AVS Historical Persons | Collin Alexander - 1990 & 1992

Collin Alexander - 1990 & 1992

Two Oral History Interviews with Collin Alexander

Interviewed by Richard Denton, October 1990 & Don Santeler 1992
DENTON: I know the first time I met you was when you were working for Bausch and Lomb in Rochester and you were working in vacuum coating. But I had been up there several times in 1942 before I was involved with that. So I'm not sure what time you went to Bausch and Lomb and how you started on vacuum and why. 

alexander.jpgALEXANDER: If I may, I'd like to go back a little bit and give you a little history on this. Bausch and Lomb was approached by the movie studios to produce anti-reflection films for their theater projections for Gone With the Wind. There were some dark scenes that did not show up well, and an agreement was made that if the theater wanted to show Gone With the Wind, they had to have anti-reflection films on their projection lenses. They started coating this in about '39. These films were very soft and very difficult to handle. 
I interviewed at the American Chemical Society at Atlantic City on Labor Day Weekend in 1941. They were interested in me because I had done some work in building my own vacuum system in school. It was a mercury pump and all that. But at least I knew the words, which was more than the other things they had. So I started work there about the first of October, 1941. At that time, most of the systems were pumped with 4-inch DPI1 glass diffusion pumps. They had one metal pump. 
In an interesting story, a fellow dropped one of those while he was trying to put it into place and took it out to DPI, which was only five miles away. His name was Billy McDonald. They fixed it up and he brought it back and put it into place, and he cracked it again. He dropped it and left and said, "I'm through." He was through for the week as far as he was concerned! It was hard work.

DENTON: This was a glass pump, I guess?

ALEXANDER: This was a 4-inch glass pump, and the gasket material was pure gum rubber. So that's where they were. At that time, there were four units in production and one in the laboratory. In addition to that, we were working at that point in time, not only on harder anti-reflection films. We had done some work and tried to analyze the magnesium fluoride that was used, which was only about two or three nines pure. We were in the middle of that when the chap from Naval Ordinance had developed the heating method and the purification of the magnesium fluoride. Both are important. With just one of those, you really don't get a good coating.

DENTON: That was Dr. Lyons.

ALEXANDER: Dr. Lyons. And of course, at Bausch and Lomb, in the old Zeiss philosophy, you never heat a lens, and there was a lot of opposition. 

DENTON: I can imagine.

ALEXANDER: And of course, all these units except for the diffusion pumps and the mechanical pumps and the gauges were all home-built - we built our own.

DENTON: How did you determine the gauges? Did you read the literature and find out how to make one and then go ahead with it, or what?

ALEXANDER: We did purchase them from DPI simply because they were only five miles away. We had a very close working relationship. But during the war, ion tubes were very difficult to get. As a matter of fact, I had some of them. I was walking out the door at DPI and they said, "Come back here a second." I went back, and they took the three ion tubes that I did have, because Oak Ridge had a higher priority. Although we thought we had the highest priority. From that point in time, we went around and took some radio tubes, 2A3's, put a hole in the end of it, put in a tubulation with a glass blower, and they were good as ion tubes of a sort. At least it told us about where we were. It would last probably about three or four days, and then we'd put in a new one.

DENTON: As long as you had a good glass blower, I guess you could do that. It's interesting, your comment. All the time I've known you. I attended that Atlantic City American Chemical Society meeting when I was still a printing chemist, before I went to the Franklin Arsenal. Maybe we bumped into each other and never knew it.

ALEXANDER: That's very possible.

DENTON: I never knew you were there. Did you actually run systems with glass diffusion pumps for a while?

ALEXANDER: Absolutely.

DENTON: Were they water-cooled or air-cooled?

ALEXANDER: They were definitely air-cooled2.

DENTON: We all had octyl or dioctyl phthalate.

ALEXANDER: Dioctyl phthalate was the standard diffusion pump fluid. As you were talking a little while ago, we had a power failure - a planned power failure, but they didn't inform us - on a weekend.

DENTON: Blackout or something of that sort?

ALEXANDER: Yes. And I got a call about 2:00 Monday morning saying that there was a problem at the plant. I went in, and here the diffusion pumps - by that time they were all metal - had heated up. And the mechanical pumps, which had to have a manual reset, had not started. We had phthalic anhydride3 all over the entire system. The job of cleaning that up was pretty tough in a small room. But we were back in business with one unit by noon. 

