Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Lasers in Modern industries Anthony ]. DeMaria The development of the ammonia beam maser in 1954 ushered in a new breed of active devices that electronic engineers could relate to and use (Gordon et al., 1954~. The ammonia beam maser was the first device to use stimulated emission from inverted-population states of quantum mechanical resonances to provide gain for an electromagnetic oscillator. The operation of this quantum mechanical device initiated the field of quantum electronics. In 1984 the field of quantum electronics was 30 years old. In 1958 Arthur L. Schawlow and Charles H. Townes pub- lished a classic paper suggesting the use of the maser principle (with appropriate modification) for the generation of coherent infrared, visible, or ultraviolet radiation (Schawlow and Townes, 19581. The operation of the first ruby laser by Theodore H. Maiman in the latter part of 1960 made available for the first time a visible light beam that had characteristics previously associated only with radio frequency and microwave radiation (Madman, 1960~. The acronym laser was formed from light amplification by stimulated emission of radiation. The year 1985 was the 25th anniversary of the laser. Laser action has now been observed in solids (crystalline and noncrystalline insulators and semiconductors), liquids, gases, and plasmas yielding thousands of discrete wavelengths vary- ing from the vacuum ultraviolet to the millimeter wavelength HISTORICAL BACKGROUND 17
ANTHONY ]. DEMAR~A portion of the electromagnetic spectrum. Dye, color centers, and lead salt lasers now provide tunability over the visible, near-infrared, and infrared spectrum. At present, the abilities of electronic and laser devices overlap for generating radia- tion in millimeter and submillimeter wavelengths. Scientists are still working toward the generation of coherent radiation at ever-higher frequencies extending to soft and hard x-ray radi- ation. Few developments in science have excited the imagination of scientists and engineers as has the laser. The laser made it possible to transport into the optical region all the basic tech- niques developed for application in the radio and microwave regions, such as harmonic generation; parametric amplification; amplitude, frequency, and phase modulation; homodyne and heterodyne detection; and chirping and pulse compression. In the 25 years since the laser was first realized in the form of pulsed coherent emission from a single ruby crystal, the field has grown at a rate rarely experienced in science. The availability of these intense, coherent optical radiation sources has made it possible for scientists to experiment with optically generated plasmas; optical harmonic generation; stimulated scattering effects; photon echoes; self-induced optical transparency; opti- cal pulses; optical pulse compression; holography; optical shocks; self-trapping of optical radiation; optical parametric amplification; optical ranging to the moon; extremely high resolution spectroscopy; refined measurements of many basic physical properties (length, the speed of light, and so forth); and ultrafast relaxation processes in atoms and molecules. A MULTIDISCIPLINARY FIELD Today, the field of laser devices encompasses numerous disci- plines. They include solid-state, molecular, and atomic physics; spectroscopy; optics; acoustics; electronics; semiconductor tech- nology; plasma physics; vacuum technology; organic and inor- ganic chemistry; molecular and atomic kinetics; thin-film tech- nology; glassworking technology; and crystallography. More recently, the field has come to encompass electron-beams, x- rays, fluid dynamics, aerodynamics, and combustion physics. In sum, even without considering applications? the field has grown so fast and proliferated so broadly that scientists are virtually required to specialize within it. As a result, probably no individ- ual today would claim authoritative knowledge over the whole field of laser devices, or even be knowledgeable about most of the significant literature.
LASERS IN MODERN INDUSTRIES 19 THE BIRTH OF THE TECHNOlOGY During the first 15 years after development of the maser, from approximately 1954 to 1969, the field of quantum electronics was in the technology birth phase. After the operation of the ruby laser in 1960, emphasis shifted from maser to laser devices. This phase was characterized by numerous scientific discoveries and inventions as well as by widely believed visions and predic- tions of numerous medical, industrial, commercial, scientific, and military applications. During this phase, many laser devices were discovered from a large variety of gases and liquids, as well as from both amorphous and crystalline dielectrics and semicon- ductor solid-state materials. Few business opportunities existed during this phase, except to sell components, materials, and devices to researchers con- cerned with developing the technology base of lasers. Some opportunities were available to sell newly discovered laser sources to researchers interested in probing the linear and nonlinear electromagnetic behavior of atoms and molecules in liquids, solids, and gases. Large corporations were funding in-house research efforts in the technology, as well as capturing significant government research and development contracts. These contracts were directed toward determining the feasibility of numerous military applications during this early develop- ment cycle of the technology. ENGINEERING DEVELOPMENT PHASE For approximately the next 15 years, laser technology entered the engineering development phase. This phase was character- ized by a noticeable decrease in scientific breakthroughs and a perceived impatience with the rate of technological progress toward applications that addressed large markets. This was the period when the statement "the laser is a solution in search of a problem" was often heard. During this phase, many companies with marginal interest in laser technology dropped out of the field. In the same era, entrepreneurs invested considerable effort in searching for markets with large growth potential. In both of these early periods, the military market was larger than the commercial and industrial markets. MA NUFA CTURING TECHNOl OG Y PHASE Laser technology has now definitively entered the manufactur- ing technology phase. Sizable markets have been identified. A -
