Oral History Interview with Robert L. Cormia (RLC)
Conducted by Ric Shimshock (RS)
Robert L. Cormia
RS: I’m here in Palo Alto, California. I have with me Robert Cormia who is the 2001 Nathaniel Sugerman Award winner from the Society of Vacuum Coaters. We have here an oral history project and an interview that will start to describe some of Bob’s accomplishments, some of his various trials and tribulations. With that, Bob, where did you start your career on the academic side? Where did you get your educational training?
RLC: After high school I spent two years at New York State Agricultural and Technical Institute because that’s all our family could afford. I got an Associate’s degree in Electrical Technology, which led me to the job opportunity with the General Electric Company in 1954, in Schenectady, where I went on a series of training programs—three months here, three months there. They called it the Technicians’ Training Program. I learned about the General Electric Research Laboratory, so I targeted that. I worked at the various locations—their Relay Division, their Bloomfield, New Jersey, air-conditioning plant—but I had targeted Schenectady, in the Research Lab; that’s where I wanted to get a training period. So I got a three-month training period there. Once I snagged that, I never let go. I started there in 1954 late in the year—November or so. In 1954, Coolidge—the inventor of the X-ray tube—was walking the halls at GE Research. I actually saw Irving Langmuir; I never met him. And I worked with Catherine Blodgett, who did some of the very early work in the development of optical coatings during WW II, although I didn’t know it at the time. Saul Dushman, author of Scientific Foundations of Vacuum Techniques, was at GE Research at the time. Many really outstanding scientists worked there. However, I realized as a technician that there wasn’t much (for me) in the future. So being local to Union College in Schenectady and RPI in Troy, I took some courses toward degrees in both of those schools, preferring Union College because I could walk to Union College, and not preferring Troy and RPI because in the winter, you couldn’t get there; they only had one snow plow. Union had an excellent program there in Liberal Arts and in Business, and reasonably good in Engineering. But I got most of my training and background in metallurgy and ceramics and physical chemistry through General Electric, while getting my degree over a 10-year period, graduating in 1966 from Union College. I think Tom Hood, currently CEO of Southwall Technologies, graduated in 1970 from Union as well. It was a good experience, and I was able to take courses in optics and physical optics, which matched nicely with some of the work I was doing at GE at the time, looking into optical analog and diffraction from crystal models, an optical analog of real solid materials.
RS: What were some of the early projects that you first started with at GE Research Lab?
RLC: The first project at the Research Lab that I had been working on a vacuum casting furnace. What we tried to do was to cast a very small alloy sample—a few cc’s of metal—melted in vacuum and then cast into molds for making metallurgical metallographic specimens for structure studies and also for mechanical property studies. Mostly aluminum alloys. Aluminum alloys were of great interest at that time because of the work GE was doing in developing alloys and corrosion-resistant alloys for nuclear power reactors. So my job was to finish a project that had been started by someone else, and that was to make a little vacuum casting furnace, resistance-heated, pumped with a mercury diffusion pump—a little steel chamber—and it was an interesting project. That led to work in corrosion studies, in particular, aluminum. I worked on the growing of single-crystal aluminum. I grew a single-crystal sphere of aluminum, which was then electropolished making a perfect little brightly polished aluminum single-crystal sphere. We grew this using a Bridgeman method. Then after polishing it we oxidized it in a controlled environment to look at the patterns of oxide growth on the surface of the sphere and show that the rate at which the oxide grew was a function of the crystallographic orientation of the metal, which was quite interesting—at least it was to the people who understood it—and I was the guy who did the crystal growth, the sample preparation, and photography for the scientists who interpreted the results.
We also grew some bi-crystals of aluminum—once again, to study the oxidation growth rate on two different orientations. I did a lot of metallographic studies—polishing and finishing.
RS: Besides the sample preparation, and the characterization, there were also a lot of fundamental studies of understanding how these materials actually grew and compounded themselves together.
RS: Was that at GE, or did you have other collaborators?
RLC: That was all at GE. Mainly because the scientific infrastructure was so powerful there. You just walked down the hall and you could contact key people. And of course there were others who would come to visit the laboratories. They just had an extraordinary cast of characters there at that time. So you could talk to people about crystal structures, and you had people who were doing the diffraction analysis of crystals for orientation. So all the services were right there, under one roof. It was amazing. In addition, there was the M&C Lab, (Metallurgy and Ceramics Laboratory), which was a great big building fully dedicated to fabrication of materials. So it wasn’t just corrosion; there were a lot of other things going on in the materials side, not only for nuclear reactors, but for jet-engine blades and super alloys work that was going on at the time. I was also doing testing for evaluation of these superalloys. I wasn’t really doing anything except learning from all these other people. But we were testing alloys for strength and durability testing. Then I got involved in some of the fabrication of alloys like iron silicon, very high silicon-content iron alloys, because if you fabricated those correctly, you could get extraordinarily strong materials and very, very corrosion-resistant materials using extrusion methods—great, big extrusion machines that could extrude an ingot for example to take an ingot 4 inches in diameter, and extrude it down to a rod 1 inch. We would then swage and wire draw them down to tensile test specimens. In addition to the work that was going on there, there was also work in ceramics and slip-casting and various methods for fabricating ceramic materials because there were studies going on in ceramics, metals, and glasses. Bob Doremus who’s now retired from RPI, I believe, was one of the leaders in glass technology. And so there were studies in materials—just very intense studies.
