Fundamentals of Device and Systems Packaging - Technologies and Applications
Fundamentals of Device and Systems Packaging - Technologies and Applications
Second Edition
McGraw-Hill Education books are available at special quantity discounts to use as premiums
and sales promotions or for use in corporate training programs. To contact a representative,
please visit the Contact Us page at www.mhprofessional.com.
Copyright © 2019, 2001 by McGraw-Hill Education. All rights reserved. Printed in the United
States of America. Except as permitted under the United States Copyright Act of 1976, no part
of this publication may be reproduced or distributed in any form or by any means, or stored
in a data base or retrieval system, without the prior written permission of the publisher.
ISBN 978-1-259-86155-0
MHID 1-259-86155-4
The first edition of this book was titled Fundamentals of Microsystems Packaging.
Information contained in this work has been obtained by McGraw-Hill Education from sources believed to be reliable. However,
neither McGraw-Hill Education nor its authors guarantee the accuracy or completeness of any information published herein,
and neither McGraw-Hill Education nor its authors shall be responsible for any errors, omissions, or damages arising out of use
of this information. This work is published with the understanding that McGraw-Hill Education and its authors are supplying
information but are not attempting to render engineering or other professional services. If such services are required, the
assistance of an appropriate professional should be sought.
To my parents: For your lifelong dedication to me as your only child, for your love
and support, and for teaching me the meaning of family and the value of education.
To my family (Anne, Dinesh, Vijay, and Suneel and grandchildren Gracen, Mason,
and Cooper): For being my love, pride, and joy to go to work for every day.
To my teachers: For teaching me how to learn well and quickly and for insisting on
my getting the best education at any cost.
To my students: For allowing me to learn with you and to teach you those principles
I believe in, in preparing you and challenging you to be the best you can be.
To NSF: For allowing me to pursue my dream of impacting the industry with the needed
strategic technologies and educated workforce.
To my academic colleagues: For accepting me into your world, for using this book,
and for working to make packaging an academic subject.
Finally, thank you to the book’s coordinator, Karen May, Georgia Institute of Technology;
the chapters editor, Prof. Raj Pulugurtha, Florida International University (FIU);
and, for graphics, Reed Crouch, Karen May, and Leonard Mendoza.
11.3.3 Receiver. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 468
11.3.4 Modulation Schemes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 469
11.3.5 Antenna. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 469
11.3.6 Components in RF Front-End Module. . . . . . . . . . . . . . . . 471
11.3.7 Filters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 473
11.3.8 RF Materials and Components. . . . . . . . . . . . . . . . . . . . . . 475
11.3.9 RF Modeling and Characterization Techniques. . . . . . . . 481
11.3.10 Applications of RF. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 483
11.4 What Is a Millimeter-Wave System?. . . . . . . . . . . . . . . . . . . . . . . . . . 486
11.5 Anatomy of a Millimeter-Wave Package. . . . . . . . . . . . . . . . . . . . . . 486
11.5.1 Fundamentals of Millimeter-Wave Packaging. . . . . . . . . 486
11.6 Millimeter-Wave Technologies and Applications. . . . . . . . . . . . . . . 492
11.6.1 5G and Beyond . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 492
11.6.2 Automotive Radars. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 492
11.6.3 Millimeter-Wave Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . 493
11.7 Summary and Future Trends. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 494
11.8 Homework Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 495
11.9 Suggested Reading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 496
12 Fundamentals of Optoelectronics Packaging Dr. Bruce C. Chou,
Prof. Gee Kung Chang, Dr. Daniel Guidotti, and Mr. Rui Zhang. . . . . . . . . . . 497
12.1 What Is Optoelectronics?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 499
12.2 Anatomy of an Optoelectronics System. . . . . . . . . . . . . . . . . . . . . . . 499
12.2.1 Fundamentals of Optoelectronics. . . . . . . . . . . . . . . . . . . . 500
12.2.2 Nomenclature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 504
12.3 Optoelectronic Technologies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 506
12.3.1 Active Optoelectronic Devices . . . . . . . . . . . . . . . . . . . . . . 506
12.3.2 Passive Optical Devices. . . . . . . . . . . . . . . . . . . . . . . . . . . . 515
12.3.3 Optical Interconnections . . . . . . . . . . . . . . . . . . . . . . . . . . . 519
12.4 Optoelectronic Systems, Applications, and Markets. . . . . . . . . . . . 527
12.4.1 Optoelectronic Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . 527
12.4.2 Applications of Optoelectronics. . . . . . . . . . . . . . . . . . . . . 534
12.4.3 Optoelectronics Markets . . . . . . . . . . . . . . . . . . . . . . . . . . . 538
