Reliability of Semiconductor Lasers and Optoelectronic Devices

By Robert Herrick

Reliability of Semiconductor Lasers and Optoelectronic Devices simplifies complex concepts of optoelectronics reliability with approachable introductory chapters and a focus on real-world applications. This book provides a brief look at the fundamentals of laser diodes, introduces reliability qualification, and then presents real-world case studies discussing the principles of reliability and what occurs when these rules are broken. Then this book comprehensively looks at optoelectronics devices and the defects that cause premature failure in them and how to control those defects. Key materials and devices are reviewed including silicon photonics, vertical-cavity surface-emitting lasers (VCSELs), InGaN LEDs and lasers, and AlGaN LEDs, covering the majority of optoelectronic devices that we use in our everyday lives, powering the Internet, telecommunication, solid-state lighting, illuminators, and many other applications. This book features contributions from experts in industry and academia working in these areas and includes numerous practical examples and case studies.This book is suitable for new entrants to the field of optoelectronics working in R&D.

• Includes case studies and numerous examples showing best practices and common mistakes affecting optoelectronics reliability written by experts working in the industry
• Features the first wide-ranging and comprehensive overview of fiber optics reliability engineering, covering all elements of the practice from building a reliability laboratory, qualifying new products, to improving reliability on mature products
• Provides a look at the reliability issues and failure mechanisms for silicon photonics, VCSELs, InGaN LEDs and lasers, AIGaN LEDs, and more

• Language: English
• Publisher: Elsevier Science
• Release date: Mar 6, 2021
• ISBN: 9780128192559

Book preview

Reliability of Semiconductor Lasers and Optoelectronic Devices

Edited by

Robert W. Herrick

Osamu Ueda

Table of Contents

Cover image

Title page

Copyright

Dedication

List of contributors

Preface

Acknowledgments

Chapter 1. Introduction to optoelectronic devices

Abstract

1.1 Introduction

1.2 Optoelectronic applications

1.3 Principles of operation for optoelectronic components

1.4 Method of fabrication

1.5 Critical metrics

1.6 Laser and light-emitting diode reliability

1.7 New technology developments

1.8 Summary

References

Chapter 2. Reliability engineering in optoelectronic devices and fiber optic transceivers

Abstract

2.1 Reliability engineering organizations and management

2.2 Developing a product: design for reliability

2.3 Developing a test plan: standards-based testing versus customized testing

2.4 Test engineering and data collection

2.5 Data collection and analysis

2.6 Failure analysis

2.7 Physics of failure and common failure mechanisms

2.8 Product release, ongoing monitoring, and postrelease support

2.9 Conclusion

References

Chapter 3. Case studies in fiber optic reliability

Abstract

3.1 Introduction

3.2 Group 1: issues that were caught before product release

3.3 Group 2: cases that had reliability issues from the start, but were not detected in the reliability qualification

3.4 Group 3: parts that were reliable after release, but later developed reliability issues during high-volume manufacturing

3.5 Conclusion

References

Chapter 4. Materials science of defects in GaAs-based semiconductor lasers

Abstract

4.1 Introduction

4.2 Characteristics of simple point defects in GaAs

4.3 Dislocations in GaAs

4.4 Epitaxial integration of GaAs-based materials on silicon

4.5 Recombination-enhanced processes

4.6 Outlook

References

Chapter 5. Grown-in defects and thermal instability affecting the reliability of lasers: III–Vs versus III-nitrides