RS: Now what role did vacuum and vacuum processing play in a lot of these studies?
RLC: Well, vacuum processing—one, making a fairly gas-free alloy for the study of these materials (aluminum, aluminum-lithium, aluminum-silicon, aluminum-palladium)—there were a bunch of materials. You could make a nice, clean melt out of it. Also, vacuum was used in the heat treating, vacuum heat treating. But at that point, that’s about it. There was no, well, thin-film coating of any sort. The only other vacuum was what they called "cathodic etching." Cathodic etching was nothing but sputtering. This was a method for preparing the surface of a metallographic specimen to study, and you could essentially sputter etch that surface using what they called "cathodic etching."
RS: To clean it, or…?
RLC: Yes, just to clean it. That’s all it was. And we didn’t have any vacuum chambers, vacuum processes. Well, there was vacuum, of course, going on there—and they were all glass vacuum systems. Mercury diffusion pumps, 30 liters per second (which was considered fast), and these were glass systems that we made ourselves. We’d do the glass blowing; we would make our own systems. And we’d do this for perhaps evaporating or refluxing some material, purifying some crystals by evaporating from one section of the glass chamber and condensing it in another; growing crystals of different materials, for a lot of work; trying to grow very high strength iron crystals because at that point people realized that single-crystal fibers and whiskers were very strong. So there was a thought of making bridges out of single-crystal iron whiskers. Of course, they never did. But that was one of the crazy things that were being done. The vacuum there at that point was limited to melting and casting and heat treatment. The rest of the heat-treating, of course, was in a controlled environment.
RS: So it was used as a tool, as a subset of trying to get the material properties you were looking for?
RLC: Exactly. It was just a tool. And especially—well, there was one other, we had little tube furnaces. We’d use platinum and palladium alloy wires, and you’d make tube furnaces out of these, wrapping the wires around an alumina tube. Then you could slip materials into quartz tubes and heat treat them in vacuum. It was done on a very small scale. Then, of course, there was the uranium work, because GE was also interested in making fuel cells. We were doing pack extrusion of uranium oxide. I had the unfortunate task of working with the uranium oxide in gloveboxes. I was also doing metallurgical work on uranium alloys—that is, depleted uranium—and swaging into wire drawing material. And I did discover that uranium burns. We started off with a rod of uranium—it was about an inch and a quarter in diameter and was about 2 feet long—that I was heat treating in an oven. I left the specific instructions for the heat treating oven to be turned off, and then after several hours, for the gas to be purged through, then we could shut it off in the morning. I came in in the morning, looked in the furnace, and the furnace was glowing red. The power was off. Someone had turned off the oven and they had turned off the gas all at the same time, so the uranium continued to burn. My uranium rod was reduced to a big pile of uranium oxide! Anyways, I guess I might have preferred to have a vacuum heating up there rather than gas! But these are some of the early experiences in materials.
After I worked in this field for a while I then got into some corrosion studies. I worked on the development of a dynamic corrosion tester, where we were pumping water at nearly critical temperatures, at very high pressures, through a flowing autoclave system to study the corrosion of aluminum in these very high-temperature water conditions. Then following that I had an opportunity to move to a different group within GE Research. I moved to studying kinetics and crystallization and materials studies with Dr. David Turnbull, who eventually went to Harvard University. He retired from Harvard several years ago. In that capacity of working at GE and getting long-range input from Turnbull, there were a lot of changes in my work. I started working in the glasses—phosphorous pentoxide glasses, glass formation, kinetics of crystallization of materials, ethylenes-polyethylenes. Then Turnbull got interested in organic semiconductors and thought organic semiconductors might be an interesting field to study because silicon studies were going on. There were alternatives to silicon, and some of them were like phthalocyanines, chloranyls, perylene-iodine, nitro benzene. These are doped semiconductors. They’re not exactly safe in terms of working with them. But one of the things that we needed to do was to make some very thin films of copper-phthalocyanine dyes. What I had to do was build a vacuum system to do that. So I built a little 7-liter-per-second mercury diffusion pump bell jar, a little bell jar about 6 inches in diameter and about a foot high, and worked on the resistance evaporation of these materials. That was my first experience, from scratch, building a small vacuum chamber. This was in the late 1950s.
RS: What was the substrate that you were coating?