20 ANTHONY J. DEMARIA RECENT LASER MARKETS strong system and subsystem development effort is in place in which laser devices either significantly lower costs or raise performance leverage over older, more mature technologies. Consequently, product developments have intensified, and many new companies are being created. In addition, large, well-established corporations promote and sell products aimed at markets that laser technology can address uniquely: telecom- munications, data processing and storage, entertainment, print- ing, material working, and medical applications. In contrast to earlier phases, the commercial and industrial markets are now larger than the military market. Consumers are also beginning to experience laser technology directly through video and audio discs, laser printers for small computers, bar code readers at checkout counters, fiber-optic telecommunications, and various medical treatments. There is evidence of consolidation among numerous small companies oriented toward markets that use laser technology. NEXT: MATURE TECHNOLOGY PHASE In the future, laser technology will enter the mature technology, or commodity product, phase, which will be characterized by cost- and volume-driven markets (i.e., economies of scale), requiring capital-intensive manufacturing plants. The laser di- ode is the first laser device to achieve the mature technology phase. As production volumes and techniques have approached those in the high-volume manufacture of integrated circuits, unit prices of laser diodes have dropped accordingly. Eventu- ally, a few large companies will address the most important laser markets, and chances are good that the surviving manufacturers will not be those known today. The dollar value of the 1984 worldwide laser market in the commercial sector and the government and military sector was approximately $2.855 billion and $1.305 billion, respectively, for a total of approximately $4.16 billion (Spectra-Physics Corp., 1983, 1984; DeMaria, 1985~. Figure 1 compares the 1983 and 1984 market dollar values. Of the commercial market, approx- imately $2.502 billion was reported to be in systems and add- ons, whereas laser devices themselves amounted to approxi- mately $353 million of the 1984 world commercial market. As reported in Lasers and Applications (1987), "Commercial sales of individual lasers reached $509 million in 1986, up nearly 14%
LASERS IN MODERN INDUSTRIES 21 COMMERCIAL $22,855 M $1 ,985 M SYSTEMS & ADD-ON $2,502 M $1,705 M LASER DEVICES $353 M $280 M GOV. & MILITARY $1,305 M $1,230 M . .. , , ............... ....... ... ..... ................. .. .. ~ ~'<~5~ · 1 // :::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::-- ............................. . ....... TOTAL MARKET 1984 -- $4,160 M ~3 1983--$3,215 M I 1 1 1 1 1 1 0 400 800 1,200 1,600 2,000 2,400 2,800 DOLLARS x MILLIONS FIGURE l Laser industry world markets. SOURCE: Spectra-Physics Corp. (1983, 1984~. Over 1985. This year's sales should increase about 10% to $559 million." From this statement we interpolate that individual laser sales in 1985 totaled approximately $447 million, up approximately 27 percent from the 1984 sales of $353 million. These estimates would lead an entrepreneur rightly to conclude that business opportunities are more plentiful with the inclusion of systems and add-one in a line of laser devices. It is important to note in Figure 1 that the military market was smaller than the commercial market in 1983 and 1984. This trend is expected to continue. A 1983 forecast predicted that the worldwide laser market would grow 23 percent annually (Spec- tra-Physics Corp., 1983, 1984), and that well over 75 laser companies contributed significantly to these world markets. According to Lasers and Applications (1987), "Overall, the laser industry continues to grow at double-digit percentage rates. However, the double digits are now in the low teens, not the low twenties as was the case in the early 1980's." In view of the slow industrial growth in 1986 and 1987, growth of the laser industry in the low teens percentage rate is very respectable. This remarkably high growth rate is compara- ble to that experienced by the microelectronic and information processing markets. Figure 2 compares the growth of laser markets in 1984 and 1983. Note that the total market in 1984 grew 29 percent over the 1983 sales. Military sales experienced only a 6 percent 1 1
22 ANTHONY J. DEMAR~A COMMERCIAL 44% SYSTEMS & ADD-ON 47% LASER DEVICES 26% GOVERNMENT & MILITARY 6% TOTAL MARKET 29% 0 10 20 30 % GROWTH FIGURE 2 Laser industry growth, 1983 - 1984. //////////////////////////,/////////A //////////, ~ ,////, 40 50 increase in this period, and the commercial market grew by a phenomenally large 44 percent. COMMERCIAL LASER INDUSTRY A more detailed look at the 1983 and 1984 commercial laser world market reveals that approximately 63 and 50 percent, respectively, of these 2 years' commercial markets ($1.25 billion and $1.44 billion, respectively) was attributed to printing and graphics associated with information and data processing (see Figure 3~. Since Spectra-Physics does not have a product line in the entertainment or computer fields, data on semiconductor lasers and subsystems associated with the video and audio discs and data storage markets were not included in its annual reports (Spectra-Physics C:orp., 1983, 1984) and thus are not included in the data shown in Figure 3. The market for audio and video disc entertainment is bringing laser technology directly into the home. Fortune forecast that music lovers in the United States would buy 15 million discs in 1985 versus 5.8 million in 1984 (Fortune, 1985~. Fortune also forecast that sales of players and discs would reach $1.3 billion worldwide in 1985. The second largest segment of the world's commercial laser market consists of laser material working, which accounted for approximately 11 percent in 1983 ($210 million) and 10 percent in 1984 ($285 million) of the total market. The third largest market segment was communication, which accounted for ap- proximately 8 percent in 1983 and 1984 ($150 million and $225 million, respectively). The medical market is the fourth largest segment, with $ 105 million in 1983 and $ 150 million in 1984,
LASERS IN MODERN INDUSTRIES 23 capturing just over 5 percent in each of these 2 years. The laser market in metrology, industrial inspection, and science was just under 5 percent in 1983 and 1984 ($90 million and $ 135 million, respectively) of the total market, and it ranks fifth in size after the medical market. The data capture sector of the market (bar code readers, for example) is the sixth largest in size, with 1983 and 1984 sales of $70 million and $85 million, respectively, and market percentages of just under 4 percent in 1983 and just under 3 percent in 1984. The 1983 numbers include a $35 million miscellaneous category that is not shown as a bar plot in Figure 3 but is included in the $1.986 million total (Spectra- Physics Corp., 1983, 1984~. The expanding applications of lasers in the medical field are a source of great satisfaction to laser researchers. One of the earliest medical applications of lasers was in retina operations. Since then, much progress has been made. Lasers are now being used or investigated for use in cataract surgery, treating bleed- ing ulcers, opening blocked windpipes, reconnecting severed nerves, removing tumors, and cleaning the plaque that clogs blood vessels. Lasers are also starting to play a role in dermatol- ogy, plastic surgery, gynecology, and podiatry (see Rodney C. Perkins in this volume). It is no wonder that the market for laser MATERIALS WORKING $285 M $210 M COMMUNICATION $225 M $150 M MEDICAL $150 M $105 M METROLOGY & INSPECTION $135 M $90 m DATA CAPTURE $1 1 0 M $75 M ALIGNMENT $85 M $70 M PRINTING & GRAPHICS $1 ,440 M $1,250 M . _,,, ' ' ' 'A , '- - - - - - - - - - - - - ' - - - - -] ///// /// .. - 2 ] .............. 3 it ...... , war TOTAL 1984 -- $2,855 M 1983 -- $1,985 M 1 1 1 1 1 1 1 0 200 400 600 800 1,000 1,200 1,400 DOLLARS x MILLIONS FIGURE 3 Commercial laser industry world market. SOURCE: Spec- tra-Physics Corp. (1983, 1984). 11 Tl
