Indium Phosphide Heterojunction Bipolar Transistor Technology for Future Telecommunications Applications

Kevin W. Kobayashi, Aaron K. Oki and Dwight C. Streit
TRW Inc., Space & Electronics Group
Redondo Beach, CA

Over the past decade there has been a significant shift in the microwave and millimeter-wave industry from space and defense applications to the rapidly expanding commercial telecommunications market. GaAs-based MESFET, pseudomorphic high electron mobility transistor (PHEMT) and heterojunction bipolar transistor (HBT) technologies have found numerous application in the wireless cellular handset, millimeter-wave local multipoint distribution services (LMDS) and fiber-optic telecommunications areas. In particular, GaAs HBTs have established a niche in cellular power amplifier MMIC applications as well as fiber-optic data communication IC chip set applications.

In the future, indium phosphide (InP) HBTs have a greater potential to leverage next-generation telecommunications applications such as the third-generation wideband CDMA (W-CDMA) cellular telephone systems, microwave and millimeter-wave MMIC power amplifiers for satellite and LMDS communications, and 40 Gbps IC chip sets for synchronous optical network (SONET) OC-768 fiber-optic data communications. This contention stems from the technology's favorable performance characteristics over the commercially established GaAs HBT technology. This article describes how InP HBT technology fits into next-generation telecommunications applications and discusses the business paradigm that can enable it to achieve the same commercial success as GaAs HBT technology.

The Technology/Business Model

During the 1980s, the development of GaAs-based ICs for microwave and millimeter-wave applications in the US was primarily led by major aerospace and defense electronic companies through the Defense Advanced Research Projects Agency (DARPA) and Tri-service Millimeter-wave and Microwave IC (MIMIC) and Microwave and Analog Front End Technology (MAFET) programs. The primary motivation for these companies was to provide high leverage advanced technology products for major systems applications such as space, defense and avionics. During the past 10 years, the use of these technologies for commercial telecommunication products has exploded providing a significant market for large aerospace companies with commercially attractive III-V semiconductor technologies. In addition, the growth of these lucrative commercial markets has spurred the startup of several smaller GaAs foundries.

In the short term, these small companies may be very profitable (and effectively make incremental technology improvements), but, in the long term, it is risky for them to commit to research and development (R&D) of next-generation process technologies like InP HBTs. In the multihundred-billion-dollar silicon industry, major companies are able to spend billions of dollars on R&D for semiconductor technology. However, since the GaAs market is still rather modest (with most companies experiencing sales on the order of $100 M), it is difficult to allocate significant resources for next-generation process technologies.

Advanced technology development for aerospace companies has been supported by DARPA and the Tri-services although the budget for RF electronics is eroding. Advanced III-V semiconductor technologies have the potential to grow into a multibillion-dollar industry if the right investments and commitments are made. In order to realize this growth, high tech aerospace and commercial telecommunications companies must work synergistically despite their dissimilar business paradigms. However, this idea of working together will remain only an idea if no commitment is involved. The commitment can be obtained by partnerships between individual high tech aerospace companies and commercial telecommunications companies in which both sides work within a common business paradigm offering mutual benefits.

Figure 1 shows a model of the aerospace/commercial telecommunications business paradigm. In this business paradigm, the aerospace company remains a major performer on various government space and defense contracts using its high leveraging technologies and is supported by government-matched R&D funding for advanced semiconductor technology development. At the same time, DARPA and government Tri-service agencies provide technology contracts, which subsidize the continued development of next-generation technologies such as InP HBTs and HEMTs. To capitalize on these evolving technologies, commercial partnerships are established to exploit the technologies for the emerging commercial telecommunications markets. These partnerships generate business as well as fabrication line volume that help to significantly reduce the costs of these advanced technologies for government space and defense programs as well as for the intended commercial markets. The commercial partners benefit from the privileged access to the latest technologies. This benefit is significant in that it provides the commercial partners with a product differentiator for the rapidly evolving commercial markets without being required to make a multimillion-dollar investment in a risky technology. Clearly the government, commercial and aerospace companies are winners in this synergistic business paradigm.

GaAs Heterojunction Technologies

From the late 1980s to the present, several major aerospace companies have developed advanced GaAs-based heterostructure technologies, including HEMT and HBT technology. The heterojunctions provide performance advantages over homojunction implementations of FET and BJT transistors. In the case of FETs, HEMT technology has the advantages of higher frequency performance (fT , fmax ) and lower noise figure than those achievable by MESFETs of similar gate length. HEMT is the technology of choice for low noise amplifiers, millimeter-wave MMIC components for LMDS and millimeter-wave automotive radar applications. On the other hand, GaAs HBT technology offers the high transconductance, high power density and power-added efficiency, and excellent threshold matching characteristics of a bipolar transistor while providing frequency and breakdown characteristics superior to advanced silicon and silicon-germanium BJT and HBT technologies.

