# Uranium haxafluoride gas **URL:** \ **Category:** Factual Questions\ **Created:** [July 8, 2004, 5:06pm UTC](https://boards.straightdope.com/t/uranium-haxafluoride-gas/254141 "2004-07-08T17:06:37Z")\ **Posts on this page:** 15\ **Page:** 1
**Author:** ![gcoochy](https://avatars.discourse-cdn.com/v4/letter/g/dfb087/32.png) [@gcoochy](https://boards.straightdope.com/u/gcoochy)\ **Post date:** [July 8, 2004, 5:06pm UTC](https://boards.straightdope.com/t/uranium-haxafluoride-gas/254141/1 "2004-07-08T17:06:37Z")
What material was used for the filters used in purifying Uranium Hexafluoride gas during the gaseous diffusion process on the Manhatten project? ---
**Author:** ![Squink](https://avatars.discourse-cdn.com/v4/letter/s/b5e925/32.png) [@Squink](https://boards.straightdope.com/u/Squink)\ **Post date:** [July 8, 2004, 5:58pm UTC](https://boards.straightdope.com/t/uranium-haxafluoride-gas/254141/2 "2004-07-08T17:58:13Z")
> [@](#): > > In the late 1930s, Miller carried out research into chemically resistant materials from which he later developed the **chlorofluorocarbon polymer** used in the first gaseous diffusion plant for the separation of uranium isotopes, a crucial factor in the development of the atomic bomb. The fissionable isotope Uranium-235 is separated from the more abundant isotope Uranium-238 by selective diffusion of uranium hexafluoride gas through barriers, or filters. [William T. Miller](http://www.news.cornell.edu/Chronicle/98/12.3.98/obit.html) ---
**Author:** ![MikeG](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/mikeg/32/4329_2.png) [@MikeG](https://boards.straightdope.com/u/MikeG)\ **Post date:** [July 8, 2004, 6:01pm UTC](https://boards.straightdope.com/t/uranium-haxafluoride-gas/254141/3 "2004-07-08T18:01:50Z")
Make sure you credit **Squink** when you turn in that there assignment now! 🙂 ---
**Author:** ![bonzer](https://avatars.discourse-cdn.com/v4/letter/b/45deac/32.png) [@bonzer](https://boards.straightdope.com/u/bonzer)\ **Post date:** [July 9, 2004, 12:00am UTC](https://boards.straightdope.com/t/uranium-haxafluoride-gas/254141/4 "2004-07-09T00:00:02Z")
Except that **Squink** ’s answer is wrong. From his cite: > [@](#): > > The problem faced by researchers was creating process materials (lubricants, gaskets, pump oil) that would not react with the highly corrosive uranium gas. Because of his expertise in organofluorine chemistry, Miller was recruited to work on the Manhattan Project. He and his colleagues developed novel polymers that did not react with the uranium gas, making possible the separation of the fissionable isotope. Nothing to do with the actual filters themselves. This is confirmed by the mention of Miller’s role in the Smyth Report (_Atomic Energy for Military Purposes_, 1945; Princeton, 1945, 10.49) in connection with “coolants and lubricants”. A highly non-trivial problem, to be sure, and significant to the success of the project, but seperate from the issue of the filter material. Interestingly, Smyth played coy on the material used in the filters. (Though he also didn’t reveal the details of what Miller and others had developed either.) However, this was declassified by the time Hewlett and Anderson came to write _The New World_ (Pennsylvania, 1962), the first volume of their official history of the AEC. The filters were made out of **nickel**. They discuss the development of these filters in some detail. Though there’s a clearer and more up to date summary in Rhodes’ _The Making of the Atomic Bomb_ (Simon & Schuster, 1986, p493-5). ---
**Author:** ![Squink](https://avatars.discourse-cdn.com/v4/letter/s/b5e925/32.png) [@Squink](https://boards.straightdope.com/u/Squink)\ **Post date:** [July 9, 2004, 12:19am UTC](https://boards.straightdope.com/t/uranium-haxafluoride-gas/254141/5 "2004-07-09T00:19:02Z")
> [@bonzer](#): > > Except that **Squink** ’s answer is wrong If wrong, it’s the result of widespread disinformation: > [@](#): > > In the late 1930s Professor Miller carried out research into the chemically resistant materials from which he later developed the chlorofluorocarbon polymer used in the first gaseous diffusion plant for the separation of uranium isotopes, a crucial factor in the development of the atomic bomb. The