Abstract

5.1 Introduction

5.2 Grown-in defects in III–Vs semiconductors for optoelectronics

5.3 Influence of grown-in defects on device reliability and degradation in III-V based optoelectronics

5.4 Grown-in defects in III-nitrides

5.5 Composition-modulated structures and ordered structures in III-V based optoelectronics

5.6 Thermal instability in III-nitrides

5.7 Conclusion

References

Chapter 6. Reliability of lasers on silicon substrates for silicon photonics

Abstract

6.1 Introduction

6.2 Early-stage prospective technologies

6.3 Heterogeneous lasers on silicon

6.4 Heteroepitaxial lasers grown on silicon substrates

6.5 Reliability results for quantum dot lasers grown on silicon substrates

6.6 Degradation mechanisms in quantum dot lasers on silicon

6.7 The path forward and future improvements for quantum dot lasers grown on silicon substrates

6.8 Conclusion

Acknowledgments

References

Chapter 7. Degradation mechanisms of InGaN visible LEDs and AlGaN UV LEDs

Abstract

7.1 Introduction

7.2 Degradation of InGaN visible LEDs

7.3 Degradation of AlGaN UV LEDs

7.4 Conclusions

References

Index

Copyright

Woodhead Publishing is an imprint of Elsevier

The Officers’ Mess Business Centre, Royston Road, Duxford, CB22 4QH, United Kingdom

50 Hampshire Street, 5th Floor, Cambridge, MA 02139, United States

The Boulevard, Langford Lane, Kidlington, OX5 1GB, United Kingdom

Copyright © 2021 Elsevier Ltd. All rights reserved.

No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Details on how to seek permission, further information about the Publisher’s permissions policies and our arrangements with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency, can be found at our website: www.elsevier.com/permissions.

This book and the individual contributions contained in it are protected under copyright by the Publisher (other than as may be noted herein).

Notices

Knowledge and best practice in this field are constantly changing. As new research and experience broaden our understanding, changes in research methods, professional practices, or medical treatment may become necessary.

Practitioners and researchers must always rely on their own experience and knowledge in evaluating and using any information, methods, compounds, or experiments described herein. In using such information or methods they should be mindful of their own safety and the safety of others, including parties for whom they have a professional responsibility.

To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume any liability for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions, or ideas contained in the material herein.

British Library Cataloguing-in-Publication Data

A catalogue record for this book is available from the British Library

Library of Congress Cataloging-in-Publication Data

A catalog record for this book is available from the Library of Congress

ISBN: 978-0-12-819254-2 (print)

ISBN: 978-0-12-819255-9 (online)

For information on all Woodhead Publishing publications visit our website at https://www.elsevier.com/books-and-journals

Publisher: Matthew Deans

Acquisitions Editor: Kayla Dos Santos

Editorial Project Manager: Isabella C. Silva

Production Project Manager: Surya Narayanan Jayachandrans

Cover Designer: Miles Hitchen

Typeset by MPS Limited, Chennai, India

Dedication

This book is dedicated to the late Dr. Robert G. Waters, my first and most important reliability mentor. Without him, I never would have gone down this very rewarding career path.

Robert W. Herrick

List of contributors

John E. Bowers

Quintessent, Inc., Santa Barbara, CA, United States

Institute for Energy Efficiency, University of California, Santa Barbara, CA, United States

A. Caria,     Department of Information Engineering, University of Padova, Padova, Italy

C. De Santi,     Department of Information Engineering, University of Padova, Padova, Italy

Qiang Guo,     Intel Corporation, Rio Rancho, NM, United States

Robert W. Herrick,     Intel Corporation, Santa Clara, CA, United States

Daehwan Jung,     Korea Institute of Science and Technology, Seoul, South Korea

Alan Y. Liu,     Quintessent, Inc., Santa Barbara, CA, United States

G. Meneghesso,     Department of Information Engineering, University of Padova, Padova, Italy

M. Meneghini,     Department of Information Engineering, University of Padova, Padova, Italy

Kunal Mukherjee,     Department of Materials Science and Engineering, Stanford University, Stanford, CA, United States

Justin C. Norman,     Quintessent, Inc., Santa Barbara, CA, United States

F. Piva,     Department of Information Engineering, University of Padova, Padova, Italy

Jennifer Selvidge,     Institute for Energy Efficiency, University of California, Santa Barbara, CA, United States

