Nuclear Museum Logo
Nuclear Museum Logo

National Museum of Nuclear Science & History

William J. Wilcox, Jr.’s Interview (2005)

Manhattan Project Locations:

Bill Wilcox was an original resident of Oak Ridge, TN, and served as the Official Historian for the City of Oak Ridge, TN. A chemistry graduate from Washington & Lee University in 1943, he was hired by the Tennessee Eastman Company on a secret project in an unknown location he and his friends nicknamed “Dogpatch.” He worked with uranium, which was referred to only by its codename “Tuballoy.” Wilcox worked at Y-12 for five years and then at K-25 for 20 years, retiring as Technical Director for Union Carbide Nuclear Division. Wilcox actively promoted preservation of the “Secret City” history through the Oak Ridge Heritage & Preservation Association and by founding the Partnership for K-25 Preservation. He also published several books on Oak Ridge, including a history of Y-12 and “Opening the Gates of the Secret City.”

Date of Interview:
September 12, 2022
Location of the Interview:


Bill Wilcox: My name is Bill Wilcox. William J. Wilcox, Jr.

Cindy Kelly: And how do you spell Wilcox?

Wilcox: W-I-L-C-O-X.

Kelly: Why was Oak Ridge chosen for the Manhattan Project?

Wilcox: It was one of the first actions of the new Manhattan Project, under Colonel [James] Marshall in the early summer of 1942. They started looking for a place to build these plants that they were going to have to have. They, of course, knew at the time that they were going to follow two courses to the bomb, the plutonium option and the uranium-235 option. But they had no idea what process was going to be used to separate the isotopes, or how big these production reactors were going to be. They just knew that they had to have a site somewhere to do all of this work.

They soon settled on the Tennessee Valley region. They started down close to what is now Watts Bar Dam. But they gradually came up the river and settled on this area that we are in today, the Oak Ridge Reservation. Their criteria were that they wanted a lot of land, because they did not have any idea how big these plants were going to be. So the reason they settled here was that they, first of all, needed about 60,000 acres.

When General [Leslie] Groves looked at it in September, he immediately decided that he would not put the production reactors here. But when it was first picked out, the site was to be for the whole Manhattan Project: the uranium enrichment, the production reactors, and the bomb laboratory.

Back in May, though, when they were really settling on this site, they wanted about 60,000 acres to do the job. They wanted a place where they had good electrical power, and we had TVA [Tennessee Valley Authority] here then. They wanted some place that was sparsely settled, so they did not have a huge problem moving out a town or moving out lots of people. Here there were just a thousand farms in this 60,000 acres. They wanted, however, to be close to a good labor supply, and they were only thirteen miles from Knoxville. That was a real plus also.

They wanted cooling water for their reactors if they needed them or for the diffusion plant, and they had the Clinch River surrounding this. They wanted good rail service, and we had two fine railroads serving the area. Those were the main criteria.

Also—a sign of the times—they wanted to be far enough from the coast, the Atlantic Coast, which during 1942 was completely overrun by German submarines. U-boats controlled the Atlantic. They wanted to be far enough from the coast so they did not have to worry about the U-boat threat.

The final selling point, the topography and the geology of the place, was just very attractive to them. Because in East Tennessee, the feature of the landscape is long, low parallel ridges about a thousand feet tall, with valleys in between. There is just one right after another. It’s like you took a Damask linen tablecloth, and put your hands on a slick surface and push it this way, and it comes up in long folds. What this meant was, they could put the town in one valley and the plants in another valley. That would make them easier to build fences and safeguards and secure a place. In the event of some untoward accident to one of the plants, it would not wipe out the town at the same time. In other words, it provided a little natural barrier.

Those are the reasons that they selected Oak Ridge. Colonel Marshall had his people studying the site all summer long. They did one study after another of this feature and that feature, and, “Could we do this, and could we do so and so?”

When General Groves took over the project in September, he came down here and characteristically decided in one day, “Let’s acquire the land.” He did not want to do any more studies, didn’t want to do any more planning. He also decided right then and there that it was entirely too close to Knoxville for the production reactors, that traces of radioactivity would undoubtedly get to Knoxville and give away the secret of what we were doing here. And so he wanted to put it somewhere.

Instead of 60,000 acres, his people went out to Washington State and got 500,000 acres for their wonderful production reactors. When [J. Robert] Oppenheimer was first approached by General Groves and told that he wanted him to be responsible for the bomb laboratory, and where would he would like to have it. “And by the way, we have this site in East Tennessee.” Oppenheimer immediately said no. He did not want that. It turns out that the [Oak Ridge] site was used primarily for the isotope separation plants.

Kelly: That is great. Learn something every day. That is terrific. And you answered the second question, too. Let us see, talk about how they decided what ways to use to get this isotope separated?

Wilcox: This problem that Oak Ridge was given, the mission of Oak Ridge, was to follow this route to the bomb, which the physicists in 1939 and 1940 felt was the most direct route to the bomb, because if they could just possibly get their hands on a couple of hundred pounds of pure U-235, they could create a chain reaction by assembling a super critical mass. The neutrons created would form a chain reaction and result in the release of energy from the nucleus—the atomic bomb. It sounds just great.

But some physicists said, “You do not know what you are talking about. Separating the isotopes of such a heavy element is probably something we will never learn how to do. Might not even be able to do it, because it is an incredibly difficult job.”

The reason is, of course, that you cannot separate isotopes the way you separate other elements, like separating iron from iron ore. You give that job to a chemist or a medical worker. It may be difficult, but it can certainly be done. People have been doing it for centuries, iron from iron ore, aluminum from aluminum ore.