fissionable isotope Uranium-235 was separated from the more abundant isotope Uranium-238 by selective diffusion of uranium hexafluoride gas through barriers, or filters, made of the polymers he developed. [Cornell Chem and Biol 1999](http://216.239.51.104/search?q=cache:UUooC_O_1doJ:www.chem.cornell.edu/alumni/pdf/news70.pdf+%22filters+made+of%22+%22gaseous+diffusion%22+%22Uranium+Hexafluoride%22+manhattan+&hl=en&ie=UTF-8) How does the nickel story go? Did they have some fancy sintering process to get small pore sizes, or were the holes drilled out somehow? I can see how they could easily develop a fluorine resistant plastic with a controlled pore structure, but it’s not obvious how they’d do that for nickel using 30’s technology. ---
**Author:** ![bonzer](https://avatars.discourse-cdn.com/v4/letter/b/45deac/32.png) [@bonzer](https://boards.straightdope.com/u/bonzer)\ **Post date:** [July 9, 2004, 1:09am UTC](https://boards.straightdope.com/t/uranium-haxafluoride-gas/254141/6 "2004-07-09T01:09:32Z")
Rather than labouriously type in extracts from either Hewlett and Anderson or Rhodes, I’ll point to [this page](http://www.nap.edu/html/biomems/hurey.html) which obviously derives from the latter: > [@](#): > > However, the barrier remained recalcitrant. Copper barriers were abandoned, and efforts were concentrated on nickel barriers. Both electrodeposited (Norris-Adler) and compressed powder barriers were tried. As 1943 wore on, it was realized that, despite heroic efforts, barriers with the properties, uniformity and ruggedness necessary for manufacture were not available. Nevertheless, a pilot plant for manufacture of the electrodeposited barrier was being completed … > In the autumn of 1943 a new type of diffusion barrier, combining features of both the electrodeposited and compressed powder barriers that had been previously developed, was proposed by the Kellex Corporation. In the spring of 1944 a plant began producing acceptable barrier material of the previously developed electrodeposited type, whose production had been urged by Urey. Ten thousand workers had been building a huge diffusion plant at Oak Ridge. Early in 1944 the Army (General L. R. Groves commanding) made the decision to rely on the barrier developed at Kellex. Fortunately, both types of barrier eventually proved satisfactory. Similarly, from [this page](http://www.udayton.edu/~hume/A-Bomb/abomb.htm): > [@](#): > > The actual K-25 plant was designed by a company named Kellex (Kelle for Kellego and X for secret). Union Carbide was responsible for the building of the K-25 plant. The Chrysler Company built the large metal diffusers for the gaseous diffusion plant. Chrysler took a large risk in building these diffusers because the diffusers had to be built out of nickel, which was in short supply, in order to withstand the strong Hexaflouride gas that occurred during the process. If Chrysler did not produce the diffusers that were needed they would lose large amounts of money as well as their reputation. Chrysler however was dedicated to helping end the war and devised a way to plate stainless steel with nickel. Mass production of diffusers began and they were given the code name X-100. The construction of the chemical plant was difficult to build because Uranium chemistry was unknown. Barriers and pumps were needed for the uranium separation process and two companies Houdaille-Hershey and Allis-Chalmers Company built them. The barriers were constructed in Decatur, Illinois and the pumps in Milwaukee, Wisconsin during the spring of 1943. None of this seems to have involved Miller. Exactly what the Kellex process involved is left vague in all sources. ---
