Nuclear Museum Logo
Nuclear Museum Logo

National Museum of Nuclear Science & History

Clarence Larson’s Interview

Dr. Clarence Larson, a chemist, began working under Ernest O. Lawrence in his lab at the University of California, Berkeley in 1942. In 1943, he moved to Oak Ridge and was appointed head of technical staff for the Tennessee Eastman Corporation. He later served as director of Oak Ridge National Laboratory and as a commissioner on the U.S. Atomic Energy Commission. During the Manhattan Project, Larson designed a process to recover and purify uranium deposits from the walls of calutron receivers at the Y-12 Plant. In this interview, he explains the importance of this innovation in producing enough enriched uranium for an atomic bomb. He also describes the challenges encountered in the Y-12 Plant’s early days, as well as Lawrence’s leadership skills and unyielding confidence.

Date of Interview:
July 13, 1965
Location of the Interview:


Stephane Groueff: Interview with Dr. Clarence Larson—L-A-R-S-O-N—head of the Union Carbide’s operations at Oak Ridge, a chemist. Dr. Larson was connected with the electromagnetic separation process during the war, and he was a personal friend of Dr. Lawrence [Ernest O. Lawrence]. He’s married to the daughter of Dr. Stafford Warren, who was also with the project. You came in 1942?

Dr. Clarence Larson: Yes.

Groueff: From where?

Larson: Well, I was the head of the chemistry department at the College of the Pacific, professor of chemistry at the College of the Pacific, which is located about sixty miles from Berkeley. I got my Ph.D. at Berkeley. I had done some work with radioisotopes as a small, independent research project at the College of the Pacific, receiving radioisotopes from the cyclotron in Berkeley.

As a matter of fact, since the isotopes were so scarce, and since the cyclotron time was so valuable and sought after by all of research, I hit on the idea of taking some of the material which held the radioisotopes for bombardment. Of course, there’d be enough scattering of the beam, and usually these were made of iron. I received some of the material in the holders which then I dissolved up. And that contained enough radioisotopes surrounding iron to work with. So I had done not a tremendous amount, but a small amount of work, in the field of radiochemistry at the College of the Pacific. So I was naturally interested in that.

When the war broke out, I joined [Ernest O.] Lawrence at Berkeley—just finishing the semester, and then joined in Berkeley in July of 1942. Then I went to work on the chemistry of the separation process which had to be used. So that’s how I happened to get into it.

Groueff: So you did not work on the building of the calutrons?

Larson:  No.

Groueff: So much, but the chemistry of the–

Larson: On the chemical processing.

Groueff: I see.

Larson:  That’s right.

Groueff: Oh, the product?

Larson: That’s right, the product. That was my particular role in the thing, but I suppose in going through all of this you receive many different impressions of the whole project. Who did what, and who is responsible for this, and who played the greater role I suppose. You mentioned about Dobie [Percival] Keith’s reaction.

Groueff: Yeah.

Larson: Well, you’ll find that with a lot of people. I’m sure on this, but I might give you some general reactions from talking to Lawrence in those early days. Number one, when I first joined the project, there was one doubt that I had in my mind: if you were able to separate this, would you get a bomb from it? And would it explode?

It’s very easy to get fissionable materials, and all of us could see how you could put it together and it’d heat up and all that sort of a thing. But there’s an awful lot of difference between a coal pile burning and dynamite going off. Both have the same amount of energy, but can you release it in the form? I can remember Lawrence saying with full confidence, “That that’s not the problem. If you get the U [uranium]-235 and get it purified, there is absolutely no doubt in my mind that it will work as a bomb.”

Groueff: Since the beginning?

Larson: He was absolutely 100 percent confident. He had no doubt whatsoever that the U-235, if purified to a high percentage, would make a bomb: absolutely no doubt whatsoever in his mind on that thing. That was the number one thing.  

The other thing that he said was that in order for this to be useful – just looking at the timetable at that time – this has got to be ready in July of 1945. So this is back in July of ’42 now. He had the target date at that time. I haven’t the slightest idea how he arrived at that. But he said, “This has to be done as of July 1945. We’ve got to have enough for a bomb at that particular time, or the project will be a failure,” and so on. Those were things which he had some insight into. I never did find out how he ever arrived at that July of ’45, but he had that.

When I joined the project, they still had just a few micrograms, or a milligram, of material in not too high a concentration separated out. A rather low concentration. It would not have been valuable for a bomb. This was just milligrams or micrograms of the material.

Groueff: Slightly enriched?

Larson: Slightly enriched, yes. Well, not slightly, but a fair enrichment, but not enough for bombs. That was where we started. But you could see with optimistic calculations on the performance, that indeed if you made so many of these units, and they performed in this way, there was complete confidence that there would be enough material in July of 1945 to make a bomb. The calculations were there: how much was needed, and so forth and so on.

Even in the days when there were just micrograms or milligrams, the calculations were made which gave—I would say he [Lawrence] was able to convince with complete confidence that this was a sure way of doing it. It would be difficult, and it would be expensive, but it would be an absolutely sure way of doing it.

Groueff: It wasn’t the most efficient way, in the sense of being too expensive?

Larson: Yeah. It was obvious that it would be very expensive, and it would be obvious it would take a lot of people. 

Groueff:  Why? Because the idea was just to multiply?