Because of the timely development of GaAs HBTs for space and defense applications (just when the wireless industry was exploding), they have made probably the biggest impact on the commercial wireless industry of all the GaAs technologies. The combination of their timely development and inherent device properties and manufacturability has made GaAs HBTs the technology of choice for current cellular handset power amplifier applications. In fact, GaAs HBT power amplifiers account for as much as 30 percent of the GaAs power amplifier IC market, and this number is growing rapidly.1 For example, Figure 2 shows the GaAs HBT power amplifier IC production rate for a major GaAs device manufacturing consortium. Since 1996, the production rate has increased tremendously to meet the present market demand. At the end of 1998, the monthly production had risen to greater than eight million power amplifiers; by the end of 2000, the monthly production rate is expected to increase to 40 to 50 million power amplifiers.

The Fundamentals of InP HBTs

Over the past decade, major aerospace companies have been developing InP-based heterostructure devices such as HBT and HEMT because of their low DC power and millimeter-wave frequency performance, which is attractive for satellite communications. InP-based HBTs and HEMTs have also been developed by commercial companies for applications in high speed fiber-optic telecommunications. The InP-based device technologies take advantage of the intrinsic material properties such as higher electron mobility, higher thermal conductivity and lower energy band-gap for low voltage operation in order to gain an advantage over GaAs-, silicon- and silicon-germanium-based semiconductors.

Figure 3 shows the electron velocity vs. electric field of various semiconductor materials and illustrates the advantage of using compound semiconductor device technology, which provides high electron velocity. It can be seen that both GaAs and InGaAs materials have peak velocities that occur under lower electric fields making them useful for applications requiring high performance under low voltages. InP material has a peak velocity under a slightly higher electric field. Because of the growth compatibility of InGaAs and InP, an optimum HBT device can be constructed by incorporating an InGaAs base and an InP collector to optimize the electron transport characteristics in the low field and high field regions of the HBT device. InP material also enables higher performance in HEMTs by allowing higher indium composition in the InGaAs HEMT's conduction channel. Silicon, on the other hand, does not possess the high electron mobility of the compound semiconductors. As a result, InP-based HEMT and HBT technologies are aggressively being developed for next-generation millimeter-wave MMIC and high speed analog IC applications in the space and defense industry. Because of the true analog bipolar characteristics of HBTs, the high speed InP-based HBTs can provide a wide breadth of circuit functions and applications in digital, analog, RF and microwave design domains at frequencies just beyond the reach of GaAs HBTs.

InP HBTs come in two basic flavors: with a single base-emitter heterojunction (SHBT) and with both base-emitter and collector-base heterojunctions (DHBT). Both of these devices are represented by their energy band diagrams shown in Figure 4 . In brief, the single HBT refers to a bipolar transistor with an emitter that has a wider energy band-gap than the base and is typically fabricated using precise growth techniques such as metal-organic chemical vapor deposition or molecular beam epitaxy. The properties of the heterojunction allow near-independent optimization of the device forward current gain b and fmax in comparison to conventional homojunction BJTs, which must compromise one device performance parameter for the other. The fundamentals of HBTs are covered in the literature.2,3

In order to improve the practical breakdown characteristics and optimize the electron transport in the high field collector region of the HBT device, a wide energy band-gap InP collector also can be employed to form a base-collector heterojunction resulting in a DHBT. The DHBT simultaneously enables high breakdown voltage and speed, which are attractive properties for future telecommunications applications. The device breakdown voltage BVceo vs. cutoff frequency fT for various bipolar technologies is shown in Figure 5 and further illustrates this advantage. For a given bipolar technology, higher device speed can easily be achieved at the expense of the device's practical breakdown voltage. However, only through the exploitation of heterojunction technology does a true improvement in the breakdown voltage-fT product result. From these data, it can be seen that III-V compound semiconductor HBTs (GaAs and InP based) have an advantage in improving this application with InP DHBT exhibiting the ultimate performance due to its optimum placement of low and high field velocity materials in the device.

In the current and next-generation telecommunications applications, the critical semiconductor IC applications will be those that require a high breakdown voltage-fT product as is evident by the need for high voltage optical modulator drivers in high data rate fiber telecommunications and the high power millimeter-wave amplifiers needed for wideband digital transmission systems such as LMDS. It is the combination of high power, efficiency and speed that gives InP-based HBTs an advantage in these next-generation applications.