**Author:** ![commasense](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/commasense/32/3017_2.png) [@commasense](https://boards.straightdope.com/u/commasense)\ **Post date:** [July 9, 2004, 1:10am UTC](https://boards.straightdope.com/t/uranium-haxafluoride-gas/254141/7 "2004-07-09T01:10:18Z")
From the previously cited _The Making of the Atomic Bomb_ by Richard Rhodes, page 494: > [@](#): > > A self-educated Anglo-American interior decorator, Edward Norris, had devised the electro-deposited mesh originally for a new kind of paint sprayer he invented; he joined the Columbia project in 1941 and worked with chemist Edward Adler, a young [Harold] Urey protege, to adapt his invention to gaseous diffusion. The barrier material that was eventually used was made by a modified version of this method developed by building contractor Kellex. See also a history of the K-25 plant here: [http://ma.mbe.doe.gov/me70/history/K-25.htm](http://ma.mbe.doe.gov/me70/history/K-25.htm) I’ve been inside K-25. It’s no longer in use, of course, but it’s an amazing facility. Thousands of pumps over this enormous building (at 42.6 acres, the world’s largest building under one roof at the time it was built!) all set to run automatically, with a minimum of operators. Take a look at the picture on the Web site in the link: each leg of that U-shaped building is half a mile long! The Manhattan Project is one of the most fascinating chapters of American history, and is well worth studying. Rhodes’ books are a good place to start, as is Groueff’s _The Manhattan Project_. Little-known factoid: uranium hexafluoride gas is explosive in the presence of grease, so a new material was needed to act as a lubricant for the valves used to route the gas around the plant. So DuPont (which had been selected to design the gaseous diffusion plant at Oak Ridge, TN) used a chemical it had discovered several years earlier, polytetrafluorethane. It remained a top secret throughout the war, and was first sold commercially in 1946, but didn’t become common in consumer products until the 1960s. We know it as Teflon. Interesting article about the history of Teflon here: [http://www.inventionandtechnology.com/2000/01/feature.shtml](http://www.inventionandtechnology.com/2000/01/feature.shtml). ---
**Author:** ![Squink](https://avatars.discourse-cdn.com/v4/letter/s/b5e925/32.png) [@Squink](https://boards.straightdope.com/u/Squink)\ **Post date:** [July 9, 2004, 1:19am UTC](https://boards.straightdope.com/t/uranium-haxafluoride-gas/254141/8 "2004-07-09T01:19:23Z")
Did a bit more digging. Nickel microtubules are used in modern gaseous diffusion plants: > [@](#): > > The filtration barrier or membrane through which the hex is pushed consists of micro-tubules of nickel about 0.01 µm in diameter. Crucially, these are resistant to corrosion by hex at 85 °C. Its manufacture has been one of the most closely guarded nuclear secrets. [Chemistry World](http://216.239.51.104/search?q=cache:DFo8mEgis1kJ:www.rsc.org/chemistryworld/features/CW00406F0046.htm+%22gaseous+diffusion%22+%22uranium+hexafluoride%22+nickel+filters&hl=en&ie=UTF-8) **However** , the IAEA wants to know about the importation of gaseous diffusion barriers or materials that can be used to make same into Iraq. Their list of proscribed materials includes: > [@](#): > > Specially prepared compounds or powders for the manufacture of such filters. Such compounds and powders include nickel or alloys containing 60% or more nickel, aluminum oxide, or **UF6-resistant fully fluorinated hydrocarbon polymers having a purity of 99.9% or more, a particle size less than 10 microns, and a high degree of particle size uniformity, which are specially prepared for the manufacture of gaseous diffusion barriers**. [Iraqwatch](http://www.iraqwatch.org/controlled/annex3text.htm) So you can make gaseous diffusion barriers out of fluorinated hydrocarbons, nickel microtubules, or aluminum oxides. Given Miller’s known involvement in the Manhattan project, I’ll bet that they used his polymers initially, then replaced it when nickel microtubules came along. There’s virtually _nothing_ available through google on how the microtubules are made. I can’t say I’m surprised. ---
