FB ) = 1 − cos ( ωτ FB ) , (16.17)

where τFB is the feedback time constant. The feedback process is governed by
secretion of GABA from the horizonal cell.43 GABA opens anion channels of the
photoreceptor.
Figure 16.8 displays the feed-forward and feedback components of the
temporal CSF. The temporal CSF is effectively modeled as

GFF ( ω) GFB ( ω)
CSF ( ω) = 135 , (16.18)
GFF ( ωP ) GFB ( ωP )

where ωP is the frequency of peak contrast sensitivity at 135. The horizontal cell
feedback defines the low-frequency response, while the photoreceptor feed
forward largely defines the high-frequency response.
Figure 16.9 displays the temporal CSF of the human vision model in this text
and the experimental data of Kelly.44 The lighting for the experimental data at
174 cd/m2 is similar to a computer monitor, which is normally 50–300 cd/m2.
The peak contrast sensitivity at 135 occurs at a temporal frequency of 15 Hz. The
high-frequency limit is near 80 Hz.
The temporal CSF model certainly makes a few broad assumptions.
However, the theoretical model does resemble the experimental data. Ergo, it has
practical value for instrument design.
Human Vision 167

Figure 16.8 Components of a temporal contrast sensitivity plot. Feed forward of

photoreceptors (PR) defines the high-frequency dependency. Feedback of horizontal cells
(HC) define the low-frequency dependency. HCs provide maximum feedback at 50 Hz.

Figure 16.9 Temporal CSF. Peak sensitivity at 135 occurs at 15 cyc/rad for the natural
pupil at 2.5 mm. The feed-forward and feedback time constants are identical at 10.2 ms.
The pupil has a natural diameter of 2.5 mm. Luminance is 175 cd/m2. Experimental data
44
points were derived from application of a CAD program to a scanned image of Kelly. The
theoretical plot occurs at multiples of 1.2 (2.0, 2.4…) cyc/rad.
Chapter 17
Optical Materials
17.1 Glass Types
The term glass implies an amorphous structure. Crystallization is defeated by
either rapid cooling or diversity of components. Quartz is a crystalline form of
silicon dioxide that is formed under intense pressure and heat. Silica is an
amorphous form of silicon dioxide that is formed during rapid solidification.
Glass is largely composed of the nine most common elements within the
earth’s crust (Table 17.1). Additional components of glass are listed in Table
17.2. Iron is not desirable for optical glass due to its green color. Arsenic (As)
and antimony (Sb) are fining agents that remove bubbles and impurities such as
iron. Figure 17.1 displays a periodic table of common elements in metal-oxide
glass.
Silica provides the bulk material for most optical glass. The liquidus
temperature of silica is 1715° C, and the melting point is 2000° C. The liquidus
temperature defines the transition from crystalline to amorphous structure. The
liquidus temperature is also called the glass temperature Tg, above which the
structure is amorphous. The melting point defines the transition from solid to
liquid.
Soda lime is an early form of glass. The liquidus temperature of silica is
reduced to 1000° C by the addition of soda Na2O. The viscosity is reduced by
addition of lime (CaO). Potassium oxide (K2O) may replace soda. A low liquidus
temperature and low viscosity are beneficial for blowing, shaping, and floating.
A trace amount of iron oxide content creates a bluish-green tint, which is
prominent at the edges of soda lime glass. A large amount of iron-oxide content
creates brown bottle glass.
Crown glass was originally formed by a rod with molten glass on the end.
The linear momentum of the rotation stretched the molten glass into a disk, but
not evenly. The process left a crown in the middle. A high transmittance for
windows was the primary goal in crown glass.
The original flint glass was comprised of silica and flint. Flint occurs
naturally as quartz with trace elements. The essential trace elements raise the
refractive index of silica without the addition of color. Naturally occurring flint
provides sparkle without color.

169
170 Chapter 17

Table 17.1 Abundant elements of glass in the earth’s crust by weight.

Natural forms
Rank Sym Fraction Formula gm/cm3 Name Purpose Name Formula
1 O 0.474
Base
2 Si 0.277 SiO2 2.6 Silica Quartz SiO2
network
3 Al 0.082 Al2O3 4.0 Alumina Stability Sapphire Al2O3

Hematite, Trace
4 Fe 0.041 Fe2O3 5.2 iron (III) element Rouge Fe2O3
oxide with color

Lower
5 Ca 0.041 CaO 3.4 Lime Calcite CaCO3
viscosity
Sodium Lower T
6 Na 0.023 Na2O 2.3 Soda ash Na2CO3
oxide glass
Potassium Lower T
7 K 0.021 K2O 2.4 Pot ash K2CO3
oxide glass
Trace
8 Mg 0.023 MgO 3.6 Magnesia Dolomite MgCa(CO3)2
element
Titania,
Electron
9 Ti 0.006 TiO2 4.2 titanium (IV) Rutile TiO2
density
oxide
Total 0.988
Data from webmineral.com and webelements.com.

Table 17.2 Rare elements of glass in the earth’s crust by weight.

Natural forms
Sym ppm Formula gm/cm3 Name Purpose Name Formula
As 2 As2O3 4.3 Arsenic Removes Arsenopyrite FeAsS
oxide iron
B 950 B2O3 2.5 Boron Lowers Borax Na2B4O7
oxide TCE :10H2O
Ba 500 BaO 5.7 Barium Electron Barite BaSO4
oxide density
Nb 20 Nb2O5 4.6 Niobium Electron Niobite Fe2+Nb2O6
oxide density
Pb 14 PbO 9.6 Lead Electron Massicot PbO
oxide density
Sb 0.2 Sb2O3 5.2 Antimony Removes Berthierite FeSb2S4
oxide iron
La 32 LaO 6.5 Lanthanum Electron Bastnasite (Ce, La)CO3F
oxide density
Data from webmineral.com and webelements.com.
Optical Materials 171

Figure 17.1 Periodic table of metal-oxide glass. Thick borders indicate common elements.
Arsenic (As) and antimony (Sb) bind to iron (Fe) in the process of removal.

Lead glass eventually replaced flint glass. The addition of lead oxide raises
the refractive index and lowers the liquidus temperature. A higher refractive
index creates more sparkle. A lower liquidus temperature extends the
temperature range for glass blowing and shaping. Lead is a common heavy metal
because three out of four heavy nuclei decay into a stable form of lead. Lead
glass employs arsenic as a fining agent that removes bubbles and iron.
Ecologically friendly glass is called ecoglass. Ecoglass is considered safe for
the environment due to the absence of lead and arsenic. Titanium oxide has
replaced lead oxide. Antimony has replaced arsenic as the fining agent in a
similar manner to arsenic. Antimony bonds with iron and reduces viscosity, thus
contrubuting to removal of bubbles. However, antimony is also similar in toxicity
to arsenic. Fortunately, antimony is les soluble than arsenic. Arsenic is soluble in
water as an arsenite anion, which is a significant problem in some water supplies.
The shift from lead to titanium was mandated by a directive of the European
Community in 2002. The mandate forbid the sale of electrical products with lead,
mercury, and hexavalent chromium from 2006 onward. The production of lead
glass is still legal, while the sale of an electrical product with lead requires an
exception to the law.
Float glass is derived from pulling molten soda lime glass across a bath of
molten tin. The temperature of the bath is graded from a high to a low
temperature. A solid glass ribbon is pulled off the low temperature end by rollers.
Most plate glass for windows is a form of soda lime float glass with its
characteristic green tint at the edges. Float glass is not appropriate for most
optical instruments due to its color and surface form error.
Commercial borosilicate comprises mostly silica (90%) with small amounts
of boron oxide (3%), sodium oxide (2%), aluminum oxide (2%), and potassium
oxide (1%). Boron is a small atom that bridges dangling bonds within amorphous
glass. Consequently, the coefficient of thermal expansion (CTE) is reduced by
the boron. Glass cookware and labware such as Pyrex® and DURAN® is
borosilicate.
172 Chapter 17

Optical borosilicate crown (BK7) comprises mostly silica (> 60%) with
significant amounts of boron oxide, sodium oxide, and potassium oxide (> 10%
each). There are also small amounts of other materials: barium oxide, zinc oxide,
aluminum oxide, titanium oxide, and antimony oxide. The diversity of metal
oxides promotes amorphous structure and neutral color. However, impurities still
create fluorescence and scattering. D263 by SCHOTT is an extremely pure form
of borosilicate for microscope slides and cover slips. Optical borosilicate crown
is substantially free of iron oxide. N-BK7 is based on titanium and antimony,
while BK7 is based on lead and arsenic.
An extremely pure form of borosilicate BSL7Y employs lead and arsenic
oxide for removal of impurities. The lead reduces viscosity, while the arsenic
removes iron. However, BSL7Y is 10 times more expensive than ecofriendly N-
BK7. Other lead glass types, such as PBM27, are essential for exposure to UV
and x-ray wavelengths. The purity of lead glass with arsenic greatly exceeds the
purity of titanium glass with antimony. Furthermore, lead glass does not
crystallize, while titanium glass does crystallize easily.

17.2 Glass Map

The glass map (Fig. 17.2) indicates refractive index versus Abbe number
[Eq. (5.11)]. In Chapter 14, a simplified version of the glass map for the Erfle
eyepiece was shown in Fig. 14.7. The glass curve describes the progression from
traditional crown glass through a range of flint glass as the concentration of lead
oxide increases. A material to the left of the glass curve is considered a fluorite
glass because these glasses were originally based on fluorite. As the Abbe
number increases, the dispersion decreases. Ergo, the Abbe number represents
the reciprocal dispersion.
Figure 17.2 displays today’s common types of SCHOTT glass. The labels are
based on German words as indicated in parenthesis in the following text. A
crown (or krone) has a low dispersion, whereas a flint (flint) has a high
dispersion. A heavy flint (or schwer flint) contains more lead oxide than a light
flint (or licht flint). The separation between crown and flint may be defined at an
Abbe number of 50. The glass curve describes the progression from crown glass
into flint glass as the concentration of lead oxide increases. Titanium oxide has
largely replaced lead oxide for environmental benefits. Antimony has replaced
arsenic as a fining agent, which reduces bubbles and impurities.
Lanthanum glass provides a high index with low dispersion. Its inner-shell
electrons are tightly bound, which exhibit low dispersion. Its outer-shell electrons
are tightly bound to oxygen. Thus, lanthanum’s electron structure resembles that
of silica, but there are many more electrons. Consequently, the dispersion of
lanthanum glass is similar to silica while the refractive index is much larger.
Lanthanum glass provides a high-index alternative to fluorite glass. The optical
properties of numerous glass types are listed in Table 17.3.
Optical Materials 173

Figure 17.2 Glass map for common materials. A diamond indicates accurate values for
both the refractive index and the Abbe number. A circle indicates an accurate refractive
index with an estimated Abbe number. The linear plot of circles is based on water and
TYPE A oil.

17.3 Fluorite
Fluorite glass can involve several materials: calcium fluoride (CaF2), phosphate
krone (PK), and fluorophosphate krone (FK). Phosphate krone and
fluorophosphate krone qualify as glass because they are amorphous. However,
calcium fluoride is not a glass; it is a single crystal. In the field of chemistry,
fluorite indicates calcium fluoride (CaF2); without exception, fluoride indicates
the anion F–, and fluorine indicates the element F. Figure 17.3 displays a periodic
table with common elements of fluorite glass.
The tightly bound electrons of fluorite glass define an absorption band that is
located deep in the UV regime. Consequently, a fluorite glass is effective for UV
excitation during fluorescence. Figure 17.4 displays a glass map of suitable glass
for application to fluorescence.
Calcium fluoride has a refractive index of 1.43 and an Abbe number of 95. It
is located at the far left of the glass curve. It has the same number of 2p electrons
as SiO2. However, the electrons of fluorite are more tightly bound due to the
increased charge of the fluorine nucleus. Consequently, the resonance is shifted
to a higher frequency [Eq. (19.11)] and a shorter wavelength [Eq. (17.1)].
174 Chapter 17

Table 17.3 Optical properties of common glass.

Relative
Type Label Glass name† nd νd λmin†† Λmax†† cost
Crown BK7 Borosilicate krone 1.52 64 330 2000 1
Crown PMMA Polymethylmethacrylate 1.49 57 ~365~ ~1060~ 0
Crown SILICA Fused silica, NIFS-S 1.46 68 ~350~ ~3500~ 5
Crown SILICA Fused silica, NIFS-U 1.46 68 ~250~ ~3500~ 10
Crown SILICA Fused silica, NIFS-A 1.46 68 ~190~ ~3500~ 30
Flint F2 Flint 1.62 36 390 2000 2
Flint SF1 Schwer flint 1.72 30 410 2000 4
Flint SF6 Schwer flint 1.81 25 420 2000 4
Flint SF66 Schwer flint 1.92 21 460 2000 16
Fluor FK51 Fluorophosphate krone 1.48 85 370 2000 26
Fluor LaK8 Lanthanum krone 1.71 54 370 2000 3
Fluor LaF2 Lanthanum flint 1.74 45 390 2000 4
Fluor LaSF9 Lanthanum schwer flint 1.85 32 420 2000 7
Fluor PK51 Phosphate krone 1.53 77 310 2500 27
Fluor PSK53 Phosphate schwer krone 1.55 63 365 2000 7
Fluorite CaF Calcium fluoride - NICF-U 1.43 95 ~225~ ~6000~ 50
Fluorite CaF Calcium fluoride - NICF-A 1.43 95 ~175~ ~6000~ 100
Fluorite CaF Calcium fluoride - NICF-V 1.43 95 ~125~ ~6000~ 400
Fluorite FK5 Fluorophosphate krone 1.49 70 310 2000 2
Fluorite FK51 Fluorophosphate krone 1.49 85 340 2500 18
† German – English: krone – crown; schwer – 1. heavy, 2. dense.
†† >90% at 10 mm thick.
~estimate.

The impurities of natural fluorite emit fluorescence; therein originates the word
fluorescence. Synthetic fluorite (a single crystal of calcium fluoride) offers high
levels of purity without fluorescence.
Calcium fluoride has a similar refractive index to borosilicate, but with a
much higher Abbe number. A positive element of fluorite in combination with a
negative element of short flint is an effective design for color correction in the
blue wavelengths.

Figure 17.3 Periodic table of common elements in fluorite glass.

Optical Materials 175

Figure 17.4 Glass map of materials acceptable for fluorescence. Suitable glass for
fluorescent applications: nd < 1.70 or νd > 50. The intrinsic fluorescence and/or absorption
of the upper right region is not acceptable in fluorescent applications. Acceptable region
defined by Misawa.22

Calcium fluoride is an excellent material for UV applications. It may

effectively correct chromatic aberration in combination with fused silica.
However, the UV and deep-UV grades of these materials are extremely
expensive. Calcium fluoride has a large CTE. The CTE of CaF2 is 2.6 times that
of BK7, and 33 times that of fused silica. It is prone to cracking from thermal
shock.
Fluorophosphate glass employs phosphate and silica as the base network
with the addition of fluorine. Phosphate krone (PK) is a mixture of phosphate and
silicate. Phosphate binds electrons more tightly than silica. The addition of
fluorine in place of oxygen creates fluorophosphate krone (FK). Fluorine binds
electrons more tightly than oxygen. Consequently, the resonant frequency of FK
extends further into the UV. The dispersion of BK or FK is less than that of
silicate.
An FK displays a positive relative partial dispersion ΔPgF. It is frequently
employed for correction of the secondary color spectrum between the g and F
lines. The positive relative partial dispersion of a fluorophosphate krone is
combined with negative relative partial dispersion of a short flint.
Fluorophosphate glass can present difficult issues. The material can melt
during polish. It reacts to atmosphere. A large CTE can promote cracking. The
CTE of PK51 is 1.7 times that of BK7. The CTE of CaF2 is 2.6 times that of BK7
176 Chapter 17

and 33 times that of fused silica. Fluorophosphate glass is difficult to manage but
well worth the effort.