DENTON: That's pretty good. I think the gauges that we all used then were far from accurate. But I guess you found ways to work around that, didn't you? Even though you weren't sure of the gauges? 

ALEXANDER: You mean the tubes or the amplifiers?

DENTON: I mean the whole thing as a readout that you really knew from the gauge that you had in the system. Were you using the Piranis4 for the relatively higher pressures. 

ALEXANDER: We were using the Pirani for the higher pressures. The first ion gauge we had was a Westinghouse gauge, believe it or not, with a galvanometer. You had to keep making adjustments all over the place every time you looked at it. Then the VG-1A came out, and that was a very good system compared to anything else that we had. We had a system, and we were putting down silver and evaporating it through a mask. Or we were evaporating cadmium through a mask in order to get reticules experimentally. We pumped the system down and we could not get below ten microns. We worked on that system for all one day and part of the next day, because we didn't dare turn on the diffusion pump because of the high pressure that was involved. But it turns out something had gone wrong with the Pirani. We were already down, but we didn't dare turn on the ion tubes because they cost $20, and that was a lot of money in those days - if you could get them. Those were the rather strange things that happened once in a while. 

1. Distillation Products Inc.
2. Air flow was by a small fan
3. From break down of octyl.
4. Pirani gauges

2. Interviewed by Don Santeler 1992

SANTELER: Good morning. This is for the American Vacuum Society historical archives on interviews with Society founders. This is February 3, 1992. We're at the Sheraton Sand Key Hotel in Florida with the Florida Chapter of the American Vacuum Society. I'm Don Santeler. I will be talking to Collin Alexander, one of the founding members of your Society. Collin had his formal education at Albion High School in Michigan, followed by a bachelor's degree at Alma College in 1937; then he did graduate work at MIT, where he studied chemical engineering. His background work experience was predominantly with Bausch and Lomb and with 3M. His vacuum interest is mostly in the field of optical coatings and general coatings. 
Let's start with the Bausch and Lomb experience. Collin, can you tell the viewing public a little bit about how optical coatings interact with vacuum technology?

ALEXANDER: I joined Bausch and Lomb in the fall of 1941, after being interviewed at the American Chemical Society meeting in Atlantic City. I was the only person they interviewed who even recognized the terms "McLeod Gauge" and "diffusion pump"! They had four 18-inch optical coating systems with diffusion pumps; two glass and two metal. When Gone With the Wind was to be shown around the United States, they found that there were certain portions of the picture which were so dark that reflections from the projection lenses would obliterate some of the parts of the scenes. So what they did is they found out that anti-reflection film - and this was some work that had been done at MIT - would help eliminate this. So they made a contract with Bausch and Lomb to provide the lenses with antireflection coatings for all movie projectors that would show Gone With the Wind. That was the beginning. Needless to say, there was no equipment available for such things. They went through a great deal of struggling in order to make the first anti-reflection films. 
So the equipment that was built was home-built. The only things that was purchased outside were some pumps and a few gauges. The pumps were from Kinney, predominantly, and the diffusion pumps were from DPI1, which was only three or four miles away. And the gauges were also from DPI and Westinghouse. 

SANTELER: What kind of vacuum levels were being used for these kinds of coatings in 1939?

ALEXANDER: This was a big struggle. In order to get below 10-4 [Torr] you have to realize that Apiezon was frequently used as a grease. There were no O-ring seals. There were no high-vacuum valves. The gaskets were primarily pure gum rubber pulled out carefully. The pumps were glass with Apiezon oil or octoil used as the liquid.

SANTELER: Did you have to develop your own valves then? What did you use?

ALEXANDER: No. There were no valves used at this time. You had anywhere from a three- to four-hour cycle in operating. The material evaporated was primarily calcium fluoride because it had a low index. But occasionally, it would come down with a high index. We never did understand that. As time went on, we had to get a more durable film in order to easily assemble binoculars and range finders, which were coming into the military needs. That was developed. We had started, however, to purify some of our magnesium fluoride. But Dr. Dean Lyons at the Naval Research Laboratories developed a method of heating the lenses and purifying the magnesium fluoride. That was adopted much against the will of certain people in the optics industry who felt that you should never heat an optic because it would distort it.