Shigetaka Tomiya

Sony Corporation, Atsugi, Japan

Tokyo Institute of Technology, Yokohama, Japan

Osamu Ueda,     Meiji University, Tama-ku, Kawasaki, Japan

E. Zanoni,     Department of Information Engineering, University of Padova, Padova, Italy

Preface

Robert W. Herrick

Semiconductor laser reliability is a subject with great depth and complexity, but one where most practitioners must become comfortable making decisions based on incomplete knowledge. Unlike many elements of optoelectronic design where modeling can predict performance with high accuracy, and measurements can be made with many significant digits of precision, semiconductor laser reliability is governed by flaws that are ill-understood, very hard to see, and often hard to measure. What is known about the subject is usually shrouded in secrecy, under the cloak of company proprietary information, so little of what is known gets published or shared with the larger technical community. Furthermore, there are only a few reliability physics graduate programs, and a few other reliability statistics programs worldwide, so there are almost no reliability engineering graduates to be hired. Over 99% of the practitioners of the field worldwide are taught through an apprentice system, often making expensive mistakes in the process, and failing to use industry best practices even after they have been on the job for several years. We hope this book can help to articulate many of the best practices and demystify the discipline of reliability engineering for those entering the profession. Hopefully it will also explain many of the reasons why the reliability standards and reliability qualifications are structured the way they are today, so that practitioners are not just executing the qualifications by rote, but understand the purpose of each test.

In this book, we have brought together several of the best-known teams in the world to cover many key areas of the subject. The book is organized into two main parts. The first two chapters are intended for new entrants into the field. While there are a small number of books addressing the subject of semiconductor laser reliability, all of them are advanced in their treatment, and require many years of practice in the field before they can be understood. The large majority of those who approach me for advice on the subject are not capable of reading and understanding the existing volumes. In this book, we have addressed that gap in the market with two entry-level chapters. The first is a simple introduction to optoelectronics with a focus on the major applications that are likely to dominate market growth in the next decade. This is written in a way that does not require understanding of solid-state physics, or energy band theory, and is intended to be accessible to anyone with a good high school science background. The second chapter is an introduction to the principles of reliability engineering. Other specialized books cover selected elements of reliability engineering, such as reliability calculations, burn-in optimization, or accelerated aging. However, this chapter is the first of its kind to cover all elements of the practice, from hiring and organization, to building a reliability test capability, to supporting released products. It is extensively cross-referenced, to where the reader is directed to the best of the existing other volumes for assistance on advanced topics. The third chapter is also unique: a set of 10 case studies in reliability that make it far easier to understand the principles of reliability engineering, and how they apply in practice. By looking at cases from companies (primarily those that are no longer in the market sector), it gives readers a peek behind the curtain to see how others have tried to solve hard reliability problems, and in some cases, where they could have done better. This third chapter should be both accessible and of interest to all readers of the book, both beginners and experts.

The second half of the book is intended for advanced readers. Chapter 4, Materials science of defects in GaAs-based semiconductor lasers, covers the material science of dislocations in optoelectronics, which is a very difficult subject to understand, and is not well-covered in material science references. Chapter 5, Grown-in defects and thermal instability affecting the reliability of lasers: III–Vs vs. III-nitrides, is a comprehensive chapter on defects in III-V’s and III-N’s by two of the most famous authors in the field, and provides experimental detail on the most important aging mechanisms from a material science characterization viewpoint. Chapter 6, Reliability of lasers on silicon substrates for silicon photonics, deals with the very cutting-edge topic of reliability of lasers on silicon, including the issue of heteroepitaxy (where III-V compounds such as GaAs or InP are grown on silicon substrates). And finally, Chapter 7, Degradation mechanisms of InGaN visible LEDs and AlGaN UV LEDs, deals with the physics of degradation of visible and UV LEDs, by a group with unique access to cutting-edge methods of studying the physics of degradation, and decades of experience doing so.

We believe this new volume will be highly instructive to readers at a wide range of expertise levels. But hopefully this book will be just a starting point, and readers will dive into the references, and learn more about the subjects that interest them most. Feel free to reach out to Professor Ueda or myself if you have questions or feedback.