17.4 Short Flint

A short flint displays a transmission spectrum with reduced transmission in the
blue regime. The short spectrum length is created by a broadened absorption-
band peak in the UV regime. The small absorption in the blue regime is created
by resistance to dipole current. This resistance reduces the refractive index in the
blue regime but not in the green or red regimes. The partial dispersion PgF in the
blue is smaller in a short flint than a normal glass. This defines a negative relative
partial dispersion ΔPgF.
There are several types of short flint. A kurz flint [(KzF), short flint] employs
antimony in place of lead. A kurz flint sonder [(KzFS), short-flint special]
employs a special glass such as borate (B2O3) in place of silica (SiO2). The term
“special” indicates a base network other than silica. A short-flint special KzFS of
the 1900s is based on lead borate, whereas a normal flint is based on lead silicate.
A lead-borate glass is reactive to the atmosphere.
Because of its importance as a short flint, the use of lead-borate glass persists
today; however, there are new materials in development. A niobium silicate
offers a short spectrum without the hazards of a lead-borate glass. However, the
anomalous partial dispersion of niobium silicate is normally less than that of a
lead borate.
Chapter 18 contains more details on short flint and short-flint special glass.

17.5 Anomalous Dispersion

Figure 17.5 displays the refractive indices of glass with both normal and
anomalous dispersion. The relative partial dispersion ΔPgF is an important metric
of glass, as it indicates the relative optical power in the blue spectrum (g to F) to
the optical power in the red spectrum (d to C). Section 19.3 provides a
mathematical definition of ΔPgF.
A normal glass is defined as the flint F2. The refractive index of a normal
glass grows faster in the g-to-F spectrum than in the d-to-C spectrum. An
anomalous dispersion departs from the slopes of F2.
The refractive index of dense flint SF2 grows more rapidly in the g-to-F
spectrum. Thus SF2 displays a positive relative partial dispersion ΔPgF.
The refractive index of a short flint KzFH1 grows at the normal rate in the d-
to-F spectrum; however, it grows more slowly in the g-to-F spectrum. Thus, a
short flint defines a negative relative partial dispersion in the g-to-F spectrum. A
short flint effectively corrects secondary color in the blue spectrum.
The refractive index of a lanthanum crown LaK21 grows more slowly over
all wavelengths. Thus, LaK21 has a large Abbe number. It also displays a
negative relative partial dispersion. However, a lanthanum crown is less effective
at color correction than a short flint because the slope of a lanthanum crown is
flatter in both the blue and red spectra.
Optical Materials 177

Figure 17.5 Refractive index of normal and anomalous glass. Glass types are indicated
by name (Abbe number and relative partial dispersion). The flint F2 defines a normal
relative partial dispersion near zero. The dense flint SF2 displays a positive relative partial
dispersion. The short flint KzFH1 displays a negative relative partial dispersion. The
lanthanum crown N-LaK21 displays a large Abbe number and a negative relative partial
dispersion. The parallelograms indicate the slope of the F2 plot.

17.6 Sellmeier Formula

The refractive index of a normal glass is modeled as the summation of two
electron states. A first resonator represents the bulk silicon dioxide. A second
resonator represents the flint component, lead oxide or titanium oxide. A sum of
the first and second resonators defines normal dispersion. The right portion of the
glass curve also represents normal dispersion. Addition of a third resonator, such
as lanthanum oxide, may create anomalous dispersion. Table 17.4 lists the
Sellmeier parameters for the glass of Fig. 17.5.
The refractive index of three resonators is expressed by the Sellmeier
formula as derived from the Lorentz Oscillator [Eqs. (19.9) and (19.11)]:

 K λ 2   K 2 λ 2   K 3λ 2 
n2 − 1 =  2 1 + 2 + 2 . (17.1).
 λ − L1   λ − L2   λ − L3 

The normal dispersion of F2 is defined by the first and second resonators. A large
L3 minimizes the effects of the third resonator in the visible regime. The resonant
wavelengths for the first and second oscillators are similar, as indicated by L1 and
L2. The concentrations of the first and second oscillators are indicated by K1 and
K2. SF2 displays a larger concentration of both oscillators by 5 to 10% over F2.
Consequently, SF2 displays a large index. The similar resonant wavelengths
indicate similar dispersion in F2 and SF2.
178 Chapter 17

Table 17.4 Sellmeier parameters of normal and anomalous glass.

Sellmeier F2 SF2 N-LaK21 Units

K1 1.345 1.403 1.227 per μm2
L1 0.010 0.011 0.006 μm2
K2 0.209 0.232 0.421 per μm2
L2 0.047 0.049 0.020 μm2
K3 0.937 0.939 1.013 per μm2
L3 112 112 88 μm2
SCHOTT KzFH1 Units
A0 2.55 × 10+00 —
A1 –1.29 × 10–02 per μm2
A2 1.84 × 10–02 μm2
A3 5.86 × 10–04 μm4
A4 –1.62 × 10–05 μm6
A5 2.27 × 10–06 μm8

The anomalous dispersion of LaK21 is defined by shorter resonant

wavelengths. This indicates a resonance deeper into the UV, which is beneficial
to UV excitation. Also, the second resonant wavelength is shorter than that of F2,
while the concentration has doubled. The visible spectrum now extends farther
from the UV.
The dispersion of glass normally indicates the location of absorption bands.
A low-dispersion glass indicates a short resonant wavelength without absorption
in the blue. A high-dispersion glass indicates a near-UV resonant wavelength
with some absorption in the blue. A high-dispersion glass might appear yellow
due to absorption in the blue.
The anomalous dispersion of the short flint KzFH1 requires the SCHOTT
formula as follows:

A2 A3 A4 A5
n 2 = A0 + A1λ 2 + + + + , (17.2)
λ 2 λ 4 λ 6 λ8

which is a mathematical formula without physical significance. However, it can

be essential for accurate description of anomalous dispersion. Table 17.4 displays
the SCHOTT parameters for KzFH1.
Anomalous partial dispersion is essential for correction of chromatic
aberration. Table 17.5 displays the Sellmeier coefficients of silica and fluorite
glass. The resonant wavelengths are smaller, and the silica and fluorite glass
display far less dispersion than a normal glass. However, they are more
expensive in cost.
Optical Materials 179

Table 17.5 Sellmeier parameters of silica and fluorite glass.

Sellmeier Silica PK51 FK51 CaF2 Units

K1 0.696 1.156 0.971 0.568 per μm2
L1 0.005 0.006 0.005 0.003 μm2
K2 0.408 0.153 0.217 0.471 per μm2
L2 0.014 0.019 0.015 0.010 μm2
K3 0.897 0.786 0.905 3.848 per μm2
L3 98 141 169 1,201 μm2

The relative partial dispersion for g and F is frequently cited as ΔPgF. A

fluorite krone displays a large and positive ΔPgF. A schwer flint displays a small
positive ΔPgF. A short flint KzF displays a small negative ΔPgF. A short-flint
special displays a large negative ΔPgF. An lanthanum krone or lanthanum flint
displays a negative ΔPgF. The positive ΔPgF of a fluorite krone is frequently
combined with the negative ΔPgF of a short flint.

17.7 Environmentally Safe Glass

In 2002, the European Community issued two environmental directives on the
use of lead and other substances. Lead is an important component of high-index
flint and short-flint special glass. Lead is even contained in some forms of brass.
The directive on Waste Electrical and Electronic Equipment 2002/96/EC
defines ten categories of electrical and electronic equipment. Two categories are
quite relevant to microscopes: medical devices and lighting equipment. Standards
for reuse, recycle, and disposal are defined.
The directive on Restriction of Hazardous Substances, 2002/95/EC, restricts
the use of six materials in electronic components and assemblies. Specifically,
these materials are lead, mercury, cadmium, hexavalent chromium,
polybrominated biphenyls, and polybrominated diphenyl ethers. Exceptions are
permitted when lead is necessary.
In most flint glass, lead oxide has been replaced by titanium oxide or
niobium oxide. The absorption of titanium oxide extends farther into the blue
than lead oxide. Titanium crystallizes more easily than lead oxide. The
crystalline inclusions of an ecoglass can be pulled from the glass during
manufacture. The “grind and shine” of ecoglass can be difficult, as the pulled
inclusions create scratches.
Some short flint glasses require lead borate for their unique partial dispersion
in the blue. A high-index glass with low anomalous partial dispersion is essential
for many instruments. A niobium silicate may replace some lead borate but not
all.
180 Chapter 17

Numerous challenges exist in the conversion to ecoglass. The demand for

cell phone cameras is pushing the further development of glass. Jademzik et al.
(of SCHOTT glass) describe the challenges and benefits of ecofriendly glass.50

17.8 Glass Code

A glass code defines a glass by its refractive index at the d line and its Abbe
number. The first three digits apply to the first three decimals of the refractive
index. The second three digits apply the first three digits of the Abbe number.
Thus, the glass code for K7 (1.511, 60.4) is 511-604. Other common glass codes
are 517-642 for BK7 (1.517, 64.2) and 620-363 for F2 (1.620, 36.3).
Manufacturers strive to provide identical glass types by glass code; however,
there are significant differences. The anomalous partial dispersions might not be
the same. Resistance to the atmosphere might vary. Other variations might apply.

17.9 Spectral Lines

An i-line glass indicates a tolerance to a UV wavelength of 365 nm. There are
numerous formats of i-line glass that are ecofriendly. They are based on titanium
oxide and antimony. An i-line glass is much more expensive than ecofriendly
borosilicate.
Ordinary borosilicate transmits at 330 nm (Table 17.3); however, there is a
significant absorption of 3% per 25 mm of glass at the i line (365 nm). A small
absorption of illumination can create substantial background fluorescence and/or
browning.i
Lead glass might be required for x-ray exposure (31 keV, 40 pm). Lead glass
has two distinct advantages in purity over ecoglass: reduced viscosity and lack of
titanium. The reduced viscosity of lead glass facilitates removal of impurities.
The absence of titanium prevents tight binding of impurities within titanium-
oxide crystals. Furthermore, arsenic promotes removal of impurities.
Other wavelengths of interest are listed in Table 17.6. The e line is very close
to the peak sensitivity of the human eye at 555 nm. The F′ and C′ lines shift the F
and C lines toward a better match to human vision. An optical design for human
vision is frequently based on the F′, d, and C′ lines.

17.10 Cost of Optics

There are two important components of cost in fabrication of optics: quality of
material and ease of processing. The cost of the bulk material is largely driven by
a few factors: bubbles, inclusions, striae, and purity. The cost of processing is
driven by numerous factors: surface-form errors, surface defects, cleaning,
coating, etc. Table 17.3 displays various glass parameters and relative cost of
BK7.

i
“Browning” indicates discoloration due to transformation of the glass.
Optical Materials 181

Table 17.6 Spectral lines of emission.

Wavelength
Line (nm) Source
i 365.0 Hg
h 404.7 Hg
g 435.8 Hg
He-Cd 441.6 Laser
F′ 480.0 Cd
F 486.1 H
Ar 514.5 Laser
e 546.1 Hg
d 587.6 He
He-Ne 632.8 Laser
C′ 643.9 Cd
C 656.3 H

Ecofriendly glass can display issues with inclusions and impurities. Lead and
arsenic are more effective than titanium and antimony for elimination of
impurities. Consequently, an ecofriendly glass might not tolerate UV or x-ray
exposure, while its lead-oxide predecessor tolerates exposure without issue.
Titanium oxide may crystallize in the form of small inclusions that leave
scratches during polish. Impurities may also create intrinsic fluorescence.

17.11 Structural Materials

Table 17.7 lists the quality and cost of common materials within an optical
instrument. The CTEs of doublets must be considered for large temperature
ranges. The higher CTE glass may crack during rapid change in temperature. The
density of glass is an important tool for experimental determination of a glass
type.
Stainless steel contains mostly iron with essential amounts of nickel and
chromium. Nickel promotes a face-centered cubic structure, which is more
flexible than other structures. Chromium creates a passive-oxide layer, which
does not grow into the material. Types 304 and 316 stainless steel are
recommended for biological applications. The natural chrome-oxide layer is an
acceptable finish. A black iron-oxide treatment may reduce reflectance. A small
amount of carbon (0.1%) strengthens stainless steel by bridging gaps in a similar
manner to boron in glass.
A lens barrel and retainer ring should be made of different metals. Stainless
steel provides a strong material for the threads of the barrel. Brass provides a
softer material for the threads of the retainer. The retainer might even deform by
design on contact with the lens. Brass will not stick to stainless steel, whereas
similar types of steel will fuse under friction and pressure.
182 Chapter 17

Table 17.7 Structural properties of common materials.

Density CTE Tensile†† Cost in 2008

Label Name (composition†) (g/cm3) (10-6/°C) (MPa) ($/kg) ($/L)
BK7 Borosilicate krone 2.5 7.1 3 8
PMMA Polymethylmethacrylate 1.2 80.0 50 1 1
NIFS-S Fused silica up to 350nm 2.2 0.6 48 21 46
F2 Flint 2.7 7.8 — 14 37
SF1 Schwer flint 3.0 9.1 — 28 85
SF6 Schwer flint 3.4 9.0 — 25 83
SF66 Schwer flint 4.0 5.9 — 112 448
LaK8 Lanthanum krone 3.6 5.6 — 21 75
LaF2 Lanthanum flint 4.3 8.1 — 25 105
LaSF9 Lanthanum schwer flint 4.4 7.4 — 49 216
FK51 Fluorophosphate krone 3.7 13.3 — 182 679
PK51 Phosphate krone 3.6 12.4 — 189 673
NICF-V Calcium fluorite 3.2 18.4 — 70 223
up to 225 nm
FK5 Fluorite krone 2.5 9.2 — 14 36
FK51A Fluorite krone 3.7 12.7 — 126 464
304 Stainless steel 18-8 8.0 17.2 586 13 107
(Fe, Cr, Ni)
316 Surgical stainless steel 8.0 15.9 621 20 163
(Fe, Cr, Ni)
416 Ferric stainless steel (Fe, Cr) 7.8 9.9 655 32 248
4340 Chromoly steel (Fe, Cr, Mo) 7.8 11.5 1,241 21 163
1095 Blue spring steel (Fe, C) 7.8 11.4 979 97 763
2024 Aircraft aluminum 2.8 22.8 441 111 306
(Al, Cu, Mg)
6061 Aluminum (Al, Mg) 2.7 23.4 241 53 143
360 Brass (Cu, Zn, Pb) 8.5 11.4 400 19 158
— Northern red oak, dry 0.6 — 98 2 0.1
†
Components at > 1% by weight
††
1 MPa = 10 kg/cm2

Aluminum 6061 is a common alloy for optical fixtures. It is softer than

stainless steel and easier to machine. However, it must be anodized, otherwise an
aluminum-oxide dust will form and distribute on the optics. Aircraft aluminum
2024 contains copper for additional strength, which is not likely necessary for an
optical instrument. MIC-6 is a aluminum alloy with minimal internal stress that
permits precise machining.
Chapter 18
Composition and
Spectra of Materials
18.1 Glass Structure
The descriptions of glass types in this book are largely based on the descriptions
by Clement.51 The Sellmeier coefficients are derived from vendor data as cited in
ZEMAX™. The wavelength increment of the Sellmeier index is 1%. The
transmission data are also derived from vendor data over 25 mm of internal
transmission, as cited in ZEMAX.
A glass is a random network of fused inorganic material. It may be
considered the amorphous form of a ceramic. Crystallization is defeated by a
sufficient rate of cooling for the viscosity of the glass. A glass with thick
viscosity may cool slowly due to the poor mobility of the components. A glass
with low viscosity must cool rapidly before crystals grow.