12.5 Summary and Future Trends. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 540
12.6 Homework Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 540
12.7 Suggested Reading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 545
13 Fundamentals of MEMS and Sensor Packaging Prof. Peter Hesketh,
Prof. Oliver Brand, and Prof. Klaus-Juergen Wolter. . . . . . . . . . . . . . . . . . . . . 547
13.1 What Are MEMS? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 549
13.1.1 Historical Evolution. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 549
13.2 Anatomy of a MEMS Package. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 551
13.2.1 Fundamentals of MEMS Packaging. . . . . . . . . . . . . . . . . . 552
13.2.2 Nomenclature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 556
13.3 MEMS and Sensor Device Fabrication Technologies. . . . . . . . . . . . 558
13.3.1 Photolithographic Pattern Transfer . . . . . . . . . . . . . . . . . . 558
13.3.2 Thin-Film Deposition. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 559
xvii
E
lectronic packaging means primarily interconnecting active and passive
components to form modules initially and systems finally. While powering,
cooling, and protecting are important functions of any package, they are
secondary. Design, materials, and processes as well as test are assumed to be part of
interconnections. The end goal of packaging is to form systems by interconnecting all
the miniaturized components with and without transistors. These systems can be
consumer products like smartphones, self-driving electric cars, cloud computers, PCs,
laptops, notebooks, and watches. All these systems currently have two parts: devices
with transistors and components without transistors. Combined, these two span 15
different technologies, as described in this book.
This book is about both device and systems packaging. The market for these
technologies is in excess of $500 billion worldwide, employing millions of people. Yet,
device and systems packaging is not considered an academic subject, except at Georgia
Tech. A typical academic subject such as microelectronics or VLSI has extensive courses,
curricula, educational tracks, fundamental books, software tools, and hands-on
facilities. Most of these don’t exist in packaging. The number of universities offering
research programs with one or more courses, however, has increased rapidly in recent
years. In the United States, there were only two universities offering any packaging
research or education in the 1980s: Cornell and Stanford. It is estimated that this number
in the United States, Europe, and the Far East is about 200 now, having grown from less
than 10 universities in 1990 and 50 or so in 2000.
Fundamentals of Device and Systems Packaging is written to meet the academic need to
educate undergraduate and graduate students in the fundamentals as well as all 15
technologies necessary to form systems. While there have been many books since my
first modern book in 1988, Microelectronics Packaging Handbook, most, if not all, are
considered reference books. They typically deal with one or more topics in a review
fashion. However, electronic systems, as described in this book, require 15 different
device and systems packaging technologies. The topics span IC packaging and assembly,
system board packaging and assembly, and all the variations in between. It also deals
with electrical design, thermo-mechanical design, thermal design and technologies,
materials, processes, properties, reliability, and so on. This was the case 10–20 years ago
when I wrote the first book. Today’s systems are more than microelectronics. They are
RF, optical, MEMS and sensors, power, etc.
This book is more than interconnecting. It includes microelectronics, photonics, RF
and millimeter wave, and MEMS and sensors. All these devices along with other
xix
components need to be packaged to form systems. This is exactly what this book is all
about. It deals with all these technologies at the fundamental level—defining each,
illustrating the key concepts behind each, and introducing the critical nomenclature in
a systematic fashion. Each topical chapter includes fundamental equations, homework
problems, and future trends.
The book introduces Moore’s Law for packaging as Moore’s Law for cramming
transistors to form large ICs begins to come to an end. It proposes the cramming of
small ICs to form large homogeneous and heterogeneous packages. In contrast to
Moore’s Law for large ICs, with the highest number of transistors compromising
performance and cost, from node to node, Moore’s Law for packaging can be thought
of as interconnecting the smallest transistors in smaller ICs with the highest performance
and lowest cost in 2D, 2.5D, and 3D package architectures with an equal or higher
aggregate number of transistors.
I am grateful to all of the 55 global authors from industry and academia. This book,
just like the first edition, is written by so many authors and yet reads as if written by
one. I am thankful to Karen May, who headed the book project from start to finish, and
to Prof. Raj Pulugurtha, who moved from Georgia Tech to Florida International
University during the course of preparing this book. They have put in endless hours,
coordinating with all of the 55 authors from around the world.