The Current State of InP HBT Technology

For several years, the aerospace industry has focused on the development of InP HBT technology toward microwave and millimeter-wave communications systems while the commercial sector has focused on using InP HBTs for fiber telecommunications. On the other hand, universities have primarily focused on fundamental semiconductor device development. During this period, aerospace and commercial industries as well as academia have developed and produced several III-V HBT fabrication technologies that have demonstrated phenomenal frequency performance. Table 1 lists some of the key HBT technologies that have been developed over recent years. Each of these HBT technologies represents the various device engineering techniques that have been developed for fabricating high fT and fmax HBTs. State-of-the-art cutoff frequencies ranging from 90 to 228 GHz have been achieved as well as fmax performance ranging from 140 GHz to beyond 500 GHz. In general, it has been determined that high fT is required for high speed analog and digital ICs while high fmax is required for millimeter-wave MMICs, including power amplifiers. InP HBTs have achieved a record fmax greater than 500 GHz using a transferred substrate technology for fabricating small geometry collector-base regions.10 By incorporating a combination of the listed device fabrication techniques, fmax performance above 1 THz will soon be possible.

Table I
Key GaAs and InP HBT Technologies

Technology

Technique

Impact

fT (GHz)

fmax (GHz)

AlGaAs/GaAs HBT4

ballistic collector with launcher

reduces base transit time; increases fT

100 to 171

148 to 192

AlGaAs/InGaAs HBT5,6

extrinsic base regrowth

reduces base resistance; increases fmax

96

240 to 280

InP/InGaAs SHBT7

submicron emitters

reduces device geometry and parasitics; increases speed

165

140

InP/InGaAs DHBT8

step graded InGaAsP quartenary collector

reduces current blocking due to bandgap discontinuity; increases current

144

267

InP/InGaAs DHBT9

hexagonal emitter

reduces device geometry and parasitics; increases speed

228

227

InP/InGaAs SHBT10

transferred substrate

reduces collector-base capacitance; increases fmax

157

> 500

Of the HBT technologies mentioned previously, InP-based HBTs have demonstrated the most impressive MMICs in the millimeter-wave regime. Table 2 lists state-of-the-art InP HBT microwave and millimeter-wave MMIC performance capability. For example, some noteworthy MMIC demonstrations have been a 108 GHz InP HBT VCO that benchmarks the lowest phase noise obtained from a fully monolithic W-band VCO;11 a DC to 85 GHz direct-coupled amplifier, which represents the widest bandwidth achieved from an analog bipolar topology;8 and a 44 GHz highly linear amplifier that has obtained the highest third-order intercept point (IP3) per DC power consumption ratio (42:1) for a MMIC amplifier in any technology.14,15 The high frequency and linearity demonstrated by these MMICs indicate the potential that InP HBTs have for millimeter-wave digital communications such as LMDS and other emerging systems.

Table II
InP HBT Microwave and mm-wave MMIC Performance Capability

Technology

MMIC Demonstration

Benchmark Signifigance

InP SHBT11

108 GHz VCO - wide turning range = 2.7 GHz

lowest phase noise fully monolithic W-band VCO

InP SHBTSchottky diodes12

94 GHz InP HBT Schottky diode mixer and frequency multiplier

6 dB lower LO driver requirement than GaAs-based Schottky diode frequency convertors

InP SHBTwith transferred substrate8

DC to 85 GHz direct-coupled amplifier

widest bandwidth achieved from an analog bipolar amplifier

InP SHBTwith transferred substrate13

80 GHz distributed amplifier

widest bandwidth achieved from an distributed bipolar amplifier

InP SHBT14, 15

44 GHz high IP3 (linearity) amplifier

best IP3 (linearity) per DC power linearity figure of merit for a mm-wave MMIC amplifier

InP SHBT16

14 GHz direct BPSK modulator with >3 Gsps

high data rate, direct BPSK modulation onto microwave carrier

InP SHBT17

DC to 20 GHz InP HBT balanced analog mixer

widest gain-bandwidth product for an active analog mixer

InP HBTs have also made some breakthroughs in analog and optoelectronic ICs (OEIC), showing capability for 20 Gbps operation and greater. Although there are obviously many commercial companies developing these next-generation ICs, a sample of the industry's capability is listed in Table 3 . Noteworthy demonstrations include a 47 GHz photoreceiver, which integrates a monolithic PIN photodetector comprising the intrinsic doped InGaAs HBT collector. This OEIC is capable of receiving 60 Gbps transmission. In addition, several 40 Gbps multiplexers (MUX) and demultiplexers (DEMUX) have been demonstrated as well as a 68 GHz static frequency divider,19 which signifies the fastest clock rate reported to date for a bipolar technology. With a conservative production InP HBT process technology, 24 GHz monolithic photoreceivers can be manufactured23 using conventional analog circuit design approaches.

