Oral History Interview with Collin H. Alexander (CHA)
Conducted by Dale E. Morton (DEM) on April 16, 2002

Collin H. Alexander

DEM: I’m Dale Morton, I’m here to interview Collin Alexander, who by now, we have to really regard as a fossil in the industry. Collin?

CHA: Correct.

DEM: Maybe we should start with your educational background. Where is it that you went to school, and what did you study?

CHA: Well, I went to school at Albion High School, Albion, Michigan; Washington-Gardener High School. I got my bachelor’s degree at Alma College in 1937 in a double major of chemistry and physics; and after that I had the privilege of going to a little engineering school and study chemical engineering in the graduate school, and it’s called MIT. I was very lucky to be able to get into this situation, because not everybody has that opportunity. But I did have a chance to have a job working in downtown Boston during this period of time. Now, what were my career aspirations? What I really wanted to do was to become a catalytic chemist in order to make high-octane gasoline, because the war was coming and it was needed for higher compression airplane engines, etc., and I had a cousin who was active in that, and I’d helped him do calculations in the Summer of 1936 earning some money this way. That’s what I really wanted to do.

DEM: Well, what did you do first? Where did you go?

CHA: Well, I went, I had an opportunity to go to a refinery in St. Louis, Illinois, and talk to the manager, and everything looked good, because I had what he needed. Then he asked me, ‘Do you have any relatives in the company, and if you say you don’t and you do, you’re eligible to be fired.’ I said, ‘I have a cousin by the name of Dana Mellett.’ He picks up the phone, calls Dana Mellett, and I can hear the conversation. He says, ‘This guy Alexander is a good man; I’d like to hire him.’ And my cousin comes right back and says, ‘You hire him, you’re fired. I don’t want any relatives in the company that I’m working in.’ And he was a pretty a high-class guy in the company; he later ended up as Manager of everything east of the Atlantic. That’s what I wanted to do, but it didn’t go. So I finally got a job with a paper company in Niagara Falls. This was fine. I really wasn’t wildly enthusiastic about it because I wanted to do experimental work. But I was given jobs that were interesting, and one of them was to tell them how many cords of wood in a pile. The pile was just wood dumped off a conveyor. It was a couple of hundred feet high. It took me a couple of days—I figured out how to convert jumbled wood into cord wood. I turned in the information, and that was fine. After a while, the information came through that there was a shortage in the wood that was in the pile. The man that I was working for was responsible for it, and the Plant Manager suggested that I leave town. He would do anything possible for me to get a job, and he would recommend me. But he was worried about this guy who would be going to jail and his friends might take care of me in a paper mill. And that’s what happened. That was my first job in a paper mill.

DEM: What did you do the next step? When was this, now?

CHA: This was in 1941. I went to the American Chem Society meeting in Atlantic City and interviewed there. What I had done previously to try to get ready for this career in catalytic converting, you have to understand something about surfaces. I was living in Chicago for a while, and I took a course at Armour Institute on catalytic converting and then it was suggested that I learn something about surface analysis, what there was of it. I took a job at—or rather, a night school thing and I learned how to make—well, the job was to analyze a gas. And the way to do that was to make it a diffusion pump, a mercury glass diffusion pump, then make a Macleod gauge, and take an unknown gas sample, and weigh it, and empty and full, and that was what amounted to knowledge of McLeod gauge and diffusion pump. And apparently Bausch & Lomb realized—and I found this out after they hired me—that I was the only one who at the American Chem Society meeting that they interviewed who knew what those words were, let alone what they meant. And that was the beginning of my work in vacuum coating at Bausch & Lomb in the Fall of 1941.

DEM: What was it that you first did when you arrived at Bausch & Lomb?