18.2 Crown
A krone, or crown (K), is basically soda lime glass:

SiO2 CaO Na 2 O , (18.1)

where silica (SiO2) is derived from sand, sodium oxide (Na2O) is derived from
soda, and calcium oxide (CaO) is derived from lime. Soda lime glass might also
contain potassium oxide (K2O), which is derived from potash (K2CO3). The
silicate is a glass network former. Sodium oxide and calcium oxide are network
modifiers that lower the melting point and the viscosity, respectively. Diversity
in glass defeats crystallization and promotes homogeneity in the glass melt. Silica
is the base network of most glass.
A borosilicate krone (BK) is frequently described as

SiO2 B2 O3 M 2 O , (18.2)

where the alkali metal oxide (M2O) is a network modifier, such as sodium oxide
(Na2O) and potassium oxide (K2O). The addition of the alkali oxides lowers the
melting point and reduces the viscosity; these features promote ease of glass
183
184 Chapter 18

manufacture. The addition of borate (B2O3) fills voids and reduces the thermal
coefficient of expansion. Borate is a network modifier in smaller concentrations,
as shown in Table 18.1.

18.3 Flint
A flint (F) contains lead oxide (PbO) and alkaline oxide (M2O):

SiO2 PbO M 2 O . (18.3)

Flinti occurred naturally as flint stone from the south coast of Britain. Lead oxide
provides higher electron density than silica. Lead oxide also shifts UV resonance
toward the visible regime. A high lead content can make glass appear yellow.
Lead oxide is weakly reactive to the atmosphere. For improved chemical
stability, titanium oxide (TiO2) and zirconium oxide (ZrO2) may replace small
amounts of lead oxide at the cost of increased absorption in the blue. Figure 18.1
displays the Sellmeier index of a crown glass (K7) and flint (F2). The addition of
lead oxide shifts UV resonance toward the blue. The F2 glass displays a larger
partial dispersion ΔPgF than the K7. The shift of resonance toward the blue also
shortens the transmission spectrum (Fig. 18.2).
Lead oxide has several important features as a network modifier. The 90
electrons of lead oxide provide a higher electron density than the 30 electrons of
silica (SiO2). This increases the refractive index. The large mass of lead (82 u)
has a lower thermal velocity than the silica (14 u) within a glass melt.
Consequently, the lead oxide defeats crystallization through its low mobility. The

Table 18.1 Composition of borosilicate glass S-BSL7 (Ohara MSDS 2009).

Compound Formula Minimum Maximum

Silica SiO2 65 75
Potassium oxide K2O 10 20
Sodium oxide Na2O 10 20
Boron oxide B2O3 5 15
Barium oxide BaO — 5
Zinc oxide ZnO — 5
Aluminum oxide Al2O3 — 1
Titanium oxide TiO2 — 0.5
Antimony oxide Sb2O3 — 0.3

i
Flint is derived from splie, meaning split. Gemanic fli-, or spli-.
Composition and Spectra of Materials 185

Figure 18.1 Sellmeier spectra of K7 and F2. SCHOTT defines normal partial dispersion
by these two glass types. The addition of lead shifts the resonant peak toward the visible.

Figure 18.2 Transmission spectra of K7 and F2. The addition of lead shortens the
spectrum in the blue.

bond length of lead oxide at 4.0 Å is much longer than silica at 1.6 Å.
Consequently, its weaker bond promotes lower viscosity. To the benefit of the art
of glass blowing, lead oxide has two important effects: its high index creates
sparkle, and its high mass extends the working temperature.
Lead is the most common heavy metal in the earth’s crust. Three of four
heavy atoms spontaneously decay into a stable form of lead. Its high electron
density creates a bright reflection in a thin coating of paint. White lead (PbCO3)
is the basic pigment of lead paint, which is now strictly regulated and controlled
in numerous countries. As existing lead paint deteriorates, it settles in soil and
ground water. The same destination applies to the lead silicate in glass parts. Fine
particles of lead and other heavy metals must be properly disposed. A few early
symptoms of lead poisoning are diffuse muscle weakness, joint pain, nausea,
diarrhea, and constipation. The effects of lead on the central nervous system are
not, as yet, reversible.
186 Chapter 18

18.4 Long Crown

A lang krone, or long crown (LgK), displays a long optical spectrum with a small
Abbe number. In fluorite krone (FK), some oxygen is replaced by fluorine. A
fluorophosphate glass composition is frequently described as

P2 O5 Al2 O3 MO ( F2 − O ) , (18.4)

where (F2 – O) represents a “fluorine minus oxygen.” The phosphate P2O5 is a

network former. The alumina (Al2O3) is a network modifier that promotes
stability, as the phosphate tends to react with atmosphere. The alkaline metal
oxide (MO) acts as a network modifier. The fluorite component (F2 – O) binds
electrons more tightly, which shifts the resonance deeper into the UV. This shift
increases the spectrum length and reduces the dispersion.
Figure 18.3 displays the Sellmeier index of a krone (K7) and a fluorite krone
(N-FK5). The addition of fluorite shifts the UV resonance away from the blue.
The fluorophosphate krone glass displays a smaller partial dispersion ΔPgF than
the K7. The fluorophosphate krone also displays a longer transmission spectrum
(Fig. 18.4).
A fluorite glass is reactive to atmosphere. A coating such as magnesium
fluoride (MgF2) is required for protection from the atmosphere. It can also melt
during grinding or polish due to its low glass temperature. It might also fracture
easily. The benefits of a low-dispersion glass have a significant cost in
production.

18.5 Short Flint

A kurz flint, or short flint (KzF), provides a short optical spectrum through
replacement of lead oxide with antimony oxide (Sb2O3):

SiO2 Sb2 O3 B2 O3 . (18.5)

The borate is a network modifier that shifts the resonant wavelengths to

shorter wavelengths. The antimony oxide provides electron density.
A short flint (KzFH1) employs shorter resonant wavelengths than flints such
as F2 in Fig. 18.5. The transmission spectrum of a short flint is shorter in Fig.
18.6. The shorter resonant wavelengths of the antimony oxide reduce partial
dispersion while maintaining the reactive index. The shortened transmission
spectrum is created by a shorter lifetime of resonance. As indicated by the
Lorentz model of Figs. 5.3 and 5.4, a shorter lifetime of resonance has two
effects: a broader absorption band and a flatter index profile. The broader
absorption band shortens the spectrum range in the blue. The flatter index profile
creates anomalous partial dispersion in the blue. The relative partial dispersion
ΔPgF of the KzFH1 is –0.008.
Composition and Spectra of Materials 187

Figure 18.3 Sellmeier spectra of K7 and N-FK5. the addition of fluorite shifts the resonant
peak from the visible regime.

Figure 18.4 Transmission spectra of K7 and N-FK5. The addition of fluorite lengthens the
spectrum in the blue.

18.6 Short-Flint Special

A kurz flint sonder, or short-flint special (KzFS), provides a short optical
spectrum due to replacement of silica with borate as the network former:

B2 O3 PbO Al2 O3 . (18.6)

The glass is frequently called lead borate. A special glass borate (B2O3) serves as
the base network in place of silica. The alumina is a network modifier for
chemical stability, while the lead silicate provides electron density. A short-flint
special (KzFS) is reactive with the atmosphere. A protective coating is required.
The lead borate of a short-flint special is reactive to the atmosphere. Storage in
dry nitrogen is recommended until a protective coating is applied. Significant
expertise is required for processing of a short-flint special glass. Ecofriendly
versions of short-flint special are based on niobium and silicate.
188 Chapter 18

Figure 18.5 Sellmeier spectra of F2 and KzFS1. Antimony in place of lead shifts the
resonant peak from the visible regime. A shortened lifetime flattens the refractive plot in
the blue.

Figure 18.6 Transmission spectra of F2 and KzFS1. Antimony in place of lead shortens
the spectrum in the blue.

The borate of the short-flint special KzFSN4 displays a resonance near 8 μm

in Fig. 18.7. The absorption band of the borate KzFSN4 at 8 μm extends to 1 μm
in Fig. 18.8, while the absorption band at 350 nm is not greatly widened. The
shortened spectrum in the blue is enough to reduce the partial dispersion of the g
and F lines. The relative partial dispersion ΔPgF of the KzFSN4 is –0.009.
The partial dispersion of a short-flint special is reduced by a shorter lifetime.
The broader absorption band shortens the spectrum range of the infrared from 10
to 8 μm. The flatter index profile creates anomalous partial dispersion in the blue.
Composition and Spectra of Materials 189

Figure 18.7 Sellmeier spectra of F2 and KzFS1. Lead borate in place of lead silicate shifts
the resonant peak from the visible regime. A shortened lifetime flattens the refractive plot
in the blue.

Figure 18.8 Transmission spectra of F2 and KzFS1. Lead borate in place of lead silicate
shortens the spectrum weakly in the blue and strongly in the infrared.

The negative partial dispersion of a special short flint is essential for color
correction in the blue wavelength range. A shorter lifetime of resonance
decreases the partial dispersion. The relative partial dispersion is similar in
magnitude to the short flint KzFSH1, while the transmission in the blue is much
improved. A short-flint special provides a negative partial dispersion without a
deep reduction of spectrum length in the blue.

18.7 Environmentally Safe Short Flint

SCHOTT still manufactures a lead borate short-flint special KzFSN5, which has
ecofriendly versions: N-KzFS5 by SCHOTT and S-NBH5 by Ohara. Table 18.2
indicates the composition of S-NBH5, which is mostly silica and niobium oxide.
The niobium oxide Nb2O5 is a network former of the same kind as the
190 Chapter 18

Table 18.2 Composition of niobium silicate glass S-NBH5 (Ohara MSDS 2009).

Compound Formula Minimum Maximum

Silica SiO2 31 41
Niobium oxide Nb2O5 16 26
Boron oxide B2O3 8 18
Sodium oxide Na2O 4 12
Zirconium oxide ZrO2 4 12
Zinc oxide ZnO 3 9
Barium oxide BaO — 3
Calcium oxide CaO — 3
Potassium oxide K2O — 3
Lithium oxide Li2O — 3
Antimony oxide Sb2O3 — 0.5

phosphate oxide P2O5. The zirconium oxide (ZrO2) is a network modifier that
shortens resonant lifetime of the niobium oxide. Zirconium has one less electron
than niobium. The addition of zirconium in place of niobium creates a holeii in
the valance band of the niobium-oxide network. A hole provides opportunity for
electron scatter and shorter lifetime of oscillation. The other metal oxides provide
diversity toward a lower melting point, lower viscosity, and less crystallization.
The antimony oxide removes iron.
The name of the Ohara glass S-NBH implies several features: S indicates an
environmentally safe glass; NB indicates the bulk composition as niobium; H
indicates a high index for this glass type.
A niobium-silicate short flint is very similar to a lead-borate short flint;
however, it is also different. The refractive index of S-NBH5 nearly equals that
of KzFSN5 in the visible regime, as shown in Fig. 18.9. The refractive indices at
the d lines are the same at 1.65412. The Abbe numbers are similar at 39.68 for
the S-NBH5 and 39.63 for the KzFSN5. The transmission of both materials is
nearly identical from 436 to 550 nm. This indicates a similar reduction in
resonant lifetime, which is essential for a short flint. However, there is significant
difference in the relative partial dispersion: ΔPgF equals –0.004 for S-NBH5, and
–0.007 for KzFSN5. The smaller ΔPgF of S-NBH5 also correlates with improved
spectrum length in Fig 18.10. S-NBH5 displays a longer spectrum with less
absorption in the blue.
In summary, the lead borate displays more relative partial dispersion, while
the niobium silicate is much more chemically stable. Environmental stability

ii
A hole is an empty electron state.
Composition and Spectra of Materials 191

Figure 18.9 Sellmeier spectra of KzFSN5 and S-NBH5. Niobium silicate in place of lead
borate replicates the spectrum in the visible regime.

normally requires some compromise in glass features. The holes responsible for
electron scatter also promote chemical reaction. Absorption and reaction are
intrinsic features of a short-spectrum flint with anomalous partial dispersion.

18.8 Dense Flint

A schwer flint, or dense flint (SF), employs a larger concentration of lead oxide
than normal flint. The refractive index of SF6 is larger than that of F2 in Fig.
18.11. The addition of more lead oxide shifts the UV resonance further into the
blue in Fig. 18.12.

Figure 18.10 Transmission spectra of KzFSN5 and S-NBH5. Niobium silicate in place of
lead borate extends the spectrum from the visible regime. Less absorption indicates
longer lifetime and less negative relative partial dispersion.
192 Chapter 18

Figure 18.11 Sellmeier spectra of F2 and SF6. The addition of lead oxide in place of silica
greatly increases refractive index and dispersion.

An ecofriendly version of schwer flint employs titanium oxide in place of

lead oxide. There is a small shift of resonant wavelength into the blue, as shown
in Fig. 18.13. The refractive-index profile is substantially the same. However the
absorption peak of the titanium oxide flint extends far into the visible spectrum,
as shown in Fig. 18.14. There is significant absorption from 400 to 580 nm. The
ecofriendly N-SF6 can appear yellow. The composition of the ecofriendly
version of SF6 is displayed Table 18.3. The S-TIH6 glass by Ohara indicates
silica and titania oxide as the major components (25–35%). Zirconium oxide and
niobium oxide defeat crystallization of the titanium oxide through diversity of
similar bonds. Other components are the typical ingredients of borosilicate glass.

Figure 18.12 Transmission spectra of F2 and SF6. The addition of lead oxide in place of
silica reduces transmission the blue.
Composition and Spectra of Materials 193

Figure 18.13 Sellmeier spectra of SF6 and N-SF6. Titanium oxide in place of lead oxide
replicates the refractive index in the visible.

An important issue with ecofriendly glass is crystallization. Titanium oxide

crystallizes more easily than lead oxide. The crystallites occur in numerous
shapes: filaments, cords, and other inclusions. Filaments and cords create an
inhomogeneous refractive index. Inclusions may scratch the surface as they are
pulled during polish. Most lead glass can be replaced with titanium oxide;
however, titanium glass is more difficult and more expensive to process.

Figure 18.14 Transmission spectra of SF6 and N-SF6. Titanium oxide in place of lead
oxide reduces transmission in the blue and green.
194 Chapter 18

Table 18.3 Composition of titanium silicate glass S-TIH6 (Ohara 2009).