Acknowledgments

Robert W. Herrick

I often reflect on how grateful I am for the many mentors who helped me in my professional development, and want to take a few minutes to thank several of them for what they taught me, before discussing acknowledgments that are specific to this book. My first reliability mentor is the most important on this list: the late Dr. Robert G. Waters. Bob was a co-worker of mine at McDonnell Douglas from 1987 to 1992, working at the Opto-Electronic Center in Elmsford, New York, and doing ground-breaking work in the physics of failure of semiconductor lasers for space-based laser communication systems. He never hesitated to share what he knew, or to take a call. Even with the limited tools available back in the late 1980s, he was able to develop ingenious experiments to understand the root causes of sudden and gradual failure in laser diodes, culminating in elimination of sudden failures by the development of compressively strained quantum well lasers. Discovering the secrets of semiconductor laser failure is a remarkably difficult challenge, given the atomic scale of the defects, the paucity of published literature, and the patience required to wait for experimental results, but he was able to open the black box and tease out many of the secrets using a multidisciplinary approach, and shared his excitement for the breakthroughs with me. He not only published some of the best papers ever written on the subject, but chaired many of the conference sessions where the breakthroughs were being first announced. I feel fortunate to have such an excellent role model to have followed, and for the very satisfying career that it led me to. I would never have followed a career path in reliability without the inspiration he provided, which is why I have dedicated this book to his memory.

To mention the most important other mentors (in chronological order), Joe Levy was my manager at my first job in McDonnell Douglas, and taught me much of what I know about semiconductor lasers. I owe him a debt of gratitude for encouraging me to take risks, to go beyond my comfort zone, and get into device simulation and design, wafer processing, characterization, mask design, and other fields that were outside both my job responsibilities and my knowledge base. When I decided to go back to school to get a PhD, Professor Larry Coldren brought me in to UCSB, funded my PhD work, and made me part of his VCSEL group for the first couple of years of graduate studies, and I will always be grateful to him for those opportunities. Professor Pierre Petroff taught me most of what I know about Materials Science, and was an excellent research advisor for the second half of my PhD. Brent Wahl hired me in to Hewlett Packard, and taught me much of what I know about transceiver reliability (including much of what is in Chapter 2, Reliability engineering in optoelectronic devices and fiber optic transceivers). Kevin Weitsman was our Department Manager, and one of the most capable and amazing people I have ever worked with. I also owe him a debt of gratitude for what he taught me about the business side of managing a quality department. Saurin Shah has been a manager or co-worker of mine for most of the past 15 years, and he deserves the biggest share of the credit for building the quality department for Intel SPPD out of almost nothing in his first year on the job. Saurin’s manager, the late Pinder Mathru, had the vision for what our group could become, and helped put me in front of many important connections throughout Intel, and played an important part in our current success—I just wish he were around to see how his vision has been fulfilled. Finally, I would like to thank my many excellent co-workers and managers along the way.

I benefitted from many others I enlisted for help in providing feedback and editing assistance on the chapters I wrote. The most helpful of these was from my co-worker Felipe Vallini; his honest feedback was very beneficial, and helped me to re-organize two of the chapters. I also want to thank my co-author Qiang Guo for reading and commenting on all of the chapters in this book, as well as his help with the first three chapters. He also managed the figure permissions for several of the chapters, and found and shared many great background articles. Andra Hioki also provided a very careful reading of Chapter 1, Introduction to optoelectronic devices, and I appreciate her copy editing help. I also enlisted family and friends without engineering degrees to read the first chapter to assess its suitability for newcomers to the field—thank you to Andy Herrick, John Herrick, and John South for your careful reading and suggestions.

At Elsevier, I would like to thank Kayla Dos Santos, who originally approached Osamu and I in 2018 and invited us to put together a book proposal on the subject of laser reliability, and helped shepherd the proposal through Elsevier management. I also would like to thank Isabella Silva for program managing the book and its production. And I would also like to thank my co-editor Osamu Ueda for also providing inspiration early in my career with his excellent published work, and for his collaboration in the past decade where I have had the chance to work together with him on many conferences and publications. Finally, I would like to thank all of the authors of this book for the work they have put into their chapters, and sharing what they know with the readers of this book. I hope this work can help to make the process of reliability engineering a less mysterious and more approachable discipline.

Chapter 1

Introduction to optoelectronic devices

Robert W. Herrick¹ and Qiang Guo²,    ¹Intel Corporation, Santa Clara, CA, United States,    ²Intel Corporation, Rio Rancho, NM, United States

Abstract

As the optoelectronics market continues to grow at 20%–40% per year in many segments, many new practitioners join the field, without the benefit of the traditional graduate school training programs in laser and light-emitting diode physics, fabrication, or operation. Our goal in this chapter is to provide an introduction to optoelectronics for those transferring into this field. This introduction is intended to share the basic principles of the design and fabrication of lasers and optoelectronics for those who are not involved in core design activities, and also to serve as an introduction for those who want to pursue more rigorous studies.