RLC: Just coated on glass. We were looking at the electrical conductivity of copper-phthalocyanine when they were exposed to various dopants. They are electron transfer complexes. So what you could do was you could inject electrons into the molecule, and then they’d become conductive. Another area of interest was nickel-phosphorous. Nickel phosphorous, as you know, is a glassy metal. Glassy materials were the other area of interest that Turnbull had—he wanted to study these in every imaginative way. I was making thin films of nickel phosphorous, starting out with an electroless nickel, I think it was around 15% or so, I can’t remember the exact amount. It was electrolessly plated onto a tungsten filament. Then you’d heat this up and evaporate it. The material would condense as a thin film, which was totally amorphous, on glass. Then we heated that up until it crystallized, and looked at the glass crystal transition. Just generally trying to look at the kinetics of this to determine how fast it would go, and this again requires vacuum, a vacuum pump. I was doing a number of other things, looking at heat treating of aluminum alloys and then the kinetics of formation of the small particles (copper, silicon, magnesium) within an aluminum matrix and the effect of the precipitation of these little particles on the electrical resistivity of the aluminum. This was just pure kinetics study. But it was kind of interesting. Then polyethylene, looking at the glass formation of polyethylene and the crystallization of tiny particles of polyethylene with the objective of determining the rate at which crystallization would occur at temperatures well below the point where it should crystallize. Because particles are so small, there are no defects in them. So if there’s no defect, there’s nothing to nucleate crystallization. They would super-cool way, way, way down, and then suddenly it would freeze. Then you’d detect the freezing by optical methods, looking at the droplet, using polarized light and using crossed polars, you could see the Maltese cross appear when the little drop of polyethylene froze, then you could tell it was crystallized. So I took seemingly infinite numbers of pictures through microscopes. I spent days and days and days with microscopes, using a little heated stage that I designed and built that could allow me to heat the sample up and cool it down and heat it up and cool it down, and melt and freeze these little tiny droplets, which I dispersed in a detergent-type solution. It was interesting work because the challenge was to study something in a small-particle state without affecting it chemically by the material in which the particles were dispersed.
RS: What was your favorite microscope?
RLC: I had two. I had a Leitz-Wetzlar and a Zeiss. And they were both really good. They were both fine. They were sitting right next to each other. So I would just take photos using the old black-and-white Polaroid and 35-mm cameras.
Then we realized we had to do some vacuum work on a larger scale. So I scouted around and found this old, ancient Kinney vacuum bell jar. It was an oil pumped, 6-inch diffusion pump, glass jar. It wasn’t truly ancient, it was just sort of set aside. But it was one of the early versions. It had been located at one of the GE operating divisions; they just shut it down. So I bought it and brought it into the Research Laboratory. Oddly enough, here we were—the most prestigious research laboratory in the world—and I had the only vacuum bell jar in the whole plant. It’s odd that the research group, being so advanced in every other way, didn’t have any of the industrial modern equipment. Soon after that the people in another group, Metallurgy and Ceramics Group, were doing electron beam evaporation of metals—niobium in particular—brought in another chamber. Now things got interesting. We started doing electron beam evaporation of niobium as a ground plane for a cryogenic computer onto which lead-tin circuits were built. So lead-tin would be the superconductor. These were thought to be potentially high-speed computers of the future and would be thin-film computers. Work proceeded on that for a while, but it really didn’t lead anywhere. Crystal growth on the lead films proved to be a show stopper. There were too many other things that were fruitful, and the technology was coming along pretty fast.
RS: So with your early bell jar, did people bring you projects now that they knew that you had that tool—coat this for me, or can you put this in the system?
RLC: Right. Well, one of the first things was to come up with a hard coat for Lexan. People in the Pittsfield GE Division wanted somebody to work on hard coating Lexan. So I worked on the evaporation of silicon monoxide. The trick was to evaporate it without heating up the Lexan. I came up with a tricky little source design that minimized the radiation heat load on the plastic. Then I found out that it didn’t make any difference; the films didn’t stick anyway. So then I started looking in a reference by Hass, Vacuum Sciences.
RS: The Fort Belvoir Reports?
RLC: Yes, in the glow discharge section. So I needed to come up with some ideas for oxygen glow discharge of the surface of plastic and did just that. In addition, I continued to glow discharge during the evaporation. I got the films to stick pretty well, actually. Because of the oxygen, I was able to change the refractive index from 1.98 down to about 1.5, and got a good match to the Lexan, so you couldn’t see the coating. I put it in my environmental tester. It was January, and so what I did was open the window and set it out on the windowsill in the cold. Then I would bring it in and drop it in warm water. If it didn’t pop off, I felt very successful. The rest of the testing was a falling-sand test; the loss of gloss was specified. So I ran those tests on it, looking for scratch-resistance. And it was a really good hard coating. It was just not economically feasible at that time to coat Lexan that way. The Lexan folks went to the chemical method instead. But we did get a patent on the source and the hard coating.
RS: Now was this your first foray into optical coating?
RS: Control of both the mechanical and the optical matching?
RLC: Then the next one was a similar one, but not exactly optical. I found it through a reference in the optical journals, but I didn’t apply it that way. We were working on hybrid circuits, and a lot of work was going into electrical circuits on ceramic substrates and glazed ceramic substrates. One of those needs for those was a really good capacitor that had a good high dielectric constant, and yet was pinhole-free. So I came up with this idea of making a capacitor with aluminum oxide, aluminum, and titanium. I took aluminum and evaporated some titanium oxide on top of it, then I anodized this, growing an aluminum oxide film underneath the titanium dioxide. And it was absolutely, totally pinhole-free. And that’s a method that you find in Haas, which was an optical method, but it occurred to me that it would be a great way to make a capacitor, and it worked out really well. So that was applied to some work by GE in their Light Military Division, located in Utica, New York. I was continually looking at applications within the operating units, the world of real people and real applications, and trying to direct some of the work in the laboratory toward addressing the problems that the people had.