Table III
InP HBT OEIC Performance Capability

Technology

MMIC
(Demonstration/Capability)

Benchmark Signifigance

InP SHBT18

100 kHz to 47 GHz monolithic photoreciever; integrated PIN detector; HBTdistributed transimpedence amplifier

highest operating frequency for a monolithic photoreciever MMIC

InP DHBT19

68 GHz static frequency divider

highest operating static frequency divider

InP SHBT20,21

40 Gbps 2:1 MUX
40 Gbps 1:2 DEMUX

SONET OC-768 capability

InP SHBT22

20 Gbps laser driver

SONET OC-768 capability

InP SHBT23

monolithic photorecievers to 24 GHz

typical InP HBT commercial photoreciever using conventional analog techniques

Power amplifiers operating from cellular frequencies to millimeter-wave frequencies just above 30 GHz (31 GHz) have also been a major target application for InP-based HBT technology. While SHBTs have demonstrated good power-added efficiencies of 66 to 76 percent with sub-3 V operation at cellular and X-band frequencies,24,25 it is the DHBTs that have been identified as the preferred technology for MMIC power amplifiers in the 18 to 31 GHz K- and Ka-band frequency range. Table 4 lists some InP HBT device power results that have been obtained in recent years.24-28

Table IV
InP HBT Power Capability

Technology

Power Demonstration

Benchmark Signifigance

InP SHBT24

f0 = 1.9 GHz
PAE = 66%
Vce = 1.9 GHz

high cellular-band efficiency under low voltage operation

InP SHBT25

f0 = 10 GHz
PAE = 76%
Pout = 126 mW
Vce = 1.9 GHz

high X-band efficiency under low voltage operation

InP DHBT26

f0 = 4.5 GHz
PAE = 53 to 60%
Pout = 2.05 W
power density  = 4.2 W/mm

high power density of 4.2 W/mm at C-band

InP DHBT26

f0 = 9 GHz
PAE = 59 to 60%
Pout = 1.2 W
power density  = 5 W/mm
peak power density  = 6 W/mm

high power density of 6 W/mm at X-band

InP DHBT27

f0 = 30 GHz
PAE = 35.5%
power density = 2.34 W/mm

high power density of 2.34 W/mm at Ka-band

InP SHBT28

f0 = 18 GHz
PAE = 54%
Pout = 1.17 W
power density  = 4.88 W/mm

high power density of 4.88 W/mm at K-band

One outstanding feature of DHBTs is their extremely high power densities, which are a measure of delivered output power in watts per millimeter of HBT emitter finger length. The power densities achieved from these DHBTs are typically greater than 4 W/mm and as high as 6 W/mm. The high power density translates into small device periphery and chip implementation and, consequently, low production costs. The power densities demonstrated by DHBTs are two- to three-times higher than those achieved from commercially available GaAs HBT power amplifiers and, thus, suggest a potential cost advantage for future power MMICs.

InP HBT Telecommunications Applications

Table 5 lists competing device technologies and how they compare regarding a few of the critical telecommunications product requirements of the future.

Table V
Key Future Telecom MMIC Applications

 

Linear power amplifiers for cellular telephones

Linear mm-wave power amplifiers for satellite and data communication systems

OEIC for > 20 Gbps fiber data links

 