Keywords

Tutorial; overview; fiber optic transceiver; datacom; telecom; semiconductor laser; diode laser; LED; light-emitting diode; fabrication

1.1 Introduction

The laser was invented in 1960, and in the years since then it has played a rapidly growing part of our lives, carrying data across the Internet, powering laser printers, barcode scanners in supermarkets, and driving imaging and proximity sensors such as computer mice. The light-emitting diode (LED) has gone past being a simple indicator lamp in its early life, to powering most colored lighting and traffic signals, and is gradually taking over the majority of the lighting market, as well as most backlighting for TVs and computer displays. Optoelectronics play an integral role in many of our daily activities, in more ways than most people appreciate. The purpose of this chapter is to give an introduction to how lasers and LEDs work, how they are fabricated, and a few of their key applications.

In the past decade (the 2010s), three large markets developed in optoelectronics, each of which resulted in billions of new semiconductor optoelectronic devices. The first was the large-scale replacement of conventional incandescent and fluorescent lighting with solid-state lighting, based on LEDs. The second was the development of the mega-scale cloud data centers, which are powered by fiber optic transceivers, where transceivers and other fiber optics now account for more than a third of the total investment. The third, and most recent, is the explosive growth of sensors for facial or gesture recognition, such as the FaceID sensor first used in the Apple iPhone X. Looking forward to the 2020s, laser-based range mapping (LIDAR, short for light detection and ranging) for autonomous driving and other advanced driver-assistance systems will add a fourth market that is expected to also ramp up quickly into the billions of devices created and sold.

All of these will create hundreds of thousands of new jobs for those designing, fabricating, testing, and assembling those systems [1–5]. While we expect that those designing the devices will continue to be trained through the same time-consuming and thorough process in engineering schools around the world, many more in the fabrication, test, assembly, market development, and management roles will transfer in from successful careers in other fields, without the benefit of a graduate-school education in optoelectronics. It is that need for retraining and continuous education that this chapter seeks to address, with an introduction that has a lower barrier to entry than previous tutorials. We will attempt to emphasize ease-of-understanding over rigorous or in-depth treatments, and introduce a minimum of theory and no equations. For many, we hope this will be enough to better understand decisions being made in engineering meetings. For others, it can help provide a larger context before reading other chapters in this book, or in other books or papers on optoelectronic design and fabrication. Some of the most popular treatments elsewhere include Coldren [6], Verdeyen [7], Kasap [8], and Culshaw [9], and these are recommended for those looking for greater depth, with the time and inclination to dive into the equations and physics.

We assume that the reader has a background in college-level physics, and basic engineering concepts. Unlike other treatments of the subject, we do not assume previous training in semiconductor devices or solid-state physics. We cover a few basics of laser design and operation, with a particular emphasis on elements important to the manufacturing and test process and (in keeping with this book) with device reliability. This chapter does not require sequential reading, so we encourage readers to skim the chapter, and pick and choose just the sections that look relevant to their industry, and to their job.

1.2 Optoelectronic applications

In the following section, we introduce three optoelectronic applications that have achieved over $1B per year in revenue: InGaN LEDs for high-efficiency lighting, lasers for sensing arrays, and lasers for data communication. On top of the fiber-optic network backbone that was built up in previous decades, which will be discussed in Section 1.3, these represent the three most important new optoelectronic markets of the past decade.

1.2.1 InGaN-based light-emitting diodes for high-efficiency lighting

The most efficient type of general lighting now available is based on blue InGaN LEDs with phosphors to down-convert most of the photons to other lower-energy wavelengths to provide broad-spectrum white light. In many jurisdictions, traditional incandescent lighting has now been banned as being too inefficient and wasteful of energy, and for contributing to climate change. Fluorescent lighting, while far more efficient than incandescent, is less efficient than LED lighting, and also has a shorter lifetime, and environmental concerns related to the disposal of the mercury used in fluorescent bulbs. In Fig. 1.1, we show images of LED-powered bulbs [10,11].