But separating uranium or other elements’ isotopes is extraordinarily difficult, because these two forms of uranium that you dig out of the ground are identical. The way they act in any chemical reaction, anything you do to it—make metal out of it, or make liquid out of it, or make gases out of it—both of the forms of uranium just go along with each other. There is no separation that occurs in any of these normal steps.

I would like to illustrate the difference between the isotopes, which is only in weight, in just how much they weigh. If you take two identical basketballs and you ask me to show you the difference in weight, all I have to do is to Scotch Tape a nickel to one of those basketballs. That is the relative difference between U-235, which you can make a bomb out of, and U-238, which you cannot make a bomb out of. How do you separate these two things? Now, these are atoms and not basketballs.

It is an extraordinary job. You have to invent some process that has them work a little bit differently, because this one is just a little bit different in weight. That had never been done in 1939 with an element anywhere near that heavy. It had been done by Professor [Harold] Urey with heavy water and regular water, heavy hydrogen and light hydrogen. But there the difference in weight is tremendous relative to the weight of the atom. Jesse Beams at the University of Virginia had separated the isotopes of chlorine-35 and 37, but there again, the relative difference is pretty huge. Here you are talking about 235, 238, where the difference is very, very, very small.

You asked, what processes did they think about? The word is “think about,” because nobody had done it. And so the question was, “How could we possibly do this?” One of the first methods that was suggested by the British was one called gaseous diffusion. There what you do is, you take advantage of the fact that in a gas where all the molecules are the same temperature, they are moving at slightly different speeds if one molecule is a little bit lighter than another.

So the U-235 in a gaseous mixture, the U-235s are going to be moving just a little bit faster. If you provide a structure where it is diffusing through small holes—and I mean tiny, tiny, tiny holes—you might be able to get a little bit more of the 235 through there than the 238. It is not a sieve. It is molecular interactions.

Another process that somebody knew about from work that had been done years before was thermal diffusion, and there you use liquid. Here you have got to use liquid uranium hexafluoride gas, which is an extremely difficult material to work with. It is almost as bad as gaseous uranium hexafluoride. But under high pressure, real high pressure, at very high temperature, if you have a thermal gradient across a liquid stream of UF6, there is a difference in the rate at which this material will diffuse because it is a little bit lighter than the liquid U238F6.

There was another process. I mentioned Jesse Beams had separated chlorine. He did that with a gas centrifuge and a long cylinder that he spun at extremely high speeds. The heavier chlorine-37 gas moves, under centrifugal force, like a whip at an amusement park. Lots of force. That 37 goes out a little farther than the 235. They said, “Well, maybe we can make that work with uranium.” In 1940, Dr. Beams at the University of Virginia started seeing if he couldn’t do it with uranium hexafluoride.

So there were those three main processes. Then in 1940, Dr. [Ernest] Lawrence at the University of California, who had invented the cyclotron, or spinning nuclear particles, suggested that that might be a way. He might be able to modify a cyclotron so that he could start spinning, if you will, or just accelerating a gas with the 235 and 238 in it, so that you could spin this under a magnetic field instead of running it around in a whirling metal cylinder like a gas centrifuge.

He suggested that we modify a cyclotron to do this. He was very enthusiastic, an optimist, and a brilliant physicist with a team of marvelous associates. By 1942, when the Manhattan Project came along and got serious about doing this on a large scale, that was added to these other three processes as a candidate.

By then, Ernest Lawrence and his team had designed a specially new physics gadget where they turned the cyclotron sort of up on its side. It had a magnetic field going in this direction, and the uranium gas—in this case uranium tetrachloride, UCL4—introduced at the bottom, ionized with an electron beam and then accelerated out of this slit, but in a magnetic field so that both of the U-235 beam and the 238 beam are forced to curve in a circle.

Then up at the top, he put a receiver with slits in it, and, again, the 238s take a larger radius of curvature and go into a slit up at the top of this eight-foot tall machine separated from the U-235 beam by maybe a quarter of an inch or so. There again, the simple theory, the simple concept, sounds not too difficult, sounds pretty reasonable. But boy, putting it into action just was an incredibly difficult job.

At the end of the year in 1942, General Groves was in a position—got his advisory council together, and he said, “Look, if this is going to make a difference in this war that President [Franklin] Roosevelt wants, we have got to make a decision. We cannot just let the scientists keep working in these different universities trying to improve and improve and improve, and making this better, better, better. We are going to have to pick how we are going to go.”

Each one of the processes had its strong advocates. They were all honest physicists, and they all had to admit that they could not guarantee success. Groves was left just having to make a decision between these competing processes, each of which had its attractions and its liabilities, each one of which was a gamble.

His choice was that, first of all, he ruled out thermal diffusion because it looked extraordinarily difficult, both experimentally and theoretically. He stopped work on that. But the U.S. Navy continued working on that. Dr. Philip Abelson kept working on that, but it was not part of the Manhattan Project.

The second one that he ruled out was the gas centrifuge. There it was not because of the theory, which looked solid and there was no question about it, but the mechanical difficulties were extraordinary. Jesse Beams’ machine in 1942 had bearings at both ends that had to be lubricated because this thing is spinning many, many miles per hour, rotational speed on duraluminum cases, and extremely beautifully machined so that they are perfectly balanced. You cannot have any bearings like you have in a ceiling fan. You cannot have any wobble. This thing has got to be almost perfectly balanced.

Then you have got these very difficult bearings at the top and the bottom, which had to be lubricated with oils, and there you’ve have got to separate that from UF6, which reacts violently with oils. The mechanical problems were just enormous.

So that left him with gaseous diffusion and the electromagnetic process, the University of California process. Of those two processes, gaseous diffusion and the electromagnetic, it seemed to him and to his advisory committee that the California process was the best bet. Because the gaseous diffusion process—which also in theory is solid—but the key to the gaseous diffusion process is this little thing I told you that has got all these holes in it, that the molecules are going to bounce through.