RS: So did you go out and visit the different facilities, or did you look at the GE Bulletins to find out what were their problems? How did you find these applications so that you could address them?
RLC: Well, the way GE was organized then was that the research was supported by assessing the operating divisions. So the operating divisions actually paid money to the research group. And because of that, the operating divisions and their advanced engineering groups were constantly visiting the labs, trying to find answers to the questions. So the questions were just popping in every day. We had a little method for transferring this information around, a paper method. And I’d see these things and I’d snag them, then I’d contact people. I’d rarely travel. Most of the time the operating division wanted to visit the labs because the labs were such a prestigious operation. Nobody wanted to go to Bloomington, Illinois, or anything like that, or Waynesboro, Virginia, or wherever the heck that was. They just didn’t want to go there. This was a wonderful opportunity for us. One of the projects was to put some metal coatings on reed switches in order that the reed switch could operate many times without sticking together. Sputtering was used to deposit a thin coating of palladium on the switches to accomplish their objective. At this point it became evident sputtering was an interesting technology and some work should be done at the labs. I’d heard about Davidse and Maissell at IBM, who developed something called RF sputtering. So I got interested in that right away. I threw away the bell jar from my vacuum system and put on a big glass cross, and came up with the world’s ugliest vacuum system. I got a police-car radio and a Heath Kit linear amplifier, and a ham radio operator who helped me build a matching network and built a radio-frequency sputtering system. It was just a little planar diode, but I superimposed a magnetic field on it and a cross-magnetic field using HelmHoltz coils. Everything was custom built there at the GE labs. We even made the ceramics. Everything was made there—metal spinning, ground shields, etc. I started playing around with sputtering dielectric materials and strontium titanate was one of them, which had a very high refractive index, a very high dielectric constant material. And these proved to be just extraordinary capacitor films. Without realizing it, of course, you’re making very high index coatings now considered valuable in the building of antireflective coatings. But I didn’t care. They were pretty. That is, they exhibited very brightly colored interference colors. And then realized that if you put a specimen on the surface of the cathode, you could sputter-etch it. I was working with a guy by the name of Marv Garfinckel, the inventor of the self-registered gate technology. That was the technology that sort of revolutionized integrated circuit manufacturing, where you put down a gate material and then you used that as the mask. We sputter etched silicon using that method. His technology used a molybdenum gate, but later it became a silicon gate technology, a CVD method to deposit silicon. The molybdenum never really worked, because they couldn’t find a clean enough way to make moly coatings without injecting sodium and fast states into the oxides. I was doing sputter etching, some of the very first sputter etching anywhere. So we were just exploring. It was just always like this. Every day was an exploration; it was always fun to come to work because you discovered something new practically every day.
RS: How big a setup was this sputter etch system?
RLC: It was about 4 inches in diameter. I have a picture of it. You remember the picture on my office wall, a glass cross? It’s starting to fade now. But I think I have another one that isn’t so affected by the light.
RS: I’m wondering if it’s possible for me to get a copy of that?
RLC: Oh yes. I’ll do that. I think I have one on a slightly smaller scale. I could scan it.
RS: They had photo shops at GE?
RLC: Oh, yes. They had the photo labs and photo operations. It ended up on the cover of one of their monthly journals at GE Research, this beautiful color plasma.
RS: It’s a long way from your cathodic surface prep, cathodic etching of your metallurgical samples.
RLC: Oh, yes. A long way.
RS: About this time you were looking to solve other problems. You had this sputtering technology. Were you looking at potentially moving? Or how did you leave GE?