Figure 1.1 (Left) An image of the construction of a high-brightness white LED package; (right) a common low-cost LED lamp. Source: (left) Copyright 2005 Society of Photo-Optical Instrumentation Engineers (SPIE) & Lumileds LLC from M.G. Craford, LEDs for solid state lighting and other emerging applications: status, trends, and challenges, in: Proc. of the SPIE, vol. 5941, Sep. 2005, p. 594101, https://doi.org/10.1117/12.625918; (right) [2018] IEEE. Reprinted, with permission, from X. Perpiñà et al., Thermal management strategies for low- and high-voltage retrofit LED lamp drivers, IEEE Trans. Power Electron. 34 (4) (2019) 3677–3688, https://doi.org/10.1109/TPEL.2018.2853119.

LED lighting really took off starting in the late 1990s, with the invention of efficient InGaN LEDs. Professor Shuji Nakamura received the 2014 Nobel Prize in Physics for the development of the method of growth used to produce these devices. It took another couple of decades for it to ramp up to ever more powerful LED chips, and to scale up production volumes and cut costs. At the time of this writing, an 800-lumen bulb (the most popular type for indoor lighting) is expected to last 20 years, with an annual energy cost only $1.20, leading to $6.08 annual energy savings when compared to the incandescent bulbs it replaced [12]. Based on a US Department of Energy forecast, LED lighting is projected to make up over 99% of the total residential lighting market by 2030 in the United States [13].

1.2.2 Lasers for sensing arrays

In 2017, Apple introduced Face ID to its phones, which was powered by a vertical-cavity surface-emitting laser (VCSEL) array with 219 VCSEL emitters [14]. The system is capable of creating a 3-D map of the face of each user, which is used as a security biometric system for sign-on to the phone. Other gesture recognition systems use a similar engine. Production of these sensors reached 197 million units in 2018 [15]. With the addition of these smartphone applications to the previous base, the total VCSEL market reached US$0.738B in 2018 [4,16]. As with many consumer products, it is not uncommon for each dollar of spending on lasers in the supply chain to enable 100 dollars of value/consumer spending in the final product (e.g., on DVD or CD lasers in disc players). Images of the dot projector that powers the FaceID are shown in Fig. 1.2 [17].

Figure 1.2 Image of (A) face recognition module and (B) vertical-cavity surface-emitting laser-based dot projector arrays. Source: Reprinted from A. Liu, P. Wolf, J.A. Lott, and D. Bimberg, Vertical-cavity surface-emitting lasers for data communication and sensing, Photonics Res. 7 (2) (2019) 121–136, https://doi.org/10.1364/PRJ.7.000121, copyright 2019. With permission from Chinese Laser Press (CLP).

1.2.3 Lasers for data- and telecommunications

Nearly everyone is aware that low-cost telecommunication is powered by fiber optic communications. However, fewer people are aware of the fact that around 45% of the cost of modern cloud data centers is the cost of optics, including both the fiber, and the components we will discuss in this section. It is useful to understand when communications are made by electronics over high-speed cables, and when it is more advantageous to make them over optical fiber. A simple summary would be to say that where it is possible to do communications through electrical signals alone, that is the normal industry practice. However, there are real limitations, in both speed and link length, to high-speed electronic communications, as shown in Fig. 1.3. Above the combination of speeds and distances shown (where the product of distance in meters and speed exceeds approximately 100 m-Gbps) [18], the signal degradation is severe, the power required is prohibitive, or both. In those cases, it is worth the additional expense of adding an electrical/optical and optical/electrical converter to the link. Photons do not suffer from the problems high-speed electrical links do with capacitance or electromagnetic radiation—they are able to pass through fiber optic cables with much lower degradation per meter traveled, and require less energy for longer links.

Figure 1.3 Graph of data rate vs. link distance. Copper links are the most cost effective for low speeds and short links (lower left), but fiber optics are used for longer or higher speed links. After Reference [19].