24 ANTHONY J. DEMARIA MATERIALS WORKING 36% COMMUNICATION 50% MEDICAL 43% METROLOGY & INSPECTION 50% DATA CAPTURE 47% ALIGNMENT 21% PRINTING & GRAPHICS 15% COMMERCIAL MARKET 44% systems for medical applications is expected to double in each of the next several years. It is an embryonic but fast-growing market. Figure 4 compares 1984 sales growth with 1983 sales for the market segments identified in Figure 3. The communications market and the metrology and industrial inspection market had an astonishing 50 percent growth in this 2-year period. The data capture and medical sections of the market had outstanding increases of 47 and 43 percent, respectively. The growth of 36 percent achieved by the materials working market and 21 percent by the alignment market would be the envy of most high-technology industries. The printing and graphics segment of the laser market had the smallest growth 15 percentof the identified market. In the automated offices of tomorrow, it has been widely forecast that video screens will replace ink and paper. Today, however, office automation is producing more rather than less paper. The printers that help computers create much of this paper have become a $2.4 billion industry, with the promise that printer sales will more than double before the decade ends (The Wall Street journal, 19841. Semiconductor laser printer technol- ogy is expected to become the major competitor against ink jet printers in the future computer printer market. One or both of i//////////// ///////////////////////~ ////////////////////~ ///////////////////~//~ /////////////////////~ /////// a//// ////////////////////~ 1 1 1 1 1 1 0 10 20 30 40 50 60 % GROWTH FIGURE 4 Commercial laser industry growth, 1983-1984.
LASERS IN MODERN INDUSTRIES 25 5,000 4,000 cn CC o C] O In By o ~ 2,000 x cn J In 3,000 1,000 o U.S. WORLD 1982 1984 1986 1988 1990 FIGURE 5 Fiber optics market: fibers, cables, transceivers, components. SOURCE: Business Week (19841. these technologies is expected to displace the typewriter tech- nology used today. Much of the laser communications market that is, the fiber optics telecommunications marketis held by large, vertically integrated corporations such as AT&T, ITT, and Nippon Telegraph & Telephone and is thus not available to other manufacturers. Undoubtedly, this accounts for the relatively small fraction of the total world market of the commercial laser industry attributable to the laser communications market, as shown in Figure 3. By 1990 the fiber optics world market is expected to exceed $4.5 billion per year, whereas the U.S. market will be approximately $2 billion per year (Figure 5~. Business Week has predicted that the end of conventional copper wire in the telecommunications industry could come as early as the turn of the century; moreover, semiconductor lasers cou- pled with fiber optics technology will make ground and under- sea cable communication so inexpensive that few commercial communication satellites will be launched in the l990s (Business Week, 1984~. There were 250,000 miles of optical telecommu- nication fibers installed in the United States in 1983. Northern
26 ANTHONY J. DEMAR~A 11 V DEO SEM - D~SCS ~ LASER ~ lASER \ PRINTING AUDIO ~ ~ DATA ~FIBER- _ 1 OPTICS | TELECOMMU- ~CATIONSJ - \ ~ - BAR CODE READING / \ FIGURE 6 Applications of semiconductor lasers. _~ Business Information, Inc., forecast that approximately 1.3 million miles of telecommunication fiber would be installed in the United States in 1986. This installation is expected to increase to 4.5 million miles in 1990 (Business Week, 19841. The development of laser communications technology is discussed in detail by C. K. N. Patel later in this volume. Over the last 20 years, three areas of laser technology have received continuous and extensive research and development support: laser weapons, controlled fusion, and semiconductor laser development. By most estimates, the practical realization of the first two is still believed to be 20 years away. Because of their importance to national security and economic well-being, they have received extensive government research and development support in many countries. The semiconductor laser, on the other hand, has been developed primarily with industrial re- search and development funds. This development has spawned many new sectors of major industries, such as telecommunica- tions, printing, video and audio discs, data recording, and bar code reading (Figure 61. In another example of the large markets generated by semiconductor lasers, Frost & Sullivan, Inc. (1986), has projected tremendous market growth over the next few years for optical disc system manufacturers and retail-
LASERS IN MODERN INDUSTRIES 27 ers for personal computers. They have forecast a market of $2.5 billion in 1990 for optical disc systems. Optical discs can offer much greater information density than current magnetic storage devices at a lower cost per byte. Consequently, they are expected to generate the next revolution in mass storage. The technologies for write-once and read-only optical discs have been well established for some time. They use a modulated laser beam to permanently engrave a submicron- size bubble or pit in the active layer of a medium. Intensive research has been focused on erasable optical discs using a semiconductor laser that causes either a phase change or magneto-optical change in a medium. Compact disc read-only memories store 550 megabytes of data that cannot be altered or erased and were the first to reach the market. Write-once discs are relatively new to the market and typically hold 1-2 gigabytes and let users store and update information without eliminating data already stored on the disc. Erasable discs are not expected to appear until 1988 and will permit continual reuse. The success of the semiconductor diode laser in bringing about many new sectors of major industries is probably attrib- utable to its compatibility with semiconductor integrated cir- cuits. Its small size, low manufacturing cost, low voltage and power requirements, and high efficiency make it compatible with modern electronic technology. The semiconductor laser is the first truly mass-produced laser. It has been reported that Mitsubishi Electronic Corporation produced 400,000 semicon- ductor lasers monthly in 1985. It is expected that 10 million diode lasers will be sold in 1987, with the vast majority going into audio disc players and low-cost printers. Laser diode production has become a commodity process with commodity pricing strat- egies. PHOTONICS VERSUS ELECTRONICS In the early days of electronics, vacuum tubes played an impor- tant role in developing the industrial base of the radio fre- quency, microwave, and millimeter wave portion of the electro- magnetic spectrum. Similarly, gas lasers and optically pumped solid-state lasers have been important in developing the indus- trial base of lightwave technology in the new field of quantum electronics. There is also a clear analogy between the roles played by the semiconductor diode laser and the transistor in establishing the industrial base of their respective spectral re-