CHA: There were some jobs needed. I’ll have to describe to you what they had in the lines of equipment at that point in time. They had four 18-inch systems, two of which were pumped with glass-diffusion pumps, and two with metal-diffusion pumps, with no valves. And what they were doing was, they were doing two things: They were coating anti-reflection films for ‘Gone With the Wind,’ for their projection lenses, and two, they were coating aluminum mirrors, evaporating chromium onto the mirror downward, so that they could get complete coverage. And that’s where they were. And some jobs were assigned to me. One of them was to make reticles for rangefinders or heightfinders. This involved evaporating cadmium through a 10 micron line slot and others of varying shapes and sizes. This was developed originally by etching the design into glass, but it took a day to measure it on a microscope, and they were unable to make mirror images of the marks that were required. This was background that came from the Battle of Jutland. The Germans had the ability to do this. So we were trying to make these depth marks on a piece of glass. We would evaporate cadmium and we learned a couple of things. First, the cadmium would go down, and sometimes it wouldn’t deposit at all! And then other times, if it did deposit, it would come down about 13 to 14 microns wide. What we did, then, was to quit doing it that way. Other people got Charlie Baer to get involved in using photolithography to make it, and he went along on that direction, and I went in the direction of anti-reflection films and all the other paraphernalia, including gradient-density goggles, wide-angle field flatteners for exposure on wide-angle lenses, because the lenses had a tendency to be overexposed in the center and not on the side. And from this we learned, and put a booster pump in behind our diffusion pumps, and we had Hills-McCanna valves and 4-inch horizontal fractionating pumps. The idea being that, if you had a booster behind it, contaminants built into the diffusion pump would be pulled out at a lower pressure. And it would run at about a micron or two. So we did this, and lo and behold, we got excellent films and a very rapid cycle. We could pump down and make a coating in about 15 minutes, which was unbelievably fast at that time. And the coatings were unbelievably hard. They were much more durable than those that had been pumped overnight. The lesson I think we learned is, that the length of time between a glow discharge and the time you coat, if it can be real short—and it is short—you can get a good coating. If you have a long period of time, you have too many monolayers of residuals striking the substrate and you have a contaminated piece of substrate. So that’s so much for what we did in terms of Inconel.

We also did a lot of things for the military. And this included gradient-density goggles. These were particularly useful in the African campaign because of the intense sunlight down there for the pilots, trying to look from intense sun outside to a dashboard that was not that brightly lit, by comparison. By having this non-uniform coating on there, it would cut out the sun to a great extent and they would be able to see. And this we did in jig time and rapidly. As time went on—and this is about 1943 we were doing this—at a later time we converted that and did a large system, but we made very sure that our distance between our source and our substrate was always the same, it was always the same—it was 7 inches. And we measured them electrically and optically during deposition.

CHA: And from this multiple-coating procedure we went in the direction of having to clean glass without heating it. The problem on that arose for optical contacts. Now, optical contacts in a rangefinder are two different types of glass, slightly wedged, then put together like Johanssen blocks, and they held together physically without any glue or anything like that. But they needed to have an anti-reflection film. So with **** Rice and I, we put together a procedure of transformer and emitter, very much like a rectifier vacuum tube, which was pretty common in those days. But a vacuum tube rectifier, the purpose was to be as efficient as possible, and generate no more heat than possible. That meant there was a low-voltage drop between the tungsten emitter and the plate. Well, we made our optics the plate and put a high voltage on it, but severly limited the number of electrons that would be available. And this gave us a high-voltage drop, that we had a voltage drop of somewhere between 5,000 and 7,000 volts. And everything was fine; it heated just the surface of the glass, got a hard magnesium fluoride coating, and everything was just wonderful in October. But the transformer we had, we had picked up off a junkyard. We knew it was a high voltage; we didn’t know how much, and during the war you took anything you could lay your hands on and used it. Well, the electrician came by and said, ‘You must ground this and you do it like this.’ So he grounded it, turned up the Variac, and blew the transformer. What a mess that was in January! Well, it took us about May—and this is May of 1944—before we could find a transformer that looked suitable. We used an instrument potential transformer and ran it backward. An instrument potential transformer is the thing that they put on high voltage, then they put a meter at 110 volts, and they can measure what’s going on, only we used it in the other direction. Well, we started to coat, and we kept getting brown coatings—absorbing coatings. They were durable, but they were terrible. And we did all kinds of things trying to straighten this situation out. So when I finally talked to my boss Ted Zak and asked him, ‘Let’s go down to General Electric and let us talk to some people down there who really know what they’re doing.’ They were the outstanding people in tube design and vacuum and so forth. I had the privilege of seeing Dr. Saul Dushman down there, and we explained to him after five minutes of generalities what we had done. Saul turned around and said, ‘You’ve got one problem: The water temperature changed from January to May.’ My boss was there and he said, ‘You’re sure about that?’ Saul said, ‘Absolutely. And this is the first guy who’s ever come in here and told me what his problem was to help him out.’ So as time went on we had a good relationship together, and we went home and made an arrangement to have a low-temperature provided for our water, so it was down around about 38° instead of up to 60°, and got good results. Problem solved.