Nobody by the end of 1942 had made a piece of this porous material that was any good. They were brittle, and they had holes in it, pinholes in it. The British were working on it, and we were exchanging technical information with the British. But the British could not do it. We could not do it either, in spite of the fact that scientists at Columbia University who started working on this porous material with the holes in it—which we have since come to call “barrier,” which is the key to this process, having a material with all of these holes in it. Even though they started in 1940, ’41, or ’42, they still had not made a barrier that was any good.

Of course, I am an old research and development man myself, and success is always around the corner. We never run out of ideas of something else to try. Of course, that is the way they talked in ’42, too.

He decided to put his main efforts on the California process, which was admittedly a gamble, but worth doing. Because if there was any way we could develop a method, put this to work, and get some U-235—a couple hundred pounds—we might make a real difference in ending the war. That is the story about the selection of those two processes.

He went to work right away in ’43 designing and building the California process plant at Oak Ridge which was codenamed Y-12, and had a thousand when we got through with Y-12, got through building it in two years. It had 1152 of these University of California designed machines, which they named “calutrons.” “Cal-” for California, and “-tron” for cyclotrons, and “-u-”, calutron, University of California. California University cyclotrons, calutrons. But designed specifically for separating the uranium isotopes. And they worked beautifully.

It was a fantastic gamble and a monumental achievement, when you think that they broke ground in February of ’43. The decision was made to go ahead with it in December of ’42, and that project started in the summer of ’42. Groves took over in September. In December, they made the decision to go with the calutron, and they broke ground in February before they even had a design for the equipment.

They got Stone & Webster in as an A-E [Architect-Engineer] and told them to build the plant. “How many buildings do you want?”

“Well, we are not really sure, but let’s build them.”

“How big?”

“They need to be big enough so that—” and so on and so on and so on. Just did a fantastic job.

The heart and soul of that machine were these huge electromagnets, eight feet tall with iron cores all wound with some conductor. As soon as they start talking about where they are going to get these magnets, somebody said, “Well, wait a minute. We got a war going on. You cannot possibly get the thousands of tons of copper wire that you have to have to wind these magnet cores. It will shut down our production of tanks and the airplanes, motors. Every motor that we put in vehicles for the war effort has to have copper windings.”

Somebody said, “Silver is a better conductor than copper. Why don’t you use silver?”

“Where are we going to get silver?”

“The government has got tons of it sitting up there at West Point Depository for silver up on the Hudson River. We have got tons of ingots sitting up there.”

“Hey, that’s a great idea!”

Colonel [Kenneth] Nichols, who was in charge of the Oak Ridge operations, went up to Washington and sat down with Undersecretary Bell of the treasury and said, “We have got a secret project. I cannot tell you what it is, but we would like to borrow some silver by borrowing silver from the vauls. We are going to keep it under lock and key, barbed-wire fences. We promise we will give it back as soon as the war is over. It will make a huge contribution to the war effort as a substitute material.”

And Undersecretary Bell said, “Of course, and we know about substitute materials. My wife is using margarine, and we do not have any butter. Silver for copper sounds like a good deal.” He asked, “How much do you want?”

Colonel Nichols said, “Well, we are thinking about somewhere like a few thousand tons.”

Bell just jumped up from his chair. He said, “Colonel Nichols, here at the Treasury Department, we do not talk about silver in terms of tons. Our units of measurement here are troy ounces.”

Colonel Nichols said, “I will figure out how many troy ounces that is.” It is about 300 million in troy ounces, or something.

We ended up at Y-12 borrowing from the Treasury 14,000 tons of silver. Those ingots were taken out and shipped under guards and so on, and rolled into sheets and made into bus bars and windings for all of these 1152 magnet cores, the cores for the magnets for these units at Y-12. Fourteen thousand tons, $300 million worth at that time out of the Treasury reserves.

General Groves insisted they keep very good track of it. All of the machine turnings of dust and so on was all collected and accounted for and weighed. After the war, was all taken apart and given back to the Treasury. General Groves in later years bragged about—I cannot remember the figure right now, but it was well over 99% of the silver that was turned back.

You have things like that going on. You have 25,000 workers, construction workers in the middle of 1943. I said they broke ground in February. By summer, they had 25,000 construction workers at the Y-12 site, building that plant and designing pieces and building it all along. What an accomplishment.

It is hard to believe, but by November, the first calutron was ready to get started in operation. Groundbreaking was in February, with the designs unsure, uncertain at all. By golly, by November that same year, they had the building up, all the infrastructure, all the diffusion pumps, all the electrical supplies, and they were actually able to start running a machine.

It was a disaster. The magnets shorted out and came to a screeching halt. General Groves had insisted that they just go ahead. They did not have time to setup and run a pilot plant of these calutron units. Groves said, “Look, this is a war. We have got to beat Germany to the bomb. We will do the experimenting on the plant.” Even though he was very unhappy with the fact that the machines did not run well the first time, he said, “Let’s shut her down, and figure out what is the matter.”

They found that the magnet coils were too close together, and there was dirt and oil that had not been cleaned well enough, and so on and so on. They figured out what the fixes would be. They sent some of the magnets back to Allis-Chalmers in Milwaukee to get them rewound. But there were lots of other magnets in the production cycle stream in the production process. They fixed those right up there at the time.

By November or December or January, they had calutrons going in somewhere else in the plant, somewhere else in the production cycle. As I said, there were over 800 of them in the big machines, in the Alpha machines, and 400 in the smaller machines. In January, they were ready with the machine that had been fixed, and they ran those close to the end of January, I think January 27th or so was the actual date. The top brass, Tennessee Eastman officials and Army officials, came and watched the first run, and it was a success. Then it was just a matter of bringing more calutrons on stream, and that happened all through 1944.