28 ANTHONY ]. DEMARIA LANDMARKS IN ELECTRONICS The field of electronics began in 1883 when Thomas Edison discovered, while working with his carbon filament lamp, that current flowed across a vacuum when he placed a positive voltage on a metal plate a small distance from a glowing carbon filament in a vacuum envelope. This phenomenon was the basis of all electron tubes, which were the foundation of electronics until the era of the transistor began in 1949. From 1883 to 1904, no one exploited the effect to make a useful device for detecting, generating, and amplifying electrical signals, even though the telephone industry during that era could have benefited from the invention of a suitable amplifier. In 1887 Heinrich Hertz transmitted and received radio signals within his laboratory and experimentally confirmed lames Clerk Maxwell's equations, which had been published in 1864. In 1901 Guglielmo Marconi propagated radio waves across the Atlantic Ocean from Poldhu, Cornwall, England, to St. John, Newfoundland, Canada, with- out the aid of electronic amplifiers and oscillators. The vacuum diode rectifier was invented by John Ambrose Fleming in 1904. The famous audion, the three-element vacuum tube, was in- vented by Lee de Forest in 1906 and was the first electronic amplifier. Just 29 years after Edison's discovery in 1883, the first electronic oscillator came into being in 1912 with Edwin Arm- strong's invention of the regenerative circuit and operation of the first coherent electronic oscillator. From that point on, the electronics industry developed rapidly, heralding the beginning of radio and modern electronics. "ions. At present, the technologies of integrated optics and guided wave optics are being developed by researchers whose goal is to obtain benefits in the optical region similar to those obtained earlier in the electronic region by planar electronic integrated circuits, hybrid circuits, and planar microwave/milli- meter strip-line technologies. The ability to realize both optical and electronic devices from compound semiconductor technol- ogy is largely responsible for the present intensive research on compound semiconductors. The field of electronics was created with the invention of the vacuum tube around the turn of this century. The heart of electronic technology is the device that controls the flow of an electron stream (electrical current) either in a vacuum (the vacuum tube) or in a solid (the transistor). Since the laser ll 1
LASERS IN MODERN INDUSTRIES 29 controls the flow of a photon stream (light), the laser can be considered the heart of quantum electronics technology. This analogy can be carried one step further by including in the new term photonics the field of quantum electronics, which includes lasers, as well as optoelectronics, electro-optics, acousto-optics, fiber optics, integrated optics, and nonlinear optics. One should not jump to the conclusion that electronics and photonics technologies compete against each other. Rather, these two fields are complementary. Photonics depends heavily on elec- tronics technology, and is useful for those tasks that cannot be performed using electronics technology. By performing such tasks, photonics creates new segments of existing industries, thereby establishing a niche for itself and also further expand- ing the base of electronic technology. How much time will be required to commercialize the field of pnotonics? Indications are that electronics developed slowly during its earlier phase and then more rapidly in later stages. Photonics, in contrast, developed more rapidly during its earlier phase because of the technical support provided by the field of electronics and is expected to continue its swift progress. As photonics continues its rapid expansion past its 25th birthday, it is also expanding the future horizons of electronics technology. Since the 1912 invention of the electronic oscillator, there has been a steady drive toward the production and use of coherent electromagnetic energy of higher and higher frequencies. This tendency results partly from the realization that an increase in transmitted information, directivity, and efficiency can be achieved by increasing carrier frequency and partly from the crowding and interference between existing frequency bands. Another important push toward generating coherent radiation of higher frequencies resulted from researchers' interest in using these waves to probe atoms in solids, liquids, and gases by employing experimental techniques such as nuclear magnetic resonance, paramagnetic resonance, and cyclotron resonance. To meet these needs, investigators have devised active electronic devices that use the flow of an electron stream in a vacuum or the flow of electrons and holes in semiconductor materials. They have been greatly improved for the generation of higher fre- quencies. Examples of such devices are vacuum tubes, transis- tors, magnetrons, klystrons, traveling wave tubes, parametric amplifiers, and tunnel diodes. With these devices, researchers have generated coherent radiation in the hundreds of gigahertz. With the use of harmonic generators, this figure has been extended by approximately one order of magnitude. Almost without exception, as soon as higher frequency devices become 1 1
30 ANTHONY J. DEMARIA Srl Framers IN DUSTRIAL APPLICATIONS available, researchers rush to use them in probing the atomic and molecular domain of liquids, gases, and solids. The physical dimensions of the resonators used to select the oscillating frequency of conventional oscillators in the higher frequency range are of the order of magnitude of the wave- length of the radiation generated. As a result, it becomes extremely difficult to construct resonators to the small dimen- sions required at submillimeter wavelengths. In the late 1940s and early 1950s, it became apparent to workers in the field that it was becoming impossible to apply the old methods of scaling down existing devices for higher frequency generations. In the search for alternate methods, researchers came to realize that natural resonators in the form of atomic and molecular systems could be used to amplify and even generate coherent electro- magnetic energy. This realization led to the invention of the ammonia beam maser and the laser and to the creation of the field of photonics. This section will