For some other reasons that I’d rather not get into, I had the privilege of having the respect of Dr. Saul Dushman, who was an outstanding research man with Langmuir—he was really Langmuir’s right-hand man. So that was the story we had for electron bombardment. And a lot of people don’t believe that it works. And there have been PhD thesis’s that have proved that it doesn’t work. The only thing is, it was used in commercial production as late as 1995. What it’s doing now, I don’t know—but it was used for cold mirrors and other such things that were made with magnesium fluoride and zinc sulfide. And that’s about as far as we got on that particular subject.

DEM: What was your next project while you were at Bausch & Lomb?

CHA: Well, our next project really was multi-layer coatings. We built a machine in 1948. It was composed of a 40-inch diameter trap, with the first dpi 6-inch fractionating diffusion pump and a 24-inch stainless steel bell jar. And with this we were depositing silver, magnesium fluoride, and silver. That was the principle—a reflector, two reflectors spaced by a spacer. Well, some of them were fairly good, but we learned a few things while we were doing this. We were both measuring the pressure on a recorder, and we also set up the possibility of measuring reflection and transmission at the same time, with a 2X recorder. And that 2X recorder used a 5,400-Angstrom mercury line source, so it was very precise into what the wavelength was. And when doing this, we noticed that the reflection was late in getting started. We looked very carefully and we found out that the silver deposited in small islands. So we tried measuring it electrically. And in truth, it was a non-conductor. So we stopped some of those and made some samples of different amounts, and we found these islands, and they would move around. So as a solution for the islalnds, we then deposited a very thin layer of magnesium fluoride from the source which we already had in the system. This thin layer, which was a fraction of a quarter of a wavelength, we could get unbelievably improved Fabry Perot filters. These filters would be about half the width of a normal filter with a regular silver, and somewhat higher in peak, because of the efficiency of the silver. And then we noticed something that took place. We would put one of these down, we’d coat it, we’d look at it, and come back the next day—and the wavelength had changed. Remarkable. So what we did is, we figured that perhaps there was something in there that was causing crystallization, and this was particularly noticeable in second-order filters versus first-order filters. First-order filter being a half a wavelength, second-order was a full wavelength. Well, Charlie Baer and I put together a little deal. We would evaporate magnesium fluoride, then we’d throw a little SiO into it, then go back to fluoride, then we put in a little SiO into it, so we’d break it up. We found that the wavelength shift was vastly reduced. Not eliminated—but reduced. We learned this as a part of what happens. And the conclusion we came to in time was this: That when you deposit a material on a substrate, the substrate has to be cleaned, and I don’t mean glow discharge it and then pump it down—that doesn’t work. You need to have it cleaned, either with a Hass glow or electron beam, and almost instantly—and I’m talking about a few seconds—start depositing your material because of the molecular bombardment of the background gases that are present in the chamber. And this is quite an education that we learned from this. And we had much better filters because of it. And that’s the background on Fabry Perot filters.

DEM: You’d made a comment earlier about refrigeration systems.

