> Alex Wellerstein > Research > Atomic Patents >

The Secret Patents for the Atomic Bomb.

During World War II, one of the many ways the United States attempted to plan for postwar control of atomic energy was to attempt to secure a patent monopoly over the entire field. Below are some of the more interesting once-secret "atomic patents" I came across while writing a paper on the subject, "Patenting the bomb: Nuclear weapons, intellectual property, and technological control" (Isis, March 2008). Another, somewhat different (more accessible, shorter) version was published in the Bulletin of the Atomic Scientists as "Inside the Atomic Patent Office" (May/June 2008). A nice story on my research in atomic patents by David Kestenbaum ran on NPR's "Morning Edition" on March 28, 2008, as "The Rush to Patent the Atomic Bomb" (the audio is available online, along with a print version, at that link). I was also interviewed on this for a segment of the PBS show "History Detectives" which aired on July 29, 2009.

The papers detail the motivations for "patenting the bomb" in the first place—a policy championed originally by Vannevar Bush, the head of the Office of Scientific Research and Development (which administered atomic bomb research), General Leslie Groves, head of the Manhattan Engineer District (the atomic crash program run by the US Army Corps of Engineers), and even supported by President Franklin D. Roosevelt himself. From 1942, when the patenting program started, until January 1947, when the United States' atomic infrastructure was taken over by the Atomic Energy Commission, the patent division of the Manhattan Project had docketed reports on over 5,600 different inventions relating to the atomic bomb, resulting in some 2,100 separate patent applications to be filed—in secret—with the US Patent Office. The Manhattan Project patenting program was a systematic attempt to acquire total legal ownership for the United States government in the entire field of atomic energy that had been developed during the war—the patents spanned 493 different subject classes of technology, "from the raw ore as mined to the atomic bomb."

The patents below represent just a handful of the "atomic patents" generated by the wartime patenting program headed by Captain Robert A. Lavender. They were all initially filed in secret under a World War I-era statute that allowed the Commissioner of Patents to declare certain patent applications secret during emergencies. The ones below were all declassified and then granted at a later point—the difference between their application date (e.g. 1944) and their granting date (e.g. 1954) indicates how long they were kept secret (anything more than five years is an unusually long time between filing and granting). I estimate (based on various Google Patents searches) that at least 500 patents from the wartime period have since been declassified and granted. Anyway, there's a lot more to the whole story, but that's what the articles and etc. are for. A lot of the below, however, couldn't really fit into the articles (I might do something separate on the plutonium patents someday).

All patents below are for the United States unless otherwise indicated. Links to the patents at the US Patent and Trademark Office (USPTO) and Google Patents are provided when possible. Google Patents is the easiest to use, but the USPTO has 300dpi TIFFs of the patent pages. For reference, I have linked many of the names to their Wikipedia articles. The patents are in no particular order.

Neutronic Reactor, figure 38 NEUTRONIC REACTOR
Inventors: Enrico Fermi and Leo Szilard
Patent Number: 2,708,656
Application filed: December 19, 1944
Patent granted: May 17, 1955
View patent: USPTO | Google Patents

Description: This 58 page patent is the major patent covering the invention of the first nuclear reactor (here called a "neutronic reactor" since it works by neutron multiplication—a nuclear chain reaction). The patent actually covers many different reactors designs, many of which also had their own separate patents covering them.

Back-story: Contrary to popular belief, Szilard did not "patent the atomic bomb" when he filed for a patent on the idea of a nuclear chain reaction in 1933—he did not know at the time about the possibility of nuclear fission as the driving force of the chain reaction, and as such his ideas were extremely theoretical. This patent, though, covers the actual use of nuclear fission to create a chain reacting pile, and you can recognize the similarities between the face of the reactor shown at right with the photos of reactors used to breed plutonium in Hanford, Washington, during the war.

Szilard had attempted to use his legal ownership of the intellectual property of the first nuclear reactors as a way to gain more clout during the Manhattan Project itself, as a way to try and gain more influence over how the project was administered. General Groves told Szilard that he had a choice between staying in the Manhattan Project at all or signing over his patent rights; Szilard, in the end, reasoned that being isolated from the project wouldn't help his cause much, and capitulated. The atomic patent program was, in part, designed to make sure that no individual scientists or contractors could use their ownership of the atomic inventions as a way to influence the US government's atomic policies.
Method for Producing, Separating, and Purifying Plutonium, figure 1 METHOD FOR PRODUCING, SEPARATING, AND PURIFYING PLUTONIUM
Inventors: Glenn T. Seaborg, Joseph W. Kennedy, and Arthur C. Wahl
Patent Number: 3,190,804
Application filed: December 27, 1945
Patent granted: June 22, 1965
View patent: USPTO | Google Patents