briefly discuss selected applications of lasers in semiconductor integrated circuits manufacturing, radar sys- tems. material cutting and drilling, and inspection of electric power cables. These examples indicate the breadth of laser applications in modern industries, such as microelectronics, · · , · · . . avionics, machining, and e. ectrlc power. MANUFA CTURING SEMICONDUCTOR INTEGRA TED CIRCUITS The technology of semiconductor integrated circuits has con- tributed greatly to the electronic, or information, revolution which may hold more enduring implications for mankind than the industrial revolution (Abelson and Hammond, 1977~. Of the large number of different integrated circuits now produced, the dynamic random access memory (DRAM) chips have the largest unit sales and volume and the greatest dependency on the manufacturing learning curve for reducing costs and increasing yields to meet the aggressive pricing strategy of competitors. DRAM chips use the most advanced processing technologies to achieve the highest density of semiconductor devices per chip. They also have one of the shortest product life cycles in the semiconductor industry; the last 25 years have seen a rapid increase in the complexity of these chips, from the first 4-kilobit (K) product up through the 16K, 64K, 256K, and the present
lASERS IN MODERN INDUSTRIES 31 1-megabit DRAM chip. The rate of progress has been breath- taking from the standpoint of the number of chips per wafer, the rapidly decreasing layout rules down to the present micron to submicron dimensions, increasing die sizes, decreasing num- ber of dies per wafer, and increasing capital cost for a wafer fabrication factory. Consequently, a manufacturer's ability to be the first to market and to increase sharply chip yields early in the production cycle can usually determine success or failure in the market. Because of the insatiable appetite of computer manu- facturers for an ever-larger memory capacity per chip, fierce competition is now under way to be the first to market with 4- and 1 6-megabit DRAM products. Laser technology has made it possible for manufacturers of DRAM chips to increase their yield and volume dramatically in the early phases of their production cycle by use of a technique often referred to as laser redundancy (Pose, 1981; Smith, 1981~. Laser redundancy enables manufacturers to design spare, normally in- active, address encoders into their memory chips. When a portion of a memory chip is found not to meet specifications during initial die testing, a pulsed 1.06-,um laser with neodymium ions in yttrium-aluminum garnet (Nd3+:YAG) is used to "explode" away the polysilicon conductor connecting the encoder addressing that portion of the circuitry, thereby disconnecting the defective por- tion from the circuit. The laser system is then used to open the polysilicon conductor, shorting out the spare address encoder and thereby connecting that encoder to the other circuitry of the chip. Figure 7a shows a small portion of an integrated circuit, including functioning polysilicon film conductor interconnect- ing lines, crystal silicon substrate, silicon dioxide insulator film, and aluminum film interconnecting lines. Figure 7b shows a polysilicon film interconnect line cut with a laser. Because the polysilicon has a higher absorption coefficient for the laser radiation than the crystal silicon, it melts selectively and evapo- rates under high-intensity, pulsed, 1.06-,um laser irradiation. The use of laser redundancy in the manufacture of 64K DRAM chips typically increased yields by 2 to 3 times during the start-up phase of manufacturing and led to an average 40-50 percent improvement in the number of good dies produced per wafer in the start-up phase. Manufacturers have been able to obtain an equivalent number of chips from one manufacturing facility as they previously obtained from two facilities. Mostek Corporation was one of the early users of laser redun- dancy in the manufacturing of 64K DRAMs. It took Mostek 1.5 years to produce 2 million 4K DRAM chips during the start-up phase when they introduced that product to the market in the early 1970s. Because of the added complexity of the 16K DRAM chip, it
32 ANTHONY J. DEMARIA 14 FIGURE 7 Top: Small portions of an integrated circuit showing the single-crystal silicon substrate (A), the silicon dioxide insulator film (B ), the aluminum film Interconnecting conductor lines (C), and the polysilicon film interconnecting con- ductor lines (D). Bottom: A polysilicon film interconnecting conductor line cut with a laser.
LASERS IN MODERN INDUSTRIES 33 took Mostek 2 years to reach the 2-million-chip production level. The process took only about 9 months for the 64K DRAM chips because of the use of laser redundancy. Based on this outstanding Mostek result, laser redundancy techniques are now used exten- sively in the manufacture of complex semiconductor chips. lASERMATERlAL WORKING The ability of a laser to deliver a high-intensity beam of radiation through the atmosphere and heat an absorbing material has drawn attention to its use in the mater~al-working industry for such appli- cai~ons as cutting, drilling, welding, heat treating, and melting. The utility of the laser in these applications can be seen by referring to the Stefan-Boltzmann law of radiation: The total energy radiated per unit area by a perfect thermal source is equal to the fourth power of the temperature times the Stefan-Boltzmann radiation constant. For instance, a power density of 1 million W/cm2 corre- sponds to a thermal source operating at 20,500 K. By means of optical focusing, the laser can easily achieve temperatures in this range and is therefore capable of providing such high energy concentrations that its focused radiation can melt or vaporize any known material. Figure 8 shows the laser radiation power density and the 1o1o ~ ~ ~ Specific energy, J/cm 1 09 _ Shock hardening ~ ~ ,,,,,,,. 1o8 NO I` CO CL 1o6 05 04 03 _ ~///~x 2 ~ 10 0 ~10 :~ in///' x~ `<Drilling, - ` _ .Nlo6 I//////,,, LASERGLAZETM:/// /' my////////// -~ region ~///~ - ;~//~/, Welding & cutting ''it 1 1 1 ~1 ~ //,///,, . N~ Transformation x hardening ~% 1 'L///~^ 10-8 10-6 10-4 10-2 10° Interaction timers FIGU RE 8 Laser beam-material interaction spectrum. Laser power density (W/cm2) and material interaction time (s) required by various material-working tasks. 1 1