RAD: What we did is, we had a Freon system in the inside of this 40-inch trap that was, I believe, 3/8 of an inch in diameter. We had F12 or F22 working in it to keep some of these undesirable gases down. Now we did have a booster pump behind our diffusion pumps, and we had all our systems from about 1945 on with booster diffusion pumps behind each diffusion pump. Sometimes we had a booster pump that would handle two diffusion pumps, but these always ran at low pressures. We thought this cleaned up the oil that was in the diffusion pump, and it worked beautifully. I think in time—maybe in the 1970s or 1980s—it was determined that it was backstreaming from the mechanical pump that was causing the problem; however, the booster pump took care of the problem for us. We did the right thing for the wrong reason and it worked! As long as it worked, we didn’t publish any of these things, we went and did something about it—we made it. Well, in there we also had a refrigeration system, as I was saying, and in that refrigeration system one day—and this is in, I believe, February of 1951—the refrigeration unit went on the blink. And we had some important stuff for Wright-Patterson Air Force Base in there, and we needed to get it out there. So we went down and dug up some liquid air, brought it up, opened up the refrigeration system coils, and we would substitute liquid air for it. Then we poured some through a funnel in there. And something strange happened. The gauge was broken or something, because here’s a recorder, instead of coming down slowly, it went vertically, straight to the bottom. We had some problems figuring that out. We let it warm up and boy, it came right back up—there was nothing wrong with the gauge. Well, it turns out it was a pretty good pump. And it tells you that we had water for sure in there. So one of the things that we needed to do—and I know what we were doing—we made an attempt to measure the rate of pumping of water on a cold surface. And we did this crudely, believe me, because we’re manufacturing people; we’re not out to present papers for publication. We wanted product out the door. So we took and measured the amount of water vapor that was introduced through a heated cylinder. In chemistry you have a little cylinder that’s graduated, and we had to keep it warm. We introduced water vapor out of that by allowing a little vacuum to pull it in. Then we had our liquid air on. We didn’t have liquid nitrogen in those days; we had liquid air. And watched it go. We made some measurements as to how far it would go down, how much water we introduced over unit time, etc., etc. I won’t go into the details of it. We came up with the conclusion it was about 100 liters per second per square inch of cold surface. Now this is an unbelievable number, it was to us. Of course, this coil is out in the middle, so to speak—it was not barricaded or hidden, it was about 2 or 3 inches from the wall and could be gotten at from all directions. We measured this, and our 100 liters per second was fascinating and a great help to us in future work to keep pressures down and we realized, at that point in time, that our major gas load was water. Now, there were no such thing, really, as mass spectrographs then. Yes, you could buy one, but I think they were available for around $60,0000. $60,000 in those days was 10 years’ salary, if you were lucky, and obviously you couldn’t get it just to run an experiment, at least not in a manufacturing operation. So that’s where we stood on the subject of high-speed liquid air pumping. And I used it, as time went on, very nicely and advantageously.

DEM: Does this mean you discovered the Meissner Trap before Meissner, or you discovered it independently of him, not realizing at the time?

CHA: I’ve talked to him about it. I think we did it before he did it. But we were not in a position to publish or anything, we were in a manufacturing operation and product going out the door was what was important, not publishing. But I have talked to him about it, and we’ve had lots of discussions about it, and he does a hell of a job, believe me. He does it because he can do it with a machine by turning a button. What we did at that time in 1951, well, was something that was not really adaptable to production work, and what they did at Bausch & Lomb—I left four or five months later—I don’t know what they did. But in time at certain cases we could use, at later date, for higher pressures chambers, we could use dry ice and acetone, and then went to low-temperature refrigeration and this was done at about 1953.

DEM: That was after you were at Bausch & Lomb, then.

CHA: That was after I was at Bausch & Lomb. And we needed it, desperately, for web coating.

DEM: Where did you do your web coating work?