Compound Formula Minimum Maximum

Silica SiO2 25 35
Titanium oxide TiO2 25 35
Barium oxide BaO 10 20
Calcium oxide CaO 10 20
Potassium oxide K2O 10 20
Sodium oxide Na2O 10 20
Niobium oxide Nb2O5 5 15
Zirconium oxide ZrO2 — 5
Antimony oxide Sb2O3 — 0.3
Chapter 19
Advanced Concepts
19.1 Wave Equation
An electromagnetic wave is governed by three important features: the spatial
 ,
curvature of the electric field ∇2E, the temporal curvature of the electric field E
and the temporal slope of electric current density J , which may also be
considered the acceleration of electric current density. These parameters are
expressed in the wave equation as follows:

∇ 2 E = μ ( ε0 E
 + J ) , (19.1)

where μ is the magnetic permeability and ε0 is the electric permittivity.

The ∇ operator defines the gradient as

∂ ∂ ∂
∇= i + j+ k , (19.2)
∂x ∂y ∂z

where i, j, and k are unit vectors. The ∇2 operator defines the Laplacian as

∂2 ∂2 ∂2
∇2 = i + j + k. (19.3)
∂x 2 ∂y 2 ∂z 2

The temporal versions of the spatial operators are

 = ∂A
A (19.4)
∂t

and

2
 = ∂ A .
A (19.5)
∂t 2

Application of the spatial and temporal derivatives yields

195
196 Chapter 19

ω2
k x2 + k y2 + k z2 = N 2 , (19.6)
c2

where kx, ky, and kz are the spatial frequencies along orthogonal axes, N is the
complex refractive index, ω is the temporal frequency, and c is the speed of light.
Each k2 term represents a spatial curvature of the wave along an axis. The ω2
term represents the temporal curvature of the wave.
Examination of Eq. (19.6) reveals an important concept: the sum of the
spatial curvatures is proportional to the temporal curvature. The scale of spatial-
to-temporal curvature is dependent on the refractive index. Thus, the spatial
curvature is dependent upon the electric current.

19.2 Refractive Index

Free and bound electrons have very different effects on spatial curvature. The
free electrons of a metal react quickly enough to cancel an external electric field.
This creates an exponential decay of the electric field within a spatial wavelength
(Fig. 19.1). The restoring force of bound electrons prevents complete
cancellation of the external electric field. Consequently, the bound electrons
drive the electric field faster toward zero but not completely to zero (see Fig.
19.2). The faster return to zero is described by larger spatial frequency or shorter
wavelength.

Figure 19.1 Exponential decay by free-electron current. The solid line is the external
electric field; the dashed line is the internal electric field. The electric field of the free-
electron current completely cancels the external electric field within a spatial wavelength.
Advanced Concepts 197

Figure 19.2 Shortened wavelength by bound-electron current. The solid line is the
external electric field; the dashed line is the internal electric field. The electric field of the
bound-electron current partially cancels the external electric field. The partial cancellation
creates a shorter internal wavelength.

The complex refractive indices of free and bound carriers are well
documented.5 The refractive index of free carriers is expressed as a Drude
model:5

 ω2 
N F2 = 1 +  2 PF  = εR , (19.7)
 −ω − iω / τ F 

where ωPF is the plasma frequency of free carriers, τF is the lifetime of a free
carrier, and εR is the dielectric constant. The plasma frequency of free carriers
may be expressed as

ρF q 2
ω2PF = , (19.8)
m ε0

where ρF is the concentration of free carriers, q is the charge of the carrier, and m
is the mass of the carrier.
The corresponding refractive index of bound carriers is expressed as a
Lorentz oscillator:5

 ω2PB 
N B2 = 1 +  2 2 , (19.9)
 −ω − iω / τ B + ωRB 
198 Chapter 19

where ωPD is the plasma frequency of the bound carriers, ωRB is the resonant
frequency of the bound state, and τB is the lifetime of the bound state. The plasma
frequency of bound carriers may be expressed as

ρB q 2
ω2PB = , (19.10)
m ε0

where ρB is the concentration of bound carriers.

If the temporal frequency is much faster than the lifetime τB, then the
refractive index of a glass becomes

 ω2  1
N B2 = 1 +  2 PB 2  = ε R ; ω >> . (19.11)
 ωB − ω  τF

This approximation is the foundation of the Sellmeier formula for refractive

index [Eq. (17.1)].
Aluminum represents a combination of Drude and Lorentz spectra. Its
refractive index and extinction coefficient are 1.15 and 7.15 at the d line,
respectively. Aluminum behaves much like silver; however, it also has a bound
oscillator at 800 nm. Consequently, there is a small dip in reflectance at 800 nm.
The reflectance of aluminum is near 92% at the d line.

19.3 Relative Partial Dispersion

A SCHOTT glass map frequently plots partial dispersion PgF versus Abbe
number. The SCHOTT glass company defined two glass types as normal in
dispersion: a crown K7 and a flint F2.
The normal glass types, K7 and F2, define a linear relationship for normal
partial dispersion:52

PgF = − ( 0.6438 ) + ( 0.001682 )Vd . (19.12)

A departure from normal dispersion is described as relative partial dispersion

ΔPgF. It represents the vertical distance from the plot of normal partial dispersion
[Eq. (19.12)]. A positive ΔPgF indicates a larger partial dispersion in the blue,
while a negative ΔPgF indicates a smaller partial dispersion in the blue. Relative
partial dispersion is an important metric for color correction.
Advanced Concepts 199

19.4 Emission
A dipole moment p defines spatial distribution of charge as

p =  ρ r ∂V , (19.13)

where ρ is the charge density, r is the spatial position, and ∂V is the differential
volume. The electric potential of the dipole is

 1   p ⋅ rˆ 
Φ=  2  , (19.14)
 4πε 0   r 

where r̂ is a unit vector. As the distance r grows, the electric field of a radiating
dipole becomes

 1   ( rˆ ⋅ p
 ⋅ rˆ ) 
E= 2   exp ( ik r ) , (19.15)
 4πε 0 c   r 

where p is the temporal curvature of the dipole moment. Exercise of the cross
products reveals

 p    sin θ 
E = 2   exp ( ik r ) , (19.16)
 4πε 0 c  r 

. The irradiance of the dipole emission is

where θ is the angle between r̂ and p

 
p
2
  sin 2 θ 
ID =   2  . (19.17)
 32 π2 ε 0 c 3  r 
 

19.5 Coherence
The axial version of the space-angle product is1

Δz Δk z = 8 . (19.18)

This may be converted to wavelength as1

4 2
Δz Δλ = λP , (19.19)
π
200 Chapter 19

where Δλ is the wavelength range and λP is the peak wavelength. The coherence
length is

 4  λ P 
2

sC =    . (19.20)
 π   Δλ 

The coherent optical length in cycles is

λP
ΛC ≈ cyc . (19.21)
Δλ

The coherence length within a refractive medium is

2
 λ  λ  λ
Λ C ≈  P  P  = P . (19.22)
 Δλ  n  n Δλ

The correlation between two points of an electric field is defined as

Γ ( r1 , t1 , r2 , t2 , ) = A ( r2 , t2 ) A*( r2 , t2 ) , (19.23)

where A is the amplitude of the electric field and A* is the complex conjugate of
A. If the correlation has a magnitude of unity, then the system is coherent. If the
correlation is null, then the system is incoherent. Figure 19.3 displays the axial
coherence of a 50-nm spectral filter.
Coherence determines the structure of interference patterns. Partial
reflections of an electric-field pattern can rejoin the transmitted electric field. The
contrast of the interference pattern is determined by the correlation of the two
waves after separation in both space and time. Coherence length becomes more
important as dimensions approach the coherence length.
The correlation of a single dipole field is completely coherent, while the
correlation between two dipole fields may be either coherent or incoherent. If
two dipoles radiate in constant phase, such as during stimulated emission, then
the correlation between dipole fields is coherent. However, if two dipoles radiate
in random phase, such as during multiple fluorescent events, then the correlation
between the dipole fields is incoherent.
Van Cittert and Zernike developed expressions for the partial coherence of an
incoherent source as the Fourier transform of the angular extent of the source.5 In
Eq. (10.2) the partial coherence of a circular incoherent source is stated as

2 J1 ( ρ k NA S )
Γ (ρ) = ,
ρ k NA S
Advanced Concepts 201

Figure 19.3 Axial coherence of a Gaussian 250-nm bandpass filter. Peak wavelength of
transmission is 500 nm. Bandwidth is 250 nm. Axial coherence is 1 μm according to Eq.
(19.22), or 1.27 according to Eq. (19.20).

where ρ is the distance between two points of the illumination field and NAs is
the numerical aperture of the source as defined below:

NA S = sin βS , (19.24)

where βs is the full angle of the source with respect to the illumination field.

19.6 Gaussian Beam Power

The irradiance profile (power per area) is

 r2 
I = I 0 exp  −2 2  . (19.25)
 rG 

The space-angle product of the Gaussian beam is

dG ( 2 ΝAG ) = 1.27λ , (19.26)

where NAG is the Gaussian NA.

The irradiance of a Gaussian profile is

 −π ΝAG2 2 
I = I 0 exp  2
r . (19.27)
 λ 

The peak irradiance of a Gaussian beam (power per area) is related to the total
power P0 (quanta per time) by
202 Chapter 19

π ΝA G2 2
I0 = 2 2
P0 = P0 , (19.28)
λ AG

where P0 is the power of the beam and AG is the area of the beam. The power
enclosed by a radius is

  r 2 
P ( r ) = P0 1 − exp  −2 2   . (19.29)
  rG  

19.7 Transfer Functions

The amplitude transfer function (ATF) and optical transfer function (OTF) of a
lens aperture apply to coherent and incoherent images, respectively.53 The
modulation transfer function (MTF) can apply either the ATF or the OTF.
The MTF of a coherent system applies to the electric field. Thus, the MTF of
a coherent system is the ATF. A laser beam is a coherent system, wherein all
points have a finite correlation by electric field. The product of consecutive ATFs
determines the MTF.
The MTF of an incoherent system applies to the irradiance. Thus, the MTF of
an incoherent system is the OTF, which is an autocorrelation of the ATF. Two
adjacent fluorophores represent an incoherent system. The phase angle between
amplitude point spreads is random over multiple emissions.
The ATF of a circular aperture is

1 k ≤ kM 
ATF ( k ) =  , (19.30)
0 kM ≤ k 

where kM is the spatial frequency of the margin in cycles per distance and the
ATF cutoff frequency is the spatial frequency of the aperture:

ΝA M 1.22
kM = = . (19.31)
λ φ Airy

The ATF describes a coherent system where the temporal coherence between
points is stationary.
The OTF is the autocorrelation of the ATF. The OTF of a lens with a circular
aperture is
Advanced Concepts 203

  2  
 2  arccos  k   −  k  1 −  k   k ≤ ( 2k M ) 
       
OTF ( k ) =  π   2k M    2k M   2k M   ,
   
0
 ( 2kM ) ≤ k 
(19.32)

where the OTF cutoff frequency is

2ΝA M 2.44
( 2 kM ) = = . (19.33)
λ φ Airy

As displayed in Fig. 19.4, the OTF closely resembles a cone with a height of
unity. This cone can be correlated to the Airy transform as follows:

 k 
OTF ( k ) ≅ tri  . (19.34)
 1.6 kM 

Figure 19.4 MTFs of an Airy pattern and a conical estimate. The solid line is the Airy
pattern. The dashed line is the conical estimate. Standard deviation of MTFs is < 0.01.
Standard deviation of the corresponding point spread is < 0.03.
204 Chapter 19

The triangle function is defined as

1 − x ( 0 ≤ x ≤ 1) 
tri ( x ) =  . (19.35)
0 (1 < x )
The marginal frequency is derived from the conical frequency as

kC
kM = . (19.36)
1.6

The standard deviation conical estimate from the Airy MTF is < 1% of the
expected value up to the marginal frequency. The marginal NA is derived from
the conical frequency as

kC λ
ΝA M = . (19.37)
1.6

Equation (19.35) is extremely useful in application to MTF data for an off-the-

shelf product. The OTF describes an incoherent system where the temporal
coherence between two point spreads is zero.

19.8 Scatter
Surface texture may create a random phase delay during reflection or
transmission. Consequently, the exiting electric field EE relates to the incident
electric field EI as

E E =E I exp iφ0 cos ( k0 x )  , (19.38)

where φ0 is the magnitude of the phase delay and k0 is the spatial frequency of the
phase delay. At a small phase delay, this becomes

E E =E I 1 + iφ0 cos ( k0 x )  . (19.39)

The scatter coefficient for the electric field is

EE
s = 1− = − iφ0 cos ( k0 x ) . (19.40)
EI

The scatter, or scatterance, applies to the irradiance as

Advanced Concepts 205

2
S= s =φ 2 , (19.41)

where φ is the rms of the phase delay. The scatterance is frequently called the
total integrated scatter (TIS).
There a three specific formats of the rms phase delay. The rms phase delay
for transmission is

 2πσ 
φ T = Δn  , (19.42)
 λ 

where σ is the rms surface texture. The rms phase delay for an external reflection
is

 2πσ 
φ ER = 2  . (19.43)
 λ 

The rms phase delay for an internal reflection is

 2πσ 
φ IR = 2n  . (19.44)
 λ 

19.9 Interference Filters

An interference filter comprises a dielectric stack of high- and low-index
materials. The repeated structure is similar to a crystalline structure wherein an
x ray is diffracted based on a lattice vector. The analogous filter diffraction is
displayed in Fig. 19.5, where the reflected wave vector kR is the sum of the
incident wave vector kIn and the filter wave vector kF:

k R = k In + k F . (19.45)

The wave vector of reflection is dependent on the angle of incidence as

2k R ⋅ k F = 2k R k F cos θ = k F2 . (19.46)

The spatial frequency of reflection obeys

1
k R cos θ = k F . (19.47)
2

At normal incidence, the spatial frequency of reflection is half that of the filter
lattice.
206 Chapter 19

Figure 19.5 Wave vectors of the interference filter. A reflected vector is the sum of the
incident vector and the filter vector. Blue shift is indicated by longer wave vectors at a
larger angle with the filter vector.

19.10 Shot Noise

The contrast of a fluorophore with a short lifetime is frequently limited by the
background shot noise. The quantum of an object with a short lifetime is
dependent on an exponential decay

 −t 
QOSL ∝ 1 − exp   ( t ≈ τF ) , (19.48)
 τF 

where t is the integration time and τF is the lifetime of the fluorophore. The
quantum of the dark current is linearly dependent on the integration time:

Q DC ∝ t . (19.49)

The quantum of read noise is independent of the integration time:

Q RN ∝ 1 . (19.50)

The quantum of the background shot noise is dependent on the square root of
integration time:

Q BSN ∝ t1 2 . (19.51)

The shot noise of the background may easily become larger than the read noise or
the object quantum.
The contrast of a fluorophore with a long lifetime is frequently limited by the
shot noise of the quanta. The quantum of a fluorophore with a long lifetime is
linearly dependent on the integration time:

QOLL ∝ t ( t << τF ) . (19.52)

Advanced Concepts 207

The contrast with dark current is stationary in time, the contrast with read noise
grows linearly with time, and the contrast with shot noise grows with the square
root of time.
Beyond increasing the integration time, the effects of background and noise
may be reduced by optical design. These expressions for contrast provide a
foundation for maximization of contrast through selection of NA and
magnification.
Appendix: Prescriptions
The prescriptions used in this book are summarized below.
The magnification M has several formats: spatial enlargement (10X), spatial
enlargement with infinity correction (10XI), spatial reduction (10XR), or angular
enlargement (10XA) .
The focal lengths are important to macro- and microlens formats. A
microscope objective is normally defined by magnification, numerical aperture,
and tube length. The tube length (TL) may specify the focal length of the tube
lens for enlargement with infinity correction.
The field of view (FOV) can be spatial (25 mm) or angular (15 deg). An
attempt is made to make the field of view equivalent the 12-mm diameter of a
typical CCD sensor.
The image convergence is specified by paraxial numerical aperture (NA),
image-space numerical aperture (ISNA), object-space numerical aperture
(OSNA), f-number (f/#), or image-space f/#, which indicates the f/# inflation.
Aspheric parameters indicate the following: conic indicates a conic constant;
A4 indicates the scale of radius to 4th power; A6 indicates the scale of radius to
6th power.