Larson: Just to multiply a small operation by thousands. Each operation was very complex, and needed a lot of people to operate it, and to service it, and to recover the material. I mean it was very obvious that it was going to take a lot of people.

Groueff: There was no way to do it in an industrial way, by a sort of assembly line technique?

Larson: Well, of course that was brought in. That did simplify it, but it didn’t lend itself to a continuous operation, such as the gaseous diffusion.

Groueff: You couldn’t make gigantic beams, which in one unit would separate by the tens of grams or something like that? It had to be done by a very small—

Larson: Very small. Well, they did, by ingenious inventions and so on, increase the output of this tremendously, so that the efficiency of each machine did go up. Many people of course—I think probably [Robert] Thornton pointed this out. There was serious doubt in many physicists’ minds that you would ever be able to get that much material in a beam, in an ionized beam, because the theory was that it would blow up. Each one of the particles being charged would repel each other and then it would just blow apart. I mean that was it. I don’t know the mechanism yet. In fact, I’m not sure whether it’s even agreed to, but the beam does not blow up. There’s neutralization there.

Groueff: So in other words you were working the theory—the fundamentals—at the same time with the experimental stage and the practice and production?

Larson: And all of it.

Groueff: All together?

Larson: All at once. The production of each individual unit went up tremendously and far beyond, oh, maybe a hundred or a thousand times, what theory would have predicted. But it was quite early that it was determined that the beam would be stabilized and would not blow up.

Groueff: So the doubts not so much of the scientists of Berkeley, but in the minds of industry and engineers and military people, because logically it didn’t sound very promising?

Larson: Even to me it looked extremely difficult. Most of us who had been in academic work could not visualize just mobilizing this much complex machinery in that short of time. You see that is where the big contribution of industry—General Electric and Westinghouse and Eastman and the rest of them. Well, they had full confidence that if you can make one of these, a thousand isn’t going to be a thousand times as much. It may be a hundred times as much work, but not a thousand times. They’ll just make more of these [inaudible].

Groueff: I can imagine their reactions when they first came, people from Stone & Webster and all these. They were quite surprised because they’d never seen anything like that? 

Larson: Oh yeah. They could visualize spending hundreds of millions of dollars in order to—and this is just strange. This complex machinery—

Groueff: To you, all this was completely new and rather scary?

Larson: It was scary. I think men like Lawrence and Thornton really didn’t have much doubt in their mind that this could be done, so they carried the day.

Now actually – I’m not sure whether you’ll get this – but it’s my understanding that the electromagnetic process, while very expensive and very difficult, was something that you could do if you had enough money and enough people. I mean it was a sure thing. Now you see the [gaseous diffusion] barrier and the seals and the probability of corrosion—it was not sure that it would ever work. I mean it might not work at all. It might fail completely.

Groueff: Yeah. It’s an important difference.

Larson: If some of these properties of nature had been against you, they just wouldn’t have worked. But we were fortunate and there was an awful lot of good work, and it did work. It took longer than people had hoped, and consequently, the contribution to the first bomb was not tremendous for the gaseous diffusion. For the second one and subsequent ones, it was tremendous of course. But the electromagnetic process was a sure thing, and there was some doubt about whether they could get a plutonium bomb to even explode.

Groueff: Implosion?

Larson: The implosion was controversial as to whether it would work or not. The only sure thing, as to whether you could make a bomb or not, was the electromagnetic process.

Groueff: It was expensive.

Larson: It was expensive.

Groueff: But sure. 

Larson: But sure. It needed an awful lot of people, and it was full of grief and so on.

I think this probably sustained the whole effort, because they knew that if just one of these approaches would work, they could afford to take the alternate. Even if just one of them worked, that would save the whole project. Then if the others came through, it’d be just that much better. Once they got the material, they were sure they could make a bomb out of it. They were sure that by July of 1945 they could make enough material for a bomb according to their calculations. They were behind schedule for a while, and then ahead and so forth and so on. That’s the sort of background as I visualize it, with regard to Lawrence’s big contribution.

Groueff: He was definitely the dominant personality on the project?

Larson: Oh yes. Certainly outside of [General Leslie] Groves: I think he probably was a strongman. Of course, you’ll find since Lawrence was an experimentalist primarily—probably one of the world’s greatest experimentalists—as far as the physicists are concerned, in certain quarters he doesn’t rank very high because he isn’t known as a great theorist. But there isn’t any question that he also understood theory very well. Therefore, I think you’ll find contradictions among the scientists about his stature and so forth.

Groueff: But contrary to many scientists, he knew how to get things done and also how to work with industry and engineers?

Larson: That’s right.

Groueff: I see that the engineers never resented him or—

Larson: That’s right.

Groueff: He knew how to talk to them?

Larson: Oh yes.

Groueff: The military too?

Larson: Yes. Well, he was a very unusual individual in that respect, because of course he had the confidence of having a large number of successful projects before the war, with the cyclotron and so on. Not to be underestimated was the fact that the group there at Berkeley under [Glenn] Seaborg worked on the chemistry of plutonium and radioactivity. If it had not been for Lawrence, that work never would have existed, you might say. 

Groueff: Because they did it in his instruments? I mean the cyclotron.

Larson: Yeah. That’s right. Seaborg and his group had been working since 1935 I believe, or something like that, on radioactivity obtained from the cyclotron. They had built up this tremendous volume of information and instruments, and everything else which was needed to make the plutonium separation.

