Kutz M. Handbook of materials selection

Подождите немного. Документ загружается.

1300 DIAMOND FILMS

3. Electroplate diamond anchors onto the substrate surface (Fig. 7). This

allows the adhesion of the diamond ﬁlm to a substrate that has no chem-

ical compatibility with the diamond such as steel or INCO 718.

4. Braze a free diamond ﬁlm onto a prepared substrate. The brazing material

is usually a low-temperature alloy that can be a copper–silver or other

easily ﬂowing alloy that contains powder of a carbide former such as

titanium or zirconium. The carbide former makes a good chemical bond

with the diamond surface and in turn bonds to the braze material also.

Figure 8 places the three critical aspects of diamond wear ﬁlms into a context

of their individual importance with respect to one another as well as how the

threes aspects interact with the bearing system in general.

BIBLIOGRAPHY

Angus, J. C., ‘‘Empirical Categorization and Naming of ‘‘Diamond-Like’’ Carbon Films,’’ Thin Solid

Films, 142, 145–151 (1986).

Angus, J. C., H. A. Will, and W. S. Stanko, ‘‘Growth of Diamond Seed Crystals by Vapor Deposition,’’

J. Appl. Phys., 39, 2915–2922 (1968).

Bachmann, P. K., and H. Lydtin, ‘‘High Rate versus Low Rate Diamond CVD Methods,’’ NATO ASI

Series B: Phys., 266, 829–853 (1991).

Banholzer, W. F., and T. R. Anthony, ‘‘Diamond Properties as a Function of Isotopic Composition,’’

Thin Solid Films, 212, 1–10 (1992).

Bauer, M., Precious Stones, Dover Publications, New York (1968).

Bernholc, J., A. Antonelli, T. M. Del Sole, Y. Bar-Yam, and S. T. Pantelides, ‘‘Mechanism of Self-

Diffusion in Diamond,’’ Phys. Rev. Lett., 61, 2689–2692 (1988).

Bernholc, J., S. A. Kajihara, and A. Antonelli, ‘‘Theory of Doping of Diamond,’’ Mater. Res. Soc.

Symp. Proc., 242, 323–333 (1992).

Bridgman, P. W., ‘‘Change of State Solid-Liquid at High Pressures,’’ Phys. Rev., 11, 401 (1913).

Bridgman, P. W., ‘‘An Experimental Contribution to the Problem of Diamond Synthesis,’’ J. Chem.

Phys., 15, 92–98 (1947).

Bundy, F. P., H. T. Hall, H. M. Strong, and R. H. Wentorf, Jr., ‘‘Man-Made Diamonds,’’ Nature, 176,

51–55 (1955).

Davies, G., Diamond, Adam Hilger, Bristol, UK, 1984.

Davies, G. (ed.), Properties and Growth of Diamond, INSPEC, The Institution of Electrical Engineers,

London, 1994.

Derjaguin, B. V., D. V. Fedoseev, V. M. Lukyanovich, B. V. Spitzin, V. A. Ryabov, and A. V.

Lavrentyev, ‘‘Filamentary Diamond Crystals,’’ J. Crystal Growth, 2, 380–384 (1968).

Derjaguin, B. V., D. V. Fedoseev, and K. S. Uspenskaya, ‘‘The Synthesis of Diamond. 1. Kinetic and

Diffusional Conditions for the Synthesis of Diamond from Gaseous Methane,’’ Russian J. Phys.

Chem., 47, 13–20 (1973).

Eversole, W. G., Synthesis of Diamond, U.S. Patent, 3,030,187, 1962a.

Eversole, W. G., Synthesis of Diamond, U.S. Patent, 3,030,188 1962b.

Field, J. E., ‘‘Natural Diamond: The Standard,’’ in Diamond and Diamond-Like Films and Coatings,

R. E. Clausing (ed.), Plenum, New York, 1991, pp. 17–35.

Hershey, J. W., ‘‘Synthetic Diamonds,’’ Trans. Kansas Acad. Sci., 32, 52–54 (1929).

Hershey, J. W., ‘‘Recent Work on Synthetic Diamonds,’’ Trans. Kansas Acad. Sci., 40, 109–111

(1937).