RLC: Oh, I didn’t want to. You see, I was getting my degree and I figured that when I got my degree in 1966 I would plan that I had done so much work for the outside industries in the operating divisions that I wanted to stay right in Schenectady, live there, and stay at the lab and lead an activity toward developing a better relationship between the labs and the operating divisions. Unfortunately, because all I had was a bachelor’s degree, the lab was determined to have Ph.D.s only. And that’s understandable. But I explained that I actually had more publications than some of the Ph.D.s that were there, and why shouldn’t I just be put on the staff? And they promised that in three years or so I might be able to do that. It was evident that it wasn’t going to happen. So instead I went out and interviewed at the various operating divisions—the magnetic tape and disk division down in Phoenix; the Sunnyvale aerospace group; the Cleveland lamp division; Syracuse semiconductors. And I decided Syracuse semiconductors was the best fit. I’d done a lot of work with them, they had a strong program toward making Shottky barrier diodes with sputtered tungsten and tungsten silver, which I eventually worked on, and then also the work in Syracuse toward the development of integrated circuits using vacuum methods. So that’s where the vacuum technology really took off. It was at Syracuse where we needed to develop some pretty sophisticated systems. I went out and I bought some equipment and I and one other guy developed—it was the other guy’s idea, not mine—Brian Corey was the gentleman’s name. He eventually ended up at Tektronix in Oregon. But his idea was to come up with a sputter cathode to drop through the hole of a chamber wall and have a big ring insulator. So that was his concept. I reduced it to practice, came up with a method for it, and we built the very first of the cathodes that went through the wall of a chamber. Everybody in the industry copied it after that. And we had it built at Veeco-Andarin, Sunnyvale. So of course I had a trip to California. I took that trip and decided, "I think I’d rather live here than in Syracuse." I had an opportunity to move to Berkeley and work for a company called Temescal. Temescal is the Indian name for a little hut, a little steam teepee sort of thing. Basically a steam room. That’s the reason they called the company "Temescal," because it was located under the Bay Bridge and they were doing vacuum melting and processing, vacuum evaporation—mostly melting and heat-treating. And it was like a furnace there. So that’s the reason they named it Temescal. Hugh Smith started it out of the Livermore Labs. I joined them with the specific goal of getting Temescal into the sputtering business and did so with the development of RF sputtering there and built some equipment for RF sputtering. But they never took it too seriously because the business was so strong in electron beam evaporation until 1972 when along came an opportunity to fund a program in Colorado for the development of a high-rate sputter cathode. All of this was with the expectation that this high-rate sputter cathode would allow you to deposit metal films at high rates over very large areas. The idea was to replace the electron beam evaporation methods for glass coating with a sputtering source. But the sputter methods were too darned slow. There was this guy up there at VTA called Ted Van Vorous, and he claimed to know how to do this. All we had to do was fund a small program, and he’d come up with cathode that sputtered copper in 30 microns per minute. And sure enough, he did. But it was John Chapin who developed the method. In December 1972, I was managing the program and extracting a monthly report from John. Each month I’d go up and I’d make sure that his notebook was kept up to date and his entries were proper, and that his laboratory notebook and his reports were consistent.
RS: Knowing John, I bet you had to write some of those reports!
RLC: Well, John was good, but I helped him. He didn’t want to write them. John actually did a wonderful job. He did a great job of his work in the notebooks. He was just a little bit behind. So all we did was get caught up. During the project I got a call from Dave Robertson, who was our sales guy in sputtering. Dave was actually the one who saw the first sputtering test. He said, "Bob, you’d better get up here. I think John’s onto something." So I got on a plane and went straight up to Boulder, and John showed me his invention, which was a fairly large copper plate with a magnetic field produced by coils of wire and a piece of steel channel behind it, within a clear glass bell jar. And he’d hooked this thing up to a telephone pole transformer backward so that he had lots of current at fairly low voltage and fired it up. It was an AC system; it wasn’t DC sputtering—it was AC. Because the anode was very small and the cathode was very large, it had the tendency to act like a DC cathode. He fired it up and whoosh, this bell jar was copper-colored in microseconds. That was impressive. So we ran a couple of tests, got some samples and glass slides, ran Sloan DekTak thickness measurements and got some thickness data. And sure enough, he was achieving the goal.
RS: He met the 30 microns per minute?
RLC: Oh yes, easily. We also went on to do some titanium work, titanium oxide with this cathode. It didn’t work too well because the arcs sparked so much and destabilized the process. After a couple of days of further lab work and then a day of skiing, I jumped on a plane and flew back to California. While I was on the plane I sketched a drawing of what I thought would be a first magnetron cathode. It was a very simple system and I gave this sketch to Richard Derrick and Nick Tsujimoto at Temescal, and they built it, fairly quickly. By March 1973 we had built it, put it into a vacuum chamber, and sold it to RCA. It was used for sputter-coating copper onto their first video disc, called a "Selectavision" product. That was the first reduction to practice that we were aware of. We put together a patent application. But it turned out that there was a little business altercation between Temescal and VTA as to who should really own that patent. Because of this, it delayed the application and the ultimate issuance of that patent. In the meanwhile, Sloan Corp. and John Corbani invented the same thing—literally, the planar magnetron, filed and had the patent issued. So now his patent was issued and our patent—Temescal and VTA—was still being prosecuted. But it appeared that we may have had the early date. So we dragged out John’s patent notebook and my signature at the bottom of the page, and sure enough, in December 1972 we proved to the satisfaction of the courts that we were four months ahead of the John Corbani invention that was his discovery. As a result, the patent went to Temescal, not to Sloan. All the more reason to keep a good notebook—I mean, it really is important. There were some rights given to VTA; they could manufacture cathodes in a small size. It worked out well for both companies. Temescal went on to build some very large cathodes. Andy Dubois, who was the engineering manager, and Nick Tsujimoto, one of the engineers in Berkely and I worked on the development of a very large cathode. We came up with two or three designs until we found one that we really liked. That became the workhorse for Temescal for many, many years. Alnico magnets were used in the industrial designs. We, of course, built cathodes with electromagnets during the development phases. The main driver for commercializing the technology turned out to be the architectural glass coating.
RS: How long did it take you to scale from this small one with RCA to larger sizes?
RLC: Just about three months. We moved quickly. We went to 54 inches first, then onto the large size—the very large size. The problem, of course, was scaleup. No, it was actually a little bit longer than that; it was more like about nine months, because we had to go through a period where we struggled with power supplies, trying to make power supplies large enough to drive the current requirements of the large cathodes.
RLC: Did you invent your own power supplies, or did you purchase them from someone?