**Author:** ![bonzer](https://avatars.discourse-cdn.com/v4/letter/b/45deac/32.png) [@bonzer](https://boards.straightdope.com/u/bonzer)\ **Post date:** [July 9, 2004, 11:23pm UTC](https://boards.straightdope.com/t/uranium-haxafluoride-gas/254141/9 "2004-07-09T23:23:25Z")
Sorry **Squink** , you’d lose that bet. I suspect you’re seriously underestimating the amount of information about the Manhatten Project that is publically available - and has been for quite some time. As we’ll see, some of the technical details remain murky, but what avenues were persued, by who, when, what the decisions were, what resources were required and what the politics were has long since largely become laid open. The debates around the design of the barriers are part of this and Miller and polymer chemistry simply have nothing to do with this. To repeat myself: it’s largely all in Hewlett and Anderson’s official history, which was pubished in 1962. They devote an entire chapter to the history of gaseous diffusion within the Manhatten Project, much of which is concerned with the design of the barriers. The proposals, the dead-ends, the leads. Virtually all of which were variations on using nickel. Norris and Adler’s method was just one of these. When it came down to the final decision on a barrier to use, the key design was that developed by Kellex; for a recent account of Kellex and their role in the barrier design, see _Racing for the Bomb_ by Robert Norris, Steerforth, 1992, p205-9. But like Rhodes, this ultimately derives from the aforementioned official history. Confirmation comes from Margaret Gowing. Gaseous diffusion was an area were the UK was particularly advanced early on. (It also helped that we didn’t have E.O. Lawrence pushing electromagnetic seperation as a competing technique.) Everybody agrees that when Groves came to make the final decision as to what the barriers at Oak Ridge would be, British input was significant. Thus the fact that Gowing’s British official history - _Britain and Atomic Energy, 1939-1945_ (Macmillan, 1964) - agrees with Hewlett and Anderson shouldn’t be ignored. And she goes through the early British research on barriers in extreme detail, quoting entire documents in her appendices. It’s all about nickel. Once the Americans adopt the Kellex process, the British government commited most of their nickel production (mainly from a mine in Wales) to barrier producion. K-25 was vast and required a lot of nickel to produce just the barriers. Further confirmation comes from Gowing’s official history of postwar British nuclear research. Cut off from American collaboration by the McMahon Act, The UK’s bomb project had to rely on what was known from our participation in the Manhatten Project. As already mentioned, when it came to gaseous diffusion, the Brits were utterly in the loop. So what happened: > [@](#): > > It was known, from collaboration with the United States, that nickel powder was the best material to use, and indeed the supplies for the American project were made by a British firm. _Independence and Deterrence: Britain and Atomic Energy, 1945-1952_, volume 2, _Policy Execution_ (Macmillan, 1972), p432. The British firm was that Welsh mine. And this is a reference to the Kellex process. So what was the Kellex process? All accounts that mention it agree that this was some sort of wonderful method of using nickel to produce barriers. It’s here that one runs into the coyness of official historians. The process produced some sort of powdered nickel that formed the barrier. Beyond that, the only statement seems to be in Hewlett and Anderson (p137): > [@](#): > > Kellex had produced the new barrier in the laboratory by a very simple hand process; if production could be multiplied by the use of thousands of employees doing piece-work, there would be no problems translating a laboratory process into an industrial one. This sounds much more like a way of grinding nickel to produce a powder than some fancy way to get a polymer. Whatever, there’s deliberately not enough information here for for some hostile power to immediately produce the requsite powdered nickel. But it’s powdered nickel. As for Walter Miller, the sources cited really don’t convince when it comes to him having been involved in the barrier design. We’ve already noted that [the first](http://www.news.cornell.edu/Chronicle/98/12.3.98/obit.html) only mentions him in connection with the “coolants and lubricants” that the Smyth Report discusses him. The [other cite](http://216.239.51.104/search?q=cache:UUooC_O_1doJ:www.chem.cornell.edu/alumni/pdf/news70.pdf+%22filters+made+of%22+%22gaseous+diffusion%22+%22Uranium+Hexafluoride%22+manhattan+&hl=en&ie=UTF-8) is an obituary in a departmental newsletter. To be fair, it’s _possible_ that the item was written by a colleague who’d discussed Miller’s wartime contributions with him. On the other hand, it’s just as likely that it was written by admin staff who’d read and misunderstood an obituary, well, like that [already mentioned](http://www.news.cornell.edu/Chronicle/98/12.3.98/obit.html). The Iraqwatch cite is interesting, but historically useless. Sixty years have elapsed since the Manhatten Project and technology has moved on. No doubt the polymers mentioned in the banned list ultimately derive from Miller’s achievements with resistent lubricants, but that’s far from evidence that these were used at Oak Ridge. ---