Hershey, J. W., and J. Hamersley, ‘‘Attempt to Prepare Synthetic Diamonds from a Large Iron

Nugget,’’ Trans. Kansas Acad. Sci., 43, 213 (1940).

Knight, D. S., and W. B. White, ‘‘Characterization of Diamond Films by Raman Spectroscopy,’’ J.

Mater. Res., 4, 385–393 (1989).

Liander, H., ‘‘Diamond Synthesis—The True Story,’’ Ind. Diamond Rev., 40, 412–415 (1980).

Loparev, V. V., V. I. Veprintsev, A. V. Manukhin, and E. V. Funtikov, ‘‘Effect of Added Boron on

the Oxidation Resistance of CB Diamonds,’’ Sverkhtverdye Materialy, 6, 11–14 (1984).

BIBLIOGRAPHY 1301

Obata, T., and S. Morimoto, ‘‘Free-Standing Diamond Films—Plates, Tubes and Curved Dia-

phragms,’’ Proc. SPIE, 1146, 208–216 (1989).

Phelps, A. W., The Vibrational Spectra of Diamond Polytypes and Defects in Diamond and Cubic

Boron Nitride, Ph.D. Thesis, Pennsylvania State University, University Park, PA, 1990.

Phelps, A. W., Durable Coatings Synthesis for Cryogenic Engine Technology, Edwards Air Force

Base, CA, Phillips Laboratory PL-TR-93-3084, 1993.

Phelps, A. W., and R. Koba, ‘‘P-type Diamond Stripper Foils for Tandem Ion Accelerators,’’ Elec-

trochem. Soc. Proc., 89-12, 38–49 (1989).

Phelps, A. W., A. C. Gonzalez, and T. W. Owens, Diamond Deposition, Edwards Air Force Base,

CA, Phillips Laboratory PL-TR-91-3052, 1992.

Plano, L. S., and F. Adar, ‘‘Raman Spectroscopy of Polycrystalline Diamond Films, Proc. SPIE, 822,

52–56 (1987).

Ramdas, A. K., S. Rodriguez, M. N. Grimsditch, T. R., Anthony, and W. F. Banholzer, ‘‘Effect of

Isotopic Composition of Diamond on Its Elastic Constants:

C Diamond, the Hardest Known

Material,’’ Phys. Rev. Lett., 71, 189–192 (1993).

Ravi, K. V., and M. I. Landstrass, ‘‘Silicon on Insulator Technology Using CVD Diamond Films,’’

Electrochem. Soc. Proc., 89-12, 24–37 (1989).

Roy, R., ‘‘Diamonds at Low Pressure,’’ Nature, 325, 17–18 (1987).

Seal, M., ‘‘Diamond Heat Sinks,’’ International Industrial Diamond Conference: Chicago, Industrial

Diamond Association of America, 1970, 191–198.

Seal, M., and W. J. P. van Enckevort, ‘‘Applications of Diamond in Optics,’’ Proc. SPIE, 969, 144–

152 (1988).

Spear, K. E., ‘‘Diamond—Ceramic Coating of the Future,’’ J. Am. Ceramic Soc., 72, 171–191 (1989).

Spitsyn, B. V., ‘‘Origin and Evolution of the Science and Technology of Diamond Synthesis in the

USSR,’’ NATO ASI Series B—Phys., 266, 855–873 (1991).

Srivatsa, A. R., J. Narayan, and K. V. Ravi, 1989, ‘‘Microstructural Features of Diamond Films on

(100) Silicon Substrates,’’ Electrochem. Soc. Proc., 89-12, 215–224 (1989).

Strong, H. M., ‘‘Early Diamond Making at General Electric,’’ Am. J. Phys., 57, 794–802 (1989).

Tither, D., A. Matthews, P. John, and J. I. B. Wilson, ‘‘A Perspective on Japanese Patents on Hard

Carbon/ Diamond Films and Related Subjects 1967–1987,’’ Surface & Coatings Tech., 38, 251–

266 (1989).

Tsai, C., J. C. Nelson, J. C., Gerberich, W. W., Heberlein, and E. Pfender, ‘‘Metal Reinforced Dia-

mond Composite Films,’’ Mater. Res. Soc. Symp. Proc., 239, 275 (1992).