Performance

Technology

Low Voltage

Linearity

PAE

Single Supply

Idle Curent

fT , fmax

Linearity

Breakdown Voltage

Reciever IC's

High Voltage Modulation Driver IC's

GaAs MESFETS

Excellent

Poor

Average

Poor

Poor

Poor

Poor

Average

Poor

Average

GaAs HBTs

Average

Excellent

Excellent

Excellent

Excellent

Poor

Excellent

Excellent

Excellent

Excellent

GaAs HEMTs

Excellent

Poor

Average

Poor

Poor

Excellent

Poor

Average

Average

Average

SiGe HBTs

Excellent

Average

Average

Excellent

Excellent

Average

Average

Poor

Average

Poor

InP SHBT

Excellent

Excellent

Excellent

Excellent

Excellent

Excellent

Excellent

Average

Excellent

Average

InP DHBT

Excellent

Excellent

Excellent

Excellent

Excellent

Excellent

Excellent

Excellent

Average

Excellent

InP HEMP

Excellent

Poor

Average

Poor

Poor

Excellent

Poor

Poor

Excellent

Poor

Linear Power Amplifiers for Cellular Telephones

The first key application is the low voltage power amplifier for cellular telephony. The present generation of cellular telephony has been marked by the introduction of spectrally efficient digital cellular systems. The next-generation (3G) systems are currently in the early phases of a system architecture study and will include a form of W-CDMA.29 Leadership in this generation of cellular telephony will determine which technology, companies and countries will dominate the future of the wireless industry. Worldwide volume for this product is expected to surpass 200 million power amplifiers per year in 2002. Presently, GaAs HBT technology supplies 30 percent of the industry's power amplifiers for handset applications as well as 80 percent of all CDMA-based handset applications.1 Key technical requirements for the 3G power amplifiers are low voltage operation to reduce power consumption, high linearity to minimize spectral regrowth, high efficiency for extended battery life and low cost. Practical requirements include single-supply operation for reduced part count and assembly costs and low idle current, which removes the need for cumbersome off-chip drain switches.

The data indicate that InP HBT technology is the best suited to replace GaAs HBTs as the dominant cellular power amplifier technology of the future. In fact, preliminary data suggest that InP HBT power amplifiers can achieve as much as 5 to 10 dB lower third- and fifth-order intermodulation distortion compared to GaAs HBTs under low (2 V) voltage operation, as shown in Figure 6 . This advantage is expected to increase at lower operating voltages and higher frequencies.

Regarding costs, InP HBTs may have an even greater advantage than GaAs HBTs have had over MESFETs. The key economic advantage of GaAs HBT technology over MESFETs is that it can realize extremely compact power amplifier chips due to its high power density. Even though MESFET technology is far cheaper per square millimeter of semiconductor area than HBT technology, the MESFET power amplifier MMIC is still more expensive due to the larger chip size. On the other hand, InP HBTs benefit from even greater power densities than GaAs HBTs, ultimately resulting in a lower cost implementation than GaAs HBTs even though their wafer costs are higher.

Linear Millimeter-wave Power Amplifiers for Digital Communications

Another key application for the future is the millimeter-wave MMIC power amplifier for LMDS and other millimeter-wave wireless digital communication systems. In particular, the LMDS is a wireless two-way broadband technology designed to allow network integrators and communication service providers to quickly and inexpensively bring a wide range of high value, quality services to homes and businesses. Potential applications include high speed Internet access, real-time multimedia file transfer, remote access to corporate local area networks, interactive video, video-on-demand, video conferencing and telephony. With the recent surge of interest in the Internet and the demand for bandwidth, fast data transmission appears to be the greatest application for this technology.

Past communication technologies focused on lower RFs because they can generate enough power for long distance transmission and building penetration. (Such is the case with television and radio signals.) However, LMDS uses lower power and higher frequency signals (27.5 to 31 GHz) over a shorter line-of-sight distance. Several technologies, including MMICs, digital signal processors, video compression techniques and advanced modulation systems, have sufficiently matured during the past few years and should enable an affordable LMDS. In particular, MMIC technology will have a big impact on these systems by reducing module complexity, part count and cost. MMICs are also highly manufacturable and can easily accommodate the high production requirements.

Among all the MMIC components, it is the power amplifier that is the most coveted MMIC and requires high gain at millimeter-wave frequencies, good power linearity and power efficiency, and high power density for realizing small and inexpensive chips. For frequencies up to 31 GHz, InP DHBTs appear to be the best suited because of their fundamentally high breakdown voltage-fT product and extremely high power densities. For frequencies above 31 GHz, GaAs HEMT power MMICs will probably remain the preferred technology due to their sufficient breakdown voltage and superior frequency performance.

OEIC Applications

The third key application area for InP HBT technology is in OEICs for fiber telecommunications. The advent of the erbium-doped fiber amplifier, wavelength division multiple access technology and high speed OEICs has enabled fiber-optic telecommunications to become the primary means of high speed data transmission for most of the US' data transmission, including Internet backbone and submarine and landline telephony. Presently, GaAs HBTs are well suited (technically and economically) for the SONET OC-192 10 Gbps application in the MUX/DEMUX- and clock-and-data-recovery-type ICs. The higher frequency capability of InP HBTs can extend these same data communications ICs to 40 Gbps optical systems such as the SONET IC-768 and beyond.

In addition, because of their compatibility with growing InGaAs PIN photodetectors, InP HBTs can produce miniature inexpensive monolithic photoreceivers for the popular 1.30 to 1.55 mm wavelength fiber systems. However, the requirements of the optical modulator driver for converting electrical signals to modulated light are more demanding. At frequencies above 1 Gbps, an indirect optical modulator is typically used. The modulating element is usually a lithium-niobate crystal that changes its optical absorption properties when an electric field is applied to it. In these applications, the modulator requires very large voltages to change the absorption characteristics and is on the order of 3 to 6 V. In a 50 W system, this voltage level is equivalent to 20 to 26 dBm of RF power.