RS: No, we had to build them. Temescal was very good at building power supplies, and so they put their electrical people to work on it and they came up with something called the SP 30 a 30 kW supply.
RS: How many cathodes were on this first machine?
RLC: Just one. And the first application was to sputter some metal—I think it was stainless steel; it might not have been, I can’t remember—onto glass for a moon roof for an automobile. So the first application actually was for automotive glazing, and they called it a moon roof. The idea was to cut down a little bit on the sunlight that came through the glass. It worked quite well.
RS: 1973, 1974?
RLC: 1974. Let’s see—no, it was 1973. Because I then got a big chunk of this glass and put it in my house out in Walnut Creek, where it stayed there for a good long time until it eventually sort of faded away. It was exterior glazing. We even worked on some of the reactive sputtering fairly early on and looked at titanium nitrides. Then all down the side of the Temescal plant we had glass windows—all different colors—sputtered to demonstrate the various colors you could get with interference coatings on absorbing, dark, stainless steel, titanium, inconel layers. This work led eventually to low-e coatings, but this was to be a long way off.
RS: The early implementations were DC?
RLC: All DC.
RS: All DC sputtering?
RLC: It was evident that the problem would be arcing when you were doing reactive sputtering. So in 1974–1975, I worked with Tsujimoto and a guy by the name of Sig Andresen—Sig built a power supply—to prove that high frequency and mid-frequency sputtering would solve the arcing problem, and we got a patent on that. It really never became a product until the dual-magnetron came out many years later. But it was used, on and off, as a way of reducing arcing. It became quite evident that by mixing DC and AC you could keep the arcing down. A lot of power-supply companies did, in fact, develop products that used that idea.
RS: So if the focus at Temescal was on these refining applications of metals, implementations and e-beam smelting, vacuum process, and you had this one cathode system for architectural, the RCA copper system, did you start to grow the area? Did business start to develop around this sputtering technology?
RLC: It didn’t until Al Grubb, who had been the Project Manager for the RCA project decided to leave RCA—he had originally worked for Libbey-Owens-Ford (LOF). He came to Temescal, and he came so that he could direct the development of the architectural glazing business. Until that happened, we were just playing around, doing a little semiconductor here, a little silicon there, but nothing very serious. Then things really took off. Under Al’s direction he grew a real business. He had a good division, and we built in-line sputtering machines and demonstrated coating at a small-scale for many other applications, including the silver titanium oxide, which became the basis for a heat mirror coating. We did the early work for SunTek Research Associates, which later was to become Southwall. I did that with a glass coater, sheets of material that we put through the machine just to prove that you could make these coatings.
RS: Did they come to you with the idea and say, "How can you help us fabricate this on an industrial scale?"
RS: And Temescal said, "You’re the world’s experts, help us to take this design."
RLC: Exactly. They had the idea of metal dielectric. I’d never heard of it. As it turns out, I heard of it quite soon after that because it was the General Mills Corporation and their consultants who wanted to make a clear heat-reflecting window for a kitchen oven.
RS: We’re just talking about some of the early heat mirror-type coatings. You indicated that SunTech had approached Temescal for interest in trying to produce a coating of this type, and you said you had just heard it from someone else.
RLC: Now this was Bob Amos as it turns out. And there was one other guy, but I’ve completely forgotten his name. The whole point was that people had been using thin gold films as heat reflecting heat mirrors on furnaces for a long time. But these folks wanted to make an oven window for commercial and household ovens. It would be clear and heat-reflective. So we made those coatings for them, and they were, in fact, dielectric, metal, dielectric coatings, with titanium oxide-silver-titanium oxide. It was this group that brought that to my attention in the early 1970s. I had no idea that MIT was doing the same thing; John C. C. Fann was working on the same thing as well as Day Chahroudi of the MIT Architectural Dept. was "inventing" the same thing. These guys all conceived it all at the same time, so they probably all knew of each other’s work. But the very first application that I was aware of was this one. My first exposure to that—and I missed this earlier in our conversation—was actually through Ted Van Vorous of VTA. So my first knowledge of it was in 1972. And this was in maybe late summer of 1972. So that was about the same time that John Fann of MIT was thinking he was inventing it for the first time. So there’s no question—had enough work gone into this, there would have been a discovery that others had invented this before MIT, and MIT may never have gotten the patent. But nevertheless, that’s all history. And the patent has expired.
RS: Day Chahroudi was with which company?