Tsai, C., J. C. Nelson, W. W. Gerberich, J. Heberlein, and E. Pfender, ‘‘Diamond–Metal Composite

Coatings on Cemented-Carbide Tools,’’ Diamond and Related Materials, 2, 617–620 (1993).

Wang, X. H., G. H. M. Ma, W. Zhu, J. T. Glass, L. Bergman, K. F. Turner, and R. J. Nemanich,

‘‘Effects of Boron Doping on the Surface Morphology and Structural Imperfections of Diamond

Films,’’ Diamond and Related Materials, 1, 828–835 (1992).

Watermeyer, B., Diamond Cutting, A Complete Guide to Diamond Processing, Perskor, Johannesburg,

SA, 1991.

Williams, B. E., H. S. Kong, and J. T. Glass, ‘‘Electron Microscopy of Vapor Phase Deposited

Diamond,’’ J. Mater. Res., 5, 801–810 (1990).

Windischmann, H., and G. F. Epps, ‘‘Properties of Diamond Membranes for X-ray Lithography,’’ J.

Appl. Phys., 68, 5665–5673 (1990).

Windischmann, H., G. F. Epps, and G. P. Ceasar, ‘‘Tensile Strength and Biaxial Young’s Modulus of

Diamond Films,’’ in Proceedings of the Second International Conference, New Diamond Science

and Technology, R. Messier, J. T. Glass, J. E., Butler, and R. Roy, (eds.), The Materials Research

Society, Pittsburgh, 1991, 767–772.

1303

CHAPTER 43

ADVANCED MATERIALS IN

TELECOMMUNICATIONS

Glen R. Kowach

Ainissa G. Ramirez

Agere Systems

Murray Hill, New Jersey

1 INTRODUCTION TO

COMMUNICATIONS 1303

1.1 Networks 1304

1.2 Transmission 1309

2 MATERIALS SELECTION FOR

SELECT COMPONENTS 1309

2.1 Packaging 1311

2.2 Solid-State Semiconductor

Lasers 1314

2.3 Photodetectors 1317

2.4 Optical Fiber 1318

2.5 Electro-Optical Materials 1323

2.6 Antireﬂection (AR) Coatings 1324

2.7 Optical Filters 1324

2.8 Isolators Incorporating Faraday

Rotators 1325

2.9 Microelectromechanical

Systems 1325

2.10 Microwave Radio-Frequency

(RF) Resonators 1328

2.11 Repeaters and Switches 1329

2.12 Multilayered Ceramics 1329

3 COMMUNICATION SYSTEM

COMPONENTS 1330

4 SYNTHETIC METHODS 1330

4.1 Introduction 1330

4.2 Sputtering 1330

4.3 Evaporation 1330

4.4 Chemical Vapor Deposition 1334

4.5 Epitaxy 1335

4.6 Crystal Growth 1336

4.7 Plating 1337

5 VISION OF FUTURE

COMMUNICATION

COMPONENTS 1337

REFERENCES 1338

1 INTRODUCTION TO COMMUNICATIONS

The communications revolution of the previous decade is based on transfer of

information. Initially, most of the information transferred between people was

via voice communication, that is, over the telephone network. Today, due to the

birth of the Internet, the transfer of data now exceeds the network capacity

demanded by voice communication even though the number of cellular phones

is increasing dramatically and the cost per minute of long distance is decreasing.

One may ask: Can the communication network meet this demand for capacity?

The capacity of any network, whether it is cable, wireless, or optical is related

to bandwidth. Bandwidth literally describes the width of the frequency band, or

window, that can be utilized to send information. Bandwidth is a limited re-

source. For wireless communication bandwidth is so valuable that it is regulated

HandbookofMaterialsSelection.EditedbyMyerKutz

1304 ADVANCED MATERIALS IN TELECOMMUNICATIONS

and sometimes sold by the government. Bandwidth in optical ﬁber now seems

limitless, this is due to the small amount of information currently being carried

along the ﬁber lines. Few people who are connected to the Internet utilize wide

bandwidth services, such as Ethernet, cable modems, and digital subscriber lines.