Groueff:  So he actually created one of the world’s most important centers for nuclear science?

Larson: Yeah. The nuclear chemistry was created there at Berkeley because the cyclotron was there.

Groueff: And it was the only one?

Larson: Yeah, it was the only one at the beginning. Of course, the first plutonium was produced with the cyclotron for working out the chemistry of plutonium synthesis. It was not produced by a reactor; it was produced by the cyclotron. It wasn’t plutonium-239, incidentally. It was plutonium-238. It’s a fact that a lot of people don’t realize.

Groueff: But how did you work on the chemistry of uranium when you didn’t have any sizable samples?

Larson: Oh. So far as the work that I was concerned with, our problem was twofold. Number one, you saw how the beam goes into the receiver, and is separated into the 235-enriched fraction and then the depleted fraction. And then also there is—I think they described how the material evaporated. The uranium was evaporated from a container and then passed through an arc.

Groueff:  In the container, it came in a sort of very impure—mixed with the graphite or something?

Larson: No. The uranium was in the form of a uranium compound. I’m not sure whether it’s classified. I think it probably is. But I’d say it’s uranium compound, which when heated up, doesn’t melt. It vaporizes. It sublimes.

Then there’s a little slit on the top of this metallic, stainless steel container where the material comes out. There’s an arc which is struck across that, which then changes the uranium into an ion—from a neutral molecule into a uranium ion which is positively charged. Then there is a high voltage placed right above that in sort of a grid, like the grid of a vacuum tube. You’re familiar with the grid of a vacuum?

Groueff: Yeah.

Larson: So there’s a high voltage grid that’s above that and it’s negatively charged. It attracts the uranium positive charge. It goes through the grid and then, at high speed, encounters a magnetic field and then goes around in a circle. Now at the end of this circle it’s separated into U-235 and U-238.

Now there are two problems. You vaporize this and put it into the arc, but only about ten percent of it gets ionized. The rest of the vapor goes out and collects on the side of the machine. So uranium, being valuable, had to be recovered from this. Then the other point was that the uranium would go in with great energy and would embed into this metallic container. Or sometimes it collected in graphite, and then this container had to be dissolved— 

Groueff: The whole container?

Larson: Well, fortunately we discovered a way that you didn’t have to dissolve the whole container.

Groueff:  But it didn’t come easily?

Larson:  No. I mean it would embed in there, and then you had to take acids and dissolve it, and then—

Groueff: That was your job?

Larson: My particular job was to devise a system whereby you recovered the uranium that deposited on the walls and also that was stuck in the receiver. Now, the physicists thought this was very easy. They thought, it’s valuable stuff, and you could do this whole thing with beakers if you wanted to. Well, it wasn’t that easy. It turned out to be a tremendously difficult task.

There were many of us who worked on the best process for doing this. Of course, there was controversy. Was the best process to use this reagent or that reagent? Or to precipitate this out first, or precipitate that out second? All kinds of variations because the uranium, when it collected on the walls, you’d always have it hot. It was very corrosive. It would dissolve part of the wall.

Groueff: The walls are glass or—

Larson: The walls were metal, and it would dissolve part of the walls. So there were a tremendous amount of impurities that you would find.

Groueff: And all of this is in very small quantities?

Larson: In small quantities. So you had a lot of impurities in a small quantity of uranium, and this had to be then recovered.

Groueff: Physically, how did you do it? Every couple of days?

Larson: About every three days, I’d say.

Groueff: So you’d stop the units?

Larson: Stop the unit.

Groueff: Take out the tube?

Larson: The receiver, and then that would be sent to a processing building.

Groueff: And put a brand new receiver?

Larson: Put a brand new receiver in. In the meantime, the material on the walls would be scraped off and dissolved off and get it off the best you can. It was a terrible operation.

Groueff: So every three days that had to be done.

Larson: That’s right.

Groueff: And the same machine can be used except for the receiver, the same walls or—

Larson:  Oh sure, yes. They eventually put stainless steel liners in to minimize corrosion. In any event, you’d have to replace some of it. It would eventually dissolve.

Groueff: We’re talking in terms of even not grams now [inaudible]?

Larson: Yes. Well, in terms of the material that is vaporized on the walls, that got to be, you might say, hundreds of grams. Then the amount in the receiver would be only ten percent of that – that would get over to the receiver. But then, you see, there was only one part in 140 of that is U-235. So you’d take ten percent of that, and then only one percent of that is–

Groueff: In the receiver at the end, which is marked U-235, after three days how much of this enriched material do you have? Do you have a few grams?

Larson: Yes. It’s in the order of grams.

Groueff: But it’s not very pure? It’s mixed with graphite and things like that?

Larson: That’s right: graphite and stainless steel and so forth and so on.

Groueff: So it wasn’t the pure material that you can use for a bomb?

Larson: No, so that was the particular job. 

Groueff: Where did you do that?

Larson: At first at Berkeley. 

Groueff: When did you come here [Oak Ridge]?

Larson: I came here just one year later, July of 1945. I’m sorry, ’43.

Groueff: Did you come with a whole group, [Wallace] Reynolds and Thornton?

Larson: Yes. I originally did my work as a part of the University of California group, but by that time Eastman Kodak had taken the contract, so I transferred over to Eastman.

Groueff: Like Thornton did?

Larson: Yes, that’s right, at that time. Because of part of the contractors—

Groueff: It was a technicality. You were on the payroll of Eastman Kodak.