I am also grateful to Senior Editors Michael McCabe and Lara Zoble and the other
editors at McGraw-Hill, who facilitated printing the book in two colors and at a very
affordable low cost for such a book that goes from wafers to systems with all the
electronic technologies and applications. I thank my wife, Anne, for her patience,
understanding, and full support during the course of the book, as I worked every day
and every weekend over 3 years, on top of my normal responsibilities of being a Professor
and the Director of the Microsystems Packaging Research Center at Georgia Tech.
My final thanks are, again, to all the authors who contributed to the book, working
nights and weekends, adding to their normal job responsibilities.
3D System Package
MEMS
DNA
Med Moore’s Law Image
BGA Package RF
for Packaging Processor
Memory
2.5D 3D
Optical MEMS Direct
Signal Programmable
Logic
I/O Processing
Moore’s Law for ICs SOC
Low Analog Memory
DRAM
Micro-
processor
CMOS-
RF
Chapter Outline
1.1.
What Is Packaging and Why?
1.1.1 What Is Packaging?
1.1.2 Why Is Packaging Important?
1.1.3 Every IC and Device Has to Be Packaged
1.1.4 Controls Performance of Computers
1.1.5 Controls Size of Consumer Electronics
1.1.6 Controls Reliability of Electronics
1.1.7 Controls Cost of Electronic Products
1.1.8 Required in Nearly Everything
1.2 Anatomy of an Electronic Packaged System from a Packaging Point of View
1.2.1 Fundamentals of Packaging
1.2.2 Systems Packaging Involves Electrical, Mechanical,
and Materials Technologies
1.2.3 Nomenclature
1.3 Devices and Moore’s Law
1.3.1 On-Chip Interconnections
1.3.2 Interconnect Materials
1.3.3 The Resistance and Capacitance Delays (RC Delays)
of On-Chip Interconnects
1.3.4 Future of Device Scaling
1.4 Electronic Technology Waves: Microelectronics, RF/Wireless,
Photonics, MEMS, and Quantum Devices
1.4.1 Microelectronics: The First Technology Wave
1.4.2 RF and Wireless: The Second Technology Wave
1.4.3 Photonics: The Third Technology Wave
1.4.4 Micro-Electro-Mechanical Systems (MEMS):
The Fourth Technology Wave
1.4.5 Quantum Devices and Computing: Fifth Wave
1.5 Packaging and Moore’s Law for Packaging
1.5.1 Three Eras in Packaging
1.5.2 Moore’s Law or SOC Era (1960–2010)
1.5.3 Moore’s Law for Packaging Era from 2010 to 2025
1.5.4 Moore’s Law for Systems Era from 2025
1.6 Electronic Systems Technologies Trends
1.6.1 Core Packaging Technologies
1.6.2 Packaging Technologies and Their Trends
1.7 Future Outlook
1.7.1 Emerging Computing Systems
1.7.2 Emerging 3D Systems Packaging
1.8 How the Book Is Organized
1.9 Homework Problems
1.10 Suggested Reading
Chapter Objectives
• Define packaging and describe its functions
• Introduce a three-level packaging to form systems
• Define and describe devices and their evolution
• Describe the evolution of packaging technologies during and post–Moore’s Law
• Introduce the concept of Moore’s Law for packaging
2
Transistor On-Chip
Level Integration
IC 1 IC 2
2 µm SOC
Package 80µm On-Package
Level 80 µm Integration
400µm SOP
Board RF
Level Digital Optical On-Board
400 µm PWB Integration
MEMS SOB
Sensors
Power
System integration can take place at one or more of the three levels, as shown in
Figure 1.1.
The current approach to end-product systems starts with on-chip transistor inte-
gration but ends up interconnecting mostly through system board level to SOB, as
described below:
1. The devices such as IC1 and IC2, which are individually-packaged on package
substrate 1 and substrate 2, are interconnected through the board, making chip-
to-chip interconnections very, very long.
2.
System integration, which includes active ICs with transistors and passive
components without transistors, can be integrated at IC level to SOC, package
level to SOP, or system level to SOB. Figure 1.1 illustrates the process of forming
these three-level systems. It starts with system design, and then wafer
fabrication processes using 750 mm thick, 300 mm diameter wafers. It contains
hundreds of ICs, which are diced from this wafer. These individual diced ICs
are then packaged either by wirebond, flip-chip, or tape-automated bonding
assembly technologies onto either ceramic, leadframe, organic laminate, silicon
or glass package substrates. These and other components such as capacitors (C)
or inductors (L) are assembled onto boards to form systems like smartphones.