34 ANTHONY ]. DEMAR~A Butt weld Tee weld with filler Butt weld ~ - 3.2mm 25.4 mm 1 1 :!5 4 mm 4.8 mm Power: 12 kW Power: 13 kW Power: 8 kW Speed: 12.7 mm/see Speed: 12.7 mm/see Speed: 21.2 mm/see No filler Filler: 0.89 mm wire, 127 mm/see No filler Matenal: Ship steel Matenal: A-36 steel Material: Low carbon steel FIGURE 9 Typical weld configurations performed with a CO2 laser under the conditions indicated. interaction time of the radiation with a material required to accomplish various important material-processing tasks (Banes and Webb, 19821. For a laser beam moving continuously across a material, the interaction time can be defined as the time required for the incident laser spot to move one diameter relative to the surface of the workpiece. For a material process requiring short pulses of laser radiation, the interaction time is the duration of the pulse, since the material can be assumed to be stationary during the short irradiation process. Figure 9 shows three typical laser welds performed with carbon dioxide (CO2) lasers (Duhamel and Banas, 1983~. A three-stage, gas-recirculating, closed-cycle, electric-discharge CO2 laser can yield 9-12 kW of continuous output power. The use of fast gas-flowing techniques to achieve several tens of kilowatts of continuous power from electrically excited CO2 lasers (DeMaria, 1973) has been responsible for placing CO2 lasers in a dominant position for large material-working appli- cations. In the next decade, laser material processing in manu- facturing is expected to increase dramatically. The Nd3+:YAG, ruby, Nd3+:glass, and CO2 lasers are expected to be the most widely used in these applications. LASER RADAR 1 1 Laser radar technology is an obvious progression of radar technology from the radio frequency, microwave, and milli-
LASERS IN MODERN INDUSTRIES 35 meter wave region of the spectrum into the optical region (i.e., infrared, visible, near ultraviolet). Laser radar technology has both advantages and disadvantages when compared with con- ventional radar technology. Consequently, laser radar systems will complement and not compete with conventional, lower frequency radar systems. Laser radar systems will be used predominantly in those applications that cannot be addressed by conventional radar systems. Range finders are the most basic radar systems of either the microwave or the laser variety. They measure the range to a target by measuring the time of flight of a transmitted and an echo pulse of electromagnetic radiation. The speed of the target can be obtained by measuring the change in range as a function of time. Range finders can also provide information about the azimuth to a target. Radars of this kind are known as incoherent radar systems. Coherent radar systems are more complex and have the ability to measure the velocity of the targets by means of the Doppler effect. This type of radar was originally used exten- sively in the early development of radar technology to detect moving targets against stationary background clutter. Coherent radar systems measure the Doppler shift of the echo radiation by comparing the frequency of the received echo signal with the frequency of the transmitted radiation. This comparison is accomplished by heterodyning, or mixing, the returned signal with the signal of the system's frequency reference (called the local oscillator) on a detector. By maintaining the frequency of the transmitter signal either above or below the local oscillator signal by a fixed value determined by electronic control circuits, and superimposing on the detector the local oscillator signal with the return signal from a stationary target, an interference pattern that modulates the amplitude of the detected laser radiation is generated on the detector. This "beat" signal is equal to the difference between the frequencies of the transmitted and local oscillator signals. Since this beat signal is arranged to occur within the radio frequency range (tens to hundreds of mega- hertz), electronic amplifiers tuned to this frequency can process the signal electronically and obtain the same signal-to-noise benefits well known in conventional heterodyne radio receivers. Measurement of the deviation of this known beat signal by the Doppler effect caused by the moving target provides a measure- ment of the speed of the target. An additional advantage of the coherent radar system is that one can increase the power of the local oscillator signal on the detector to achieve the theoretical detector performance, which is the quantum noise-limited sensitivity. A laser radar, using CO2 lasers, typically operates in the 1 0.6-,um wavelength region 1~1
36 ANTHONY J. DEMARIA (Silverman, 19821. At this wavelength, HgCdTe detectors at present provide the optimum sensitivity. A figure of merit for detectors is usually given in terms of noise-equivalent power, or NEP. Heterodyne NEPs of 2 x 10-~9, 5 x 10-~8, and 2 x 10-~7 WHz have been measured with HgCdTe detectors operating at 1 GHz at 77 K, i95 K, and 300 K, respectively. Table 1 compares some of the relevant parameters of x-band and CO2 laser radars. The laser radar operates at a frequency 3,000 times higher (or at a wavelength 3,000 times shorter) than an x-band radar. The large difference in wavelengths between the CO2 laser radar and the x-band radar results in large differences in reflection characteristics of targets for the two technologies. Variation in target surface dimensions (i.e., sur- face roughness) typically are greater than the wavelengths of CO2 laser radars, but less than the wavelengths of x-band radars. Since man-made targets usually have smoother surfaces than natural targets, even small man-made targets such as wires have a larger cross-section than natural targets for laser radars. Figure 10 shows the ratio of detector signal current to noise current (is/in) as a function of range for various natural and man-made targets irradiated with a pulsed, coherent CO2 laser radar system having 400 mW of average power and an HgCdTe detector. TABLE ~ Comparison of Basic Radar Parameters CO2 Laser Radar Characteristic Radar x-Band Radar Frequency, Hz 3 x 10'3 10~° Wavelength, cm 10-3 3 Beamwidth, AID, radians 10-3/dia. 3/dia. Doppler sensitivity, Sv/A, Hz 2,000 x velocity 2/3 x velocity Photon energy, joules 2 x 10-20 6.6 x 10-24 Note: D= diameter, in cm; v = velocity, in cm/s. Since the beam divergence varies directly with wavelength and indirectly with transmitting aperture, CO2 laser radars have 3,000 times smaller beam divergence than x-band radars with the same aperture. Since the Doppler shift varies inversely with wavelength, the CO2 laser radar has three orders of magnitude higher Doppler sensitivity than an x-band radar (see Figure 1 1~. Figure 11 shows, for instance, that for a radar frequency of 30 THz (CO2 laser frequency), a target moving at 0.5 km/in (about 1/10 the speed of a person walking) yields a Doppler signal of 100 kHz, whereas a 30-GHz microwave radar would yield a Doppler signal of 0.1 kHz.
LASERS IN MODERN INDUSTRIES 37 100 _ - <,, 10 1 ~ ~ GRASS WIRE \ WD-1 NORMAL°\ O VHF ANTENNA \POLE O O TV ANTENNA ~ ° CABLE \ No WIRE WD-1 160° \ HOUSE 0 \ TREES SNOW ~ \ 1 00 1 ,000 1 0,000 RANGE, m FIGURE 10 Signal-to-noise ratio of a CO2 laser radar (400 mW average power, 75 W peak power) for various natural and man-made targets. 1 10-1 10-2 co ID ~ 10 ~7 10 10-5 10-6 10-7 300 MHz / / /~-0, // . ~/~ / CO2; f ~ 3 x 1 o1 3~ 2/V/~// -2 1o~1 1 10 1o2 103 10 2 x velocity -- km/hr ~ RF Microwaves Millimeter waves 1 Infrared Tvisible luv FIGURE l l Comparison of Doppler sensitivity: Doppler frequency shift as a function of radar wavelength and target velocity.