Prescription 4.1 Spherical aberration of a spherical lens.

(M = 0, NA = 0.25, f = 20 mm, FOV = 0 deg)

Surf Radius Thickness Glass Diameter nd νd

OBJ ∞ ∞ 0 - -
1 ∞ 20 10 - -
STO 10.37 30.20 BK7 12 1.517 64.2
IMA ∞ 0.524 - -

Prescription 4.2 Coma in an aspheric lens.

(M = 0, NA = 0.25, f = 20 mm, FOV = 1.0 deg)

Surf Radius Thickness Glass Diameter Conic nd νd

OBJ ∞ ∞ 0.00 - -
STO ∞ 0.00 10.00 - -
2.00 10.34 30.33 BK7 12.00 –0.4337 1.517 64.2
IMA ∞ 0.80 - -

209
210 Appendix

Prescription 4.3 Astigmatism and lateral color in a spherical lens.

(M = 0, NA = 0.02, f = 20 mm, FOV = 15 deg)

Surf Radius Thickness Glass Diameter nd νd

OBJ ∞ ∞ 0 - -
STO ∞ 15.00 0.7 - -
2 10.34 22.19 BK7 12 1.517 64.2
IMA ∞ 0.8 - -

Prescription 4.4 Axial color in an aspheric lens.

(M = 0, NA = 0.25, f = 20 mm, FOV = 0 deg)

Surf Radius Thickness Glass Diameter Conic nd νd

OBJ ∞ ∞ 0.00 0 - -
1 ∞ 20.00 10.00 0 - -
STO 10.34 30.34 BK7 20.00 –0.4334 1.517 64.2
IMA ∞ 0.07 0 - -

Prescriptions 7.1–7.8 5X doublet from OptoSigma, Part No. 026-0270.

(M = 5XI, TL = 200 mm, NA = 0.12, f = 40 mm, FOV = 2.4 mm)

Surf Radius Thickness Glass Diameter Comment nd νd

OBJ ∞ ∞ 0.00 - -
STO 23.40 4.10 BK7 12.70 026-0270 1.517 64.2
2.00 –17.60 2.00 SF2 12.70 1.648 33.8
3.00 –60.31 36.84 12.70 - -
IMA ∞ 2.412318 - -
Prescriptions 211

12
Prescription 9.1 10XR double gauss.
(M = 10XR, f/# = 1.4, Image-space f/# = 1.5, f = 52 mm, FOV = 12 mm)

Surf Radius Thickness Glass Diameter nd νd

OBJ ∞ 530.86 120.8 - -
1 ∞ 10.00 15.0 - -
2 79.02 5.43 LaFN21 44.0 1.788 47.5
3 –387.15 0.25 44.0 - -
4 32.41 4.57 LaFN21 38.0 1.788 47.5
5 62.73 3.63 36.0 - -
6 ∞ 2.30 SF15 34.0 1.699 30.1
7 27.90 5.17 30.0 - -
STO ∞ 5.97 27.0 - -
9 –23.48 2.30 SF15 30.0 1.699 30.1
10 –650.00 6.57 LaFN21 34.0 1.788 47.5
11 –32.30 0.25 36.0 - -
12 –218.50 3.56 LaFN21 38.0 1.788 47.5
13 –58.10 0.25 38.0 - -
14 128.60 3.64 LaFN21 38.0 1.788 47.5
15 –175.70 43.05 38.0 - -
IMA ∞ 12.3 - -
212 Appendix

13
Prescription 9.2 10XR microlens, from Matsui.
(M = 10XR, f/# = 2.9, Image-space f/# = 3.2, f = 105 mm, FOV = 12 mm)

Surf Radius Thickness Glass Diameter nd νd

OBJ ∞ 1094.38 120.0 - -
1 ∞ –20.00 39.6 - -
2 ∞ 6.00 Silica 50.0 1.458 67.8
3 ∞ 14.00 50.0 - -
4 88.80 4.30 E-LaK13 44.0 1.694 53.2
5 –4656.70 7.70 39.1 - -
6 33.12 6.00 E-LaK14 36.0 1.697 55.5
7 85.73 0.70 32.0 - -
8 169.57 2.00 E-SF2 34.0 1.648 33.8
9 27.68 23.48 30.0 - -
10 ∞ –3.00 24.1 - -
STO ∞ 3.00 24.0 - -
12 –29.10 2.00 E-F2 24.0 1.620 36.3
13 –420.00 7.50 E-LaK13 27.0 1.694 53.2
14 –42.41 0.10 30.0 - -
15 258.00 4.30 E-LaK13 31.0 1.694 53.2
16 –65.23 8.35 31.0 - -
17 –92.06 3.50 E-SF6 31.0 1.805 25.4
18 –56.03 12.96 31.0 - -
19 –41.90 2.00 E-LaSF013 26.0 1.804 39.6
20 83.22 1.00 30.0 - -
21 105.00 5.00 E-LaF04 32.0 1.757 47.8
22 –89.47 44.50 32.0 - -
IMA ∞ 12.0 - -
Prescriptions 213

13
Prescription 9.3 2XR microlens, from Matsui (1998).
(M = 2XR, f/# = 2.9, Image-space f/# = 4.3, f = 105 mm, FOV = 12 mm)

Surf Radius Thickness Glass Diameter nd νd

OBJ ∞ 240.23 23.0 - -
1 ∞ –20.00 32.3 - -
2 ∞ 6.00 Silica 50.0 1.458 67.8
3 ∞ 14.00 50.0 - -
4 88.80 4.30 E-LaK13 44.0 1.694 53.2
5 –4656.70 7.70 32.0 - -
6 33.12 6.00 E-LaK14 36.0 1.697 55.5
7 85.73 0.70 32.0 - -
8 169.57 2.00 E-SF2 34.0 1.648 33.8
9 27.68 16.93 30.0 - -
10 ∞ –3.00 24.3 - -
STO ∞ 3.00 24.0 - -
12 –29.10 2.00 E-F2 24.0 1.620 36.3
13 –420.00 7.50 E-LaK13 27.0 1.694 53.2
14 –42.41 0.10 30.0 - -
15 258.00 4.30 E-LaK13 31.0 1.694 53.2
16 –65.23 35.80 31.0 - -
17 –92.06 3.50 E-SF6 31.0 1.805 25.4
18 –56.03 12.96 31.0 - -
19 –41.90 2.00 E-LaSF013 26.0 1.804 39.6
20 83.22 1.00 30.0 - -
21 105.00 5.00 E-LaF04 32.0 1.757 47.8
22 –89.47 44.50 32.0 - -
IMA ∞ 12.0 - -
214 Appendix

13
Prescription 9.4 1X microlens, from Matsui (1998).
(M = 1X, f/# = 2.9, Image-space f/# = 5.5, f = 105 mm, FOV = 12 mm)

Surf Radius Thickness Glass Diameter nd νd

OBJ ∞ 149.72 12.1 - -
1 ∞ –20.00 29.2 - -
2 ∞ 6.00 Silica 50.0 1.458 67.8
3 ∞ 14.00 50.0 - -
4 88.80 4.30 E-LaK13 44.0 1.694 53.2
5 –4656.70 7.70 28.9 - -
6 33.12 6.00 E-LaK14 36.0 1.697 55.5
7 85.73 0.70 32.0 - -
8 169.57 2.00 E-SF2 34.0 1.648 33.8
9 27.68 11.76 30.0 - -
10 ∞ –3.00 24.5 - -
STO ∞ 3.00 24.0 - -
12 –29.10 2.00 E-F2 24.1 1.620 36.3
13 –420.00 7.50 E-LaK13 27.0 1.694 53.2
14 –42.41 0.10 30.0 - -
15 258.00 4.30 E-LaK13 31.0 1.694 53.2
16 –65.23 64.69 31.0 - -
17 –92.06 3.50 E-SF6 31.0 1.805 25.4
18 –56.03 12.96 31.0 - -
19 –41.90 2.00 E-LaSF013 26.0 1.804 39.6
20 83.22 1.00 30.0 - -
21 105.00 5.00 E-LaF04 32.0 1.757 47.8
22 –89.47 44.50 32.0 - -
IMA ∞ 12.0 - -
 Prescriptions 215

14
Prescription 9.5 2XR TSL, prescription patent pending by Seward (2008).
(M = 2XR, f/# = 3.5, Image-space f/# = 3.9, f = 193 mm, FOV = 12 mm)
(Glass types by Ohara Inc: lanthanum with low-index LAL, titanium with medium-index
TIM, lanthanum with medium-index LAM.)

Surf Radius Thickness Glass Diameter Conic A4 A6 nd νd

OBJ Infinity 414.12 24.44 - -
1 Infinity 11.80 Silica 60.00 0 1.22 × 10–08 1.458 88.3
2 –249.50 15.80 60.00 - -
STO Infinity - 54.00 - -
4 Infinity 6.00 Silica 60.00 1.458 88.3
5 Infinity 15.80 60.00 - -
6 161.70 15.60 S-LaL7 60.00 1.652 76.7
7 –161.70 10.00 S-TIH4 60.00 1.755 36.7
8 Infinity 281.60 60.00 - -
9 21.00 11.40 S-LaL7 30.00 0 –1.52 × 10–06 –4.58 × 10–09 1.652 76.7
10 Infinity 6.60 S-LaM7 22.00 1.749 46.8
11 18.30 11.40 16.00 - -
12 Infinity 1.00 Silica 14.00 1.458 88.3
13 Infinity 1.00 14.00 - -
IMA Infinity 12.02 - -

Prescription 10.6 Aspheric plus singlet. Production parts from OptoSigma.

(M = 4XR, ISNA = 0.25, f = 22 mm, FOV = 1.2 mm)

Surf Radius Thickness Glass Diameter Conic Comment nd νd

OBJ ∞ 93.12 4.9 0 - -
STO ∞ 11.54 13.0 0 LS - -
2 ∞ 4.20 BK7 30.0 0 011-2280 1.517 64.2
3 –51.90 5.00 30.0 0 - -
4 13.90 11.00 B270 30.0 –0.544 023-2390 1.523 58.6
5 ∞ 18.21 11.3 0 - -
6 ∞ 1.00 BK7 10.0 0 Slide 1.517 64.2
7 ∞ 20.00 1.3 0 - -
8 ∞ –20.00 12.2 0 VLS - -
IMA ∞ 1.3 0 - -
216 Appendix

Prescription 10.7 Aspheric plus doublet. Custom doublet, production aspheric from
OptoSigma. (M = 4XR, ISNA = 0.25, f = 19 mm, FOV = 1.2 mm)

Surf Radius Thickness Glass Diameter Conic Comment nd νd

OBJ ∞ 72.95 4.8 0 - -
STO ∞ 7.11 13.0 0 LS - -
2 –25.82 3.00 SF2 20.0 0 Custom 1.648 33.8
3 16.27 10.00 BK7 20.0 0 Custom 1.517 64.2
4 –16.27 5.00 20.0 0 - -
5 13.90 11.00 B270 30.0 –0.5443 023-2390 1.523 58.6
6 ∞ 19.95 12.1 0 - -
7 ∞ 1.00 BK7 10.0 0 Slide 1.517 64.2
8 ∞ 20.00 1.3 0 - -
9 ∞ –20.00 11.4 0 VLS - -
IMA ∞ 1.3 0 - -

Prescription 10.8 Abbe illumination. Production parts from OptoSigma.

(M = 4XR, ISNA = 0.25, f = 19 mm, FOV = 1.2 mm)

Surf Radius Thickness Glass Diameter Comment nd νd

OBJ ∞ 54.05 4.8 - -
1 10.38 5.20 BK7 15.0 011-1180 1.517 64.2
2 ∞ 5.00 6.8 - -
STO ∞ 9.87 5.6 - -
4 5.19 3.40 BK7 8.0 011-0470 1.517 64.2
5 ∞ 2.50 8.0 - -
6 ∞ 1.00 BK7 10.0 Slide 1.517 64.2
7 ∞ 20.00 1.4 - -
8 ∞ –20.00 11.5 - -
IMA ∞ 1.4 - -

Prescription 10.9 Aspheric Abbe. Two custom aspheric elements of B270. Conic
constant is the only aspheric parameter.
(M = 4XR, ISNA = 0.25, f = 18 mm, FOV = 1.2 mm)

Surf Radius Thickness Glass Diameter Conic Comment nd νd

OBJ ∞ 56.01 4.8 - -
1 10.38 5.20 B270 15.0 –0.852 Low Tg 1.523 58.6
2 ∞ 5.00 7.1 - -
STO ∞ 10.45 5.6 - -
4 5.19 3.40 B270 8.0 –0.633 Low Tg 1.523 58.6
5 ∞ 2.50 8.0 - -
6 ∞ 1.00 BK7 10.0 slide 1.517 64.2
7 ∞ 20.00 1.2 VLS - -
8 ∞ –20.00 11.7 - -
IMA ∞ 1.2 - -
Prescriptions 217

Prescription 10.10 TIRF illumination, derived from U.S. Patent No. 6,519,092 by
Yamaguchi.23 (M = 60XI, TL = 200 mm, ISNA = 1.0, f = 3.3 mm, FOV = 60 deg)

Surf Radius Thickness Glass Diameter Dec Y

OBJ ∞ –15.81 0.0 nd νd
STO ∞ 0.00 1.6 - -
2 - 0.00 - 4.22 1.000 -
3 ∞ –4.90 30.0 - -
4 ∞ 1.00 25.0 - -
5 ∞ 1.00 12.0 - -
6 8.17 3.00 E-SF6 10.4 1.805 25.4
7 204.67 2.60 E-LaSF013 10.4 1.804 39.6
8 4.80 1.70 7.1 - -
9 ∞ 1.70 6.6 - -
10 –4.82 5.00 E-LaSF013 7.1 1.804 39.6
11 238.40 5.20 N-PSK58 15.0 1.569 71.2
12 –8.78 0.15 15.0 - -
13 –25.25 1.00 KF9 16.4 1.523 51.5
14 48.48 6.30 CaF2 19.0 1.434 95.0
15 –11.82 1.60 KzFH1 19.0 1.613 44.4
16 –18.67 0.10 21.0 - -
17 28.66 8.00 CaF2 22.0 1.434 95.0
18 –14.80 1.10 KzFH1 22.0 1.613 44.4
19 –47.67 0.15 24.0 - -
20 16.27 9.40 CaF2 23.6 1.434 95.0
21 –20.05 1.00 KzFH1 23.6 1.613 44.4
22 84.75 0.15 23.2 - -
23 13.453 6.8 N-PSK58 20.8 1.569 71.2
24 –34.23 1.00 F5 20.8 1.603 38.1
25 27.891 0.10 20.4 - -
26 7.247 3.75 N-PSK58 13.6 1.569 71.2
27 13.716 0.10 12.4 - -
28 3.716 3.60 LaSF35 7.4 2.022 29.1
29 1.332 0.65 TYPE A 2.3 1.515 41.6
30 ∞ 0.14 TYPE A 6 1.515 41.6
31 ∞ 0.17 D-263-C 10.000 1.523 54.6
IMA ∞ D-263-C 0.352 1.523 54.6
218 Appendix

Prescription 11.1 Perfect 10X objective in air.

