Redactions

The blue flash

by Alex Wellerstein, published May 23rd, 2016

This last weekend was the 70th anniversary of Louis Slotin's criticality accident. One slip of a screwdriver; a blue flash and wave of heat; and Slotin had a little over a week to live. It's a dramatic story, one that has been told before. I tried to give it a little bit of a fresh look in my latest piece for the New Yorker's Elements Blog: "The Demon Core and the Strange Death of Louis Slotin."1

Demon Core New Yorker Screenshot

In researching the piece, I looked over a lot of technical literature on the accident, as well as numerous accounts from others who were in the room at the time. A few things stuck out to me that didn't make it into the piece. One was that it was remarkably non-secret for the time. Los Alamos put out a press release almost immediately after it happened (by May 25th, five days before Slotin's death, it was in national newspapers), and followed it up with more after Slotin's death. For mid-1946, when the Atomic Energy Act had not yet been signed and the future of the American nuclear infrastructure was still very much in question, it was remarkably transparent. The press release was where I saw the phrase "three-dimensional sunburn" for the first time.

I also went over the account of Slotin's case that was published in The Annals of Internal Medicine in 1952.2 Slotin isn't named, but he's clearly "Case 3." Harry Daghlian, who also died from an accident with the same core, is "Case 1," and Alvin Graves, who was the nearest person to Slotin during his accident, and later became a director of US nuclear weapons testing, is "Case 2." The article is long and technical, and ends with some of the most disturbing photographs I have ever seen of the Daghlian and Slotin accidents. There is a photo of Daghlian's hand that has been reproduced many places (including in Rachel Fermi's Picturing the Bomb), but I'd only previously seen it in black and white. It is much worse in color — the contrast between the white blistered skin and the pink-red stuff under the cut-away area is dramatic and disturbing. There are others in the same series that are just as bad if not worse: blackened, gangrenous fingers. Slotin's photos in that article are comparatively tame but still pretty unsettling. Blisters. Cyanotic tissue. A photograph of his left hand — the one that was closest to the reacting core — on the ninth day of treatment (his last day alive) looks almost corpselike, or even claw-like. It is unsettling. I will not post it here.

An anonymous e-mail tipped me off that there were more photographs, and more documents, at a collection at the New York Public Library. These were part of a collection deposited by Paul Mullin, who authored the Louis Slotin Sonata, a very interesting, very curious play about Slotin from the late 1990s. I haven't seen the play, though I had seen mentions of it for awhile. Mullin's materials were fascinating and very useful. There were two boxes. The first was mostly notes relating to the creation of the play. It is always interesting to see how another researcher takes notes, much less one whose end-product (a play) is very different from the sort of thing I do. It does not take much glancing at his notes to see that Mullin got as deep into this topic as anyone has. The second box contained research materials: four folders of documents obtained from Los Alamos under the Freedom of Information Act, and a folder of photographs.

The hands of Louis Slotin, shortly after admission to the Los Alamos hospital. Source: Los Alamos National Laboratory, via the New York Public Library (Paul Mullin papers on the Louis Slotin Sonata).

The hands of Louis Slotin, shortly after admission to the Los Alamos hospital. Source: Los Alamos National Laboratory, via the New York Public Library (Paul Mullin papers on the Louis Slotin Sonata).

The photographs were, well, terrible. They included the ones from the Annals of Internal Medicine article, but also many more. Some showed Slotin naked, posing with his injuries. The look on his face was tolerant. There were a few more of his hand injuries, and then the time skips: internal organs, removed for autopsy. Heart, lungs, intestines, each arranged cleanly and clinically. But it's jarring to see photographs of him on the bed, unwell but alive, and then in the next frame, his heart, neatly prepared. The photo above, of just his hands, is one of the tamest of the bunch, though in some sense, one of the saddest (there is a helplessness, almost like begging, in the position). I didn't make copies of the really awful ones. History is often very voyeuristic — I joke with students that I read dead people's mail for a living — but, as I commiserated with Mullin over Twitter, at some point you start to almost feel complicit, as silly as that notion is.