The first production run really that worked was in January of ‘44. But when you look now in hindsight, where they started in February or even January of ’43, to us it almost seems unbelievable. A remarkable accomplishment.

I arrived in October of 1943 before they started the calutron operation, but I was in the chemistry end which was in an entirely different part of Y-12. I knew nothing at all about the calutrons, the magnets, the silver. There were just those big red buildings, and I had no idea what was going on in those big red buildings.

I was a chemist, and I was in the chemistry building, which was finished in October of ’43. What I mean by “finished” is the outside and inside walls and the heating. There wasn’t any cooling. The heating system and sewers and so and so on. That was all ready for us in October.

Those of us who were hired out of the college spring graduating classes of 1943, the spring classes. Tennessee Eastman and the other Manhattan Project contractors all over the country scoured the graduating classes of 1943 and picked up all the chemists, chemical engineers, physicists, electrical engineers, mechanical, and anybody with technical backgrounds. They hired us all. I went to work in May of ’43.

As I said, groundbreaking here for the first plant was in February. There wasn’t anywhere for us to come in Oak Ridge, so they shipped us all somewhere else. Tennessee Eastman’s people all went to Eastman Kodak in Rochester, New York. That is where I spent the summer along with, I guess, forty or fifty other college graduates. We were working in Eastman Kodak’s laboratories under lock and key in their beautiful research building they called Kodak Park in Rochester. The Kodachrome developing laboratories for the whole country were on the first floor, ground floor, and we were up on the second floor.

We spent the summer learning uranium chemistry, developing our own understanding of how to separate it from—when we get impure uranium, get the uranium-235 that is enriched in these calutrons, you cannot make bomb-grade material in one step in a calutron. You get a lot of separation in that first step. But the highest enrichment you can achieve there in that first step was about 15 percent, 10 to 15 percent. Not enough to make a bomb out of.

We had these big eight foot units, over 800 of those in a series of four different buildings. Then we had a second stage called the “Beta” calutrons. Those were half as big. You put the 15 percent material into that, and what you get out is 80 enriched plus, 80 to 90 percent. That is what you want for the bomb. You have these two stages.

But the chemists are needed because when this stuff comes out of the Alpha units, you have to scrape it or dissolve it or dig it or something to get it out of the units. Then it gets all junked up or contaminated. If you use nitric acid, for example, to make sure you get every bit off the stainless steel, what that means is, you end up with a solution. Your U-235—if you use nitric acid—your U-235 nitrate, uranyl nitrate, comes along with contamination by iron, nickel, copper, molybdenum, and all the other elements that are in stainless steel.

So the chemists have to purify them from the chemical standpoint. Get the uranium away from those other elements, make it real pure, and then convert it to uranium tetrachloride. It is a long series of chemical steps to purify it and convert it to uranium tetrachloride, which is exactly like the uranium tetrachloride over here when you go to the Alpha calutron. Chemically, it is exactly the same. But isotopically, this is now 15 percent instead of seven-tenths of a percent. The product is what you want, the bomb-level highly enriched uranium.

That is what we learned to do all summer, and then came down here and went to work doing it—getting ready to do it on a production scale. But as I said, I got down here in October. We had to buy all the laboratory equipment, get it all set up, and continue our research and development before we ever started getting any materials to really work with until January and February of the next year.

It was a major effort. By the time Y-12 got into full production in the spring of 1945—that’s when we made the bulk of the material for the first atomic bomb. There were trickles all through ‘44 going to Los Alamos, but most of it went in the fall of ’44 and the spring of 1945. That is when most of the U-235 was produced and shipped out to Los Alamos.

At that time, Y-12’s employment cadre was up to 22,400 people in one plant. Goodness sakes, it was more than work in the plants in Oak Ridge today, all of Y-12. It was a busy, busy, busy place, and many of them—I do not know what the fraction is. I am tempted to say about half of them were involved in the chemical end of it, half in production and support.

That is Y-12’s story. It was just an amazing place. The materials and the supplies that went in there to Y-12 to build the buildings just came in boxcar after boxcar. Somebody said that the total railroad shipments into Oak Ridge Reservation in those early wartime years was 3,000 boxcars—full of lumber, concrete, supplies, iron, steel, and everything else—3,000 boxcars a month, and they all went out empty. The railroad people kept wondering, “What was going on here? Month after month after month after month, and we don’t get anything out!” 

Of course, people in town did not see any trucks going out. The people in Oak Ridge would have been terribly disappointed to learn that the total product of this huge outfit was going out in an attaché case a couple of times a week, chained to the wrist of a military security lieutenant dressed in plainclothes with a couple of plainclothes guards going with him.

 Of course, the reason is that what you are doing is just taking that small amount of U-235 in the original uranium ore that you dig out of the ground—if you dig out 1,000 pounds, you only get seven pounds of 235. In a thousand pounds! So the damn stuff is scarce as hen’s teeth. When you get it pure, you process up lots and lots of this stuff coming in. Then you’ve got process inefficiencies, so you do not get anywhere near 100 percent of the stuff out and you have a lot of process losses all along. You really feed a tremendous amount of stuff into this process, and get a tiny little bit out.

By the spring of 1945, every bit that went to Los Alamos was going out in these little nickel cans about the size of a coffee cup. A little nickel can heavily plated with gold on the inside, and it went in the form of uranium tetrafluoride, a salt. Pretty green, blue-green salt crystals. Highly purified from a chemical standpoint as well as from an isotopic standpoint. Eighty percent plus U-235.