**Author:** ![commasense](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/commasense/32/3017_2.png) [@commasense](https://boards.straightdope.com/u/commasense)\ **Post date:** [July 10, 2004, 1:21am UTC](https://boards.straightdope.com/t/uranium-haxafluoride-gas/254141/10 "2004-07-10T01:21:47Z")
**Bonzer** : I’m not entirely sure it is powdered nickel. Rhodes characterizes it as “combin[ing] the best features of the Norris-Adler barrier and the compressed nickel-powder barrier.” (p. 495) The powdered-nickel method was developed by Foster Nix, and he was only able to come up with a few square inches at at time, when what was needed was square miles of the stuff. A Union Carbide engineer named Frazier Groff came up with a mass production method, something that had been considered impossible. Groueff (p. 161) reports him as visiting Nix’s lab and asking " Look, why not try to do it the way we would in industry?" Groueff adds a footnote at this point: “The technical details of the process developed are still classified.” However, the separating qualities of Groff’s barrier were unsatisfactory, and it seems that it was not the process used. That credit goes to Clarence Johnson, a lab director at Kellex, who, with his assistant Tony Suleski, “at least five times in recent experiments…had produced some kind of barrier; but something had always failed – the piece would break, or the porosity would be insufficient, or the holes would be too large. But now they were going to try again.” Groueff goes on to say, in essence, that they tried something else, and it worked. But no additional details, except to mention that Johnson “had been talking to Frazier Groff, whose new industrial techniques threw further light on Johnson’s own experiments.” (p. 185) Johnson’s successful barrier was extremely good news, because construction on the enormous K-25 plant and on the Houdaille-Hershey plant that was to manufacture the barrier material were both well on their way. Failure to find a workable barrier would have been a hugely expensive failure and damaging to the reputations and careers of many prominent people. The story of the Manhattan Project is just full of cases like that, where extremely talented and knowledgeable people, working under tremendous pressure, managed to produce scientific and technical breakthroughs in the nick of time. BTW, **Bonzer** , your contributions have been terrific, but would it kill you to spell Manhattan correctly? ---
**Author:** ![Squink](https://avatars.discourse-cdn.com/v4/letter/s/b5e925/32.png) [@Squink](https://boards.straightdope.com/u/Squink)\ **Post date:** [July 10, 2004, 2:08am UTC](https://boards.straightdope.com/t/uranium-haxafluoride-gas/254141/11 "2004-07-10T02:08:54Z")
> [@commasense](#): > > BTW, **Bonzer** , your contributions have been terrific, but would it kill you to spell Manhattan correctly? Agreed (to both 🙂 ) Yet I’m still not entirely convinced that Miller’s fluorocarbon polymers didn’t fit into early diffusion barrier work. Nickel technology obviously won out, yet still ~60 years later we find references to a viable fluorocarbon process? Where did that come from? Surely after inventing the wheel, no one went back and developed another top secret wheel? Why? History is written by the winners, and those who developed nickel technology may simply have not been inclined to mention other approaches that eventually lead to satisfactory results. ---