The documents were invaluable. They mostly covered the period immediately after the accident — people checking in on Slotin's health, the complicated legal aspects of dealing with the death of a scientist (and with his distraught family), the questions of what to do next. An inordinate amount of paperwork was generated in dealing with the disposition of Slotin's automobile (a 1942 Dodge Custom Convertible Coupe). The Army's interactions with Slotin's family appeared sympathetic and generous. There appears to have been no cloak-and-dagger regarding the entire affair. Slotin was, after all, a friend to many of those at Los Alamos, and a key member of their "pit crew."

One of the accounts that I found most fascinating was that of the security guard, Patrick Cleary, who was in the room when the accident happened. Cleary was there because you don't just keep a significant proportion of the nation's fissile material stockpile unguarded. He seems to have understood little about what risks his job entailed, though:

When the accident occurred, I saw the blue glow and felt a heat wave. I knew something was wrong, but didn’t know exactly what it was, when I saw the blue glow and somebody yelled. ... Our instructions are also to keep in sight of all active material that is around, except in the case of a critical assembly, but [I] am not sure about that. I did not actually know what the material or sphere was at the time, or anything about it.3

When Cleary saw the flash and heard yelling, he literally took off for the hills, running. He was called back, as the scientists tried to reconstruct where people were standing for the purposes of dosage calculation. Cleary, in fact, was the last person to leave, because security guards can't walk off the job — he had to wait until a replacement came.

Close-in shot on the Slotin accident re-creation. The beryllium tamper is on top; the plutonium core is the smaller sphere in the center. Notice in this particular shot, they have a "shim" on the right. Slotin removed the shim right before his fatal slip.

Close-in shot on the Slotin accident re-creation. The beryllium tamper/reflector (they called it a tamper) is on top; the plutonium core is the smaller sphere in the center. Notice in this particular shot, they have a "shim" on the right. Slotin removed the shim right before his fatal slip. The scientist re-creating the photograph is physicist Chris Wright. I wonder if they took extra precautions in making this particular set of photos?

For a long time I had been wondering what happened to the so-called "demon core," which was also known as "Rufus," something that strikes me as just too strange to be anything but true. It has been reported many times that it was used at Operation Crossroads, at the Able shot. I found some documentation that suggested this was very unlikely. For example, shortly after the accident (Slotin was still alive), lab directory Norris Bradbury wrote to a few other scientists at Los Alamos about how the accident had affected the forthcoming Crossroads tests. He notes that the sphere in question was getting "its final check" during the accident — so it was definitely slated for Crossroads. But he continues:

Obviously Slotin will not come to Bikini. [Raemer] Schreiber will come although the date of special shipment was postponed one week to allow us to pull ourselves together. Only two shipments will be made at this time as I see no courier for the third. The sphere in question is OK although still a little hot but not too hot to handle. We will save it for the last in any event if it is needed at all.4

Which seemed pretty suggestive to me that they weren't going to use it: only two shipments were going to be made early on, and "the sphere in question" was not one of them. It would be saved for the "last event." Which in this case was the "Charlie" shot — which was cancelled.

I wanted some more confirmation, though, because a plan isn't always a reality. I e-mailed John Coster-Mullen, who I knew had done a lot of research into the Slotin and Daghlian accidents. (John is the one that provided me with these wonderful high-resolution photographs of the Slotin re-enactment, and some of the documents in his appendices to Atom Bombs were very useful for this research.) John suggested I get in touch with Glenn McDuff, a retired scientist at Los Alamos who was also one of the consultants on Manhattan (he drew the equations on the chalkboards, among other things). This turned out to be a great tip: Glenn has been working on an article about the fate of the first eight cores. There is much still to be declassified, but he was able to share with me the fate of the core in question: it had not been used at Crossroads, it had been melted down and the material re-used in another core. Glenn says there was no particular reason it was melted down. It was old, as far as cores went, and they were constantly fiddling with them in those days — the days in which they still gave bomb cores individual nicknames, because there were so few of them.