The research chemists at Y-12 took that precious product and converted it to uranium tetrafluoride, which would make it very easy for the Los Alamos people to reduce it to a metal, turn it into uranium metal, what they needed to machine to make the parts for the “Little Boy” bomb. We sent it as uranium tetrafluoride in these little cans, screwed a lid on top of them, and put two of them in one of these attaché cases, if they had that much material to deliver, then send it out.

I think most people at Oak Ridge would be quite distressed to know all this effort was producing just that little dab of stuff. But of course, over a period of six, seven, eight months—we produced some all through 1944, and then the rest of it through the first six months of 1945.

By July, Los Alamos finally had enough to make the first atomic bomb with, the Little Boy bomb. Y-12 achieved its almost impossible mission that General Groves set forth in separating the isotopes of uranium, and doing this in two and a half years. An almost incredible mission.

Cindy Kelly: Can you talk a little bit about K-25?

Wilcox: Yes. As I said, go back to December ’42 when Groves is picking between the processes. The people at Columbia University and the British overseas had urged that gaseous diffusion, the theory, was better known—or not better known, but let’s say well-known. If they could just get their hands on this barrier material, it would be probably an even better process than the calutron.

From an engineer’s standpoint, the calutron process is what we call a batch process. I said you had to put some uranium tetrachloride in the bottom, heat it up, and vaporize it, and then it comes out at the top. This machine runs for about three, four, or five days, and then you have ended up vaporizing all the UCl4 down there. You have to take the machine apart and break the vacuum, pull the machine out. You have got 800 of these things, so you have got people running all over the place taking these things apart, putting them back together, and so on. People, people, people, people. It’s a batch process because now you have to load them up again, run it for four or five more days, and stop it and charge it up again, and so on.

The gaseous diffusion process, the concept is that you put this gas through this membrane, this barrier material, and you get a separation, and that’s nice. If everything works right, you get a little bit of separation at each stage, but that’s the hooker—you get a very small amount of separation. So what you have to do is to take that gas you get out after passing it through one time, and put it through another stage. When you put it through another stage, every time you go through the barrier, you lose pressure. You have to put pressure here to get the gas to come through. Now you have to take this gas, and pump it up again and run it through another barrier and so on and so on.

How many times do you have to do this? It turns out about 3,000. You have got to have a huge number of these tanks with these barriers in it, and compressors and valves to control the pressure, and gas coolers to take the heated compression out of the gas at each stage so it does not keep getting hotter and hotter. You end up with a mammoth plant, because you have to do this 3,000 times.

“If they could ever the barrier,” General Groves said, “This plant could be better, because all we do is switch on the plant and it just runs. We do not have to take this apart and fix it every time. Here is the great benefit, if we can make this work.”

General Groves and his top policy committee decided that what they would do is, get the Y-12 Plant in there. That was their first sure bet. That was surely their first good bet. I want to emphasize how much a risk these people—Vannevar Bush, J. B. Conant, Groves, and the rest of these intellectual giants and leadership giants—were taking, because it was just a real gamble.

What they were saying was that, “Y-12 is our best bet, and then let’s hedge that by starting the gaseous diffusion plant. Since these guys cannot make one of these barriers yet, let’s give them as much time as we can.” He did not start it [building the K-25 plant] for another year after Y-12. Actually, ground was broken in September of ’43. Ground for Y-12 was broken in February, and I mentioned that the first units were actually started up in November.

K-25 was started in September [1943], groundbreaking, but its first units did not start operating until January of ’45. It was a year behind Y-12. When Groves started talking to the engineers about this K-25 plant, he found out they were talking about building a building to house all of these tanks that are going to have this barrier material in them. They called them diffusers, 3,000 of them. The building was going to take forty acres of ground, and it all needed to be under one roof, and it all needed to be hot, so that the UF6 did not condense and turn into a solid. That is just the beginning of the difficulties that these engineers faced in building the gaseous diffusion plant at just the beginning.

But they did take that risk. They made that decision. They started building. The civil engineers, the people that built the building, they could do that. They leveled out the ground perfectly, because they could not have any settlement cracks going on with all these pipes in there. 

But the barrier material—that is, the key to the whole thing—was never developed until the summer of ’44. By then, the building was already completely committed to, and it was about almost half built. They had to go ahead and build all this stuff, and order these compressors and order the tanks and all this stuff, and the guys in the laboratory are still not able to make this barrier material.

This is an incredible job. I hate to keep using that word, but it just jumps out at you when you look at the job that they had to do. This barrier material has got to have holes in it, very small holes. Very, very small holes. Small enough so that in a square centimeter—say the size of your thumb nail—you have over 100 million holes in area the size of your thumbnail. You have got to have acres of this stuff. The holes all have to be the same size. If these tiny little holes are too big, the gas just flows through them, and you get not separation at all. If they are too small, the gas condenses on the walls and flows through the holes as a liquid, and you get no separation at all. So they have got to be the same size. You cannot have any pinholes. A terrible job.

You have got to have it be flexible, be able to move it or bend it around, make different shapes out of it. It has got to be strong, in other words. You have got to have strength so that you can put pressure on one side and pump it through something. Oh, and by the way, this has to be chemically resistant to uranium hexafluoride, which is a real nasty, corrosive material. Because if you have any corrosion taking place over a period of time, those little holes will get a little bit bigger, or maybe they will plug up, so you cannot have that.

Oh, and I forgot to mention, by the way, we cannot have any air leaking into this plant. This is all going to be under a vacuum. When they told the construction contractor that this whole building with 300 miles of piping and all this stuff had to be just as vacuumed tight as a thermos bottle, the construction contractor says, “I cannot imagine anything like that.” 