Unlike most microwave amplifier applications, which are narrowband, the optical modulator essentially operates at baseband, which has an operational frequency range from a few kilohertz to tens of gigahertz. This multidecade frequency range makes the optical modulator driver IC extremely challenging. Obviously, the technology of choice requires high speed, a large breakdown voltage and good analog characteristics that can only be provided by bipolar devices. From the data it can be seen that InP DHBTs are the best suited for this application and become the preferred technology as the data rate requirements are increased. For this particular driver MMIC, the market demand is expected to be one million units per year by the end of 2000. Although the volume of this product is somewhat modest, it is a premium fiber telecommunications chip commodity analogous to the cellular MMIC power amplifier.

A Summary

In summary, InP HBT technology is preferred over other technologies for high linearity power amplifiers operating from cellular frequencies to LMDS frequencies (28 to 31 GHz). It is also the technology of choice for next-generation (> 20 Gbps) fiber telecommunications systems due to its intrinsic speed, practical breakdown voltage and good analog bipolar characteristics. Figure 7 shows the relative volumes expected for each of these key InP HBT MMIC telecommunications product applications.

For cellular power amplifiers, the volume is large and will be determined by the market environment. The price and volume for InP HBTs will depend heavily on their relative costs compared to competing technologies like GaAs HBTs. For this market the performance advantages of InP HBTs like higher linearity and lower voltage operation may be a significant factor considering the more demanding requirements of the 3G telephony systems. However, the higher power density inherent to InP DHBT technology may give it a cost advantage over existing GaAs HBTs in the future.

For commercial LMDS and other digital mm-wave applications in the 18 to 31 GHz frequency range, InP DHBT technology is expected to offer superior performance and lower costs compared to MESFET and HEMT technologies in realizing high efficiency and high linearity MMIC power amplifiers. Medium volumes of InP DHBT MMIC power amplifiers are expected for this application and will be driven by the proliferation of the LMDS and similar systems.

Based on inherent performance only, it is believed that InP HBT (SHBT and DHBT) will be the dominant technology for affordable high speed OEIC applications. The volume of this market will depend on the need for higher capacity and transmission rates, which are growing at a tremendous pace. For satellite payload electronics applications, the volume of InP technologies (in general) will always be maintained at a fairly low level, but will continue to exist due to the constant demand for better performance.

Conclusion

The nature of the present telecommunications industry is such that the next-generation microwave and millimeter-wave technology development will more than likely be led by aerospace companies. InP HBT technology will find significant commercial applications in the third-generation cellular telephony, LMDS and millimeter-wave digital radio communication and fiber-optic data link markets.

Acknowledgment

The authors wish to acknowledge the contributions of the many managers, engineers and technicians at TRW, including Augusto Gutierrez-Aitken, J. Cowles, Liem Tran, Frank Yamada, Li Yang, Eric Kaneshiro, Chris Grossman, Ken Sato, Matt Hoppe, Donald Umemoto, Tim Naeole, Scott Olson, Greg Leslie, Mike Lammert, Rich Lai, Y.C. Chen, Mike Wojtowicz, Mike Barsky, Ron Grundbacher, William Jones, Peter Chou, Jane Lee, Rosie Dia, Eve Ahlers, Y.Y. Tu and Hung Nguyen; the researchers and project managers at the DARPA, Air Force, Army and Navy research laboratories; and their partners at RF Micro Devices and Multilink Technology Corp. 

References

1.         "The Road Ahead: Things to Watch in 1999," Compound Semiconductor, Vol. 5, No. 1, January/February 1999.

2.         Herbert Kroemer, "Theory of a Wide-gap Emitter for Transistors," Proceedings of the IRE, Vol. 45, No. 11, November 1957, pp. 1535-1537.

3.         Herbert Kroemer, "Heterostructure Bipolar Transistors and Integrated Circuits," Proceedings of the IEEE, Vol. 70, January 1982, pp. 13-25.

4.         Matsuoka Yutaka and Eiichi Sano, "High Speed AlGaAs/GaAs HBTs and Their Applications to 40 Gbps-class ICs," 1994 IEEE GaAs IC Symposium Digest, Philadelphia, PA, pp. 185-188.

5.         Yasuyuki Suzuki, Hidenori Shimawaki, Yasushi Amamiya and Kiyoshi Fukuchi, "An HBT Preamplifier for 40 Gbps Optical Transmission Systems," 1996 IEEE GaAs IC Symposium Digest, Orlando, FL, pp. 203-206.