RLC: Day was with MIT in the Architectural Division. Of course, we had the oil crisis and everybody was sick and tired of spending all this money on oil. We needed to save energy and there should be a way to do this. Day thought that an insulating window would be the way to go. He and John Brooks, Shawn Wellesley Miller, Beth and Blair Hamilton, Charlie Tilford—these were the people back at MIT who started a company called Energy Research Associates. But they started a small company and got some funding toward the development of a transparent heat-reflecting film on plastic. The idea was to stretch the plastic in an air space, then you got a substantial improvement in the insulating value of a window. And that became the basis for the Southwall heat-mirror window. After many years of work we put it into production and practice, and solving the seal problems and all this other stuff. But the initial coating was titanium oxide-silver-titanium oxide. So when I made the first samples for SunTek, I met Day, John Brooks, and others, and they were sort of an interesting bunch of people. I wasn’t sure I wanted anything to do with them, but I thought it was a neat idea. Then they hired this guy, Mel Hodge. He had originally been the Research Director for the Lockheed R&D Labs. He decided to become President of this little company because he saw something there. I thought, "Oh, this is different." And I was ready to make a change. I thought it would be kind of fun to get into a small company. So in the summer of 1978, I joined SunTek. I got over there and now they opened the doors and all the little secrets came out, and it was very primitive. They had built a small web coater and had an extraordinarily primitive magnetron cathode, which I quickly disposed of. I put in some power supplies and got some equipment and got it working. But they had already made a pretty nice-looking film: clear, heat-reflective film. Beautiful. So we made a window and it looked great. I asked how well did it hold up in the sunlight? And the answer was that nobody knew. So I put it on the roof and it lasted about three weeks; it all turned yellow. And it turns out that the titanium oxide-silver system wasn’t particularly durable. The silver denucleated and it kind of fell apart in sunlight. So it wasn’t long before I switched from titanium oxide to indium oxide. I was well aware that the system—bismuth oxide silver- bismuth oxide would produce a well-nucleated silver layer. And it was right there in these old 1950s technology. We well knew that bismuth oxide was an ideal nucleating surface. But bismuth oxide was sort of water-soluble and it didn’t look like a good candidate to be part of the heat-mirror stack. But indium oxide did. Easy-to-make targets—you could cast the individual targets easily. It was a matter of just a few weeks and we made some indium oxide-silver-indium oxide films, and they were just beautiful. They didn’t deteriorate at all in the sun. We built a little homemade weatherometer, tested everything, and the indium oxide system became the basis for the design of a new vacuum chamber. This was a large sputter coater that Leybold-Heraeus built for us. We kind of designed it and sketched it out, then they came up with some ideas, built the cathode, power supplies, winding system, and in the 1979-1980 time frame, built this very first multi-cathode sputtering machine for the manufacture of indium oxide-silver-indium oxide. But we didn’t want to tell them that indium oxide was the material. So we left them trying to make to make a good titanium oxide-silver-titanium oxide, keeping the indium oxide a big secret. Well, titanium oxide was extremely hard to sputter on a very large DC cathode because of arcing and sparking and disappearing anodes, which nobody knew about at that time. Indium oxide, because the oxide was more conductive, was less susceptible to these problems and so as soon as we got the machine out of Germany and back in Palo Alto, we quickly put indium targets in and off we went.
RS: Now is Southwall an entity at this point, or was it still SunTek?
RLC: Southwall was an entity in January 1979, incorporated in January 1979. SunTek continued to exist until October 19, 1980, at which point SunTek dissolved, and the people from SunTek became part of Southwall. We went on from there. So we consolidated everything in Palo Alto. Then it wasn’t long before we realized one machine wasn’t going to be enough, and it was too limited anyway. So leveraging input money from 3M and others, we bought a second machine—bought this in England from the General Engineering Group—with a much better design in terms of more pumping, better handling of the cathodes, more sophisticated winding system for plastic film. It was a success. And that became a manufacturing machine that ran for about 10 years, making heat mirror films before it was then converted into a machine for other applications, antireflective coatings. The trick was the sputtering was difficult, but the web handling and the film handling proved to be really the toughest job of all. As we tried to wind this film through the machine, we just couldn’t seem to do it without wrinkling it. We seriously thought that we had a business failure in the early 1980s—that we would never be able to make a film for a window that looked good because of what we called "heat creases." It was a combination of learning how to handle the film and getting better-quality polyester, that we ultimately resolved that problem. It remains an occasional problem, but not a fundamental one. And it did appear to be a fundamental. So here we thought we could master the vacuum technology, the sputtering, the control of materials, the environmental issues, the making of a window—we couldn’t make this film without wrinkling it. Very frustrating.
RS: And that substrate was polyester-based, then?
RLC: Yes, polyethylene teraphthalate. And of course, it was UV-stabilized. That was made by ICI in England, and later, of course, with all the consolidation of ICI, Teijin, and DuPont. But at that time there were two or three different manufacturers: DuPont, Teijin, and ICI. We stuck with ICI for a long time. They had a good film and yet their equipment was a little bit old and it proved to be uneconomical for them to continue with the business, so we had to develop a new source. And all this time trying to develop a polyester that would be more durable in exposure to ultraviolet, because what happens is, polyester embrittles and it cracks and it fractures. And we had some windows that had done that—the plastic film had failed up in Colorado and the high country. So we were aware that we had a potential problem. So that was quickly resolved with the introduction of the slightly thicker—but far more durable, UV-stable—polyester, developed and invented by Teijin. And that continues to be the best film that I’m aware of.
RS: After this heat-mirror for architectural, there were forays into automotive?