As more of these services are installed in businesses and neighborhoods, the

demand for enormous data transfer rates will increase in the form of streaming

audio and video. The maximum capacity of a network, which is the sum of all

of the individual users’ data transfer rates, is directly proportional to the band-

width of the network.

Information can be transmitted in analog or digital format and with a wide

variety of encoding techniques: modulation by amplitude (AM), frequency (FM),

or phase (PM).

Regardless of the technique used to encode the information, the

capacity of a network, also referred to as the data transfer rate, can be estimated

by Shannon’s equation,

where the capacity of the system, given in bits per

second, is equal to the bandwidth in hertz for a system where the signal-to-noise

ratio approaches 1 (i.e., the weak signal limit). Therefore, if the bandwidth is 1

MHz, then the capacity is equal to 1 million bits per second (1Mb/s).

The goal is to manufacture components that operate at high frequencies and

over large frequency ranges in order to increase bandwidth. Based on these

components, system engineers will be able to design high-capacity networks.

The current conﬁguration of the communication network is illustrated sche-

matically in Fig. 1. The three dominant communication networks are presented:

cable, optical, and wireless. These networks are divided into local (metro, wire-

less) and long-distance (also known as long-haul) sections. Depending on the

network design and transmission protocol (modulation, multiplexing, access),

speciﬁcations are assigned for devices and components. Some devices are high-

lighted in Fig. 1, and their components are listed in the ﬁgure key. This illus-

tration provides an overall view of the telecommunications business: markets,

networks, devices, components, and materials. Envisioning dominant markets

for the future and deciding which networks will evolve to meet capacity demands

will aid in the determination of materials research directions.

This chapter will describe brieﬂy the communication networks, illustrate in

depth several key network components with attention paid to the materials used

in these devices and the reasons for their selection, list some materials and their

respective components, summarize various synthetic techniques for making these

materials, and ﬁnally look toward the future for communications technology.

1.1 Networks

Cable

The original telephone network was initially assembled for use with telegraphs.

Since then with the expansion of nearly a phone in every home, the telephone

network links the world. This initial copper cable network was based on twisted

wire pairs and coaxial cable. The twisted wire cables were assembled with 6–

4200 pairs wrapped in a protective sheath either composed of polymer insulation

with a copper core or a cost-effective alternative, an aluminum core.

Coaxial

cable, pictured in Fig. 2, was used in the backbone of the wired network. A -

–

in. coaxial cable with 300 MHz bandwidth demonstrates 4 dB attenuation over

1 INTRODUCTION TO COMMUNICATIONS 1305

Fig. 1 Communication network.

100 m while operating at a center frequency of 300 MHz.

Coaxial cable has

almost entirely been replaced by optical ﬁber and analog transmission was com-

pletely phased out by the year 1990.

Cables of twisted copper wire pairs still connect the ‘‘last mile’’ of the tele-

communications network. The last mile refers to the distance from the end of

the ﬁber-optic backbone to the home. Due to encoding techniques, the standard

telephone network has a maximum data transfer rate of 56K bits/s even though

the bandwidth of a twisted pair, which varies with the type of wire and its length,

is 2 MHz over a one mile span.

Recently, this bandwidth limitation has been

overcome by digital subscriber loop/line (DSL) technology.

Primary concerns of the copper network are loss (attenuation of the signal

from resistance), crosstalk (mirror of another conversation through induction),

and delay distortion (data transfer rate limitation due to capacitance of twisted

wire pair). These are being addressed through an evolution toward optic ﬁber

rather than incremental (and costly) advances to improve resistivity with silver

1306 ADVANCED MATERIALS IN TELECOMMUNICATIONS

Fig. 2 Schematic of copper cable of twisted pairs showing layer structure and materials.

wire and lower dielectric constant insulation while still maintaining high break-

down ﬁelds.