6.         Kazuhiko Honjo, Hidenori Shimawaki, Shinichi Tanaka and Yasuyuki Suzuki, "High-fmax AlGaAs/InGaAs HBTs and Their Application in Millimeter-wave Ranges," 1997 Asia-Pacific Microwave Conference Proceedings, pp. 733-736.

7.         Madjid Hafizi, "New Submicron HBT IC Technology Demonstrates Ultra-fast, Low Power Integrated Circuits," IEEE Transactions on Electron Devices, Vol. 45, No. 9, September 1998, pp. 1862-1868.

8.         S. Yamahata, K. Kurishima, H. Nakajima, T. Kobayashi and Y. Matsuoka, "Ultra-high fmax and fT InP/InGaAs Double-heterojunction Bipolar Transistors with Step-graded InGaAsP Collector," 1994 IEEE GaAs IC Symposium Digest, Philadelphia, PA, pp. 345-348.

9.         Shoji Yamahata, Kenji Kurishima, Hiroshi Ito and Yutaka Matsuoka, "Over-220-GHz-fT -and-fmax InP/InGaAs Double-heterojunction Bipolar Transistors with a New Hexagonal-shaped Emitter," 1995 IEEE GaAs IC Symposium Digest, San Diego, CA, pp. 163-166.

10.       Mark Rodwell, Q. Lee, D. Mensa, R. Pullela, J. Guthrie, S.C. Martin, R.P. Smith, S. Jaganathan, T. Mathew, B. Agarwal and S. Long, "48 GHz Digital ICs Using Transferred-substrate HBTs," 1998 IEEE GaAs IC Symposium Digest, Atlanta, GA, pp. 113-116.

11.       K.W. Kobayashi, J. Cowles, L.T. Tran, A. Gutierrez-Aitken, T. Block, F. Yamada, A.K. Oki and D.C. Streit, "A Low Phase Noise W-band InP HBT Monolithic Push-push VCO," 1998 IEEE GaAs IC Symposium Digest, Atlanta, GA, pp. 237-240.

12.       E.W. Lin, H. Wang, K.W. Chang, L. Tran, J. Cowles, T. Block, D.C.W. Lo, G.S. Dow, A. Oki, D. Streit and B.R. Allen, "Monolithic Millimeter-wave Schottky-diode-based Frequency Converters with Low Drive Requirements Using an InP HBT-compatible Process," 1995 IEEE GaAs IC Symposium Digest, San Diego, CA, pp. 218-221.

13.       B. Agarwal, R. Pullela, Q. Lee, D. Mensa, J. Guthrie and M.J.W. Rodwell, "80 GHz Distributed Amplifiers with Transferred-substrate Heterojunction Bipolar Transistors," 1998 IEEE MTT Microwave Symposium Proceedings, Baltimore, MD.

14.       K.W. Kobayashi, J. Cowles, L.T. Tran, A. Gutierrez-Aitken, M. Nishimoto, J. Elliott, A.K. Oki and D.C. Streit, "High IP3/Low DC Power 44 GHz InP-HBT Amplifier," 1998 IEEE GaAs IC Symposium Digest, Atlanta, GA, pp. 29-32.

15.       K.W. Kobayashi, J. Cowles, L.T. Tran, A. Gutierrez-Aitken, M. Nishimoto, J. Elliott, T.R. Block, A.K. Oki and D.C. Streit, "A 44 GHz/High IP3 InP HBT MMIC Amplifier for Low DC Power Millimeter-wave Receiver Applications," IEEE JSSC, Special Issue on the 1998 GaAs IC Symposium, 1999.

16.       R. Desrosiers, J. Cowles, C. Hornbuckle, A. Gutierrez-Aitken and J. Becker, "Monolithic 14 GHz Wideband InP HBT BPSK Modulator," 1998 IEEE GaAs IC Symposium Digest, Atlanta, GA, pp. 135-138.

17.       K.W. Kobayashi, A. Gutierrez-Aitken, J. Cowles, B. Tang, R. Desrosiers V. Medvedev, L.T. Tran, T.R. Block, A.K. Oki and D.C. Streit, "15 dB Gain, DC-20 GHz InP HBT Balanced Analog Mixer and Variable Gain Amplifier with 27 dB of Dynamic Range," 1999 IEEE RFIC Symposium Digest, Anaheim, CA.

18.       K.W. Kobayashi, A.K. Oki, J. Cowles, A. Gutierrez-Aitken, L.T. Tran, T.R. Block and D.C. Streit, "InP-based HBT Technology for Next Generation Light-wave Communications," Microwave Journal, Vol. 41, No. 6, June 1998, pp. 22-38.

19.       B. Tang, J. Notthoff, A. Gutierrez-Aitken, E. Kaneshiro, P. Chin and A. Oki, "InP DHBT 68 GHz Frequency Divider," to be published in IEEE GaAs IC Symposium Digest, Monterey, CA, October 17, 1999.