CHA: Well, I did the web coating at 3M, and this was a great asset and help, because what was happening was, they were heating the web in order to get rid of the water, and instead of that, what we did is, we took the water that was there and got rid of it in a hurry and made it a lower pressure to go into the next chamber to be coated. But I would like to go back to something else that happened. It really shook me up. In 1946 the military and the government went over to Europe and looked at what German technology had and so forth. And they brought over to Western Electric Hawthorne Works, Cicero, IL, condenser tissue manufacturing equipment. I won’t go into what they had put on as a coating, lacquer on the condenser paper. But they did something that was really amazing to me and shook me up. It took me a long time to figure out. What they coated—the condenser tissue at high speed, I think somewhere around 300 meters per minute—and the system was clean. There was no dirt in the system, to speak of. And the only pumping facility was a two-stage mechanical pump with a dry ice/acetone trap. It started me to thinking on a long-term basis that vacuum is really not a vacuum. It is really a low-pressure and has molecules available in high degree; even at high vacuums at that time you had somewhere in the neighborhood of 1015 atoms of gas per liter. You stop and think about that, and that’s more people than on the face of the earth per liter, and you thought you had an empty space in a vacuum. So it changed my philosophy at the time to a completely different direction, and that is the philosophy was developed that there are three things in coating: one is what’s going on at the source; what’s going on at the substrate; and what is the battle that’s going on between the gas molecules that you have evaporated and the gas molecules present in the chamber. And it is a battle. And in many things, it’s quite reactive. But the zinc was a deal that impressed me no end and helped get me started in thinking in these directions. And that has been a basis of much of my coating experience beyond that and into the 1970s and the 1980s and the 1990s. Those principles are basic and fundamental. There are three, and I’ll repeat them: the source, the substrate, and what goes on in between. And there’s a lot of action that goes on in the middle, in between the source and the substrate.

DEM: In this example, though, didn’t they have the zinc source very close to the web, so that there was no opportunity for the other molecules to get in there?

CHA: Well, it wasn’t that close, it was on top of that. The web was running flush on top of a liquid bath of zinc. Now you could see it rise a little bit, just a tad when it got up to temperature. And the conclusion was that the vapor pressure of the zinc was enough to lift the web. So what they had done—gently—is that they had put it in such a way that actually, in my opinion, the rollers were a little lower than the substrate, yes, and the substrate would go under it, and therefore it would be contacting the zinc. But they couldn’t do that. So they had a shutter in there that jammed itself in between, and they could pull it aside after they got the web up to speed. Tremendous learning operation—to me it was. You know, it’s when you don’t have something… Now the Germans did not have oil-based diffusion pumps. They did everything in all their work with mercury. And the story behind that, they tell me, is that an outfit in England developed Octoil and other diffusion-pump oils. He was a friend of some people in Germany. And they wanted to sell them the rights, no how, with a royalty. And the Germans wouldn’t accept that. That’s what the story says. I can’t prove it, I wasn’t there—but I’ve heard it from reliable people that it was true.

DEM: You were at 3M for quite a while. About how long?

CHA: I was at 3M from 1951 to 1969. After I was involved in vapor coating of ScotchLite, I was given the responsibility of building a laboratory for ThermoFax. It went from 10 people to a couple of hundred. I was interested in going in one direction, and the management was interested in going in another direction, so the conclusion was, ‘If you can’t go the way I want to go, I’ll leave.’ So I left.

DEM: What did you do after that?

CHA: Then I took my wife and we made a tour of the world and I investigated Europe and Asia and several places, and came back, and I had some jobs in consulting set up at that point in time. And I was consulting ever since. However, as of late, I’ve tried to get out of this business because I’m getting tired of it, frankly. I have worked five years in my life. The rest of the time I’ve had fun doing what I’ve been doing. And I think it’s time now to go back and put into society some of the blessings which I have received, both in know-how and in funds that are available to colleges and other institutions of interest. That’s how I feel.

DEM: You’ve mentioned to us quite a few interesting things that you’ve learned. Is there any other important, interesting thing that you’ve learned over this vast experience?