Description: This is one of over a dozen patents relating to fundamental plutonium and fission product chemistry developed by Glenn Seaborg's team (which also included Emilio Segrè, along with the other co-inventors above). These patents covered many of the processes undertaken at the Hanford site of the Manhattan Project used to produce the plutonium for both the "Trinity" gadget as well as the Nagasaki "Fat Man" bomb. The patent at right specifically describes the steps needed to extract and purify plutonium bred in a nuclear reactor of the type invented by Fermi and Szilard (above).

Back-story: The plutonium patents presented a major problem for the Manhattan Project patenting program: Seaborg's team had done most of the fundamental work before they were on an OSRD contract, and so OSRD/Manhattan Engineer District did not automatically own the patents. In fact, it wasn't entirely clear who did own the patents: as the work was done at the University of California, Berkeley, the Regents of the University of California felt that they had a stake in the patent rights as well (even though in this case, only a couple of them had any idea what the patents covered, since they were not on the whole privy to what the bomb project was about). The inventors wanted to keep the patent rights for themselves, the Regents wanted them for the UC, and Vannevar Bush desperately wanted them for the OSRD/US government.

The issue took a decade to work out. In the 1950s, the Atomic Energy Commission set up a patent compensation board to reward inventors who had made key contributions to the atomic infrastructure, and after a lot of back-and-forth decided to give each of the inventors $100,000 apiece—not bad money in 1955, but still not anything as much as the patents would have been worth in a commercial market. Immediately after the settlement was announced, UC President Robert Sproul wrote to the inventors proclaiming that all the Regents were so happy at the result, and especially at the news they had heard that the inventors were planning to give a significant amount of the money back to the university. Segrè, never a fan of Sproul, wrote in his autobiography that the UC President "must have been disappointed when we did not give him anything."

(Segrè wasn't the only one to find Sproul difficult; at one particularly low point in the wartime deliberations, Vannevar Bush put a cross-reference sheet in his files: "Regarding: Patent rights being assigned to the U. of California. See: Dictator.")
Methods of and apparatus for separating materials, figure 26 METHODS OF AND APPARATUS FOR SEPARATING MATERIALS
Inventors: Ernest O. Lawrence
Patent Number: 2,709,222
Application filed: October 9, 1944
Patent granted: May 24, 1955
View patent: USPTO | Google Patents

Description: A 91-page patent covering the invention of the Calutron, the electromagnetic method for separating uranium-235 from its isotope uranium-238, invented by Ernest Lawrence at the Radiation Lab in Berkeley, California. The basic idea is to shoot ions of uranium into a C-shaped chamber, using a massive magnet to deflect the beam into a collecting cup. Because the isotopes of uranium weighs slightly different from one another, running a lot of uranium through many of these machines in parallel would, over time, separate out a considerable amount of uranium-235 for use in a bomb.

Back-story: The Calutron system was very complicated and a major source of research and development at Berkeley during the war. Lawrence used his experience with particle accelerator research towards this end, and many of the drawings in the patent are clearly modeled after the prototype work done with the 184-inch cyclotron in Berkeley. During World War II there were over 150 different patents developed for it alone for the Manhattan Project; most of which were declassified and granted in a flurry of activity following changes in the Atomic Energy Act of 1954 which attempted to release information that might be useful for more peaceful applications. Basically, after declassification rules changed, the AEC would review back-logged secret patent applications that seemed like they would be relevant to the declassifications, and then decide whether or not to push them forward to the US Patent Office for examination (see the very last patent in this list for the most extreme example of a delay between filing and granting).

This particular patent at right is the major one covering the Calutron apparatus itself—the other patents included all manner of improvements, variations, the idea to use many of them in tandem, different types of ion sources, collectors, gauges, etc.
Magnetic shims, figure 16 MAGNETIC SHIMS
Inventors: J. Robert Oppenheimer, Stanley Phillips Frankel, and Eldred Carlyle Nelson
Patent Number: 2,719,924
Application filed: December 28, 1945
Patent granted: October 4, 1955
View patent: USPTO | Google Patents

Description: J. Robert Oppenheimer's only (non-secret) patent, for the use of magnets to focus Calutron beams. In the figure at the right (a Calutron tank cross-section), they are the oddly shaped bits in between the walls and the inner pieces.