Larson: That’s right.

Groueff: But the important boss was still Lawrence? And you would report to him?

Larson: Yes. He would always be in on it. Of course, he kept traveling between Berkeley and Oak Ridge and then other parts, Washington and so forth. He would also lead the contractors. We would see him maybe once every three or four weeks or something like that.

Groueff:  They explained to me how those girls were trained to operate the unit, the calutrons. But on the other end, when the receivers were filled, who did that job? Specialized chemists?

Larson: Yes. In general, that was specialized chemists. Now this is where we ran into two very major difficulties. When the process was worked out, as I said, we had several competing processes. We had about three different groups and each one had different ideas on the processing, so there was controversy as to which one would be the best one.

I’ll describe the difficulties in the receiver first as an example. When the material was put into the receiver, it was thought that with a stainless steel container, all you would have to do is put nitric acid in it and dissolve out the uranium. Stainless steel does not dissolve in uranium, so it’d be very simple. The uranium would come out, but not the stainless steel. Well, it turns out that when you have uranium irons at great energies, it varies itself into the stainless steel.

Groueff: It doesn’t come out?

Larson: It doesn’t come out. It’s a great consternation when you go to dissolve it. Less than half of it came out. There was great consternation at the thing. I was afraid of that particular thing, so I had done some work on copper-plating the receiver—copper-plating the stainless steel—so that now, if the uranium comes in, it buries itself in copper and does not reach.

There’s a rather moderately thin coat of the copper. Copper has the property of dissolving in nitric acid, so it’s easy to get it out. So then you would get out the copper and uranium in solution. Then it’s quite easy to separate the uranium from the copper. The dissolved stainless steel is just an impossibility, and it would ruin the receivers too. You’d have to have a new receiver each time and they were very expensive.

That’s how we solved that. There’s an awful lot of things and arguments that went back and forth with merits, but that’s eventually how we did it. I told your other—let’s see. Who was it that came down?

Groueff: Ms. [Shirley] Tawse?

Larson: Yeah, Ms. Tawse. I told her about this particular part of the story. This worked out quite well, and was the standard way of recovering the material all the way through. So it worked.

Groueff: Because there is no use separating it if you can’t take it out—

Larson: If you can’t recover it.

Groueff: But physically, they would take all those receivers by hand?

Larson: That’s right. They would take it over to a special building, which was equipped with nitric acid sprays that would dissolve this material and recirculate it until it was all dissolved out. It’s just special equipment.

Groueff: Is it true that you couldn’t keep too many of them close to each other for fear that they may develop some criticalities?

Larson: Yes, that’s right. However, there were rather small amounts at first, and we did not have to take special precautions until we got into the next stage. I think I’ve described how you first performed this operation and how you got the material in the receivers, but that wasn’t quite pure enough as far as isotopic ratio was concerned. The U-235 was not purified. It had too much U-238 in it. So then you had to take that and put it back into another machine called the beta process. It was sort of a two-stage process. You would take the material that you collected first and then—

Groueff: Which was enriched, but not enough?

Larson: It was enriched partially, but not enough. Then you would take that in a smaller machine. Since that material is now very valuable, it had special liners and a lot of refinements in it to make sure you didn’t lose anything. Then you’d put it through again. The second time you received it was enough for weapons.

Groueff: The first time it was about fifteen percent?

Larson: Yeah, roughly.

Groueff:  And Thornton worked mostly on the beta [calutrons]?

Larson:  Yes, that’s right.

Groueff:  Now he told me that it was a big problem. His waste problem was tremendous. In the first stage, in alpha, even if you lose something it’s not important, but in beta it is.

Larson: In beta, if you lose it, you’re gone. The point of it is that only ten percent of it gets into the beam. You see the ninety percent just goes all over the wall. If you lose about ten and you’re only ninety percent efficient in recovering that, you’re losing as much as you’re preparing almost. 

Groueff: Only ten percent goes to the receiver?

Larson: Yes. It was a tremendous problem to recover as much of that material as possible. Of course, that’s where the second difficulty came in. When they went to recover this material, they found that this was one of the competing processes. I lost out on the original. They decided to use a competing process, which looked good. I had no particular objection to it. But it turned out it was very difficult in operation, and it held up so much material in the process. It was also a rather long-winded process with so many steps. You lost a little material at each step. It got to be almost an impossibility. It looked like we were going to just lose all the material. 

If you lose a little bit at each step, you finally end up with nothing. So it looked like an impossibility. I had worked out a process for after this material was in beta—the ninety percent that got on the walls and was washed down and had all these impurities in it. The process that I had worked out was a direct precipitation of the peroxide. 

Uranium has a very unique property: it’s precipitated by hydrogen peroxide, just the type that you buy in the drug store. If you add that to uranium, you’ll get a precipitator. It’s one of the very few elements that precipitates as a peroxide. Thorium does and uranium, but only those up in the very high range. So it precipitates as a peroxide. Since almost nothing else precipitates as a peroxide, it was a good way to separate the uranium from all the rest of these impurities.

Now there’s one disadvantage: iron is a catalyst for the decomposition of peroxide. All the experiments that we did originally had difficulty, because the peroxide would decompose and it would just froth all over the room, because the iron was a catalyst for the decomposition of the peroxide. So you add peroxide and the uranium precipitates nicely, but in a few minutes it starts to decompose. Then it heats up as it decomposes, which makes the reaction go faster. Finally it just switches out; it just goes all at once. 