Then they told him, “Oh, and by the way, we cannot have any oil or organic material.” For example, no fingerprints on the inside of that piping. So every time you turn around this corrosive UF6 gas, you cannot have any holes in the piping anywhere that air will leak in, because air brings moisture in. Moisture reacts instantly with UF6 vapor and turns it into uranium oxyfluoride, UO2F2, which is a beautiful yellow material, but it is a solid. In other words, if you let any air—just the moisture in this room’s air—if you let it into a tank of UF6, it hydrolyzes UF6, turns it into this nice yellow smoke, and it plugs the holes. So we cannot have any holes in this forty-acre plant.

I think it is just amazing that Groves would take that chance. The engineers just did an absolutely brilliant job, Kellex Corporation. Manson Benedict and Dobie [Percival] Keith were probably the lead engineers. They just did an absolutely fantastic job of designing this K-25 U [shaped plant]. J. A. Jones, a construction contractor from Charlotte, North Carolina, did a fantastic job of building it. It cost $512 million dollars to build that plant. Probably six or seven billion dollars today. It was done in eighteen months. And by golly, right from the beginning, it worked. Absolutely beautiful.

Groves says after the fact, he said one time that it was probably his greatest gamble of a war. He thought that if they had not come through with that barrier at the last minute, he said the Congressional investigation probably would have gone on for the rest of his life. But by golly, the thing worked.

Because it was a continuous process, they turned that switch on. Then they kept putting pieces, these stages, they put more and more on stream over a period of a year and a half. By the end of 1946, the top product enrichment that K-25 [made] was now at bomb level, and Groves just instantly shut down Y-12. The cost of making the stuff at K-25 was less than 10 percent of what it was at Y-12. It was just a much cheaper process.

Of course, the improvements that were made in the next decade in the expansion program, the arms race with Russia, brought the cost way, way down below what it was for that wartime plant. That was made possible with civilian nuclear power reactors, which today would not be economically feasible at all with the original K-25 process efficiencies.

There was a third isotope separation plan at Oak Ridge, but it had a lifetime of only about one year. Some of us here at Oak Ridge feel like it’s our forgotten story. But in terms of our total costs, it really didn’t figure largely in the Manhattan Project.

Oak Ridge spent 1.1 billion dollars by the end of 1945, which is about 60 cents out of every Manhattan Project dollar. Biggest part of the Manhattan Project. Now, at Oak Ridge, when you look at the expenses, Y-12 was $478 million, so it is a huge plant, nearly half a billion dollars in wartime dollars. K-25 was a little more than that, $512 million.

X-10, which was the code name given to the production reactor that was built here by the University of Chicago Met Lab, Arthur Holly Compton and his great team at the University of Chicago. They first wanted to build that production reactor right there in Argonne Forest. Groves and others, DuPont, talked them out of that. They put it down here on the Oak Ridge Reservation, Clinton Engineer Works.

That was the production reactor that they had to have in order to get their hands on some gram quantities of plutonium, so that [Glenn] Seaborg and his people, chemists, could develop the separation processes for the Hanford plant, the separation of plutonium from all the fission products from the reactor slugs.

They had to have that, so that graphite reactor was built here. It was our first success. Just like Y-12, ground was broken in February of ’43, right away. Of course, the Chicago people knew how to design that. They had done the Stagg Field [experiment], and the just souped that up, but the theory was there.

Some people wanted it built as a pilot scale for the pilot reactor for Hanford, but it was an air-cooled reactor. As things worked out, they went to water-cooling [at Hanford], so it really wasn’t a pilot plant, but it was a production plant. It went critical in November, November 4th, the same time that later that month that the first calutron was tried and it failed. But the graphite reactor went critical November 4th, and it was a success from then on.

All through the war, they produced plutonium. They produced a total of about seven-tenths of a pound over the next couple years, two and a half years. The present Oak Ridge National Laboratory evolved out of that particular X-10 operation. That is where Oak Ridge National Laboratory is today. But during the war, compared to the two half-billion dollars plants, ORNL cost $27 million.

The plant that I was starting you the story about, the S-50 plant, was $16 million. So it was a small piece at Oak Ridge. Nevertheless, it was our third isotope separation plant. The town cost $96 million. That makes up our total of a little more than 1.1 billion dollars.

To get back to S-50, I mentioned earlier that when Groves was looking with his advisory committee at the possibilities, thermal diffusion was one, as a process. The Navy kept a small program going on thermal diffusion through the war. It was not part of the Manhattan Project. But they were very interested—right from the beginning—in the possibility of using a nuclear reactor to power submarines. They knew that they were going to have to have highly enriched uranium as a fuel. So that was their interest in thermal diffusion.

Phil Abelson, a brilliant researcher out of University of California, one of Lawrence’s protégé’s, kept that work going. They got to the point in 1944 where Abelson started building a pilot plant of a number of big thermal diffusion columns in the Philadelphia Navy Yard, up in Philadelphia. J. Robert Oppenheimer, who had kept track with a lot of his old friends from California, heard about what Abelson was doing.

Of course, Oppenheimer was extremely interested in doing whatever he could to increase the quantities of U-235 that he could get for the first weapon. It occurred to him that, “By golly, we might use a thermal diffusion plant to augment the output of the other two plants at Oak Ridge.”

He went to Groves, and Groves was talking to him all the time in the course of the operations. The subject came up that Oppie said, “Well, General Groves, I understand you have got this power plant built up there for the K-25 gaseous diffusion plant, but it is going to be done six months before your gaseous diffusion plant needs the power.”

Now, Oppie was not talking about electric power. He was talking about the superheated steam that turns the generators to make the power. He said to Groves, “Hey, you are going to have this capability of putting out this 900 degree steam. I know just who could use that steam to give you a little bit of enrichment.”