Optical

Optical ﬁber possesses nearly inﬁnite bandwidth, or so it seems. Currently, all

Internet trafﬁc can be carried on a single optical ﬁber. The only reason this is

still true is due to slow modems using conventional phone networks, which are

limited to below 56K (56,000 bits/s) data transfer rates, in addition to the limited

number of people connected to the Internet at one time. However, this is chang-

ing daily. Cable modems and digital subscriber lines are now available in many

areas in the United States yielding much higher data transfer rates up to 500

KBytes/s (500,000 bytes/s

⫽ 4,000,000 bits/s). In the near future, optical ﬁber

to the home with extremely fast connections using transponders (optical mo-

dems) will greatly increase demand for capacity and thus bandwidth.

What is the maximum bandwidth of optical ﬁber? The usable wavelength

range (related to frequency range) is from 1280 to 1620 nm (1.28 to 1.62

␮

m),

a portion of the infrared electromagnetic spectrum. Restating this in terms

1 INTRODUCTION TO COMMUNICATIONS 1307

Fig. 3 Optical ﬁber transmission spectrum.

of frequencies (2.34 ⫻ 10

Hz to 1.85 ⫻ 10

Hz) and taking the difference

(0.49

⫻ 10

Hz) yields a maximum bandwidth of approximately 50 THz. There-

fore, a conservative estimate of the capacity is roughly 50,000,000,000 bits/s,

or 50 Tb/s. Recently, based on nonlinear effects in optical ﬁber, the maximum

transmission rate is estimated at 100 Tb/s.

This value for capacity is based on the acceptable wavelength range, which

is determined by the transmission medium: silica ﬁber. Within this range, how-

ever, absorption due to hydroxyl (OH) groups in the SiO

ﬁber prohibits use

within a band around 1.4

␮

m (Fig. 3).

In addition, losses due to Rayleigh

scattering and vibrational absorption in the infrared place bounds on the wave-

length range. For this reason, two main wavelength regions are used for tele-

communication centered about 1.3 and 1.55

␮

m. The region at 1.55

␮

mis

divided into three bands: C-band (common) from 1.53 to 1.56

␮

, the L-band

(long wavelengths) from 1.56 to 1.62

␮

m, and the S-band (short wavelengths)

from 1.48 to 1.53

␮

m, which is commonly used for pump lasers. Further im-

provements have been made in ﬁber with the announcement of Lucent’s Allwave

ﬁber in which the hydroxyl concentration is negligible and therefore the entire

wavelength range can be used to transmit data. Unfortunately, due to the absence

of a broadband optical ampliﬁer, the use of Allwave is limited to metro appli-

cations where repeaters are unnecessary. See Section 2.4 for more details on

optical ﬁber.

The most ﬂexible format with which to utilize the optical spectrum of ﬁber

is wavelength division multiplexing (WDM). WDM is a type of frequency di-

vision multiplexing where each channel is separated in frequency by 200 GHz

for coarse WDM or as small as 50 GHz for dense WDM (DWDM). Currently

available optical networking systems can transmit data at 40 Gb/s per channel

over 80 channels. Predictions indicate that systems with 320 channels at 80 Gb/

s each are approaching quickly; that is a capacity of 25.6 Tb/s.

Perhaps the

elegance of DWDM is the ability for each channel to be independently encoded.

1308 ADVANCED MATERIALS IN TELECOMMUNICATIONS

Therefore, a system can be upgraded sequentially such that some signals can be

modulated at 2.5 Gb/s (currently employed networks) simultaneously with signal

at 40 Gb/s. In fact, each channel can also be encoded via a different format,

such as SONET, ATM, or IP.

Since optical communication has the beneﬁt of incredibly high data transfer

rates of terabits per second (Tbit/s), new encoding techniques were necessary

to accommodate this bandwidth: SONET (synchronous optical network)/SDH

(synchronous digital hierarchy) and ATM (asynchronous transfer mode).

A for-

mat that can accommodate voice, data, and image (television) information si-

multaneously will be the encoding technique in the future. Currently, ATM has

this capability; however, Internet protocol (IP) will be favored if Internet con-

nections can be made with greater reliability.

For the 2 billion subscribers worldwide (as of the year 2000), bandwidth

demand for telephone, Internet connections, and high-deﬁnition television

(HDTV) will continue to increase exponentially if cost can be decreased by

integrating devices and by transmitting along one medium—optical.