38 ANTHONY J. DEMARIA 11 Since the photon energy of the CO2 laser radar is 3,000 times higher than that of the x-band radar, the laser radar beam has 3,000 times fewer photons per unit of energy than the x-band radar. If one photon in unit time is the minimum detectable signal, then the operation of a CO2 laser radar is limited to a smaller field of view than the x-band radar for the same transmitted power. Consequently, the laser radar is not suited to wide-area search applications, but is well suited to applications requiring ultrahigh sensitivity in range, azimuth, Doppler shift, image resolution, and small field of view. It is important to point out that laser radar suffers from poorer propagation characteristics through the atmosphere than conventional microwave radar because of higher back- scatter from rain, snow, haze, and fog and because of higher absorption by water in the atmosphere. Consequently, in the earth's atmosphere, laser radars have a shorter range than microwave radars. Fortunately, the operating wavelengths of CO2 lasers falls within one of the best atmospheric windows when compared with other laser wavelengths. Consequently, for applications in the atmosphere, the relatively long wavelength of 10.6 ,um for CO2 lasers over other lasers, such as Nd3+:YAG, ruby, and semiconductor lasers, makes the CO2 laser radar the system of choice for most applications. Figure 12 shows general areas of applications of various radar technologies. Ladar is a commonly used acronym for "laser radar," and was formed from laser detection and ranging following the example of the word radar, which was originally an acronym formed from radio detection and ranging. (Lidar, light detection and ranging, is also used.) Figure 13 compares a telescopic photograph of a control tower at a range of 1.2 km with an image taken by a 15-year-old coherent CO2 laser radar. The radar system used a binary (black and white) gray scale and was not intended to produce a photographic-quality image. The output power of the early radar used to produce this image was 0.25 W at a pulse repetition rate of 30,000 pulses/s. The trees in the far background of the scene do not show up in the laser radar image because of a range-gating technique used by the radar system. The evergreen trees at the bottom of the photograph do show up in the radar image because their return signals fell within the time window of the time-gated receiver. Since glass is absorbing at 10.6 ,um, the windows appear black in the radar image. The antennas on top of the control tower are difficult to see against the sky in the photograph but are easily visible in the laser radar image. The ability of CO2 radars to detect small obstacles such as wires,
1o2 1o1 E ,o - 8 ~ -1 ~ 10 - o 3 rsynt~ Ire Microwave S~irface/airborne/space surface radar ~ \ Micrwave _~ J / Tactical lADAR As/ / Surfacelairborne = ~ 1 10-2 10-3 - 4 10 Strategic LADAR Surface/space ~1 , ~ 11 1 1 10 4 10 3 10 2 10 1 0 1 10 Field of view, sterad LASERS IN MODERN INDUSTRIES 39 ~ 1o2 . 104 ~ 106 <~) ._ n 0 8 010 Z FIGURE 12 Active sensors capabilities: general areas of applications of various radar technologies from an angular resolution and field-of-view perspective. FIGURE 13 Comparison of a photographic image (taken through a telescope) with a black-and-white binary scale CO2 coherent laser radar composite image. The trees in the far background were not recorded in the radar image because of range-gating techniques used in the radar system. Range: 1,200 m; average power: i/ W; pulse rate: 30,000/s.
40 ANTHONY ]. DEMARIA poles, and antennas makes them ideally suited for obstacle and terrain avoidance applications in helicopters. They are also compatible with the 8- to 12-,um passive night-viewing avionic systems now in use. One of the most exciting potential applications of radar is in the measurement of wind velocities in the upper atmosphere by measuring, from the space shuttle, the Doppler shift from backscatter off naturally occurring aerosols in the upper atmo- sphere. The operation of such a system is expected to improve greatly the accuracy of weather forecasting. ~~ , ElECTRlC CABlE INSPECTION When it was established that polyethylene (PE) and cross-linked PE (XLPE) material had intrinsically high dielectric strengths, on the order of 800 kV/mm, the electric power industry ex- pected 40-year lifetimes for underground electric power cables in distribution systems using such materials. Consequently, in the 1960s, the electric power industry began to make extensive use of underground cables using PE and XLPE as the dielectric between the inner and outer conductors. The expected lifetime was not achieved even at average stress levels of 2-4 kV/mm, even though such stress levels provided 200-400 times smaller voltage gradients than the intrinsic dielectric strength of the material. The failure rate for cables put into service since the 1960s reached a level that disturbed the electric power industry. It led the Electric Power Research Institute, the U.S. Depart- ment of Energy, and cable manufacturers to launch a research and development program in the 1970s to solve the problem of the premature failures of PE and XLPE cables. The cables are produced in a continuous operation. A central conductor of stranded copper wire passes through an extruder that coats it with a smooth, thin semiconducting shield consisting of PE filled with carbon black. Over this opaque semiconducting surface, the white PE insulation is extruded and then cross- linked with heat, ultraviolet radiation, or electron bombard- ment. A second semiconducting shield of PE and carbon black is then extruded over the insulation, followed by a mesh of stranded copper wires and finally a protective plastic coating. The research and development programs indicated that the aging of the insulation generates branched channels caused by dielectric breakdown, which in turn causes an electrical short circuit between the outer ground conductor and the inner conductor (see Figure 141. The branched channel structure, or "trees," of dielectric breakdown in the insulation is believed to be
LASERS IN MODERN INDUSTRIES 41 Nn',tral Arraign - Microvoids Void Contaminant , ~°~` /N Voids and j' EN at interface a\ OWN Conductor in contact with insulation A/ ~ Undisoersed antioxidants kY~ ,_ ~ contaminants ~ Insulation and \~/ shield eccentricity Electrical tree Neutral wire embedded in insulation shield Strand shield fall-in ~ Bow be Bee 3 /~ nits ~ ' Electrochemical trees · Loss of shield conductivity · Damage during installation · Loose fitting insulation shield FIGURE 14 Common causes of electric power cable failures. caused by surface imperfections at interfaces within the cable and by irregularities in the insulating materials. These irregu- larities are caused by gas- or vapor-filled voids, contaminating particles, inhomogeneous variation of density in the materials, and other defects. Unfortunately, visual inspection is not possi- ble during the manufacturing process, where these defects arise, because PE is normally a milky white, opaque material except when immersed in hot oil. Corona testing is a nondestructive inspection technique that can detect 50-,um-diameter voids in 500-ft lengths of cable. The disadvantage of the technique is that it cannot detect contami- nants and flaws, voids filled with liquids or vapors, and micro- voids 1-10 ,um in diameter. The technique also does not provide on-line inspection during the manufacturing process, nor can it locate the position of the defect. The inspection procedure now commonly used is to cut out a 2-in. piece of manufactured cable every 10,000 ft. slice it into 0.5-in. portions, slice these portions into 0.020-in. wafers, make these wafers transparent by immersing them in hot oil, and then inspect the wafers visually under a 15-power microscope. The obvious disadvantages of this technique are that it is neither an . . . . . On- 1ne, rea -time inspection process nor a none .estructlve pro- cedure. One promising approach to an on-line, real-time technique tor nondestructive inspection of cable is based on the fact that PE and XLPE are almost transparent in the far infrared.