**Author:** ![commasense](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/commasense/32/3017_2.png) [@commasense](https://boards.straightdope.com/u/commasense)\ **Post date:** [July 10, 2004, 2:40am UTC](https://boards.straightdope.com/t/uranium-haxafluoride-gas/254141/12 "2004-07-10T02:40:18Z")
I think the confusion arises from a misreading of the quote you first provided: > [@](#): > > In the late 1930s, Miller carried out research into chemically resistant materials from which he later developed the chlorofluorocarbon polymer used in the first gaseous diffusion plant for the separation of uranium isotopes, a crucial factor in the development of the atomic bomb. The fissionable isotope Uranium-235 is separated from the more abundant isotope Uranium-238 by selective diffusion of uranium hexafluoride gas through barriers, or filters. I think the phrase “for the separation of uranium isotopes” modifies “gaseous diffusion plant,” not “chemically resistant materials.” In other words, the material was needed for the plant, but it wasn’t used to separate the isotopes. My reference to Teflon is an example of the kind of applications of chlorofluorocarbon polymers used in the gaseous diffusion process, _although not for the actual separation_. This page: [http://www.chem.cornell.edu/alumni/pdf/news67.pdf](http://www.chem.cornell.edu/alumni/pdf/news67.pdf) (PDF!) suggeests that Miller’s contribution involved materials other than Teflon, but presumably other polymers might have been needed for other purposes around the plant. BTW, Miller is not mentioned in either Rhodes or Groueff. ---
**Author:** ![Squink](https://avatars.discourse-cdn.com/v4/letter/s/b5e925/32.png) [@Squink](https://boards.straightdope.com/u/Squink)\ **Post date:** [July 10, 2004, 2:59am UTC](https://boards.straightdope.com/t/uranium-haxafluoride-gas/254141/13 "2004-07-10T02:59:07Z")
> [@](#): > > I think the confusion arises from a misreading of the quote you first provided That’s a distinct possibility, especially after reading your PDF. Yet I’m still left wondering about the who and when of the fluorocarbon-plastic barrier process referred to by the IAEA. ---
**Author:** ![commasense](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/commasense/32/3017_2.png) [@commasense](https://boards.straightdope.com/u/commasense)\ **Post date:** [July 10, 2004, 3:10am UTC](https://boards.straightdope.com/t/uranium-haxafluoride-gas/254141/14 "2004-07-10T03:10:13Z")
Well, I’ll admit that gave me pause, too. But it might refer to a process developed much later. For instance, on my trip to Oak Ridge back in the 1980s, I learned that the main method for isotope separation these days uses lasers, although I wasn’t told (or at least can’t remember) much more than that. ---
**Author:** ![bonzer](https://avatars.discourse-cdn.com/v4/letter/b/45deac/32.png) [@bonzer](https://boards.straightdope.com/u/bonzer)\ **Post date:** [July 10, 2004, 12:17pm UTC](https://boards.straightdope.com/t/uranium-haxafluoride-gas/254141/15 "2004-07-10T12:17:06Z")
> [@commasense](#): > > BTW, **Bonzer** , your contributions have been terrific, but would it kill you to spell Manhattan correctly? Apologies - I’m aware that I have a tendency to substitute vowels when spelling. > [@](#): > > **Bonzer** : I’m not entirely sure it is powdered nickel. Rhodes characterizes it as “combin[ing] the best features of the Norris-Adler barrier and the compressed nickel-powder barrier.” (p. 495) But the Norris-Adler barrier was itself made of nickel - as Rhodes states twice at the top of p494. Since he’s getting most of his information about this from Hewlett and Anderson, like them Rhodes does indeed never quite say that the final design involved _powdered_ nickel. That’s partly why I quoted Gowing’s _Independence and Deterrence_ in the last post: this is an example of a detail leaking via the British involvement. Those developing the postwar British diffusion plant were privy to how things had been done at K-25 and, rationally, copied its design. Gowing and her official reviewers were, by 1972, evidently prepared to be a little more open than their American equivalents had been a decade earlier. > [@](#): > > BTW, Miller is not mentioned in either Rhodes or Groueff. Similarly, the only reference to his involvement I’ve found is that one in Smyth. To reiterate, his wartime contribution was significant and related, it just didn’t impact the choice of design for the barriers.