RLC: What happened there was it was evident—a guy by the name of Dennis Hollars worked with Steve Meyer at Southwall, and with Tom Hood. There was a lot of work going on in the Defense Lab. It was quite evident that Fabry-Perot optical structures, multilayers—where you repeatedly deposited the dielectrics and metals to get narrow-pass, sharp-cutoff, optical stacks—had great value. And Dennis Hollars said, "We should just do repeated depositions of the heat mirror stack." It was Steve Meyer and Tom Hood, however, who realized that if you did this in a certain way that you could make a very clear and very effective, near-infrared, reflector-visible transmitter. If you designed it correctly, you could design it such that it would be neutral, colorless, literally invisible—if you laminated that within a window; that is, within the PVB polyvinyl buteral, the glue that holds a car windshield together. You put this film in between that, it would disappear and it would give you a heat-reflecting laminated window. That was in 1985, in combination with some work with LOF. It was several years before it became a real product. But it was evident that it would be the windshield or the windscreen of choice for automobiles, without a doubt. An alternative technology is to sputter directly on glass. And that continues to be an alternative. So sputtering has proliferated from that 1972 invention into coating glass and coating plastic and many other substrates as well. It’s one of the most ubiquitous technologies that you can imagine. I mean, it’s pretty rare that you look at a new, high-tech product of any sort and there isn’t some kind of sputtered coating in it somewhere. It’s interesting to note that one of the largest single applications for magnetron sputtering is in the coating of CDs, the very reason for the development of magnetron sputtering in the first place.
RS: Bob, what do you think that this sputtering technology will go with certain applications? Displays?
RLC: Obviously the new technologies will come along and eclipse this at some point. But I think that sputtering in general has a very long, bright future because of the ease with which you can control the processes and the flexibility of materials and the creative genius of the people who are finding ways to use it. There are people still developing new methods of plasma refinement, efficient use of target materials. There are still a lot of challenges. But it does seem to keep growing, and if you look at the miniaturization and the micro-miniaturization of semiconductors, they are finding ways now of embossing circuits in silicon so that densities can be further increased. There’s a growing technology in chemical vapor deposition and plasma-enhanced CVD that looks pretty exciting, but the ability to deposit atom by atom with sputtering—there are still major benefits there.
RS: Bob, one other thing you’ve speculated on from time to time is the viability of the stand-alone sputtering business entity. The strengths of rigid technologies, independent operation as opposed to focused industrial process. What do you think about the fact of this?
RLC: I think it’s been about 14 years since I’ve looked at that for the Board of Directors at Southwall, saying that as a stand-alone business, sputtering as a service is not viable. Sputtering, I believe, will always be a very important step in some manufacturing process. That sputtering often, while it can prove to be rather expensive, can allow you to make films, coatings, devices quickly and easily—that you can demonstrate and prove the viability of some new product fairly easily. And yet the first thing you do is say, "Gee, can I do this coating any cheaper somehow?" So often it becomes a bridge, as a bridge technology. In the defense business, for example, that was the case. There were a lot of thin-film coatings used for special shielding applications that could be made easily, quickly by thin-film sputter deposition. But the cost was so darned costly that they quickly moved to other technologies to make a good-enough coating for a lot less money. So there are cases like that. You know, I think that the bridge technology issue is not as big an issue anymore. That sputtering becomes the ideal solution—or the technology of choice—rather than the alternative is more the norm. So maybe in the late 1980s there was more bridging, bridge technology was sputtering to prove a point, then you go off and do it some other way. Now I think that it’s becoming the very method that you prefer. I mean, just take a look at the compact discs and the DVDs. You wouldn’t have these things if it weren’t for the fact that to sputter—I don’t know what the time is now—but I think it was like 2/10 of a second; some tenths of a second spent in front of the sputter cathode. Whoosh! And you have the coating. It just wouldn’t have happened. And it is interesting to think that that was the first application of the sputtering, and is now one of the biggest, I believe. I don’t know if anyone is evaporating coatings onto CDs. Probably not. Then one of the other first applications of sputtering was architectural glass. That still remains one of the biggest single applications. And the other first—which I failed to mention—it was called "low-energy sputtering." Magnetron sputtering generates very few fast states in the oxide due to damage of the dielectric. Electron beam evaporation damaged the dielectrics; sputtering didn’t. And so semiconductors benefited right away from sputtering of aluminum and aluminum-silicon because it didn’t—that’s the gate metallization—it didn’t damage the oxide. So it made a better semiconductor right away. Those were the three firsts. And they still remain some of the primary applications. Of course, all of the new display technology is the rest of the story. But that is antireflective coatings and materials for displays—there’s no question that sputtering is a major participant in all that. I think that optical light-emitting diodes, for example, may prove to be one of the very best technologies that have come along in a long time. I think that’s going to require sputtered films, almost certainly. And barrier coatings will be part of that, and it could be that sputtering will be part of that, but I’m not sure. Physical vapor deposition doesn’t really lend itself to effective barriers for OLEDs.
RS: What are the new applications?
RLC: We can only imagine. I think display technologies. And you know, every day you can just let your mind wander around a bit, and you could probably come up with a couple of new applications. If you take a look at the AIMCAL memos that come out—you get them via e-mail, perhaps—there are people who are always looking for new methods of putting down coatings, anti-counterfeiting for example. I think that’s done e-beam right now, but it could well be that sputter coating will do it.
RS: Thank you Bob, I appreciate your time.