–6
[M = 10XI, ISNA = 0.25, f = 20.0 mm, FOV = 1.2 mm,1-nm defocus (1 × 10 )]

Surf Radius Thickness Glass Diameter Comment nd νd

OBJ ∞ –20.00 1.2 Virtual object - -
STO 20.00 20.00 10.0 Lens stop - -
2 ∞ 0.00 D263M 10.0 Cover 1.523 52.3
IMA ∞ 1.2 Image - -

Prescription 11.2 10X objective with cover glass in place of air.

(M = 10XI, TL = 200 mm, ISNA = 0.25, f = 20.0 mm, FOV = 1.2 mm)

Surf Radius Thickness Glass Diameter Comment nd νd

OBJ ∞ –20.00 1.2 Virtual object - -
STO 20.00 19.89 10.0 Lens stop - -
2 ∞ 0.17 D263M 10.0 Cover 1.523 52.3
IMA ∞ D263M 1.2 Image - -

Prescription 11.3 10X objective with microscope slide in place of air.

(M = 10XI, TL = 200 mm, ISNA = 0.25, f = 20.0 mm, FOV = 1.2 mm)

Surf Radius Thickness Glass Diameter Comment nd νd

OBJ ∞ –20.00 1.2 Virtual object - -
STO 20.00 19.35 10.0 Lens stop - -
2 ∞ 1.00 D263M 10.0 Cover 1.523 52.3
IMA ∞ D263M 1.2 Image - -

Prescription 11.4 40X objective with silica cover in place of glass.

(M = 40XI, TL = 200 mm, ISNA = 0.75, f = 5.0 mm, FOV = 0.3 mm)

Surf Radius Thickness Glass Diameter Comment nd νd

OBJ ∞ –0.17 D263M 0.3 Virtual object 1.523 52.3
1 ∞ –4.89 0.5 - -
STO 5.00 4.88 8.6 Lens stop - -
3 ∞ 0.17 SILICA 10.0 Cover 1.458 67.8
IMA ∞ SILICA 0.3 Image 1.458 67.8

Prescription 11.5 40X objective with tilted cover glass.

(ISNA = 0.75, f = 5.0 mm)

Surf Radius Thickness Glass Diameter Comment Tilt nd νd

OBJ ∞ –0.17 D263M 0.3 Virtual object 1.523 52.3
1 ∞ –4.89 0.5 - -
STO 5.00 4.89 8.6 Lens stop - -
3 - 0.00 - 0.50 - -
4 ∞ 0.17 D263M 10.0 Cover 1.523 52.3
5 - 0.00 - –1.00 - -
IMA ∞ D263M 10.0 Image 1.523 52.3
Prescriptions 219

Prescription 11.6 60X objective with silica cover in place of glass. Oil must vary in
thickness for compensation of silica in place of glass.
(M = 60XI, TL = 200 mm, ISNA = 1.4, f = 3.3 mm, FOV = 0.2 mm)

Surf Radius Thickness Glass Diameter Comment nd νd

OBJ ∞ –0.17 D263M 0.2 Virtual object 1.523 52.3
1 ∞ –4.88 TYPE A 1.0 Oil 1.515 41.6
STO 5.00 4.87 TYPE A 9.3 Lens stop 1.515 41.6
3 ∞ 0.17 Silica 10.0 Cover 1.458 67.8
IMA ∞ Silica 0.6 Image 1.458 67.8

Prescription 12.1 Aplanatic front end of the scaled-divergence kind.

(M = 2X, ISNA = 0.75, f = 8.2 mm, FOV = 0.0 mm)

Surf Radius Thickness Glass Diameter Conic Comment nd ν d

OBJ ∞ –0.67 0.0 0 - -
STO ∞ –9.95 0.5 0 - -
2 4.22 6.00 N-BK7 7.0 0 S1 1.517 -
3 ∞ 1.00 N-BK7 4.0 0 Oil 1.517 -
IMA –7.94 N-BK7 4.0 0 1.517 -

Prescription 12.2 Aplanatic front end with three aplanatic surfaces.

(M = 3.5X, ISNA = 1.00, f = 8.8 mm, FOV = 0.0 mm)

Surf Radius Thickness Glass Diameter Conic Comment nd ν d

OBJ ∞ –0.71 0.0 0 - -
STO ∞ –24.50 0.5 0 - -
2 10.01 5.00 N-BK7 16.7 0 S3 1.517 -
3 11.62 1.00 13.2 0 S2 - -
4 4.22 6.00 N-BK7 8.4 0 S1 1.517 -
5 ∞ 1.00 N-BK7 4.0 0 Oil 1.517 -
IMA –4.47 N-BK7 4.0 0 1.517 -
220 Appendix

Prescription 12.3 10X plan achromat. Derived from Fukatake,21 U.S. Patent No.
6,128,139. (M = 10XI, TL = 200 mm, NA = 0.25, f = 20 mm, FOV = 1.2 mm)

Surf Radius Thickness Glass Diameter Comment nd νd

OBJ ∞ ∞ 0.0 - -
1 ∞ 13.38 10.2 Flange - -
STO ∞ 6.20 11.4 - -
3 19.30 3.10 SSK1 12.0 1.617 54.0
4 –19.30 1.85 E-LaSF013 12.0 1.804 39.6
5 53.53 17.45 11.0 - -
6 42.00 2.55 E-LaSF015 12.0 1.804 46.6
7 –42.00 0.60 12.0 - -
8 11.54 5.35 E-SSK5 10.0 1.658 50.9
9 –20.10 2.95 SF6 10.0 1.805 25.4
10 11.04 6.57 8.0 - -
IMA ∞ 1.2 - -

Prescription 12.4 40X fluor, derived from Misawa,22 U.S. Patent No. 5,699,196.
(M = 40XI, TL = 200 mm, NA = 0.75, f = 5 mm, FOV 0.3 mm)

Surf Radius Thickness Glass Diameter nd νd

OBJ ∞ ∞ 0.0 - -
1 ∞ 5.00 10.0 - -
STO –8.84 1.80 E-FK5 10.0 1.487 70.4
3 10.94 4.76 LLF6 13.0 1.532 48.8
4 –16.02 30.00 13.0 - -
5 27.49 1.50 F5 13.0 1.603 38.0
6 12.55 5.60 N-FK56 13.0 1.434 95.0
7 –11.24 1.50 SF8 13.0 1.689 31.2
8 –39.44 0.23 13.0 - -
9 29.31 3.61 E-FKH1 13.0 1.498 82.5
10 –29.75 1.58 13.0 - -
11 6.72 5.48 N-FK56 11.0 1.434 95.0
12 –16.61 1.00 K3 11.0 1.518 59.0
13 15.13 0.18 8.0 - -
14 3.36 3.33 N-LaF34 6.6 1.773 49.6
15 2.25 1.09 3.0 - -
16 ∞ 0.17 D-263-C 3.0 1.523 54.6
IMA ∞ 0.4 - -
Prescriptions 221

Prescriptions 12.5–12.7 60X TIRF, derived from Yamaguchi,23 U.S. Patent No.
6,519,092. (M = 60XI, TL = 200 mm, ISNA = 1.4, f = 3.3 mm, FOV = 0.2 mm)

Surf Radius Thickness Glass Diameter Comment nd νd

OBJ ∞ ∞ 0.0 - -
1 ∞ –4.90 30.0 Flange - -
2 ∞ 1.00 25.0 Barrel - -
3 ∞ 1.00 12.0 Lens stop - -
4 8.17 3.00 E-SF6 10.4 L8 1.805 25.4
5 204.67 2.60 E-LaSF013 10.4 1.804 39.6
6 4.80 1.70 7.1 - -
STO ∞ 1.70 6.6 - -
8 –4.82 5.00 E-LaSF013 7.1 L7 1.804 39.6
9 238.40 5.20 N-PSK58 15.0 1.569 71.2
10 –8.78 0.15 15.0 - -
11 –25.25 1.00 KF9 16.4 L6 1.523 51.5
12 48.48 6.30 CaF2 19.0 1.434 95.0
13 –11.82 1.60 KzFH1 19.0 1.613 44.4
14 –18.67 0.10 21.0 - -
15 28.66 8.00 CaF2 22.0 L5 1.434 95.0
16 –14.80 1.10 KzFH1 22.0 1.613 44.4
17 –47.67 0.15 24.0 - -
18 16.27 9.40 CaF2 23.6 L4 1.434 95.0
19 –20.05 1.00 KzFH1 23.6 1.613 44.4
20 84.75 0.15 23.2 - -
21 13.45 6.80 N-PSK58 20.8 L3 1.569 71.2
22 –34.23 1.00 F5 20.8 1.603 38.1
23 27.891 0.10 20.4 - -
24 7.247 3.75 N-PSK58 13.6 L2 1.569 71.2
25 13.716 0.10 12.4 - -
26 3.716 3.60 LaSF35 7.4 L1 2.022 29.1
27 1.332 0.65 K3 2.3 1.518 58.9
28 ∞ 0.141505 K3 6.0 Oil 1.518 58.9
29 ∞ 0.17 D-263-C 10.0 Coverslip - -
IMA ∞ D-263-C 23.37486 1.523 54.6
222 Appendix

Prescriptions 12.9–12.10 Aplanat, from Smith.24

(M = 100X, TL = 180 mm, ISNA = 1.3, f = 1.8 mm, FOV = 0.12 mm)

Surf Radius Thickness Glass Diameter nd νd

OBJ ∞ 180.00 12.1 - -
STO 4.68 1.70 CaF2 4.6 1.434 95.0
2 –3.86 1.04 F2 4.6 1.620 36.4
3 41.30 0.09 4.3 - -
4 3.51 1.36 CaF2 4.2 1.434 95.0
5 –5.10 0.66 BaF2 4.2 1.570 49.4
6 –310.90 0.24 4.2 - -
7 1.93 1.08 BK1 3.4 1.510 63.5
8 7.71 0.02 3.0 - -
9 0.77 0.93 BK7 1.5 1.517 64.2
10 ∞ 0.10 TYPE A 1.3 1.515 41.6
11 ∞ 0.17 K5 1.3 1.522 59.5
IMA ∞ K5 0.1 1.522 59.5

Prescription 12.11 Schwarzschild, after geometric relationships in Ref. 25.

(M = 10XI, TL = 200 mm, ISNA = 0.25, f = 20, FOV = 10 mm)

Surf Radius Thickness Glass Diameter nd νd

OBJ ∞ ∞ 0.0 - -
1 ∞ 10.00 13.0 - -
2 ∞ 40.00 10.0 - -
STO 24.72 –40.00 Mirror 10.0 1.00 -
4 64.72 40.00 Mirror 44.4 1.00 -
5 ∞ 44.72 25.1 - -
6 ∞ 0.00 1.2 - -
IMA –20.00 10.00 - -

Prescription 12.12 Internal parabola, after Larson et al.26 and Krogmeier et al.27
(M = 20X, TL = 200 mm, ISNA = 1.0, f = 10, FOV = 0.6 mm)

Surf Radius Thickness Glass Diameter Conic constant nd νd

OBJ ∞ ∞ 0.0 - -
1 ∞ 10.00 0.0 - -
2 ∞ 1.00 Silica 40.0 1.458 67.8
3 ∞ 13.63 Silica 40.0 1.458 67.8
STO –29.26 –13.63 Mirror 40.0 –1.000 1.458 67.8
5 ∞ –1.00 Silica 3.3 1.458 67.8
IMA ∞ Silica 0.7 1.458 67.8
Prescriptions 223

Prescription 13.1 Production doublet as a tube lens, from Linos Photonics No. 322271.
(M = 1X, TL = 200 mm, f = 200 mm, ISNA = 0.025, FOV = 12 mm)

Surf Radius Thickness Glass Diameter Comment nd νd

OBJ ∞ ∞ 0.0 - -
STO ∞ 200.00 5.0 L-stop - -
2 92.39 5.30 N-BK7 31.5 322271 1.517 64.2
3 –84.75 2.80 F4 31.5 1.617 36.6
4 –1074.60 0.00 31.5 - -
5 ∞ 194.71 27.0 - -
IMA ∞ 21.9 Sense - -

Prescription 13.2 Custom doublet as a tube lens.

(M = 1X, TL = 200 mm, ISNA = 0.025, f = 200 mm, FOV = 12 mm)

Surf Radius Thickness Glass Diameter Comment nd νd

OBJ ∞ ∞ 0.0 - -
STO ∞ 200.00 10.0 L-stop - -
2 131.27 8.00 N-BK7 31.5 322271 1.517 64.2
3 –67.26 5.00 F4 31.5 1.617 36.6
4 –238.12 194.31 31.5 - -
IMA ∞ 12.0 Sense - -

Prescriptions 13.3–13.4 Doublet-pair tube lens, from Misawa,22 U.S. Patent No.
5,699,196. (M = 1X, TL = 200 mm, ISNA = 0.025, f = 200 mm, FOV = 12 mm)
Required 5X scale of Table 2.