Wireless

The wireless market has exploded. Extremely high growth rates of 67% in 1999

and 47% in 2000 have demonstrated the demand from consumers. However,

available bandwidth is quickly being utilized. The total frequency range available

for wireless is 1 Hz to 100 GHz, thus providing a maximum bandwidth of 100

GHz. Only a fraction of this, however, is available for public communication

with the remainder used exclusively by the government for military and scientiﬁc

purposes. In addition, much of this spectrum is attenuated by water vapor or

requires line-of-sight operation. The common bands for wireless telecommuni-

cation are the following in terms of center frequency and bandwidth in paren-

theses: 800 MHz cellular (25 MHz bandwidth), 1.9 GHz digital personal

communication system (PCS; 200 MHz), 50 GHz satellite (500 MHz and 2.5

GHz).

Due to limited wireless bandwidth, data transfer with currently available

portable devices is very slow. The so-called 2G (second-generation) wireless is

rapidly becoming available across the United States and is already implemented

throughout Europe and Japan. Typical data transfer rates are less than a 56K

modem. Due to the challenges presented in 3G (third-generation) technology,

an intermediate solution, 2.5G, will be necessary to satisfy market demand for

lower cost networks and handsets. Although bandwidth is limited, wireless has

evolved greatly from the ﬁrst-generation analog to digital (2G) service, which

uses digital signal processing to reduce static and cancel white noise in addition

to optimizing access.

The method used to divide the wireless spectrum for individual users is known

as access. Three types of access have been developed: frequency division mul-

tiple access (FDMA), time division multiple access (TDMA), and code division

multiple access (CDMA). Current wireless networks utilize TDMA and CDMA

due to the efﬁcient use of bandwidth (i.e., more phones can access the same

bandwidth). The promise of 3G technology is likely to enable a society with the

world at their ﬁngertips. With fast data transfer rates, the internet will be instantly

accessible. In 20 years from now, telecommunications will have evolved into a

seamless communication network accessed by wireless technology.

2 MATERIALS SELECTION FOR SELECT COMPONENTS 1309

1.2 Transmission

Voice

The current telephone network provides highly reliable dedicated connections

between telephones based on a circuit-switched system. This is obvious in the

case of wired connections where each phone call is connected when the handset

is answered. For wireless communication, both analog and digital generations

dedicate bandwidth for each individual user. In a wireless system given an al-

lotted bandwidth segment, part is used for transmitting to the mobile user and

part is used for receiving, thereby providing a dedicated connection.

Dedicated line connections yield the quality expected in a telephone network,

for example, minimal delays, unnoticeable echoes, and high clarity.

Data

The Internet has spawned a relatively new form of transmitting information via

broadband multimedia networking. Data packets are sent over the network using

Internet protocol (IP). This allows for multiple routing and error correction rou-

tines that differ dramatically from voice networks. Foremost, IP allows for very

inexpensive networking. However, in terms of telecommunications, packet de-

lays makes a conversation between people extremely frustrating. Just in the same

manner as delays from live satellite interviews seem like brief eternities. Typical

IP connections produce longer and unpredictable delays, which is the main rea-

son that the Internet is not currently used to carry voice trafﬁc. When a more

reliable connection can be made using IP, then clearly it will be preferred for

the transfer of voice, data, and video.

The details of the transmission of information over a communication network

are not of obvious importance to the materials scientist or hardware engineer.

However, the designer of future systems must be aware of the limitations of

current networks and access protocols in order to anticipate the evolution of

communication hardware. Current materials and devices are outlined in the fol-

lowing section.

2 MATERIALS SELECTION FOR SELECT COMPONENTS

Selecting a material for a particular component is an extremely difﬁcult task. In

an ideal world, the person designing a particular network realizes that the effec-

tive frequency range is, for instance, 1 GHz. In their design, black boxes rep-

resent transmitters and receivers. However, no details are presented on how to

fabricate such a device. Therefore, the electrical engineer is asked to design a

receiver for a 1-GHz device with 10 MHz bandwidth. The picture thus drawn

includes ﬁlters, ampliﬁers, capacitors, resistors, and the like. The ﬁlter perform-

ance requirements, however, surpass any component on the market. Therefore,

the engineer speaks to a physicist who determines that the only way to make a

large bandwidth ﬁlter is to utilize a piezoelectric material with a sufﬁciently

large electromechanical coupling coefﬁcient. Finally, the physicist speaks to a

materials scientist who lists the materials with such coefﬁcients or embarks on

a search for a new material with such characteristics. The materials scientist also

must select the best method to produce (or synthesize) such a material in the

1310 ADVANCED MATERIALS IN TELECOMMUNICATIONS

Table 1 Properties of Currently Used Electro-Optical Material, Lithium Niobate (LiNbO

and Desired Properties for New Electro-Optical Material Incorporated into Silicon Optical