So it was decided not to use it on that basis, but I worked out some inhibitors for the decomposition. You’ve heard of the iron, and that was a little uncertain. So again, it wasn’t adopted. Then I thought of an idea, since in some of my research work I had encountered this same problem in connection with the isolation of some biological materials, from complex chemistry of biological materials. I think you may have heard that body fluids and plant proteins and things like that are very susceptible to decomposition during a process of chemical isolation. So what they do many times is carry out those reactions in the cold. When you keep them cold, there is very little decomposition. So I thought, well, maybe if we use the same principle there, you wouldn’t get decomposition of the peroxide. That turned out to be the case.

Groueff: It reacted like biological?

Larson: Yes. This is true with almost any chemical. I mean to anybody who sat down and thought about it afterwards, it’s obvious. This is the thing that anybody should have been able to think of, but for months it escaped all of us.

Groueff: So, in other words, you put it in the freezer?

Larson: Well, actually, we carried this out in large vessels, which had a double wall, and we circulated cold brine around that. That kept the solution cold, and that was all that was necessary.

Groueff: So the peroxide remained quiet?

Larson: Yeah, it remained absolutely quiet. No trouble at all.

Groueff: That’s the way you treated the—

Larson: The beta.

Groueff: The ninety percent which were on the walls.

Larson: That’s right.

Groueff: So if the efficiency was only ten percent, it was very important that the wall material was recovered?

Larson: Yes, completely recovered, as much and as simply as possible. You couldn’t afford to have a process that had a large number of steps in it, because you lose a little bit at each step. This is the nature of it. 

Groueff: How did you scrape it from the walls?

Larson: The walls were stainless steel in the beta process. They were smaller units, and they were more expensive and difficult to construct. They were washed with nitric acid to in order to dissolve it.

Groueff: Again every three days or so?

Larson: Yeah.

Groueff: That will be enriched material, but like the one coming from alpha, it’s not yet separated?

Larson: Not yet separated. Just exactly the same.

Groueff: So what you do? Put again through beta?

Larson: Then you put it again through beta.

Groueff: And again and again, and ninety percent of that will—

Larson: And in the meantime, more is coming to make up that ten percent from alpha, you see. And so it keeps on—

Groueff: So the same material sometimes will be reprocessed about ten times? 

Larson: That’s right. Some of these molecules have probably gone through a hundred times, you see.

Groueff: So it’s a tremendously painstaking operation?

Larson: Yes, that’s right. Those are the general difficulties. The product from the beta process was bomb strength material. 

Groueff: That went to Los Alamos?

Larson: That went to Los Alamos. That’s right.

Groueff: All this happened in a vacuum?

Larson: Yes. The calutrons of course are operated in high vacuum. That’s right, just like you saw over there. I think you saw how those—

Groueff: Yeah. I saw the— 

Larson: So you can see there are two major chemical difficulties. Actually, the second one was even more serious than the first, to recover that material from the beta process. In the first one, we used the inefficient process, but just nothing came out the other end. Number one, we lost a lot of the material because of this very complicated, multistep process. The second thing was that it took so long to get it back in.

So this direct peroxide precipitation saw both difficulties, but got the material out in pure form in a one-step process. Then it could be put back into the cycle. They would shift that to this peroxide precipitate, and then it was decomposed to the oxide, dried, and converted back into a compound. The compound would go into the bottle to be vaporized again.

Groueff: All this was done in another building?

Larson: Yes. Well, actually it was done in another section of the same building.

Groueff: But employing many people?

Larson: Oh yes. There must have been five to ten thousand people in the chemistry.

Groueff: For recovery?

Larson: For recovery of both the alpha and beta before we got through. Oh, it was a tremendous operation.

Groueff: The impurities came from what? The walls?

Larson: The walls.

Groueff: No matter how clean they were?

Larson: Well, they would corrode. The material is very corrosive; it just dissolved the stainless steel.

Groueff: Did you have some major cleanliness problems, like in K-25?

Larson: No. This was not the same type of problem.

Groueff:  And the vacuum problems? Tightness?

Larson: That was always a problem of getting better vacuum. Getting higher vacuum, and getting it in less time—getting these units pumped down so you don’t lose. It took a long time to reach the proper vacuum, so you could then turn on the arc. If that took too long it was a tremendous loss. It was very important that the vacuum systems worked properly all the time. But I didn’t work directly with the vacuum. I know that in general the problems and the improvements they made. They always had a big group working on the vacuum.

Groueff: Were usually all receivers filled exactly, with more or less the same thing? Or would some operators move, and it would be empty?

Larson:  Oh yeah. It was tremendous.

Groueff: You had a lot of [inaudible].

Larson: There were differences. If the vacuum wasn’t quite good or went down, you’d get very little separation. If the insulator failed after—sometimes after five hours—you had to take it out and wash it anyways. There’d be very little in the receiver, but you had to recover it. So there were a lot of variations, a lot of troubles and a lot of variations.

Groueff: Especially in the beginning?

Larson: Oh yes.

Groueff: With all those girls who didn’t know their business very well yet.

Larson: Yes, all kinds of problems in the operation.

Groueff: Do you remember the startup of alpha [plant], when they had the big trouble and nothing worked?