Groves perked up his ears. The upshot was that Groves said, “Hey, instead of building that pilot plant in the Philadelphia Navy Yard, let’s build it down here at Oak Ridge, right by the steam plant. Instead of what you’re talking about there, let’s build 20 of them, because I do not want pilot plant, I want a plant.”

He went to one of his great contacts that he had known before in the construction business, H. K. Ferguson, a great design architect engineer in Cleveland, Ohio. H. K. Ferguson’s widow was running the firm. He talked to her. Groves does his, “Great national interests, vital war works speech” to her. “How about designing this plant for us? This guy will tell you how to build these columns and so on and so on and so on. It is another we need that plant on stream in 75 days.” They got this plant going, the first column, started in 69 days in 1944.

Right along the banks of the Clinch River, across from what’s now being developed as “Rarity Ridge”—a big black tall building with 2,000 columns of that thermal diffusion columns. The specs were so tight that even though they sent them out to something like 20 companies in the United States, there were only two companies that even responded.

In the middle, they have a nickel column, 45-feet long. This is a nickel pipe 45-feet long, completely straight, absolutely straight. Outside of it is a copper tube, absolutely straight. The gap between the nickel and the copper has to be exactly so many tenths of an inch. I mean, this is very, very small. The difference between the success of our plant, which was successful, and the Japanese that wasn’t and the German that wasn’t, was that their gaps were a lot bigger than ours. Ours were a very small gap.

But in that gap, you are going to put your liquid UF6 under high pressure. Inside the nickel tubes, you are going to run 900-degree steam, high-pressure steam. Then on the outside of the copper, you are going to run water, from the river, say. Whew. So you have a tremendous temperature gradient across that. The UF6 up at the top, liquid UF6, turns out will have a little higher U-235 concentration than the UF6 at the bottom. Forty-five feet, 2,000 of these. Incredible! All packed into this one building.

I talked to somebody this summer that was here for our Secret City Festival. He worked in the plant, and he said, “The noise was tremendous.”

I said, “Noise? You didn’t have any compressors in there.”

He said, “It was leaks from the high pressure steam. When that high-pressure steam finds a little bit of a hole and goes out, it either screams or it hollers. A terrific noise.”

The startup problems were huge. Even though they got it started in September or October of 1944—they built it during the summer, the 69 day period was over, I think, in September. But they didn’t really have the cascade on stream and start putting out product until early ’45. Then the product concentration in U-235 enrichment was a little under one percent. This doesn’t have a very large separation factor, not like the calutron. Or it wasn’t really tiny like the gaseous diffusion, but it was somewhere in between.

But the plant put out product of about one percent, and that was fed to Y-12. Then when K-25 got going, it went instead to K-25. K-25’s product started going to Y-12 also in the spring of 1945. But again, this was part of the big cascade. The enrichment level in say March or April of ’45—which is the time that Y-12 was sending the urgently needed 235 to Los Alamos, trying to build up enough of an inventory for that first bomb—it was one or two percent.

So K-25 and S-50, although you have to say that they contributed to the U-235 that went into the first bomb, it wasn’t a major contribution. Colonel [Kenneth] Nichols, and Richard Hewlett’s great history of the Department of Energy, Hewlett and [Oscar] Anderson, he says that Colonel Nichols asked his production control committee to figure out after the war, “How much of a contribution did S-50 make?”

Their answer was that it speeded up the time in which the total amount of material required for the first bomb was achieved. It speeded that up by about nine days. I think that is what is in Nichols’ book. Hewlett and Anderson, I think, say about a week. When you look at speeding it up by say a week, and you figure the cost of the war per day, and compare that to 16 million dollars, there is no question that S-50 was extremely worthwhile.

The difficulties in running S-10 were sufficient, so that one month after the war ended in August—in September—that is to say, in September they pulled the chain on S-50 and shut it down, in September of ’45, you see, after it had run a total of a year. But only part of that was really putting out good product.

But that is the S-50 story. The only relic that is down there today that some of us preservationists are trying to get saved, the only relic is two small smokestacks. They are the remnants, the only thing left of even the K-25, the powerhouse. The powerhouse is completely gone. The S-50 plant is a green field. But there are two smokestacks up there.

I was telling about how Oppenheimer got Groves’ interest aroused in using this plant, before it was needed for K-25. They did that. They accomplished that. But the time did come halfway through that production period when the K-25’s team had to go to K-25 power to make K-25 power. So then what are you going to do? Are you just going to after this S-50 plant is all up and running, now you got to shut it?

The U.S. Navy sent in three boilers from destroyers, and made a powerhouse to produce steam and set up a tank farm for diesel oil. There must be ten or twenty of these huge tanks with dikes all around them. That fed the three boilers from destroyers that made the high-pressured steam, when the K-25 could no longer give them steam. They made their own steam.

These two smokestacks are right at the end of their powerhouse. It is the only thing that is left. It is a relic. But I would like to save them, so that we can mark the site of that third fantastic accomplishment at Oak Ridge.

I had one traumatic experience in my employment. I came within a whisker of getting fired. It was in February of ’44. I had been very fortunate right from the beginning, because I got in on the ground floor. I learned there was uranium chemistry up in Rochester. We couldn’t get it out of books. There weren’t consultants that knew all about uranium chemistry. There wasn’t any internet. Everything that had been published on uranium fit on two or three pages. We really learned by experimental uranium chemistry—how to test for it, spot tests and how to measure it, how to analyze for it. It was a very intense learning experience that summer.

When we got to Oak Ridge, we started hiring other people, of course. We ended up training all of them and ended up being the supervisors, the front line foremen, even though we were kids. Our experience qualified us.

But I was chosen to handle what I think now, looking back on it—I was not told at the time—was probably the first real quantity of enriched uranium to come out of the Alpha calutrons, the ones that started in January. I think it was sometime in February, late February. It went right through my laboratory, the laboratory where I worked.