CHA: Well, I think one of the things that I’ve learned is how to make an evaluation of this residual gas that’s present. I don’t care how good your background pressure is, there are gas molecules there and they can do things to you. What I think I’ve learned is that you consider that as a positive pressure. It is a positive pressure compared to ultimate vacuum, and I think we’re using the word ‘vacuum’ incorrectly; we should be using a ‘low pressure’ of some sort. Just to go on and on, if we’re talking about temperature of liquid air or liquid nitrogen, do you know what the temperature is at either Celsius or in Fahrenheit? I don’t.

DEM: I know what it’s supposed to be.

CHA: What is it supposed to be?

DEM: Liquid nitrogen? About –196°C. But I’m supposed to ask you the questions, you’re not supposed to ask me!

CHA: Well, my point is, liquid nitrogen is 77° Kelvin. And I’m looking at pressures in terms of Kelvin, rather than in terms of Centigrade. And that’s a lesson I’ve learned, and I haven’t been able to get very many people to adopt that philosophy. That’s enough for the time being. Anything else?

DEM: Well, you’ve already mentioned several people who’ve been important. Are there any other people that you’d like to mention at this time who were important in your work history and learning experience?

CHA: Well, the people who worked with me predominantly were Charlie Baer, Harold Schroeder, and Dr. George Hass, whom I have the utmost respect for, and we have done a lot of things together over the years. One of the things that we learned—I learned—from him is that you glow discharge or clean a substrate in a vacuum. And if you’re going to do anything, you’d better deposit something on it yesterday, not 10 minutes later, but at once if not sooner. His glow discharge we used, and what we did is, we got this from him. We would pump it down to our ultimate, which would be on the five scale, raise the pressure, and do the glow at about 7 microns. And at 7 microns, when we got done with the glow, we would have our filament already warm, and we’d turn on the high-vacuum diffusion pumps and bang, we’d coat within seconds. In essence, this had a clean surface, there were no monolayers that had a chance to form, to speak of, and the same thing was when we were bombarding with electrons. We’d bombard the electrons and started to coat. The substrate was clean, all the way. And this we learned back in the early 1940’s. Harold Schroeder and Charlie Baer were very close in the work that we did together at Bausch & Lomb, and Hass would come and talk to us on occasion and we’d visit him. Those are the important people, in my opinion, whom we traded back and forth the work that we had done. This included such things as making grating replicas and coating a stack of 20 14-inch mirrors bombarding it with electrons and then putting on aluminum and then a protective film on top. We did some of it with induction heating at 27 megawatts, and we ran out of holders for holding the aluminum. We did some other things. We actually turned around and made some mirrors using Hass’s technique of putting on three times the required aluminum requirements past opacity. Then we could anodize it in a solution that he had published, and we had a really nice, hard coating. Very good coating. We didn’t progress that too far for two reasons. One is we didn’t have a good aluminum source, and that was why we got involved with the high-frequency crucible heating. The other thing was, we didn’t know when we had three times opacity. We hadn’t figured that one out. By that time, the projection TV system was no longer desirable and they were going to 7-inch and 10-inch CRTV tubes in your TV. Didn’t need projection. The early ones we made—simply, in a bell jar—we made for radar for the Navy. And that was the beginning of that type of thing. From that we developed, particularly with Charlie Baer, we stacked 21 mirrors in the chamber. But the first one never was any good. It took a long time before we figured out, ‘Forget it, just run a dummy,’ and we put a dummy in all the time, but why did it come out poorly? Well, I think it was due to the background of the disturbed, undesirable molecules of whatever in the chamber, and not having any measuring equipment in 1946, we didn’t worry about it; we just went on to the next step. Those were some of the things that I’ve learned, I think, and I hope they’re of value to others who come along.

DEM: Well, Collin, I want to thank you very much for an interesting discussion and congratulate you on a wonderful career. And if you ever really do decide to retire, I hope that it’s a good one.

CHA: Thank you ever so much. Take care. For the record, I was born 12/27/1916, Carsonville, Michigan.

Return to Oral History Index