Figures 1.2 and 1.3 show the technologies involved and their manufacturing flow
to form systems. As can be seen from Figure 1.2, there are two major parts to packaging
a system:
Figure 1.2 Packaging starts with design, then device and packaging, and ends up with a system
like a smartphone.
Applications
Software
Algorithms
Boards
Architectures
Circuits
Package Components
Package Substrates
Devices
Structures
Materials
Figure 1.3 System includes many technologies, starting with materials and structures to form
devices at device level; package substrates and passive components to form functional circuits
and architectures; system components on boards; and software at the system level.
Interconnect Power
IC Interconnections IC
Package or Board
Power Planes
Power Power Signal
Protect Cool
Figure 1.4 Definition of packaging: interconnecting, powering, cooling, and protecting all system
components.
Package or Board
Material Sciences Chemical Sciences
• Dielectrics • Lithographic Processes
• Conductors Power Planes • Microstructure vs.
• Magnetics Power Power Signal Properties
• Encapsulants
• C, L, R, Materials
Bioelectronics
• Biocompatible
• Components
• Interconnections
device and device integration technologies as well as with the changes in end-product
systems, such as a smartphone. The end goal of packaging is a final heterogeneous sys-
tem such as a smartphone with all the system components interconnected. A better defi-
nition of packaging, therefore, is interconnecting, powering, cooling, and protecting all
the system components that make up that entire system. Figure 1.4 defines packaging
this way, and Figure 1.5 shows the interdisciplinarity of packaging, requiring electrical,
mechanical, thermal, materials, chemical, and bioengineering disciplines.
Packaging @ Device
Device 1 Device 2 Level - 2µm BEOL
RDL Packaging
Package Package @ Substrate Level
Substrate 1 Substrate 2
RDL Packaging
@ Board Level
System Board
Removal of Heat
Encapsulation
IC1 IC2
Signals IC
L C B QFP L C B
PWB, Flex
IC Package
Since an electronic system involves more than one IC, effective communication
between one transistor on one IC and another transistor on another IC, all the way
through the system-level board with the required signal quality, is required. Signal
communication, however, does not start until an appropriate power is supplied to each
and every transistor. Such power distribution, however, poses a whole set of challenges
that include voltage drop as a result of long and high-resistive wiring from the power
supply to the transistor through all the levels of packaging. Simultaneous switching off
millions of transistors poses yet another challenge, in that the current drawn from the
power supply results in what is referred to as delta-I noise. Signal distribution poses a
different set of problems such as “cross-talk” between lines, as well as distortion, reflec-
tion, and alternation of signals. Electromagnetic radiation, as a result of all this radiated
energy, is yet another electrical challenge.
1.2.3 Nomenclature
3D Bare ICs that are stacked one on top of the other and interconnected on one substrate,
in contrast to packaged dies that are stacked in POP.
Embedding Embedding typically means embedded active die inside a substrate in contrast
to assembled die on top of the substrate.
Interconnect Interconnect is the length of wiring or joints between two active or passive
components.
?
1.E+10
1.E+09
1.E+08
Number of Transistors
1.E+07
1.E+06
~
2x/2 yrs.
1.E+05
1.E+04
1.E+03
1970 1980 1990 2000 2010 2020
Year of Introduction
Figure 1.8 Moore’s Law for doubling of transistors about every two years.
Node N-1
Node N
Node N+1
Cost
1
2
3
Past Future
Figure 1.9 Moore’s Law for cost reduction from node to node.
fingernail) can contain ~30 miles of interconnect “wires” in stacked levels? These wires
function like highways or pipelines to transport electrons, connect transistors and other
components to each other, and make them functional. And just as the speed we can
drive our sport cars depends (at least in part) on how clogged the freeway is, chip per-
formance depends on the ability to move signals and power through these ultra-tiny
wires. In fact, as the shrinking of feature dimensions, called scaling, has continued,
interconnects are now becoming the speed bottleneck in today’s most advanced chips.
During the first portion of chip-making in the front-end-of-line, the individual
components such as transistors or capacitors are fabricated on the wafer. In the back-
end-of-line, these components are connected to each other to distribute signals, as well
as power and ground. There is simply no room on the chip surface to create all those
connections in a single layer, so chip manufacturers build vertical levels of intercon-
nects. While simpler ICs may have just a few metal layers, complex ICs can have 10 or
more layers of wiring, as shown in Figure 1.10.