Surf Radius Thickness Glass Diameter nd νd

OBJ ∞ ∞ 0.0 - -
STO ∞ 140.00 10.0 - -
2 75.05 5.10 <MODEL> 30.0 1.623 57.1
3 –75.05 2.00 <MODEL> 30.0 1.750 35.2
4 1600.73 7.50 30.0 - -
5 50.26 5.10 <MODEL> 30.0 1.668 42.0
6 –84.55 1.80 <MODEL> 30.0 1.613 44.4
7 36.91 168.43 30.0 - -
IMA ∞ 12.0 - -
224 Appendix

Prescription 13.6 Doublet-pair tube lens with real glass, from Misawa,22 U.S. Patent No.
5,699,196. Real glass types of the 1990s are not available in 2009.
(M = 1X, TL = 200 mm, ISNA = 0.025, f = 200 mm, FOV = 12 mm)

Surf Radius Thickness Glass Diameter nd νd

OBJ ∞ ∞ 0.0 - -
STO ∞ 140.00 10.0 - -
2 75.05 5.10 SK10 30.0 1.623 56.9
3 –75.05 2.00 LaFN7 30.0 1.750 35.0
4 1600.73 7.50 30.0 - -
5 50.26 5.10 BaSF6 30.0 1.668 41.9
6 –84.55 1.80 KzFSN4 30.0 1.613 44.3
7 36.91 169.25 30.0 - -
IMA ∞ 12.0 - -

Prescription 13.9 Cube within a finite conjugate. (Strehl ratio at the d line is 0.9.)
(M = 1X, TL = 200 mm, ISNA = 0.025, FOV = 12 mm)

Surf Radius Thickness Glass Diameter nd νd

OBJ ∞ –200 12.0 - -
STO ∞ 150.00 10.0 - -
2 ∞ 50.00 Silica 50.0 1.458 67.8
3 ∞ 15.72 11.8 - -
IMA ∞ 12.0 - -

Prescription 13.10 Tilted plate within a finite conjugate. (Strehl ratio at the d line is 0.9.)
(M = 1X, TL = 200 mm, ISNA = 0.025, FOV = 12 mm)

Surf Radius Thickness Glass Diameter Tilt nd νd

OBJ ∞ –200 12.0 - -
STO ∞ 150.00 10.0 - -
2 - 0.00 - –45 - -
3 ∞ 1.38 Silica 50.0 1.458 67.8
4 ∞ –1.38 50.0 - -
5 - 50.42 - 45 - -
IMA ∞ 12.9 - -
Prescriptions 225

Prescription 13.11 Thick plate warped within ∞ correction. (Strehl ratio at the d line is 0.9.)
(M = 1X, TL = 200 mm, ISNA = 0.00, FOV = 0 deg)

Surf Radius Thickness Glass Diameter Tilt nd νd

OBJ ∞ ∞ 0.0 - -
STO ∞ 150.00 10.0 - -
2 - 0.00 - –45 - -
3 –580.62 3.00 Silica 50.0 1.458 67.8
4 –583.62 –3.00 50.0 - -
5 - 50.00 - 45 - -
IMA ∞ 0.0 - -

Prescription 13.12 Thin plate warped within ∞ correction. (Strehl ratio at the d line is 0.9.)
(M = 1X, TL = 200 mm, ISNA = 0.00, FOV = 0 deg)

Surf Radius Thickness Glass Diameter Tilt nd νd

OBJ ∞ ∞ 0.0 - -
STO ∞ 150.00 10.0 - -
2 - 0.00 - –45 - -
3 353.15 1.10 Silica 50.0 1.458 67.8
4 352.05 –1.10 50.0 - -
5 - 50.00 - 45 - -
IMA ∞ 0.0 - -

Prescription 14.2 Kellner eyepiece, comprising production parts from Linos Photonics
No. 312315 plano-convex 50 × 22.4 and 322285 doublet 30 × 12.5.
(M = 10XA, TL = 160 mm, OSNA = 0.025, f = 25 mm, FOV = 15 mm)

Surf Radius Thickness Glass Diameter Comment nd νd

OBJ ∞ –160 15.0 - -
STO ∞ 160.00 8.0 LS - -
2 ∞ 5.89 15.0 Image - -
3 ∞ 4.50 N-BK7 22.4 PC50 1.517 64.2
4 –26.42 19.09 22.4 312315 - -
5 19.67 3.50 N-SK2 12.5 322285 1.607 56.7
6 –15.51 1.50 SF10 12.5 DB30 1.728 28.4
7 –70.29 16.11 12.5 - -
IMA ∞ 1.8 Exit pupil at eye - -
226 Appendix

Prescription 14.3 Plössl eyepiece 10X telecentric, comprising production parts from
OptoSigma 026-1140 doublets 50 × 30.
(M = 9.4XA, TL = 1000 mm, OSNA = 0.025, f = 25 mm, FOV = 26.6 mm)

Surf Radius Thickness Glass Diameter Comment nd νd

OBJ ∞ –1000.00 25.0 - -
STO ∞ 1000.00 50.0 LS - -
2 ∞ 16.85 25.0 FS - -
3 62.73 3.00 SF2 30.0 1.648 33.8
4 23.40 11.10 BK7 30.0 1.517 64.2
5 –32.50 0.20 30.0 026-1140 - -
6 32.50 11.10 BK7 30.0 026-1140 1.517 64.2
7 –23.40 3.00 SF2 30.0 1.648 33.8
8 –62.73 15.78 30.0 - -
IMA ∞ 3.2 Exit pupil - -

Prescription 14.4 Erfle eyepiece 1923, from U.S. Patent 1,478,704.54

(M = 10XA, TL = 1000 mm, OSNA = 0.025, f = 25 mm, FOV = 25.2 mm)

Surf Radius Thickness Glass Diameter Comment nd νd

OBJ ∞ –200.00 25.0 - -
STO ∞ 200.00 10.0 LS - -
2 ∞ 12.24 25.0 Image - -
3 –169.70 1.50 F2 30.0 L11 1.620 36.4
4 25.00 13.80 BK7 36.0 L12 1.517 64.2
5 –34.00 0.20 36.0 - -
6 65.40 7.00 BK7 36.0 L2 1.517 64.2
7 –65.40 0.20 36.0 - -
8 28.70 10.70 BK7 32.0 L31 1.517 64.2
9 –28.70 1.50 F2 32.0 L32 1.620 36.4
10 786.00 18.70 28.0 Eye relief - -
IMA ∞ 2.4 Exit pupil - -
Prescriptions 227

Prescription 14.5 Erfle eyepiece 1997. Glass from prior art in U.S. Patent No. 5,691,850
by Arisaka55 and optimization by Seward.
(M = 10XA, TL = 1000 mm, OSNA = 0.025, f = 25 mm, FOV = 25.0 mm)

Surf Radius Thickness Glass Diameter Comment nd νd

OBJ ∞ –200.00 25.0 - -
STO ∞ 200.00 35.0 LS - -
2 ∞ 11.48 25.0 Object - -
3 –50.50 3.10 SF6 32.0 L11 1.805 25.4
4 37.10 11.60 N-SK16 36.0 L12 1.620 60.3
5 –37.10 0.30 36.0 - -
6 112.00 6.20 LaK8 36.0 L2 - -
7 –112.00 0.30 36.0 - -
8 28.10 12.30 LaK8 32.0 L31 - -
9 –64.50 3.10 SF6 32.0 L32 - -
10 417 18.80 28.0 Eye relief - -
IMA ∞ 2.07 Eye - -
Works Consulted
G. R. Fowles, Introduction to Modern Optics, 2nd Ed., Dover Publications, New
York (1989).
Matrix methods for reflection and refraction. Reflection coefficients.
Maxwell’s laws in point format and complex format. Refractive index is
derived from conduction of free and bound electrons. Math: algebra, complex
numbers.
W. H. Hayt, Engineering Electromagnetics, 7th Ed., McGraw-Hill, New York
(2005).
Broad introduction to electromagnetics. Vector operations. Point format of
Maxwell’s equations.
B. H. Walker, Optical Engineering Fundamentals, 2nd Ed., SPIE Press,
Bellingham, Washington (2009).
Numerous descriptions of vision systems with some prescriptions. Several
magnifiers and eyepieces. Math: simple algebra.
R. E. Fischer and B. Tadic-Galib, Optical System Design, 2nd Ed., McGraw-Hill,
New York (2008).
Practical design considerations. Some discussion of materials.
The American Heritage Dictionary of the English Language, W. Morris, Ed.,
American Heritage Publishing, Rockville, Maryland (1973).
Origins of words.

229
Recommended Reading
W. J. Smith, Modern Optical Engineering, 4th Ed., McGraw-Hill, New York
(2007).
Introduction to lens design. Paraxial methods. Geometric aberrations.
Practical design considerations. Math: algebra, trigonometry.
A. Beiser, Concepts of Modern Physics, McGraw-Hill, New York (2002).
Wave propagation. Diffraction. Electron states. Molecular structure.
Absorption and emission. Wave-particle duality. Math: algebra, calculus,
statistics.
E. Hecht, Optics 4th Ed., Addison-Wesley, Reading, Massachusetts (2001).
Excellent reference for the field of optics. Maxwell’s equations in both point
and integral format. Refractive index. Wave propagation. Geometric optics.
Polarization. Reflection coefficients. Fourier optics. Quantum nature of light.
Math: algebra, vectors, calculus.
J. W. Goodman, Introduction to Fourier Optics 3rd Ed., McGraw-Hill, New York
(2004).
A classic text book on summation of electromagnetic waveforms. Essential
knowledge for an optical engineer. Math: calculus, Fourier methods.

231
References
1. G. H. Seward, “Two-dimensional space-angle product of a Gaussian beam,”
Opt. Eng. 40(2), pp. 1959–1962 (2001).
2. B. H. Walker, Optical Engineering Fundamentals, 2nd Ed., SPIE Press,
Bellingham, WA (2008).
3. W. J. Smith, Modern Optical Engineering, 2nd Ed., McGraw-Hill, New York
(1990).
4. H. R. Phillipp, “Silicon dioxide glass,” Fig. 2.4 in F.-T. Lentes, “Refractive
index and dispersion,” in The Optical Properties of Glass, H. Bach and N.
Neuroth, Eds., 2nd corrected printing, Springer, New York (1998).
5. E. Hecht, Optics, Addison-Wesley, Reading, MA (1987).
6. F. B. Hildebrand, Advanced Calculus for Applications, Prentice-Hall,
Englewood Cliffs, New Jersey (1976).
7. A. Beiser, Concepts of Modern Physics, McGraw-Hill, New York (2003).
8. H. Du, R. A. Fuh, J. Li, A. Corkan, and J. S. Lindsey, “PhotochemCAD: a
computer-aided design and research tool in photochemistry,” Photochem.
Photobiol. 68, pp. 141–142, (1998).
9. R. Y. Tsien, “The green fluorescent protein,” Annu. Rev. Biochem. 67, pp.
509–544 (1998).
10. G. H. Seward, “Image contrast of subpixel objects,” Opt. Eng. 37(2), 710–
716 (1998).
11. E. F. Schubert, Light-Emitting Diodes, Cambridge Univ. Press, Cambridge
(2006).
12. G. H. Smith, Camera Lenses, SPIE Press, Bellingham, Washington (2009).
13. S. Matsui, “Long focal length microlens system,” U.S. Patent No. 5,831,775
(1998).
14. G. H. Seward, “Spectroscopy lens for telecentric sensor,” U.S. Patent
Application 12,333,488 (2008).
15. E. M. Nelson, “Critical microscopy,” J. R. Microscop. Soc., pp. 282–289
(1910).
16. C. Zeiss, “Description of Professor Abbe’s Apertometer,” Proc. Royal
Microsc. Soc. 28(4), pp. 19–22 (1877).
17. A. Kohler, “Zeitschrift fur wissenschaftl,” Mikroskopie, 10, pp. 433–440
(1893),

233
234 References

18. A. Kohler, “A new system of illumination for photomicrographic purposes,”

Proc. Royal Microsc. Soc. 28(4), pp. 181–185 (1993).
19. ANSI Z136.1, American National Standard for Safe Use of Lasers (2007).
20. T. Marshall and M. Pashley, “LED collimation optics with improved
performance and reduced size,” U.S. Patent No. 6,547,423 B2 (2003).
21. N. Fukutake, “Microscope objective lens,” U.S. Patent No. 6,128,139 (2000).
22. J. Misawa, “Microscope objective lens and microscope,” U.S. Patent No.
5,699,196 (1997) .
23. K. Yamaguchi, “Immersion microscope objective lens,” U.S. Patent No.
6,519,092 B2 (2003).
24. W. J. Smith, Modern Lens Engineering, 2nd Ed., McGraw-Hill, New York
(1992).
25. M. J. Riedl, Optical Design Fundamentals for Infrared Systems, SPIE Press,
Bellingham, Washington (2001).
26. J. W. Larson, G. H. Seward, G. R. Yantz, D. Johnson, and J. Krogmeier,
“Systems and methods for measurement optimization,” U.S. Patent No.
7,262,859 B2 (2007).
27. J. R. Krogmeier, I. Schaefer, G. Seward, G. R. Yantz, and J. W. Larson, “An
integrated optics microfluidic device for detecting single DNA molecules,”
Lab. Chip. 7, 1767–1774 (2007).
28. P. Muller, “Optical system,” U.S. Patent No. 1,977,689 (1929).
29. R. E. Bitner, “Catadioptrical lens,” U.S. Patent No. 2,215,900 (1940).
30. Y. Suehiro,“Reflective type light-emitting diode,” U.S. Patent No. 6,641,287
B2 (2003).
31. H. Sakai and T. Kawamura, “Light-emitting diode,” U.S. Patent No.
4,698,730 (1987).
32. V. Medvedev and W. A. Parkyn, “Beam-forming lens with internal cavity,”
U.S. Patent No. 5,575,557 (1998).
33. G. C. Holst and T. S. Lomheim, CMOS/CCD Sensors and Camera Systems,
SPIE Press, Bellingham, Washington (2007).
34. J. R. Janesick, Scientific Charge Coupled Devices, SPIE Press, Bellingham,
Washington (2001).
35. Eastman Kodak Company, KODAK KAI-1020 IMAGE SENSOR , Revision
5.0 MTD/PS-0205 (2007).
36. Eastman Kodak Company, KODAK KAC-00401 IMAGE SENSOR , Revision
1.1 (2009).
37. Eastman Kodak Company, KODAK PROFESSIONAL T-MAX FILMS, F-
4016 (2007).
38. R. L. De Valois and K. K. De Valois, Spatial Vision, Oxford Univ. Press,
New York (1990).
References 235

39. T. D. Lamb, S. P. Collin, and P. N. Pugh, “Evolution of the vertebrate eye:

opsins, photoreceptors, retina and eye cup,” Nature 2008(8), pp. 960–975
(2007).
40. J. Sjostrand, N. Conradi, and L. Klaren, “How many ganglion cells are there
to a foveal cone?” Graefe’s Arch. Clin. Exp. Ophthalmol. 232, pp. 432–437
(1994).
41. P. Artal and R. Navarro, “Monochromatic modulation transfer function of the
human eye for different pupil diameters: and analytical expression,” J. Opt.
Soc. A 11(1), pp. 246–249 (1994).
42. M. Tessier-Lavigne, “Photo transduction and information processing in the
retina,” in Principles of Neural Science, E. R. Kandel, J. H. Schwartz, and T.
M. Jessell, Eds., McGraw-Hill, New York (2000).
43. P. Sterling, “Retina,” in The Synaptic Organization of the Brain, G. M.
Shepherd, Ed., Oxford Univ. Press, London (1996).
44. D. H. Kelly, “Visual responses to time-dependent stimuli. I. Amplitude
sensitivity measurements,” J. Opt. Soc. A 51(4), pp. 422–429 (1961).
45. D. H. Kelly, “Motion and vision. II. Stabilized spatio-temporal threshold
surface,” J. Opt. Soc. A 69(10), pp. 1340–1349, (1979).
46. R. M. Steinmann,“Effect of target size, luminance, and color on monocular
fixation,” J. Opt. Soc. A 55(9), pp. 1158–1165 (1965).
47. R. L. DeValois, H. Morgan, and D. M. Snodderly, “Physophysical studies of
monkey vision—III. Spatial luminance contrast sensitivity tests of macaque
and human observers,” Vis. Res. 14, pp. 75–81 (1974).
48. M. S. Banks, A. B. Sekuler, and S. J. Anderson, “Peripheral spatial vision:
limits imposed by optics, photoreceptors, and receptor pooling,” J. Opt. Soc.
Am. 8(11), pp. 1775–1787 (1991).
49. B. B. Boycott and H. Kolb, “The horizontal cells of the Rhesus monkey
retina,” J. Comput. Neur. 148, pp. 115–140 (1973).
50. R. Jademzik, B. Hladik, and P. Hartmann, “Tailored properties of optical
glasses,” Proc. SPIE 5965, 59651D (2005).
51. M. Clement, “The chemical composition of glass,” Chapter 2.2 in The optical
properties of glass, H. Bach and N. Neuroth, Eds., Second corrected printing,
Springer (1998).
52. SCHOTT North America, Inc., “TIE-29: Refractive index and dispersion,”
Duryea, Pennsylvania (2007).
53. J. W. Goodman, Introduction to Fourier Optics, McGraw-Hill, New York
(1968).
54. H. Erfle, “Ocular,” U.S. Patent No. 1,478,704 (1923).
55. M. Arisaka, “Eyepiece,” U.S. Patent No. 5,691,850 (1997).
Index
∞ correction, 9 background
fluorescence, 59
radiance, 4, 74
Abbe resolution, 95 Bessel function, 46
absorbance, 53 bipolar, 161
absorption, 54 cells, 160
coefficient, 43, 53 BK7, 172, 174, 175, 176, 180
achromat, 113, 129 blink, 99
active pixel sensor (APS), 151 borate, 176, 179, 184
Airy boron, 170, 171
contrast, 82 borosilicate, 171
pattern, 46 borosilicate crown, 142
radius, 66 optical (BK7), 172, 174, 175,
aluminum, 40, 60, 99, 171, 182, 198 176, 180
angular magnification, 7, 22, 24 bound electrons, 37, 196
anode, 153 brass, 181
anomalous partial dispersion ΔPgf, Brewster angle, 42
176 bright field, 12
antimony, 169, 172, 176 bubbles, 169
aplanat, 113, 121 buried channel, 148
apochromat, 129
arsenic, 169, 172
artificial C line, 64
numerical aperture, 159 calcium fluoride, 118, 173, 175
pupil, 139, 158–160, 164 camera, 157
aspheric cardinal points, 23
lens, 91 cell, 12
surface, 90 center-surround antagonism, 161
astigmatism, 3, 27, 29, 31 central numerical aperture, 3, 4, 81,
atmosphere, 129, 175, 176, 180 82, 119, 122
axial charge-coupled device (CCD), 147
color, 3, 29, 32 chief ray, 8, 10, 12, 18
magnification, 7, 20 chip, 72
ray, 23 chrome, 132, 181
chromium, 171, 179
chromophore, 57, 60
237
238 Index