Bench Applications (SiOB)

Properties LiNbO

New Material Requirements

Linear electro-optical coefﬁcient r

⫽ 31 pm / V r ⬎ 2pm/V

Optical attenuation (at 1.55

␮

␣

⬍ 1dB/cm

␣

⬍ 10 dB / cm

RF attenuation (10–100 GHz) Low related to

␣

DC resistivity

␳

⬃ 10

⍀䡠cm

␳

⬎ 10

⍀䡠cm

Index of refraction (at 1.55

␮

m) n ⫽ 2.17 1.2 ⬍ n ⬍ 2.5 (lower is better)

Dielectric constant (at RF)

␧ ⫽ 29 (along a) 44 (along c) ␧ ⫽ 10–100 (lower is better)

Waveguide contrast (

⌬n) maximum 0.5% for Ti-diffused 1%

Materials compatibility with SiOB No Desired

Postprocessing temperature 500

⬚C ⬍1100⬚C for SiOB

Thermal stability T

⫽ 1165⬚C (Curie temp.) ⫺60 ⬍ T⬍ 100⬚C

Breakdown ﬁeld E

⬎ 10 MV / m E ⬎ operating ﬁeld (⬃1MV/m)

Depoling ﬁeld E

⬎ 8MV/matT ⫽ 130⬚C E ⬎ operating ﬁeld (⬃1MV/m)

Creep or fatigue No Typical for most Pb-based ferroelectrics

appropriate geometry for the component, such as, single-crystal, polycrystalline

monolith, epitaxial thin-ﬁlm, amorphous thin-ﬁlm, or polymer coating.

This design chain is not the end, though. The above picture is not complete

for devices, components, and networks in the real world, which must satisfy a

long list of technical and cost requirements. For instance, an electro-optic ma-

terial, which demonstrates a change in index of refraction (n) with applied elec-

tric ﬁeld (E), could be used for a number of applications if it meets the

requirements given in Table 1. As a materials scientist, tuning or optimizing one

variable is challenge enough. However, to meet a list such as given in Table 1,

thousands of choices are available, yet none may be feasible. For example, most

electronic and optical devices utilize either pure elements, alloys, or simple bi-

nary compounds (two elements only, e.g., GaAs). The semiconductor laser ma-

terials, however, incorporate three or four elements in order to tune the bandgap.

This compositional variation occurs in a pseudo-binary structure. Even though

many elements are present, there are only two unique sites in the atomic structure

for the atoms to reside. The structures for silicon (Si) and indium phosphide

(InP) are illustrated in Fig. 4. Immediately, many similarities are observed such

as the following: cubic symmetry, equal numbers of atoms in the unit cell, and

tetrahedral coordination. However, only one atomic site is unique in the structure

of Si, whereas two atomic sites are present in InP (A site and B site). Both sites

are tetrahedrally coordinated, but the (0, 0, 0) position at the corner of the unit

cell is fully occupied by P and the symmetry position at is fully occupied

111

–––

(, , )

444

by In. In the case of a gallium indium arsenide phosphide [(Ga, In)(As, P)], the

elements are substituted in these positions as a random statistical population of

each chemically equivalent element on a given atomic site. To elucidate, Ga and

In are chemically similar in that both have

⫹3 oxidation state with similar va-

lence shells; and therefore, both reside on the site. In the same manner,

111

–––

(, , )

444

As and P are similar possessing ⫺3 oxidation state and residing on the (0, 0, 0)

site.

Due to this discrepancy between tailored properties and ease of manufacture,

many materials currently in devices is very limited. Whereas, the number of