Interconnects close to the transistors need to be small, as they attach to the com-
ponents that are themselves very small and often closely packed together, as shown
in Figure 1.10. These lower-level wiring lines—called local interconnects—are usually
thin and short in length. Global interconnects are higher up in the structure; they travel
between different blocks of the circuit and are thus typically thick, long, and widely
separated with lower resistance. Vertical connections between interconnect levels,
called vias, allow signals and power to be transmitted from one layer to the next.
Metal Line
Via Global
Interconnects
Local
Interconnects
Transistors
Figure 1.10 On-chip interconnections: local and global. (Courtesy of Larry Zhao.)
Aluminum Interconnect
Aluminum
Copper Interconnect
Dielectric
much more complex, and a whole new manufacturing scheme has to be developed
for this new technology inflection, as shown in Figure 1.11. The copper interconnect
process starts with deposition of an insulating dielectric material, for example silicon
dioxide, followed by the creation of trenches. The trenches are then filled with copper
using chemical/electroplating technologies, and the excess is removed to form flat
surfaces for subsequent processing.
Core (Copper)
Line Width
Copper
0 10 20 30 40 50 60 Dielectric
Line Width (nm)
Figure 1.12 Resistance of on-chip copper wiring as a function of line width.
The electrical resistance (R) of a material describes how difficult it is to move elec-
trical current through a particular cross-section of that material, which is a function
of the orientation and proximity of the material’s atoms. Capacitance (C) refers to a
material’s ability to store electrical charge. The product of resistance and capacitance (RC)
needs to be low to create fast chips, since device speed is inversely proportional to RC
(lower RC = faster devices).
Looking at the “R” side of this challenge, higher-resistance lines carry less current,
which slows device speed. This is because higher resistance reduces electron flow, so it
takes longer to build up the minimum charge (number of electrons) or “threshold voltage”
at a transistor’s “gate” to turn it on. While transistor speed continues to improve with
scaling and by reducing the distance electrons must travel, the challenge for inter-
connect scaling is to not become a bottleneck and lose that performance improvement
by slowing the flow of electrons between transistors.
On the “C” side, capacitance is a function of the insulating dielectric material
around the metal lines and the distance between them. Higher capacitance slows elec-
trons and can create unwanted “cross-talk,” where the signal in one metal line influ-
ences the signal in a neighboring line and causes the device to malfunction. In addition
to maintaining a suitable distance between lines, the development of “low-k” dielectric
materials (capacitance is a function of a material’s “k value”) has significantly lowered
capacitance. Today’s dielectric materials average around k = 2.5, compared with k = 3.8
for pure silicon dioxide. Various methods exist to achieve lower values; however, the
resulting ultra-low-k films become increasingly fragile as the k value decreases, posing
additional challenges in manufacturing.
Core Metal
Width Simple Shrink Leads to Lower Line Resistance
Line Resistance Issues Options
Not Enough Room Alternate No
for Core Metal Barrier Barrier
Copper
Liner
Barrier
bulk copper metal. A key area of research focus is identifying strategies for improving
the barrier/liner/seed layers to lower the overall resistance and to enable smaller lines
by creating “space” for the bulk copper fill.
One strategy is to make the high-resistance barrier and liner thinner. However,
opportunity for further thinning of these layers is limited. Both need to be continuous
with no gaps or voids in the film in order to achieve good reliability performance. This
requires a minimum thickness of about 1.5 to 2 nm for each layer, which leads to a com-
bined thickness of 3–4 nm on both sides of the trench structures.
A potential alternative being studied is a new type of “self-forming” barrier that
reacts with and forms on the dielectric surface adjacent to the copper line, which allows
more room for copper. Also, new liners made of cobalt and ruthenium are being devel-
oped to replace tantalum. They adhere better to the copper seed, enabling it to be more
conformal, eliminating voids, and thinner. Already, new technologies are in place to
achieve void-free copper fill of small trenches. Around the 5-nm technology node, how-
ever, copper as the primary conducting metal will ultimately have to be replaced with a
conducting material that does not require a barrier for these ultra-thin lines.
While much attention is placed on the metals, there are also many challenges requir-
ing investigations into improvements on the dielectric side. Here, the Holy Grail is to
reduce the dielectric constant as much as possible, the ultimate being k = 1, which is
air. Indeed, novel developments using “air gaps” have been developed, but both fab-
rication and production cost challenges are significant. Thus, many of the ideas being
explored for interconnect scaling involve the development of new metals, designs, and
processes.