To take advantage of fiber optic communication, a pair of core components are needed: one to convert high-speed electronic data into high-speed optical data (E-O), as well as a companion converter of optical signals back to electrical signals (O-E). The former E-O converter is known as a transmitter or Tx, while the latter, O-E, is known as a receiver or Rx. These components were first deployed in the Bell Telephone network back in 1977. In modern application, the Tx and Rx are combined into a single unit called a transceiver with a tightly defined set of mechanical, electrical, optical, and software specifications that are part of a multisource agreement or MSA document that assures interchangeability. The most popular types of transceivers are made by dozens of companies, and have seen the same sort of rigorous price competition seen in other electronics commodity industries, such as hard drives or RAM memory modules. The net outcome has been international phone calls that used to cost dollars per minute are now too cheap to meter, and free streaming video is available to homes and cell phones.

The transceiver, which is approximately the size of a finger, normally contains a few optical components, as well as a number of integrated circuits (ICs). A typical layout is shown in Fig. 1.4 [20]. Of special interest is the optical subassembly [or transmitter optical sub assembly (TOSA) on the Tx side]. The receiver optical components tend to be less challenging, although the ICs that are used in the receiver are at least as challenging as those on the Tx side. The TOSA in turn usually provides three functions. First, it must provide a source to generate photons. Second, it monitors the average light out, to allow adjustment to a predetermined set point in response to degradation or changes in the thermal environment (although some very simple designs may dispense with this and operate an open loop). Third, it needs a way to turn the stream of light on and off rapidly—either using a modulator, or directly by powering the laser up and down.

Figure 1.4 Photo of a fiber optic transceiver, including key internal components on both the top and bottom of the PC board (upper part of the figure). The laser is embedded in the subassembly labeled TOSA. In the graph at the bottom, the transceivers are sorted by bandwidth density, with the older types on the left, and newer ones on the right. The dotted line shows the substantial reductions that have been obtained in terms of reduced power dissipation per bit, and the solid line shows how bandwidth density has increased. Source: Reprinted from J.S. Eng and C. Kocot, Optical Fiber Telecommunications VIA: Chapter 14. VCSEL-Based Data Links, Elsevier Inc., 2013, copyright (2013), with permission from Elsevier.

Infrared wavelengths (especially 1310 and 1550 nm) are typically used in optical communication due to the available laser sources and low transmission loss of optical fiber. Fig. 1.5 shows that this region can be further subdivided into multiple wavelength bands [21]. The O-band, at 1310 nm, gained popularity because it is at the dispersion minimum, and thus works best for many multimode fiber applications. The O-band is also widely used within data centers and for shorter distance communication. The C-band at 1550 nm is widely used for long-haul optical communication because of a mature erbium-doped fiber amplifier (EDFA) technology and the lowest transmission loss in this band. Near the C-band, the L-band and the S-band (so-named because they are slightly longer or shorter in wavelength than the original center band) also show sufficiently low absorption loss to support an increased number of channels. Thus, wavelength-division multiplexing (WDM) can be applied across these multiple bands to boost transmission capacity. WDM is a technology which can fully utilize the bandwidth of optical fiber by transmitting optical signals using different wavelengths at the same time. It can be divided into coarse wavelength division multiplexing (CWDM) and dense wavelength division multiplexing (DWDM). The main difference between CWDM and DWDM is channel spacing. CWDM has wider spacing between channels (e.g., 20 nm), whereas DWDM has very close channel spacing below 1 nm. Although DWDM allows more data per fiber, it has higher cost and power consumption.

Figure 1.5 Communication bands used for medium- and long-distance optical communication. Source: Reprinted from K. Grobe, Chapter 5—Optical wavelength-division multiplexing for data communication networks, in: C. DeCusatis (Ed.), Handbook of Fiber Optic Data Communication (4th Ed.), Academic Press: Oxford, 2013, pp. 85–122, copyright (2013), with permission from Elsevier.