Back-story: Even though Oppenheimer was the "father of the bomb", he is not listed on any other granted patent (Google Patent searches can be misleading; his brother, Frank Oppenheimer, has numerous patents from his work on the Calutrons at Berkeley). This is one of his only major immediate direct technological contributions to the bomb developments; on a Manhattan Project biographical form he filled out in July 1945, he noted that in June 1942, during the secret summer conference, he "Developed theory of pre-detonation and preliminary theory of radiation. Asked as consultant to Berkeley Radiation Laboratory. Invented magnetic shims." The patent itself is pretty dull, but an interesting contrast to many of the other Calutron patents: it is clearly the invention (and patent) of a theoretical physicist, filled with pages and pages of complicated physics equations.

It is probably not meaningful, but this patent was granted about a year after Oppenheimer had his security clearance revoked.
Pressure Sensitive Switch, figures 5 and 6 PRESSURE SENSITIVE SWITCH
Inventors: Alan N. Ayers
Patent Number: 3,358,605
Application filed: January 4, 1946
Patent granted: December 19, 1967
View patent: USPTO | Google Patents

Description: This patent claims it is a general invention for a switch that could detonate any "airborne ordnance devices" (bombs of any sort) on the basis of pressure. In fact, it is none other than the barometric detonation system used in the first atomic bombs.

Back-story: Atomic bombs, unlike "normal" bombs, are most effective when they detonate a few thousands feet above their target, not when they hit their target. So a few clever devices were used to make sure the bomb would go off at an optimal height above the target; the barometric switches relied on the changes in pressure the bomb would undergo as it rapidly changed altitude, falling from the bomber.

But the patent cannot say it is useful only in an atomic bomb; all bomb patents have to appear to be "general" in nature, useful for potentially anything, because the Atomic Energy Act prohibits the patenting of anything that was used exclusively for the development or production of atomic weapons. So this patent, like a few others in this list, "generalizes" its contents.

But its true nature can be easily seen from the drawing itself. Compare the cable harness seen in figures 3 and 4 (at right) with the cable harness used on the Trinity and Fat Man weapons: it's the same one, clearly drawn with reference to the same bombs. This is what (part of) an atomic bomb looks like to a patent lawyer.
Low Impedance Switch, figure 2 LOW IMPEDANCE SWITCH
Inventors: Donald F. Hornig
Patent Number: 3,956,658
Application filed: November 28, 1945
Patent granted: May 11, 1976
View patent: USPTO | Google Patents

Description: Another "general" patent, used "for effecting the simultaneous closing of a plurality of electrical circuits within very short intervals of time"... that is, a useful little switch for, well, anybody who needed to set off, say, 32 detonators at the same time within a margin of error measured in microseconds. You know, something that everyone can use... if they happen to be looking for a detonator for an implosion bomb.

Back-story: This is part of the triggering switch used in the "X-Unit" of the Fat Man and Trinity bombs. Plutonium implosion bombs like the sort used at Trinity and Nagasaki (and in contrast to the enriched uranium bomb used at Hiroshima) work by the simultaneous spherical compression of a plutonium core to high densities using specially shaped high-explosives. The margin for error is extremely small: if one part of the core is compressed faster than another part, the plutonium will not detonate and simply be scattered about (a mess, but not a nuclear explosion).

As Donald Hornig (the inventor listed on the patent) later put it: "The waves are made spherical by explosive lenses, but each explosive lens, and there were thirty-two of them, had to be initiated with a high degree of simultaneity. The problem I worked on was to get all of the explosive lenses reliably started within a time interval of a ten-millionth of a second." Today this would be known as a triggered spark gap switch. It's a very clever way to make sure that a large electrical signal get sent out to many places at pretty much exactly the same time.

I was able to get in touch with Hornig (who went on to be a prominent science administrator, and is now a professor emeritus at the Harvard School of Public Health) and he told me that he couldn't recall being informed that the patent was granted.

Low Impedance Switch, figures 1-2 ELECTRIC INITIATOR WITH EXPLODING BRIDGE WIRE
Inventors: Lawrence H. Johnston
Patent Number: 3,040,660
Application filed: November 8, 1944
Patent granted: June 26, 1962
View patent: USPTO | Google Patents

Description: This patent describes a type of detonator that also has very low time tolerances — they consist of very fine wires in tubes which, when a high-voltage electrical signal is sent through them, vaporize and convey heat and shock into the explosive lenses with very precise timing (in the sub-microsecond range).