Resistors Integration IC
of Transistors
Circuit
Board
Figure 1.14 The invention of the first integrated circuit.
Figure 1.15 illustrates the famous Moore’s Law. In 1965, three years before he
co-founded Intel with Bob Noyce, Gordon Moore published an article in Electronics
magazine that turned out to be uncannily prophetic. Moore wrote that the number of
circuits on a silicon chip would keep doubling every year. He later revised this to every
18–24 months, a forecast that has held up remarkably well over several decades and
countless product cycles. He also described cost reduction as the second important
parameter in Moore’s Law.
108
1 billion
transistors
Number of Transistors
106
80386 80486 Pentium III
80286
105
Pentium II
4004
8086 Pentium Pro
104
Pentium
8080 Projected
103
‘70 ‘75 ‘80 ‘85 ‘90 ‘95 ‘00 ‘05 ‘10 ‘15
Year
Figure 1.15 Moore’s Law predicts the IC integration to double every 18–24 months.
The secret behind Moore’s Law is that every 18–24 months or so, chipmakers double
the number of transistors that can be crammed onto a silicon wafer the size of a finger-
nail, as shown in Figure 1.15. They do this by etching microscopic grooves onto crystal-
line silicon with beams of ultraviolet radiation. A typical wire in a Pentium chip by Intel
was 1/500 the width of a human hair; the insulating layer is only 25 atoms thick. They
are now scaled by another factor of 5X in 2018.
But the laws of physics suggest that this doubling cannot be sustained forever.
Eventually, transistors will become so tiny that their silicon components will
approach the size of molecules. At these incredibly tiny distances, the bizarre rules of
quantum mechanics take over, permitting electrons to jump from one place to another
without passing through the space in-between. Like water from a leaky fire hose,
electrons will spurt across atom-size wires and insulators, causing fatal short circuits.
Transistor components are fast approaching the dreaded point-one limit—when
the width of transistor components reaches 0.1 micron and their insulating layers
are only a few atoms thick. Recently, Intel engineer Paul Pakan publicly sounded
the alarm in Science magazine, warning that Moore’s Law could collapse. He wrote,
“There are currently no known solutions to these problems.”
The key word is “known.” The search for a successor to silicon has become a kind
of crusade; it is the Holy Grail of computation. Among physicists, the race to create
another Silicon Valley for the next century has already begun.
The economic destiny and prosperity of entire nations may rest on one question:
can silicon-based computer technology sustain Moore’s Law beyond 2020? Moore’s
Law is the engine pulling a trillion-dollar industry. It’s the reason kids assume that
it’s their birthright to get a video-game system each Christmas that’s almost twice as
powerful as the one they got last Christmas. It’s the reason you can receive (and later
throw away) a musical birthday card that contains more processing power than the
combined computers of the Allied Forces in World War II.
But microsystems are more than microelectronics. The microelectronics based on
the transistor building block technology is one aspect of today’s electronic systems,
such as personal computers. But other systems, such as modern fiberoptic telecom-
munications, are based on photons, the fundamental properties of which are more
superior in some respects, providing the needed higher bandwidth for today’s Internet
traffic. We call this the photonic wave. There are other systems that are not transistors
based that are beginning to play a major role in modern electronics. These are RF and
MEMS waves.
Frequency (Hz)
104 105 106 107 108 109 1010 1011 1012
FM radio and TV
AM radio
Wireless cable
radio communications worldwide. This earned Marconi the Nobel Prize for physics.
The fundamental technology behind today’s mobile phone is the same. A whole new
industry has emerged with applications that span from AM and FM radio to cellular
phone to satellite to microwave communications, as illustrated in Figure 1.16, across the
entire electromagnetic spectrum.
The main advantage of wireless is the fact that it cuts the cables, thus liberating
the user from the tether to the network. It allows communications anywhere anytime.
Of course, that is only realistic if the wireless equipment is small enough that it can
actually be carried around everywhere. This is where systems packaging applies.
Wireless technology is also increasingly used for non-communications functions.
Top-of-the-line Mercedes cars, for example, are now equipped with a collision avoidance
system that is based on radar. Navigational global positioning systems (GPS) are being
integrated into even more consumer items, where size and cost are paramount. For all of
these products, a small, low-cost RF module is required. Another RF/wireless applica-
tion is electronic toll taking on bridges and roads.