Larson: Well yes. That was of course the impurities in the magnet. That was a dark day because there was doubt in some people’s minds that it really could be fixed.

Groueff: So the whole process risked—

Larson: Oh yes. It was sort of blank despair, you might say, at that time.

Groueff: All over the place?

Larson: Oh yes. Well, all these big units, everybody sitting around ready to operate, and nothing could go. I mean you could just imagine thousands of people, sitting around, waiting.

Groueff: And it lasted for what? For an hour or for days?

Larson: I’ve forgotten now, but it was a couple of weeks in my impression.

Groueff: Because they had to send it back to Milwaukee?

Larson: That’s right. Some of them [magnets] had to be sent back. I think they did install filters, and get some of it out without having to send it back. I mean it was a variety of treatments.

Groueff:  But in the meantime, thousands of people stood idly every day for weeks? 

Larson: That’s right.

Groueff: And morale?

Larson: Oh, morale was just terrible at that time. But fortunately, I think it was within perhaps a week that we knew that we could get out of our difficulties.

Groueff: But did it happen at the beginning? Or did it work for a few days first?

Larson: Well, as I remember, it worked partially well, and then you’d get [electrical] shorts across. The impurities and the iron, all the iron junk they had in there, I think it was just sort of erratic. Then finally it just sort of operated very poorly, and then finally—

Groueff: There was no separation and no arc?

Larson: Well, the magnets had nothing to do with the arcs, but if you don’t have a magnet, the arc won’t go into the right place. So you couldn’t really do anything.

Groueff: Was Lawrence here [Oak Ridge] at that time? 

Larson:  Oh yes, he was here. I was so busy with the chemical process at that time. I was in on some of the things with regard to the analysis of what the material was, but not directly. I mean I was working twenty-four hours a day on the chemical difficulties. We knew we were into them very early on in this thing.

Groueff: At that time they had also changed the manager here. It was Williams, I think, and they put in [Frank C.] Creedon?

Larson: Yeah. Well, there were some changes there.

Groueff: Groves describes in his book that he was quite mad at these things.

Larson: Well yeah, Groves had that particular thing. Of course, there were a lot of these things which shouldn’t have happened, but of course, they will. They do in almost any process.

Groueff: So during first few weeks, or first few months, there were a lot of breakdowns, and difficulties, and things didn’t work?

Larson: Well, there was just one thing right after the other. Things were so hectic. Number one, of course, came the magnets. Number two: we couldn’t get the material out of the alpha receivers—in the stainless steel—so we had to change that process. I remember with that we were faced with getting only an average of about thirty to forty percent out of the stainless steel receivers, no matter how hard we washed them with the acid. I had anticipated that we might run into difficulty, and so I had some experiments going on with regard to the best ways to copper-plate the stainless steel receivers. They were rather odd-shaped.

Groueff: You didn’t do it on the pilot plant stage or at Berkeley?

Larson: No, this was done afterwards.

Groueff: Did you have the same problem in Berkeley?

Larson: No. In Berkeley, one of the problems was that the energy of the beam was not as great. Also, there was so much uncertainty with regard to how much should be in there. Perhaps we were only getting maybe fifty or sixty percent out. But we wouldn’t know it, because we didn’t know how much should be in the pocket because the operation out of Berkeley was very erratic.

You can predict how much should be in there, because they put an electrode in there and then ran a meter back, so that every time one atom would go into the receiver, this would read on the meter. Then you could predict on the meter readings how much should be in that receiver. When we got down here, we found out that it was less than half of it. But at Berkeley, they’d run the meter reading for fifteen minutes, then close down. You really didn’t know what— 

Groueff: You also didn’t have this problem of a very strong beam.

Larson: That’s right.

Groueff: It wasn’t embedded into the receiver?

Larson: That was part of it, but I think there were both things. Even if you got some uranium out, it wasn’t realized that we might have been getting out only part of it. But since we didn’t know how much was in there, it was—

Groueff: Actually, you were learning during the production stage?

Larson: Yes. Well, the production went into effect without the copper-plating, and I can remember talking with Lawrence on this particular problem: one of these crises again. He said, “Obviously the only way to get out of this difficulty is to copper-plate.” 

I said, “Yes. We have all the data.” We had determined the proper shape of the electrodes and the currents and the thickness. I said, “Well, gee, if we get complete control of all of the forces and fabrication, we could probably get this in a couple weeks.”

He said, “Hell, we want this done and started tomorrow!”

We rigged up very elementary things. We didn’t have any proper vessels. I found that there were a lot of these big stoneware sinks that had been installed in several places throughout. We jerked all those out, and made those into the baths for the electrolyte.

Groueff: You did the plating yourself?

Larson: Oh yes. Well, we went into an area and took over that area.

Groueff: But the same day you mean?

Larson: We worked all day and all night, but the next day we were—

Groueff: Wasn’t there something that you would send somewhere else? To another city?

Larson: Oh no.

Groueff: You’d do it right there and then?

Larson: Well, you couldn’t afford to hold up those very expensive things.

Groueff: Are you supposed to know, as a chemist, how to plate with copper and all those things?

Larson: Oh yes. But this is something—

Groueff: And you had to do it for hundreds of—

Larson: We had to determine how thick to make the plate and— 

Groueff: Yeah. 

Larson: Also, you read in the textbooks it’s not easy to paint stainless steel and we found that out. But we solved the problem without too much trouble actually.