SANTELER: It makes sense, I guess.

ALEXANDER: That's right. From that, we got hard coatings. As time went on and along about 1943, we had a large number of systems. We were up to about two dozen systems at that time. We put in our own homemade high-vacuum valves using a Wilson seal, which was described in the RSI2 about 1937. Then we did one other thing at that time. We took and used one mechanical pump to back a booster pump, which in turn backed two diffusion pumps for two chambers. The reasoning for this was partly due to the fact that pumps were unavailable. And we had a problem with gauges. We couldn't get ion tubes. I had some ion tubes taken away from me for Oak Ridge just when I was walking out the door of DPI! What we did then was we used two A-3 tubes, which were standard radio tubes; we cut a hole in the top, had a glassblower add a tubulation. Of course, these things would last about two, maybe three days. What we did in order to understand our pressures, we started the design and built from another article we'd seen in RSI on a Knudsen gauge. So we started building Knudsen gauges for our own use. 

SANTELER: This whole thing that you've been saying, does that also tie into your work on gradient filters and TV mirrors in pretty much the same-- 

ALEXANDER: Those came on, but by the time we got to TV mirrors, the war was over and we could get the tubes and some of the things that we wanted. I would like to, at this point, say that most of this work that was done, if not all of it, was really the responsibility of three people: myself; Harold Schroeder, who was also a founding member; and Charlie Baer of Baer Associates. 

SANTELER: At some point in here, you decided to leave Bausch and Lomb?


SANTELER: And ended up going with 3M in 1951. But now, the surprising thing is the whole field switched for you at that point.

ALEXANDER: Well, it wasn't that much of a switch. I'd like to go into some of the other things that we did, if I may, at Bausch and Lomb. Those things were - We had a requirement to coat optical flats. They put two optical flats together to make an optical wedge. These, then, were ground and polished to make the wedge, but they needed an anti-reflection coating. So we developed a procedure of bombarding the substrates, namely the wedges, with electrons at about 5,000 volts and what amounted to about two-tenths of a watt per square centimeter. This gave us an opportunity to take care of the range finder needs in the United States. 
We did this and we found out that this was useful not only for magnesium fluoride, but other materials such as zinc sulfide. Some of the other things that came out of this were - as you have already mentioned - gradient density filters for optical flatteners for wide-angle lenses, gradient density filters for ray bans, which are quite common today. And in addition to that, we built a record changer in 1946 for Schmidt Mirror Aluminizing, which was a popular way of having a large TV screen other than the seven-inch screens, that were highly prized at that time. But two feet by three feet was a much nicer-looking image. This was essentially a method of putting in a stack of 20 mirrors and picking off one mirror. We would bombard it. We would coat it and stack it and continue until we did all 20. This was something, I think, that was rather new, but produced a large number of these mirrors. We ran off with most of the business in that area. 
In addition to that, we made Fabry-Perot filters composed of some seven or more layers. We coated draft glasses, which are used for protecting movie film from the radiation from the arc. We also made some cold mirrors. Cold mirrors were really mirrors that reflected the visible and transmitted the infrared and prevented the burning of film, as which you've probably all seen years ago when the black and white films, you'd see a little black thing going up and down. Well, that was because the film was on fire. This took care of that problem.

SANTELER: It seems like the Bausch and Lomb experience was the one that you were most involved with. Before we take a look at 3M, what did you learn from all these experiences with these many development phases with different kinds of products?

ALEXANDER: We learned the following. One is that vacuum is one item. But you can have an excellent or the best vacuum in the world, but a process is a completely separate operation. If you don't have a process, you don't have a product. But in order to have a process, you also have to have a vacuum. The second thing is that equipment suppliers can produce excellent vacuum equipment today, but that does not necessarily mean that you have a process that can be used in that. Third, the carry-through from development to research, if it's necessary, to scale-up, to pilot plant, to production, when carried through by a team all the way through, your successes can be rapid and very satisfying to the people. Products are made quickly and easily without time delays and high expenses. And last but not least, when you do the scale-up, the most important thing is that no process change is made. That sounds very simple, but such things as distance from source to substrate, such things as pumping systems - you can't make those changes and expect to go back and learn from your pilot plant early stages and adapt that. So what you have to do is to exactly follow what was going on in the development areas when you want to go to production.