5G: The next big wireless communication is 5G, which started in 2018.
we should become quite comfortable with the power of the exponents and what it
means at the service levels. Fortunately, we have long ways to go to reach the data
transport limit of fiber. For example, with the current state of device technology,
a good laser source can emit 1016 photons/s, and a good detector—which can detect
a bit with 10 photons—can detect 1 Pb (1015 b/s) on a single fiber. Nevertheless, the
device technology is going to get better with time, further pushing the limit of optical
fiber capacity. Thus, fiber optics is a future-proof technology. With wavelength-division
multiplexing (WDM), it is now possible to transmit different colors of light over the
same fiber, which has provided another dimension to increasing bandwidth capacity
and channeling raw data capacity into smaller chunks of bandwidth. This advance is
akin to a self-expanding highway, where you open another channel when the traffic load
increases without laying a new fiber. Thus, WDM optical networks offer, among other
capabilities, flexibility, scalability, and capacity. Current systems are capable of deliv-
ering more than one Gb/s channel on a 100-channel system. By 2030, a 100-channel
capability, each at 10 Gb/s as illustrated in Figure 1.17, optical interconnections can be
expected to provide lot more than terabit capacities. In addition to bandwidth increase,
new technologies also enable more functionality in optics. If history is any indicator,
these capabilities will only get better and richer in features. Optical networks are already
being deployed, not only in the backbone of networks, but also in regional, metropolitan,
and access networks. Thus, optics will play a key role in next-generation network modes
and eventually at customers’ premises.
1000 1000
Total
Throughput
Number of Channels
Data Rate (Gbit/s)
100 100
Number of
Channels
10 10
Data Rate
Per Channel
1 1
To my parents: For your lifelong dedication to me as your only child, for your love and support, and for teaching me the
meaning of family and the value of education.
To my family (Anne, Dinesh, Vijay, and Suneel and grandchildren Gracen, Mason, and Cooper): For being my love, pride, and
joy to go to work for every day.
To my teachers: For teaching me how to learn well and quickly and for insisting on my getting the best education at any cost.
To my students: For allowing me to learn with you and to teach you those principles I believe in, in preparing you and
challenging you to be the best you can be.
To IBM: For giving me endless opportunities to learn, contribute, grow, and become an IBM Fellow, my ultimate goal.
To Georgia Tech: For allowing me to pursue my dream of being part of the new generation of industry-centric academicians.
To PRC team: For realizing my dream of educating interdisciplinary individual students, for exploring the SOP vision, and for
educating the world about it through books and classes.
To NSF: For allowing me to pursue my dream of impacting the industry with the needed strategic technologies and educated
workforce.
To my academic colleagues: For accepting me into your world, for using this book, and for working to make packaging an
academic subject.
Finally, thank you to the book's coordinator, Karen May, Georgia Institute of Technology; the chapters editor, Prof. Raj
Pulugurtha, Florida International University (FIU); and, for graphics, Reed Crouch, Karen May, and Leonard Mendoza.
© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
B. About the Editor
Prof. Rao R. Tummala is a Distinguished and Endowed Chair Professor at the Georgia Institute of Technology, USA. He is well
known as an industrial technologist, technology pioneer, and educator.
Prior to joining Georgia Tech, he was an IBM Fellow and the Director of IBM's Advanced Systems Packaging Technology Lab.
He pioneered major technologies such as plasma display in the 1970s; and the first three generations of 100-chip MCM
package integration, based on the first LTCC, HTCC, and thin-film RDL, introducing the original MCM concept behind today's
2.5D packages for servers, mainframes, and supercomputers.
As an educator, Prof. Tummala was instrumental in setting up at Georgia Tech the largest academic center funded by NSF,
the Microsystems Packaging Research Center, of which he is Director. He pioneered a vision for an integrated approach to
research, education, and industry collaborations with companies in the USA, Europe, Japan, Korea, Taiwan, India, and China.
The Center has produced more than 1200 Ph.D. and M.S. packaging engineers, supplying the entire electronics industry.
Prof. Tummala has published 850 technical papers and invented technologies that resulted in over 110 patents. He is the
author of the first and best-selling microelectronics packaging reference book, Microelectronics Packaging Handbook, the
definitive work in the field; the first undergraduate textbook, Fundamentals of Microsystems Packaging; and the book that
introduced the SOP concept, Introduction to System-on-Package. Prof. Tummala has received more than 50 industry,
academic, and professional society awards. He is a Member of NAE, an IEEE Fellow, and past president of IEEE EPS and
IMAPS.
© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
C. Contributors
Mulugeta Abtew Sanmina Corporation, USA (Chap. 16)
© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
Nithin Nedumthakady Georgia Institute of Technology, USA (Chap. 10)
Markondeya Raj Pulugurtha Florida International University, USA (Chaps. 5, 6, 7, 11, 18, 19, 21)
Rao R. Tummala Georgia Institute of Technology, USA (Chaps. 1, 9, 10, 11, 15, 16, 17, 18, 19, 20, 21, 22)
Pengli Zhu Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences (Chap. 14)
© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
D. Preface
Electronic packaging means primarily interconnecting active and passive components to form modules initially and systems
finally. While powering, cooling, and protecting are important functions of any package, they are secondary. Design, materials,
and processes as well as test are assumed to be part of interconnections. The end goal of packaging is to form systems by
interconnecting all the miniaturized components with and without transistors. These systems can be consumer products like
smartphones, self-driving electric cars, cloud computers, PCs, laptops, notebooks, and watches. All these systems currently
have two parts: devices with transistors and components without transistors. Combined, these two span 15 different
technologies, as described in this book.
This book is about both device and systems packaging. The market for these technologies is in excess of $500 billion
worldwide, employing millions of people. Yet, device and systems packaging is not considered an academic subject, except at
Georgia Tech. A typical academic subject such as microelectronics or VLSI has extensive courses, curricula, educational
tracks, fundamental books, software tools, and hands-on facilities. Most of these don't exist in packaging. The number of
universities offering research programs with one or more courses, however, has increased rapidly in recent years. In the United
States, there were only two universities offering any packaging research or education in the 1980s: Cornell and Stanford. It is
estimated that this number in the United States, Europe, and the Far East is about 200 now, having grown from less than 10
universities in 1990 and 50 or so in 2000.
Fundamentals of Device and Systems Packaging is written to meet the academic need to educate undergraduate and graduate
students in the fundamentals as well as all 15 technologies necessary to form systems. While there have been many books
since my first modern book in 1988, Microelectronics Packaging Handbook, most, if not all, are considered reference books.
They typically deal with one or more topics in a review fashion. However, electronic systems, as described in this book, require
15 different device and systems packaging technologies. The topics span IC packaging and assembly, system board
packaging and assembly, and all the variations in between. It also deals with electrical design, thermo-mechanical design,
thermal design and technologies, materials, processes, properties, reliability, and so on. This was the case 10–20 years ago
when I wrote the first book. Today's systems are more than microelectronics. They are RF, optical, MEMS and sensors, power,
etc.
This book is more than interconnecting. It includes microelectronics, photonics, RF and millimeter wave, and MEMS and
sensors. All these devices along with other components need to be packaged to form systems. This is exactly what this book
is all about. It deals with all these technologies at the fundamental level—defining each, illustrating the key concepts behind
each, and introducing the critical nomenclature in a systematic fashion. Each topical chapter includes fundamental equations,
homework problems, and future trends.
The book introduces Moore's Law for packaging as Moore's Law for cramming transistors to form large ICs begins to come to
an end. It proposes the cramming of small ICs to form large homogeneous and heterogeneous packages. In contrast to
Moore's Law for large ICs, with the highest number of transistors compromising performance and cost, from node to node,
Moore's Law for packaging can be thought of as interconnecting the smallest transistors in smaller ICs with the highest
performance and lowest cost in 2D, 2.5D, and 3D package architectures with an equal or higher aggregate number of
transistors.
I am grateful to all of the 55 global authors from industry and academia. This book, just like the first edition, is written by so
many authors and yet reads as if written by one. I am thankful to Karen May, who headed the book project from start to finish,
and to Prof. Raj Pulugurtha, who moved from Georgia Tech to Florida International University during the course of preparing
this book. They have put in endless hours, coordinating with all of the 55 authors from around the world.
I am also grateful to Senior Editors Michael McCabe and Lara Zoble and the other editors at McGraw-Hill, who facilitated
printing the book in two colors and at a very affordable low cost for such a book that goes from wafers to systems with all the
© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
electronic technologies and applications. I thank my wife, Anne, for her patience, understanding, and full support during the
course of the book, as I worked every day and every weekend over 3 years, on top of my normal responsibilities of being a
Professor and the Director of the Microsystems Packaging Research Center at Georgia Tech.
My final thanks are, again, to all the authors who contributed to the book, working nights and weekends, adding to their
normal job responsibilities.
© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
1. PART 1: Fundamentals of Packaging
CHAPTER 2 Fundamentals of Electrical Design for Signals, Power, and Electromagnetic Interference
© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.