circular aperture, 46 dark

coefficient, 36 current, 1, 5, 73, 148, 151
absorption, 43, 53 field, 15
extinction, 37, 43, 53, 198 decay, 36, 44, 57, 165, 171, 196
polynomial for asphere, 33 defocus, 27
reflection, 42 dense flint, 191
refraction, 42 depth of focus, 3
Sellmeier, 178, 183 detector noise, 4, 75
scatter, 204 diaphragm, 98
thermal expansion, 171 diffuser, 95
coherence length, 46, 200 digs, 5, 72, 107
coherent, 94, 95, 98 diopter, 8
system, 49 dipole, 43, 58
collection efficiency distortion, 27, 29, 144
angular, 77 DNA, 58
spatial, 78 double Gauss lens, 70, 85
color correction, 91, 118, 131, 142, Drude, 40
143, 174, 176, 189, 198 model, 37, 197
coma, 3, 27, 29, 30 spectra, 198
complementary metal-oxide dynode, 152
semiconductor (CMOS), 147,
148, 151
condenser, 93, 96, 100, 101 earth’s crust, 169
cones, 157 ecofriendly, 187, 189
conic constant, 33, 209 ecoglass, 171, 179
conical edge spread, 69
approximation, 69 electric field, 37, 46, 49, 53, 195,
frequency, 68, 204 202
contrast, 73 polarization, 43
sensitivity, 158 electron, 54
plot, 165 multiplication CCD (EMCCD),
cost, 1, 5, 91, 129, 141, 142, 174, 149
178, 180 electron volts, 55
critcal illumination numerical electrons, 147
aperture, 94 emission, 54, 99
crown glass, 169, 172, 183 encircled-energy plot, 68
crystal, 173, 174, 183 entrance pupil, 18, 139
crystalline, 169 environment, 171
crystallize, 172, 179 environmental, 172, 179
curvature, 162, 196 environmentally safe, 189
of electric field, 44 erect image, 20
étendue, 79
evanescent field, 44
d line, 64 exit pupil, 11, 12, 18, 139
D263M, 107–112, 172
Index 239

extinction coefficient, 37, 43, 53, Franck-Condon diagram, 56

198 free electrons, 37, 196
extraordinary spatial frequency, 45 fringes, 71, 98
eye, 7, 157 full-frame transfer, 149
relief, 11, 12, 139 device, 150
safe, 99 full-pixel contrast, 79
eyepiece, 11, 139 full-well capacity, 1, 147, 149

F2, 41, 176–178, 180, 184–185, 198 GABA, 158, 166

F line, 64 ganglion, 157, 161
f/#, 18, 209 gate, 147
f/# inflation, 18, 70, 85, 209 Gaussian
far point of human vision, 7 beam, 45, 48
feedback, 161, 166 power, 201
feed-forward, 161, 166 depth of focus, 3
Fermi level, 147 estimate, 3, 50
field lens formula, 19
lens, 140 numerical aperture, 2, 45, 81, 201
stop, 18 geometric
fifth-order aberration, 29 point-spread, 64
filament, 95 radius, 66
filter, 51, 87, 88, 89, 90, 129 German, 172
fining agent, 169, 171 glass
finite tube, 8 borosilicate krone (BK), see
flame, 94 borosilicate
flange, 9, 90 calcium fluoride (CAF2), 173
flint, 41, 142, 169, 172, 177, 180, crown (K), 169, 172, 183
184, 198 flint (F), 41, 142, 169, 172, 177,
float glass, 171 180, 184, 198
fluor, 113, 115 fluorophosphate krone (FK), 175
fluorescein, 60, 61 krone (K), 172, 183
isothiocyanate (FITC), 60, 62 kurz flint (KzF), 176
fluoride, 173 kurz flint sonder (KzFS), 146,
fluorine, 173 187
fluorite, 113 lang krone (LgK), 186
fluorophore, 55, 57, 97 lanthanum (LaK, LaF), 113, 143,
fluorophosphate glass, 173, 175 170, 172
flux, 73 lead, 171, 180
focal licht flint (LF), 172
length, 21 long, 41, 186
point, 9, 17, 19, 23 phosphate krone (PK), 173–175
ray, 23 schwer flint (SF), 143, 172, 191
fovea, 157 schwer krone (SK), 129, 143, 174
foveola, 157 short flint (KzF), 39, 176, 186
240 Index

glass (continued) Jablonski diagram, 55

short flint special (KzFS), 176,
186
silica, 169 K7, 41, 180, 184, 198
soda lime, 169 Köhler, 97
super-schwer krone (SSK), 115 illumination, 95
glass map, 142, 145, 172, 175, 198 krone, 172, 183
glass temperature, 169 kurz, 186
green fluorescent protein, 60, 62 flint, 176
sonder, 176, 187

heavy flint, 172

hemispherical collection efficiency Lambertian, 76, 98
(HCE), 82, 91 emittance, 76
holes, 147 lang krone, 186
horizontal cell, 158, 160, 165 lanthanum glass, 113, 143, 170, 172
lateral
color, 3, 29, 31
i line, 180 inhibition, 157, 160, 161
illumination magnification, 19
field, 93 lead, 170, 179, 185
lens stop, 13, 93, 96 borate, 118
numerical aperture, 13, 93, 97 glass, 171, 180
image oxide, 171, 184, 191
conjugates, 19 poisoning, 185
distance, 23 silicate, 187
image-space lens stop, 8, 9, 18
f/#, 209 licht flint, 172
numerical aperture (ISNA), 209 lifetime, 37, 55, 96, 98, 186
immersion lens, 117 liquidus temperature, 169
impurities, 169, 181 long crown, 41, 186
inclusions, 181 Lorentz, 38, 40
incoherent, 94, 95 model, 37
indole, 57 oscillator, 197
infinity correction (∞), 9 spectra, 198
interline transfer, 149 luminance, 73
internal reflection, 21
intrinsic fluorescence, 181
inverted image, 20 magenta ring, 100
iris, 97, 98, 157 magnification, 20, 209
iron, 132, 169, 170, 190 of an objective lens, 8
irradiance, 42, 43, 46, 48, 53, 199, magnifier, 11, 24, 140
201 marginal
profile, 48 numerical aperture, 1, 18, 82,
isotropic emittance, 76 111, 119, 122, 155, 204
Index 241

marginal (continued) object

ray, 2, 9, 10, 18 distance, 23
merit function, 106 numerical aperture, 139
metric, 2, 4, 18, 40, 63, 100, 176, signal, 4
198 object-space numerical aperture,
microlens, 150 105
mirror image, 20 objective lens, 8, 9
modulation index, 69, 74 ocular, 139
modulation transfer function (MTF), lens, 11
4, 68 optical
coherent, 49 axis, 9
incoherent, 49 distance, 19, 22
of a Gaussian pattern, 50 path length, 23, 111
of an Airy pattern, 50 power, 7, 21
ordinary spatial frequency, 44
oscillation, 37, 44, 190
natural pupil, 11, 139, 158, 159, 165 oscillator, 40, 177, 198
near point of human vision, 7, 11 outward curvature, 44
negative partial dispersion ΔPgf,
176–177, 189, 198
nickel, 37, 132 parabola, 123
niobium, 170, 176, 179, 187 paraxial
niobium-silicate short flint, 190 numerical aperture, 2, 9, 119, 122
nodal optics, 17, 22
point, 17, 23 partial dispersion, 40, 176
ray, 23 peak irradiance, 201
normal dispersion, 40 period
normal glass types, 176, 177, 198 spatial, 35
numerical aperture (NA), 94, 209 temporal, 36
artificial, 159 periodic table, 171
central, 3, 4, 81, 82, 119, 122 periphery, 158, 165
critical, 94 Petzval, 27, 29
Gaussian, 2, 45, 81, 155, 201 phantom hydrogen, 28, 60
illumination, 13, 93, 97 phosphate krone, 173, 175
image-space, 209 phosphorescence, 55
marginal, 1, 18, 82, 111, 119, photocathode, 152
122, 155, 204 photodiode, 151
object, 139 photomultiplier tube (PMT), 152
object-space, 105 photoreceptor, 157, 160
paraxial, 2, 9, 119, 122 plan achromat, 114
tube, 140 plan objective, 113
vision, 13 Planck’s law, 55
plane wave, 36
plasma frequency, 37, 197
Plössl eyepiece, 141
242 Index

point relay lens, 23, 93, 96, 140

source, 98 resolution, 4
spread, 66, 94 resonant, 177, 178
Poisson distribution, 75 frequency, 37, 175, 198
polarization, 42, 161 lifetime, 41, 190
electric field, 43 peak, 187
positive relative partial dispersion retina, 99, 157, 160
ΔPgF, 175–177, 198 rms radius, 66
power rods, 157
optical, 7, 21
prescription, 70
principal saccade, 99, 162
plane, 9, 19, 105 saccation, 157, 162
point, 17, 23 sagittal, 65
surface, 2, 17 plane, 29, 65
propagation vector, 41 scatter, 51, 204
pupil, 158 schwer
pyrrole, 57 flint, 143, 172, 191
krone, 143
scratch, 5, 72, 107, 179
quantum Seidel aberrations, 17
efficiency, 73, 148 Sellmeier, 177
number, 54 coefficients, 178, 183
formula, 198
semiapochromat, 113, 129
radial magnification, 7 short flint, 39, 41, 176, 186
radiant flux, 43, 74 short-flint special, 187
radius of the Gaussian wavefront, 49 shot noise, 5, 75, 149, 155
Rayleigh distance, 2 silica, 38, 104, 109, 110, 132, 134,
ray, 17 169
ray-intercept plot, 3, 64 silver, 37, 154, 198
reactive, 129, 176, 184 halide, 154
read noise, 1, 73, 148, 151 sinc function, 48
real singlet state, 54
focus, 103 skirt, 3, 120, 121
image, 20 slide tolerance, 112
reduction, 20 Snell’s law, 41
reflection coefficient, 42 soda lime, 169
refractile, 12 source, 94
refraction coefficient, 42 space-angle product,
refractive Airy, 3, 47
index, 2, 36, 172, 177, 196, 197 axial, 199
object, 12 Gaussian, 2, 45, 48, 201
relative partial dispersion ΔPgF, 143, geometric, 98
175, 179, 188, 198 slit, 48
Index 243

spatial tolerance, 10,

curvature, 195 assembly, 93, 123
distance, 22 budget, 27, 87–89, 91, 105
frequency, 35, 196 fabrication, 96
period, 35 field, 33
wavelength, 23 specimen, 1, 105–108, 111, 112
special glass borate, 187 surface, 5, 33, 71, 124
spherical wavefront, 64, 135
aberration, 3, 27 wavelength, 99, 180
collection efficiency (SCE), 82, total
121 integrated scatter (TIS), 5, 51
spherochromatism, 29 internal reflection, 44, 109
spin, 54, 55 fluorescence (TIRF), 103, 117
spot diagram, 66 power, 47
stainless steel, 181 toxicity, 171
stop, 18 triplet state, 55, 56
Strehl ratio, 66, 82, 111 tryptophan, 59
subpixel, 80 tube
sun, 99 length, 8
surface lens, 9
form error, 5 numerical aperture, 140
quality, 5, 72
texture, 51, 72, 124
tolerance, 135 virtual
focus, 103
image, 20, 93
tangential, 65 object, 105
plane, 65 viscosity, 169, 170, 171, 183
pupil, 29 visibility, 74, 158
telecentric, 10, 90 vision
configuration, 96 field, 93, 140
temporal lens stop, 13, 93, 96
CSF, 166 numerical aperture, 13
decay, 153 volume defects, 72
feedback, 162
frequency, 35, 196
period, 36 water, 171
thick-lens model, 24 wave, 195
thickness tolerance, 111 wavefront error, 63
thin-lens model, 19, 23 wavelength tolerance, 116
third-order aberration, 29 well, 147
tilt magnification, 20 working distance, 9, 118
titanium, 170, 171, 172, 179
oxide, 171
ZEMAX, 92, 121, 183
 George Seward is a consultant in optical design. His
instruments are known for easy assembly and consistent
performance. His mathematical models define configurations
for maximum performance. He developed his expertise in
optical design through formal education, practical
experience, and continuing education.
His formal education occurred at Tufts University. His
bachelors program provided a solid foundation in physics,
mathematics, and engineering, while his master’s program
provided expertise in physical optics and charge transport.
His practical experience is broad. At Innovative Imaging Systems, he
developed micro-optics, microscopes, and telescopes. As a consultant, he
developed optics for biomedical applications: fluidics, fluorescence, genomics,
cytology, and pathology. Many of his optical designs resemble microscopes.
His continuing education developed by necessity. SPIE provided gems of
wisdom through seminars, articles, and books. Writing articles and conducting
seminars for SPIE extended the depth of his expertise.
His family contains numerous engineers. His father Harold worked at the
MIT Instrumentation Lab on guidance systems for the Apollo space program. His
brother David developed software at Digital Equipment Corporation. His brother
William developed circuitry at IBM. His brother James developed embedded
systems at Raytheon. His mother Janet encouraged education in all forms. His
great uncle William Seward served as Secretary of State under U.S. presidents
Lincoln and Johnson during the abolition of slavery and the purchase of Alaska.