SANTELER: You're talking like 1940s and ‘50s, but you know, to a large extent that's still true today, all those things that you enumerated.

ALEXANDER: They're absolutely true, but frequently not recognized. 

SANTELER: It's forgotten also. When we do learn the lesson, we forget.

ALEXANDER: That's right.

SANTELER: I think we have a little bit of time left here, yet. I'd still like to go back and explore why you switched from Bausch and Lomb to 3M, because to me it seems like a totally different experience. Roll coaters are big, bulky things, wet, lots of water load, different kinds of pumping systems. So could you tell us a little bit of that? Particularly in contrast to optical coatings.

ALEXANDER: Well, we had problems with water loads in optical coatings. That was not necessarily recognized by roll coaters. The early roll coaters were 6 and 12 inches wide, and that's not a very large system. Some of the early systems, at least that I encountered in roll coating, were, oh, two feet by three feet for the coating chamber, and the degassing chambers were about the same size.

SANTELER: How many pounds of paper would you be putting in something like that? Or was it paper, always?

ALEXANDER: Well, it's not necessarily paper. It's plastics, which are sometimes worse than paper. But the size that we liked for a pilot plant - and I still like it for a pilot plant - is something that a man can manhandle and put into a chamber. In other words, it's not over 60 pounds. You can run experiments rapidly that way because it's easy to clean up a system by comparison. When you have a large system, let's say 48 to 80-inches wide, you have to have a crew to use forklift trucks, and opportunities to run the experiment are time-consuming, and clean up is also a problem. You've got many people involved. When you have one or two men running something like this, it's very easy and rapid to run experiments and set your process conditions.

SANTELER: I assume your comments about taking carry-through from the research space through to the development of the final production operations that you described for optical coatings still applies for roll coaters.

ALEXANDER: They still apply for roll coaters. If you don't do it, you get into all kinds of problems. It's a communications problem. One of the things that we did in the roll coating aspect was to, when we did go into production, insist that the technicians went into production as foremen, for two reasons. One is, they had gone through all the mistakes - or most of them at any rate - and knew what not to do. And if they did have some new problems, they knew who to go to on a first-name basis to ask the questions. 

SANTELER: I assume that these were all batch-type operations, or predominantly batch-type operations. Have you had any experience in continuous production in roll coaters? Or how do they get the roll in and out of the vacuum cases like that?

ALEXANDER: You're referring to air to vacuum to air?

SANTELER: Yes. I was just wondering if that's a solution. Because it would take a long time to load by bulk these large rolls, particularly the large rolls. 

ALEXANDER: The large rolls, you can have a turn around time of 20 minutes for a large roll.

SANTELER: That includes the vac back to air and pump down again?

ALEXANDER: No. That includes the actual time of putting the roll and splicing it up and cleaning up.

SANTELER: So actual process to process is a bit longer?

ALEXANDER: That's right. But if you go from air to vacuum to air, you enter into a host of new problems, primarily which is not recognized is the utility cost of pumping, because you have the leaks and they're part of the game.

SANTELER: You've got some kind of a seal there that you've got to work through.

ALEXANDER: That's right. But even though you have a seal, you're running several 50-100 horsepower mechanical pumps to take care of that load. And that cost will eat you up. 

SANTELER: I assume that these are predominantly mechanical pumps on mechanical booster pump systems as opposed to diffusion pumps or something more exotic in the high vacuum range?

ALEXANDER: No. These by and large are roughed with a large mechanical pump or a mechanical pump and blower. And then the diffusion pumps are large, and they are backed by blowers, or what's commonly called an ejector-diffuser pump, of which only one company makes those. They are outstanding pumps and do a tremendous job. 

SANTELER: Well, thank you very kindly, Collin, for coming down here today and sharing your experiences in this big field of vacuum that we're all working in, and sharing it with the viewing audience. We'll look forward to hearing from you again.

ALEXANDER: It's been fun. It's fun talking to you. It's fun talking to the camera. And I've had a lot of fun in my life doing the things that I have enjoyed.

SANTELER: Good. Thank you.

1. Distillation Products Inc.
2. Reviews of Scientific Instruments

return to top