42 ANTHONY J. DEMAR~A ~ .~ , 250 Am void Microvoid background ~ AN devoid Add-- ~ - _~ _~ _ -~ Detector noise FIGURE 15 Signals of a far-infrared laser inspection system from voids in cross-linked polyethylene insulation used in electric power distribu- tion cable (25 KV moving cable'. Investigation of lasers that emit radiation in the far infrared (Chang et al., 1970) led to a nondestructive technique for continuously monitoring the quality of electric power cables in real time (Cantor et al., 1981~. A monitoring system based on this technique scans around the cable before the outer ground conductor and its protective coating are extruded onto the cable. The laser radiation scattered by voids, contaminants, or other defects in the dielectric is collected and detected, and its mag- nitude is digitally recorded. The speed of the cable through the system is monitored to maintain a complete record of signal amplitude caused by the scattered radiation as a function of cable position. Figure 15 shows typical signals obtained at a 118-,um laser wavelength and 0.1-W laser power with a germa- nium-doped silicon detector cooled with liquid- helium. Figure 16 shows an experimental arrangement of such a system. It is important to note that because of the long laser wavelength (submillimeter wavelengths), the mirrors are fabricated from finely machined aluminum and do not require extensive polish- ~ng. at, It is too early to determine the practical effect of laser inspection in the manufacture of electrical power cable, but it is already apparent that the technology will provide useful re- search instrumentation for the industry.
LASERS IN MODERN INDUSTRIES 43 I r ; _ _,..._ ._ .~ FIGURE 16 Optics used in a far-infrared laser inspection system for electric power cable insulation. Laser technology is young and robust, with a highly promising and exciting future. It is now spawning new products and opening major new segments of basic industries that will ensure its growth well into the next century. The fields of fiber-optic telecommunications, optical audio and video discs, optical data storage, optoelectronics, lasers for material working (cutting, welding, heat treating, hole drilling, and scribing), laser appli- cations in medicine, laser instrumentation, and military applica- CONCLUS~ONS
44 ANTHONY J. DEMA~A REFERENCES tions are still in their infancy; thus, considerable growth is yet to come. The most serious challenge in laser technology is the continu- ing shortage of photonic engineers required to develop the numerous new and rapidly evolving products the technology is generating, to continually advance the state of the art required to meet new product needs, and to work at the interface between electronics and photonics technologies. An engineer in this field needs a background in optics and electronics and in quantum electronics. Most engineers working ire the field today are either physicists who have learned some electronics or electronic engi- neers who have learned some optics. The offering of a formal undergraduate engineering curriculum in photonic engineering would be a big boost to this important emerging field of technology. Abelson, P. H., and A. L. Hammond. 1977. Science 195(4283):1085. Banas, C. M., and R. Webb. 1982. Proc. IEEE 70(june):556. Business Week. May 21, 1984. 181. Cantor, A. I., P. K. Cheo, M. C. Foster, and L. A. Newman. 1981. IEEE l. Quantum Electron. QE- 17(April):477~89. Chang, T. Y., T. J. Bridges, and E. G. Burkhardt. 1970. Appl. Phys. Lett. 17(Sept. 15):249-251. DeMaria, A. i. 1973. Proc. IEEE 61(lune):731-748. DeMaria, A. I. 1985. Optics News 10: 15. Duhamel, P. F., and C. M. Banas. 1983. 1983 ASM Conference on Applications of Lasers in Material Processing, Los Angeles,. 24-26 January. Reprint 8301-020. Metals Park, Ohio: American Society of Metals. Frost & Sullivan, Inc. 1986. PC Optical Disk Market in the U.S. New York: Frost & Sullivan, Inc. Fortune. July 8, 1985. 104. Gordon, I. P., H. l. Zeiger, and C. H. Townes. 1954. Phys. Rev. 95:282. Lasers and Applications. January 1987. 65. Maiman, T. H. 1960. Phys. Rev. Lett. 4:564. Posa, J. G. 1981. Electronics 28(luly):117-120. Schawlow, A. L., and C. H. Townes. 1958. Phys. Rev. 112:1940. Silverman, B. B. 1982. Proc. IEEE 1982 National Aerospace and Electronics Conf. 2:568-575. Smith, R. T. 1981. Electronics 28( July): 131-134. Spectra-Physics Corp. 1983. Annual Report. Spectra-Physics, Inc., San Jose, Calif. Spectra-Physics Corp. 1984. Annual Report. Spectra-Physics, Inc., San Jose, Calif. The Wall Street Journal. March 16, 1984. 29.