Groueff: It was done in a matter of days?

Larson: In twenty-four hours, we had the first one from the time the order was given. We had the first ones being plated.

Groueff: And it worked?

Larson: It worked. That was the end of that particular difficulty.

Groueff: That gives me a very good example of the whole atmosphere, because now, with a case like that, you would write somebody to make it for you.

Larson: That’s right. And then you’d get the engineering done for the proper size bus bars, and you’d order the right kind of rectifiers.

Groueff: And some red tape also.

Larson: We happened to have some direct current motor generator sets for another purpose. We just took and moved over to them because you need direct current. That’s the sort of thing that was—

Groueff: The kind of pioneering spirit.

Larson: Yes. Due to the fact then that we introduced copper, some other changes were required in the chemical process, which I then went to work on and so forth. After we got that solved, we ran into trouble with the beta process. It was just one thing right after the other like that.

Groueff: It wasn’t at all a smooth beginning.

Larson: No, it was a very rough beginning. I was concerned about this particular thing, and voiced my concern to Lawrence that we really should have done more pilot plans and practicing. He said, “Really, there’s such a small amount that if necessary you can do this in beakers.”

It was very difficult to convince him that the chemistry needed a sophisticated approach and sophisticated engineering to in order to obtain the results. In spite of the fact, supposedly, the product was only going to be a very small amount. Before he got through, he was very appreciative of the chemists. In fact, I think he thought the chemists were a very minor role at first, because he felt that the physics problems dominated. In a sense he was right. Because the chemistry problems, while difficult, we knew that we could overcome them. 

Groueff: They were not without precedent. They were in the field of the non-chemist.

Larson: Yes. That’s right. But the thing that made it difficult is that there was no real chance to run complete pilot plant experiences to iron out these difficulties. So the difficulties had to be ironed out right at the time at the site.

Groueff: Yeah. That’s very typical for the whole project.

Larson: Yes, that’s right.

Groueff: How did he react in the case of a crisis and big difficulties? When a difficulty comes and everything stops, how would Lawrence react?

Larson:  Well, he always had this confident approach, responsible but confident.

To get back to the interviewer, you were pointing out what is the reaction of Lawrence. I mean naturally, he was very concerned, but I would say always reflected confidence.

Groueff: He never lost his faith?

Larson: No. For the rest of us, there were some desperate things. Our morale would be way down here.

Groueff: Not his?

Larson: Not his, no.

Groueff: Even when the magnet thing stopped for several days?

Larson:  I mean he reflected confidence that we would get out of it. At least that was my impression.

Groueff: Did he get upset in the sense of shouting and getting furious or mad at people?

Larson: No, no, no. He would get very angry with people, but I would say he always kept himself under control.

Groueff: All right.

Larson:  He would get very angry with people who didn’t—

Groueff: I understand that he always demanded the maximum of everybody.

Larson:  Yes.

Groueff: He was giving the maximum of himself?

Larson:  Yes, that is right.

Groueff: Also, the machines—they tell me he would turn everything. His radio or his car—he would press them. The accelerator.

Larson: Oh, yeah. He is a great experimentalist at trying to maximize things. He demanded terrific and high performance of everybody. Of course, one of the things that was amazing: he was able to attract such high-type people. Of course, you know the number of Nobel Prizes that have come out of Berkeley.

Groueff: A lot.

Larson: Just a reflection, and there are probably another one-half dozen who are probably almost as well qualified as those who got the Nobel Prize. A tremendous reservoir of talent that he would recruit; it was just fantastic. The same with the graduate students that come out of Berkeley: they were attracted to him. Then those would become leaders in physics throughout the world, so this was one of his characteristics.

Groueff: But there was some wrong interpretation [of nuclear fission] they had, no?

Larson: Well yeah.

Groueff: I mean Lise Meitner and [Otto] Frisch helped with the interpretation.

Larson: Yes, that is right.

Groueff: They discovered it without scrutinizing.

Larson: Well he [Otto Hahn] discovered this, and pointed out that he had isolated barium, which was about half the atomic weight. He said, “As a chemist, I know it is there, and I know that it is barium, but as a scientist, I cannot believe it,” or something like that.

Groueff: Because he expected the heavier element, no?

Larson: Yes, that is right. He expected the heavier element, because it was against all theories and so on. So it was Frisch and Meitner who came out with the theory that would account for the presence of the barium.

Groueff: She was the first one, I understand, who had this very revolutionary idea: what if the laws are wrong and Hahn is right?

Larson: Yes, that’s right.

Groueff: [Enrico] Fermi before him also made a mistake in the interpretation.

Larson: Well yes, of course. Fermi had worked with this for about four years. As a matter of fact, he had undoubtedly made tremendous amounts of barium in his researches, but he did not recognize it. If he and his chemists had gone after and tried to identify the barium, they could have done it, but they did not realize it, so it led to all kinds of impossible interpretations.

Of course, once you know it can be done, then it is very simple. In fact, in our classes, we repeat Hahn’s experiment. Isolate the barium.

Groueff: So now it is too easy, no?

Larson: Now you can do it with students, but Fermi had missed it for four years. Having it right there in his test tubes right before him all the time.

Groueff: What was the difference between Fermi’s and Hahn’s experiment? Did Hahn do the same thing as Fermi to bombard the uranium?