Back-story: Another very clever device needed for the precision detonation of the plutonium bomb (see the LOW IMPEDANCE SWITCH above). EBW detonators were used long after the war (a more modern EBW detonator from an export control handbook).
Folded Dipole, figure 1 FOLDED DIPOLE
Inventors: George William Schivley and Paul W. Springer
Patent Number: 2,452,073
Application filed: November 17, 1944
Patent granted: October 26, 1948
View patent: USPTO | Google Patents

Description: This patent describes a small directional antennae to be put on the tail wing of an airplane in order to warn the pilot if an enemy silently slips in behind him. It is a variation of a Yagi antennae design, easily distinguished by its distinctive folded (U-shaped) element in the middle.

Back-story: This is not technically an "atomic patent" in the same sense as the others — it was not specifically developed by anyone in the Manhattan Project and was not subjected to much secrecy (hence it was declassified and granted by 1948, before the Soviet Union even had the bomb). The antennae here were a "tail warning device" designed to be put on the backs of airplanes to warn a pilot if another aircraft was within a few thousand feet of the back of the plane.

This seemingly unrelated invention is featured here because actually one of the little iconic touches on the photos of the original Fat Man and Little Boy bombs, the so-called "Archie" antennae that were used to detect when the bomb was a certain altitude. You can see them on the nose of the Little Boy bomb; compare with the patent drawing at right.

When dropped out of the plane at 30,000 feet or so, the barometric switches (the PRESSURE SENSITIVE SWITCH, above) would trigger at around 7,000 feet above the target and get the bomb ready to detonate, while the at 1,800 feet the "Archie" antennae would detect the ground below and actually detonate the weapon. These antennae were adapted from their original purpose to that of setting off the atomic bombs.

It's not surprising the patent was released quickly — the actual invention is clever but not essential or specific to nuclear work, and only the precise distances between the antennae on the actual bombs would have been secret information (if you knew them, you could trick the bomb into thinking it was at its target long before it actually was).
Method of Determining the Extent to which a Nickel Structure has been Attached by a Fluorine-Containing Gas, figure 1 METHOD OF DETERMINING THE EXTENT TO WHICH A NICKEL STRUCTURE HAS BEEN ATTACHED BY A FLUORINE-CONTAINING GAS
Inventors: James P. Brusie
Patent Number: 6,761,862
Application filed: September 24, 1945
Patent granted: July 13, 2004
View patent: USPTO | Google Patents

Description: This patent relates to some obscure chemical process in the enrichment of uranium (uranium hexafluoride is the "fluorine-containing gas" in question), probably in relation to the gaseous diffusion method used to enrich uranium at Oak Ridge, Tennessee, during the Manhattan Project.

Back-story: Look at the difference in the application date and the granting date — nearly 60 years! That means this patent was applied for before the Cold War really began, and granted quite some time after it had ended! That's some delay!

US laws on patent secrecy spell out that a patent application has been declared "secret" then it is not examined and granted until after the information in it has been declassified. As a result, many of the patents which were under secrecy orders (including most of those above) have relatively long dates between being applied for and granted, well beyond what such a thing should normally take.

So here is a patent which was "applied for" (by the Manhattan Engineering District patent office) shortly after the end of World War II, and then was "put to sleep" (as a participant in the patenting project termed it) for almost six decades before being awoken and released. Only after being granted does the patent's "clock" begin ticking down, so it will still be in force for some time!

From what I can tell this patent has been secret the longest time (so far) out of those which were originally made secret during the Manhattan Project. There are probably many other secret patents out there, so in the future its entirely possible than some will be declassified and granted.

According to the USPTO's "Transaction History" for this patent (which goes back only until 1990), its secrecy order was renewed annually until September 2003, when it was rescinded. (It then took a little less than a year for all of the other paperwork to go through.)

I was in touch with one of the Department of Energy legal counsel in charge of this particular patent, he said that today they annually review declassification orders that might apply to patents, and that they usually feel it worthwhile to push through the old applications:

"Our feeling has been that a significant taxpayer investment was made to create the inventions and to prosecute the patents so that payment of the issue fee finalizes the effort to provide a property right arising from the government funding. Of equal merit is the recognition provided to the inventors. When the patent issues we make a small good faith effort to find the inventor or a surviving spouse and notify them of the issuance of the patent. When notify someone, they are usually deeply moved by the recognition provided for their long ago secret efforts."

In this case, the inventor, James P. Brusie, had been some time deceased by the time the patent was issued.