1.2.4 The history of the laser

The idea of the laser dates back to the 1930s, and it is widely believed that all the technology elements were available at that time. Many of the most important inventions and technologies are shown in the timeline in Fig. 1.6. However, the theory was misunderstood at the time, and it was thought that the necessary population inversion (explained later) required a negative temperature, and was thus impossible to obtain. It was not until successful demonstration of microwave amplification of stimulated emission of radiation in the late 1950s that it was thought that optical stimulated emission might be possible, including by Townes and Schalow [22,23]. Theodore Maiman at Hughes Research Labs demonstrated the first laser in 1960. The first semiconductor lasers came a couple of years later in 1962, and the first visible LED was also demonstrated by Nick Holonyak at General Electric in 1962, as shown in the timeline in Fig. 1.6. The first large-volume laser applications were material processing using carbon dioxide (CO2) lasers beginning in 1967, and helium-neon (HeNe) lasers for barcode scanners in retail scanning systems in the 1980s. Telecommunication lasers for fiber optics started being produced in small volumes for ultra-long-haul submarine communication in 1988 [24,25], but they cost a few thousand dollars each at the time.

Figure 1.6 Timeline showing many of the most important technologies and applications for the LED (top) and for the laser (bottom). The details of many of the innovations are discussed later in this chapter. DWDM, dense wavelength division multiplexing; EDFA, erbium-doped fiber amplifier; LED, light-emitting diode; MOCVD, metal organic chemical vapor deposition; OMVPE, organo-metallic vapor phase epitaxy; VCSEL, vertical-cavity surface-emitting laser.

The first low-cost laser was the gallium arsenide (GaAs)-based 780 nm CD laser, which exceeded 10 million units produced in 1987 [26], and is pictured in Fig. 1.7 [27]. The cost per unit fell to just a few dollars each by 1990. Of course, CD lasers and the DVD lasers that superseded them have been largely replaced by cloud delivery of video content (also made possible by optoelectronics), as the storage and data communication to stream the files on demand have become so cheap on a cost-per-bit basis that it has become less expensive than the previous procedure of printing and shipping discs to transfer the information.

Figure 1.7 CD laser assembly in a transistor outline can (TO can) package. Source: After Newport Corporation, Laser Diode Technology. https://www.newport.com/t/laser-diode-technology, 2019 (accessed 26.10.19), permission to use granted by Newport Corporation. All rights reserved.

The next major chapter in the development of the semiconductor laser was the development of the vertical cavity laser, which enabled even lower cost, lower power consumption through a very small pumped area, and improved beam quality. VCSELs were first suggested by Professor Kenichi Iga, and low-temperature pulsed operation was demonstrated in the laboratory [28]. The first practical VCSEL was demonstrated by Jack Jewell of Bell Labs in 1991, followed shortly by Professor Larry Coldren’s group at UCSB the same year. By the late 1990s, companies such as Hewlett Packard had introduced the first low-cost VCSEL-based gigabit fiber optic transceivers (for $300 each, while previously gigabit per second transceivers typically cost $5000 each), which enabled data center and storage server communications, and were instrumental in the dot-com explosion of the late 1990s.

Another important development was the high-power 980 nm laser for pumping of EDFAs, pictured in Fig. 1.8, which were the main focus of the telecom bubble of 1998–2001 [25,29]. At the time, telecommunication was moving to DWDM, where 40–80 different wavelength channels could be sent down the same fiber, with different information encoded on each one. However, the information attenuated sufficiently within 80–100 km which previously had needed to go through splitting, detection, and retransmission, which required large capital investment of very expensive components. However, EDFA-based systems which simply restored the signal amplitude periodically through optical amplification were found to be capable of transmitting the information once, and crossing the entire distance (across the Atlantic Ocean, for example) with no repeaters. This created huge reductions in the cost per bit that were essential for the development of the Internet and unmetered phone communication [24,25,30]. Finally, in the past decade, we have seen widespread replacement of incandescent and fluorescent lighting with solid-state LED lighting, and also the rise of large-scale cloud computing powering social media and remote meetings. We will have more to say about these technologies in applications in the sections to come. For additional laser history information, refer to Hecht’s excellent books and papers [22,31–33].

Figure 1.8 Schematic of a fiber link with erbium-doped fiber amplifiers to restore the signal periodically, which was critical in the development of too cheap to meter phone calls and data transmission for Internet service.

1.3 Principles of operation for optoelectronic components

In this section, we discuss first LEDs, then lasers, and finally photodiodes which are used in receivers. Lasers and LEDs have many common design features,