I remember the boss impressing on me that this particular lot, which I think was about 200 grams—half a pound, plus or minus—had to be handled with very, very great care. They wanted to be certain it didn’t get contaminated.

Of course, what he was talking about was not just chemical contamination. What he was talking about was contaminated with other uranium because we had uranium all over the place, but you can’t tell whether it was enriched by looking at it. You have got to label it and keep it separate and so on. We kept this separate. This was in a hood all by itself. No other uranium anywhere close to it, and so on and so on.

I followed it through the purification cycle. That involved extraction and precipitation and so on. Finally, we ended up precipitating it with hydrogen peroxide. It produces a UO4, which looks like a lemon colored, lemon clustered, beautiful precipitate. Then you filter it out. For filters, in order to keep it pure and clean, we used what the chemists call a Buchner filter, which is Pyrex glass, and used brand new ones.

We had all this brand new equipment, of course, to make sure that this doesn’t get contaminated in any way. But it is a glass funnel about this big around, and about that tall. Then it has a centered Pyrex fret in it, a porous material, very, very small holes, so that the water goes through and this lemon custard sits up here on top of it. I got that all done. Sucked the water out of it, and got it ready. Then the next step is to put this in a real hot oven and calcine it, to convert it from UO4 to UO3. Then it goes into the next step.

Well, I did all of that just fine. I can’t remember exactly what the boss said, but I think he suggested that I stay with it all night to make sure that when it calcined, everything worked well. I had done this many, many times. We had rehearsed all of these things with just regular uranium, so we knew the chemical properties and so on and so on.

I had done this many, many times. I thought, “Boy, this is going to work good. To make real sure that nothing breaks, I am just going to just provide a little extra conservatism. I am going to put this in a stainless steel can.” What we usually did was in those early days was, we put everything that was in glass, we put it in stainless cans underneath it, so that if anything happens.

Well, the Buchner funnel has a bead at the top of Pyrex glass, sort of a swollen ring around it at the top to make it a little stronger and then one down here at the fret. When it’s set down in the stainless steel funnel, this band here at the bottom, this little extension here on the bottom of the funnel, sat right down on my stainless steel. I said, “Well, that is perfect.” Then the tube sat down in the stainless steel.

I stuck it in the oven, and calcined it. It turned a nice color like it does, from that lemon custard to a nice deep yellow, like powder, but real pretty. I said, “Well that’s just fine. Hell, I’m going home. I got a date.” I turned off the oven.

Then the next morning I came in and somebody saw me and said, “You better get the hell out of here. The boss is looking for you, and he’s so mad he can spit. They’re even talking about giving you the axe.”

I said, “Oh, my goodness! I am going to go up and talk to him.”

He said, “I wouldn’t do that if I were you.”

I said, “Why? What the hell happened?”

He said, “Something happened to that thing you calcined, that batch you had last night.”

I said, “What could have possibly happened?”

Anyway, I went and talked to the boss. I was just horrified that anything had happened. What had happened was, the hot funnel put into the stainless steel can to protect it, the stainless steel can expanded. The funnel dropped down into the can. Then when everything got cold, the can came down and crushed the Buchner funnel, so that it just powdered all that glass.

There wasn’t a bit of the material lost. But there were two days lost, or one day, in getting that stuff—I am sure now it was going to go out to Los Alamos for its precious batch, for cross section measurements or all kinds of other things. They were dying to get some enriched uranium. Nobody had grams of it to work with before.

It cost at least a day. But they had to dissolve it up and precipitate it again and calcine it again. So my name was mud, but all was forgiven. I just felt terrible about it.

That is fortunately as close as I came in forty plus years to getting fired, at least that I know of. [Laughter]

People sometimes ask me, “How did you guys in Oak Ridge react to the dropping of the bomb on Hiroshima, all of those casualties?” My answer is, “We reacted with the same great surprise and relief that the rest of the country did.”

We had 75,000 people here. I think probably 72,000 or 73,000 had no idea of what we were doing here. There were very few people that really understood the whole magnitude of what was going on at Oak Ridge, thanks to General Groves’ compartmentalization policy, which really did work. We knew what our jobs were, but we didn’t know the whole picture.

So the reaction was, “My goodness sakes! So this is the vital war work we have been doing. My goodness! How relieved I am that it looks like it’s going to really have an impact. I do not see how Japan can continue.”

Nobody that I know gloried in the deaths of the 100,000 people in Hiroshima, any more than they gloried in the deaths of that same number in the bombing of Tokyo, the firebombing the night of March 9th and 10th, when about that same number of people were killed and 16 square miles of downtown Tokyo were burned out, compared to the four square miles at Hiroshima. It was terrible, but the firebombing of 60 other Japanese cities was terrible, too, and Dresden, Germany. That was terrible, too. It was in the context of the war, however.

The relief that we felt turned to exuberant joy just a week later, when we woke up to our morning Knoxville newspaper and there was an eight-inch high banner headline that said, “Peace.” This blessed peace that we had been praying and working so hard for, for all these years, was finally a reality.

It seemed very clear to us that the success of the Manhattan Project efforts, the two bombs—the first one of which did not bring the Japanese to surrender, no. It took two of them. But the success of the Manhattan Project is what drove their reluctant Emperor Hirohito to go to his diehard militarists and insist on bringing to an end this war that they started.

That’s what we ended up being proud of, and that’s what we celebrate today, that we had a real role in bringing peace to a world that had been torn by the worst war in history for six long years. 

Copyright 2018 The Atomic Heritage Foundation. This transcript may not be quoted, reproduced, or redistributed in whole or in part by any means except with the written permission of the Atomic Heritage Foundation.