Other interesting Manhattan Project patents

These patents aren't as visual or don't have as complicated back-stories so I will just link to them briefly:
  • GB 630,726 - Improvements in or relating to the transmutation of chemical elements (filed in the UK in 1934, granted in 1936, but kept secret until 1949) by Leo Szilard. This is the famous patent on the chain reaction by Hungarian physicist Leo Szilard. This is sometimes erroneously described as the original patent on the atomic bomb. It is really no such thing: it is a speculative patent which says that if a neutron-based chain reaction could be created, it would be useful for liberating nuclear power or creating more radioactive elements. It was entirely theoretical when filed: Szilard had no idea whether a nuclear chain reaction was possible, or what materials might sustain one. The British government was rightly skeptical of Szilard's claims to the patent's importance and secrecy — that Szilard turned out to be right (with the discovery of fission five years later) doesn't mean that anyone at the time should be faulted for being unclear of its likelihood. It's an interesting patent, and demonstrates Szilard's knack for being ahead of his time, but a patent for an atomic bomb it is not. It is possible, for example, that even if a chain reaction was possible, that a bomb would not be possible (because of the various technical requirements of a bomb-based chain reaction, which are different than a reactor in several important ways). I think it is quite wrong to say that Szilard was the "inventor of the atomic bomb": he was one of many "inventors," for the bomb has no single "inventor," and no single "father."
  • GB 614,156 - Apparatus for the Production of Energy By Nuclear Fission (filed in the UK 1939, granted 1948): one of the French atomic energy patents filed in the UK by French physicist refugees. A similar patent application was filed in the US and was the impetus for Vannevar Bush establishing a patent censorship program (see my Bulletin of the Atomic Scientists article above). This particular patent describes the concept of a reactor; this one describes methods of controlling the subsequent reaction.
  • FR 971,324 - Perfectionnements aux charges explosives (Improvements in Explosive Charges (filed in France in May 1939, granted in 1950): another French atomic energy patent. This one is explicitly about the use of fission as a means of explosive energy, including for war.
  • 2,206,634 - Process for the Production of Radioactive Substances (filed in 1935, granted in 1940) by Enrico Fermi, Edoardo Amaldi, Bruno Pontecorvo, Franco Rasetti, and Emilio Segrè. This is an interesting patent that was filed long before the war but involved a long dispute over compensation by the US government over its use. Eventually Fermi and his associates received $300,000 for use of the patent by the government. Simone Turchetti has written a very interesting article on this story.
  • 3,208,197 - Diffusion Separation of Fluids (filed 1944, granted 1965): patent for the gaseous diffusion method of uranium enrichment, granted to Francis Simon, Rudolph Peierls, and, amusingly, Klaus Fuchs, who was by then working in East Germany having been convicted of being a Manhattan Project spy in 1950 and serving nine years of prison time.

Other items of note

These are things that didn't fit completely into the above categories that I thought were worth noting somewhere:
  • 2,637,536 - Method of Dispersing Materials in Water: AEC patent filed in 1947, granted in 1953, describing the best way to poison a water supply with botulism toxin or radiological materials: "It is an object of this invention to provide an improvement in the art of toxic warfare... wherein bodies of water are the targets and wherein it is desired to effectively contaminate same." Nice! I wouldn't be surprised if this one mysteriously disappeared from the Patent Office website after not too long... (update: an interesting analysis of the feasibility of the method advocated in the patent can be found here.)
  • 2,604,042 - Detonating Explosive Charge: This is a patent for the arrangement of fast and slow explosives to create explosive lensing — however, only in the limited context of making impressions upon metal plates. Interesting though is that this is not a government patent, and was instead filed by a Scottish inventor in 1948 (granted in 1952). John von Neumann had of course developed this method during the Manhattan Project for use in the implosion device, but that hadn't been declassified in the United States until 1951 (during the Rosenberg trial). I don't know if there's an interesting back story here or if it was some sort of bureaucratic snafu — my (uninformed) guess is that they let it slide because it didn't appear to have anything to do with their method, though that amusingly makes it the only public patent on the basic principle behind the implosion bomb, though the author clearly didn't know that.
  • 3,956,039 - High Explosive Compound: AEC patent for "low detonation velocity" explosives, of the sort used in the explosive lenses of implosion weapons (filed 1956, granted 1976). What's also interesting is that this patent cites the one listed previously here as the only patent to describe explosive lensing. It contains quite a lot of data on different lens properties based on different explosive compositions.
  • Patent secrecy order from 1944: Letter to two du Pont scientists making a patent application of theirs secret. It just illustrates what one of these form letters looked like; I discuss the wartime patent secrecy effort in detail in both my Isis and Bulletin of the Atomic Scientists articles linked to at the top of this page.

Last updated August 2011. All images are in the public domain; text is © Alex Wellerstein.

> Back