Larson: Yeah, I would say it was the same. They were both amazed at all of the different—it did not dawn on them. They would isolate all of these very peculiar radioisotopes. They could not understand that there could be that many heavy elements.

Groueff: Mrs. [Irène Joliot] Curie also did something similar?

Larson: Oh, yeah. They were working on it too.

Groueff: And they all missed?

Larson: They all missed.

Groueff: So it is difficult in cases like that to say who did it. Because without Lise Meitner, probably, we still would not recognize it. But she couldn’t do it without Fermi’s ground.

Larson: Yes. Hahn came out with that very vital piece of information that beyond any shadow of a doubt, the barium came from uranium.

Groueff: Yeah.

Larson: Or it certainly was there.

Groueff: He was the kind of extremely meticulous chemist, so you would—

Larson: You would have full confidence in his results.

Groueff: That is what I think she [Meitner] said. He would not be the man that would have barium by some sloppy thing or by forgetting something.

Larson: Yeah, that is right. It was a very significant piece of work, of course.

Groueff: It is an amazing case to have to choose between the established laws of science and the work of a man, and to decide if he was right.

Larson: Yeah, that is right. This is an extremely difficult type of thing. Incidentally I might add, as just a sideline on this, the man who coined the name “fission” is actually an employee of ours in the biology division.

Groueff: Somebody mentioned that, yes—what was his name?

Larson: Dr. [William] Arnold.

Groueff: Arnold. 

Larson:  He was there at Copenhagen when—

Groueff: Niels Bohr?

Larson: With Niels Bohr and when Meitner came there. She was describing her theory as to how the atom was split apart. He said, “Well, that is just like the fission of an amoeba.”

Groueff: Because the word was not used in physics and chemistry?

Larson:  No, that is the first. I think he was the first one to point out the word.

Groueff: Fission, yeah.

Larson:  Yeah, defined the word fission. Of course, I know that fact has never been published.

Groueff: No.

Larson: That is very interesting.

Groueff: Because now it becomes one of those words that—

Larson: That everybody thinks has always been in physics. But, of course, in biology, it has been used for one hundred years, I would say.

Groueff: Were there mostly physicists among the first great pioneers of the projects, rather than chemists? Who were the chemists?

Larson: I think you must remember Seaborg.

Groueff: Seaborg, yeah.

Larson: Seaborg was the exception there.

Groueff: But Lawrence, Fermi, [Arthur] Compton—

Larson: Yes, as far as I know, they are all physicists. [Eugene] Wigner—

Groueff: Wigner, [Leo] Szilard—

Larson: Szilard. They are all physicists.

Groueff: Now I was told a story about the plutonium work. Because it was a completely new element— the chemistry of plutonium I understand is extremely complicated—all the work was done at the beginning. They had to start with something, and they had the assumption that plutonium would behave more or less like uranium-235. And they did a lot of work on that. 

Larson: Wait a minute. Uranium-235 has the same [inaudible] as U-238.

Groueff: I see, as uranium.

Larson: Yeah.

Groueff: They worked on that, and everything worked well, and the bomb was built and exploded. By now they realize that the whole assumption actually was configured wrong, because I understand that plutonium has very different characteristics. In other words, the results were achieved not because of a theoretical assumption, but for some different reasons. And they were looking at this work the wrong way.

Larson: Of course, it is a little difficult. And I guess you would probably have to get the real story. Are you planning to interview Seaborg by any chance?

Groueff: Yes. 

Larson: You are. So you will get the whole story. He loves to talk about it incidentally. 

Groueff: But I talked to a chemist in Argonne [National Laboratory], Dr. [Joseph] Katz. He was one of the assistants. 

Larson: Yes, he was there with Seaborg.

Groueff: When we discussed about all those things—how did you know this and microchemistry, which sounds rather complicated to a layman. How do you work with non-existent things and just a few molecules? So after a few questions, he said, actually, “We did not know.”

Then he told me that some of the assumptions were proven later to be wrong, but luckily, it did not make a difference. Also in the metallurgy of plutonium, he told me that they had some wrong assumptions because when heated, plutonium has different phases.

Larson: Yes.

Groueff: What was the story? I have it from some of the metallurgists. They used the extrusion knifing to make – what was the other one?

Larson: They must have got those in alpha, beta, and gamma phase.

Groueff: They tried the one and did not succeed. So they took the other metal that they thought was the worst, and luckily they did so, because they did not know that by heating it more, they would have run into a problem and all the slugs would have burst and expanded.

Larson: Yeah. 

Groueff: So there were a lot of those lucky—

Larson: Oh, yeah, extremely lucky things there. There was a lot of luck connected with it, and there was a lot of rather brilliant insight you might say. So it was a whole combination of that. Well you will get the full story of the plutonium. That is a fascinating story.

Also, I am sorry you did not get to talk to Dr. [Stafford] Warren about going into Hiroshima immediately after the bomb was dropped. You see, before the war was ended actually, they made special arrangements with the military so that he could go in there and start getting the data from Hiroshima. He does not like to talk about it with his immediate family, but I have understood from others that he received the sword of surrender from the garrison because he was a colonel at Hiroshima. A few years ago, when he went back to Japan, he took the sword back with him and presented it to the family. In the meantime, the colonel or the general had died.

Copyright 1965 Stephane Groueff. From the Stephane Groueff Collection, Howard Gotlieb Archival Research Center at Boston University. Exclusive rights granted to the Atomic Heritage Foundation.