20.       J. Godin, P. Andre, J.L. Benchimol, P. Desrousseaux, A.M. Duchenois, A. Konczykowska, P. Launay, M. Meghelli and M. Riet, "A InP DHBT Technology for High Bit-rate Optical Communications Circuits," 1997 IEEE GaAs IC Symposium Digest, Anaheim, CA, pp. 219-222.

21.       Thomas Swahn, Thomas Lewin, Mehran Mokhtari, Hannu Tenhunen, Robert Walden and William E. Stanchina, "40 Gbps, 3 Volt InP HBT ICs for a Fiber Optic Demonstrator System," 1996 IEEE GaAs IC Symposium Digest, Orlando, FL, pp. 15-128.

22.       W.E. Stanchina, J.F. Jensen, R.H. Walden, M. Hafizi, H.C. Sun, T. Liu, G. Raghavan, K.E. Elliott, M. Kardos, A.E. Schmitz, Y.K. Brown, M.E. Montes and M. Yung, "An InP-based HBT Fab for High-speed Digital, Analog, Mixed-signal and Optoelectronic ICs," 1995 GaAs IC Symposium Digest, San Diego, CA, pp. 31-34.

23.       Dwight C. Streit, Augusto Gutierrez-Aitken, John C. Cowles, Li-Wu Yang, Kevin W. Kobayashi, Liem T. Tran, Thomas R. Block and Aaron K. Oki, "Production and Commercial Insertion of InP HBT Integrated Circuits," 1997 IEEE GaAs IC Symposium Digest, pp. 135-138.

24.       K.W. Kobayashi, A.K. Oki, L.T. Tran, J. Cowles, L. Yang, D.K. Umemoto, C. Grossman, T.R. Block and D.C. Streit, "HBT IC Technology for Wireless Applications beyond 2000," 1997 IEEE GaAs IC Symposium Short Course on IC Technologies for Wireless Application, October 12, 1997.

25.       A.K. Oki, D.C. Streit, R. Lai, K.W. Kobayashi, A. Gutierrez-Aitken and T. Block, "Future Technologies for Commercial and Defense Telecommunication Electronics," 1999 IEEE MTT-S Digest, Anaheim, CA.

26.       M. Hafizi, P.A. Macdonald, T. Liu and D.B. Rensch, "Microwave Power Performance of InP-based Double Heterojunction Bipolar Transistors for C- and X-band Applications," 1994 IEEE MTT-S Digest, pp. 671-674.

27.       Hin-Fai Chau, Hua-Quen Tserng and Edward A. Beam, III, "Ka-band Power Performance of InP/InGaAs/InP Double Heterojunction Bipolar Transistors," IEEE Microwave- and Guided-wave Letters, Vol. 6, No. 3, March 1996, pp. 129-131.

28.       Robinder S. Virk, Mary Y. Chen, Chanh Nguyen, Takyiu Liu, Mehran Matloubian and David B. Rensch, "A High-performance AlInAs/InGaAs/InP DHBT K-band Power Cell," IEEE Microwave- and Guided-wave Letters, Vol. 7, No. 10, October 1997, pp. 323-325.

29.       Gene Heftman, "Cellular ICs Move toward 3G Wireless - Gingerly," Microwaves & RF, February 1999, page 31.

Kevin W. Kobayashi received his BSEE from the University of California at San Diego and his MSEE from the University of Southern California in 1986 and 1991, respectively. In 1986, he joined TRW where he has been focused on the development of HBT MMIC design technology for military and commercial applications. In 1997, Kobayashi was named a TRW Technical Fellow for his contributions to the advancement of HBT MMIC design. He is also an associate editor of the IEEE Journal of Solid-state Circuits and a member of the GaAs IC Symposium executive program committee.

Aaron K. Oki received his BSEE from the University of Hawaii and his MSEE from the University of California, Berkeley in 1983 and 1985, respectively. Since 1985, he has been involved in the research, development and production of advanced GaAs and InP HBT technology for commercial and high reliability space applications at TRW. In 1995, Oki was named a TRW Technical Fellow for his contributions to the advancement of HBT device technology. In 1997, he became the assistant manager of TRW's microelectronic products and technology development department.

Dwight C. Streit received his PhD in electrical engineering from the University of California, Los Angeles in 1986. He is currently manager of the microelectronics product and technology development department in TRW's RF Products Center and principal investigator for several research and development projects related to III-V materials, monolithic HEMT-HBT integrated circuits and quantum effect devices. Streit is a TRW Technical Fellow of the compound heterostructure materials.