For nuke nerds, this is the big "reveal" of my New Yorker piece, the one thing that even someone very steeped in Los Alamos history probably doesn't know. (For non-nuke nerds, I doubt it registers as much!) And even though it is a bit anticlimactic, I actually prefer it to the version that the core was detonated shortly after the accident. The part about them immediately re-using the core in a weapon just always seemed a little suspicious to me — it almost implied that they had done it due to superstition, and that didn't really jibe with my sense of how these scientists viewed the accident or these weapons. And even the anticlimax has a bit of a literary touch to it: the "demon core" wasn't expended in a flash, it was melted down and reintegrated with the stockpile. Who knows whether bits of its plutonium ended up in other weapons over the years, whether any of that core is still with us in the current arsenal? There's perhaps something even a bit more "demonic" about this version of the story.

Notes
  1. A few small errata to the piece, based on a few questions I got: 1. Should the beryllium hemisphere be called a tamper or a reflector? In most contexts today we would call it a neutron reflector, because that's the property that you use beryllium for in a bomb (a tamper's job, generally, is to hold the core together as long as possible while it reacts, and so heavy, dense metals like uranium are used). But in this case, the scientists at the time referred to it universally as a "beryllium tamper" so the editor and I just decided to keep things simple and call it that, rather than call it a "reflector" and then clarify that it was the same thing as the "tamper" that was cited in the quotes. (This is the kind of linguistic hair-splitting that goes into these pieces — a balance between the historical language, the present-day language, the technical aspects, etc. We try to come to sensible decisions.) 2. At one point, it refers to the "pits" at Hiroshima and Nagasaki. This is just meant in a colloquial way here to refer to their fissile material cores. The Hiroshima bomb of course was a different design, made of two different pieces, called the Projectile and the Target in the documents at the time. It seemed unnecessary to introduce all that complexity to make a point that they didn't give it any kind of colorful moniker. 3. There was one legitimate typo in the piece as published, which was my fault. It misstated the amount of time between the Daghlian and Slotin accidents (three months instead of nine). I'm not sure how that got in there — I actually re-looked up the date differences at the time I wrote it, and know the months cold. One of those strange disconnects between the head and the fingers, I suppose, and somehow I missed it in re-reading the drafts. Very frustrating! It's the little things you aren't worried about getting wrong that can get you, in the end. It has been fixed. []
  2. Louis H. Hempelmann, Hermann Lisco, and Joseph G. Hoffmann, "The Acute Radiation Syndrome: A Study of Nine Cases and a Review of the Problem," Annals of Internal Medicine 36, no. 2 (February 1952), Part 1, 279-510. []
  3. Patrick Cleary, account of the Slotin accident (29 May 1946). Copy in the Paul Mullin, "Production materials for the Louis Slotin Sonata, 1946-2006," New York Public Library. []
  4. Norris Bradbury to Marshall Holloway and Roger Warner (undated, ca. 24-29 May 1946). Copy in the Paul Mullin, "Production materials for the Louis Slotin Sonata, 1946-2006," New York Public Library. []

Tags: , , ,


6 Responses to “The blue flash”

  1. Bill Mullins says:

    The security guard, Patrick Cleary, ended up going to Korea and fighting there. He lost his life in combat on Sept. 3, 1950, about a month shy of turning 25 years old.

  2. Mark M says:

    I guess I can see a technical argument to melt down the “Rufus” core, simply because it was to some extent contaminated by fission products, (many are neutron absorbers) and was also more radioactive than a “clean” core would be. Mixing the material with other “clean” material would dilute the fission products. The neutron emission rate might have been increased as well, although I’m not certain that would be a long-term concern.

    Always have wondered what Louis Slotin was thinking when he did this, of all people he certainly understood the risks. Seems like more of a stunt than an experiment.

    • I think the key is that it was more a “demonstration” than an “experiment.” Slotin was leaving the lab soon for Bikini, and after Bikini he was going to be starting up at the University of Chicago. The accounts of what the day was meant to be are somewhat sketchy but it sounds like he was showing people around and Alvin Graves asked if he’d show them how he did the criticality work. The sense I get from reading the incident accounts is that it was not a very well-planned thing, just something tacked on to other activities. This is speculation, but it seems like the very irregularity may have increased the odds of an error being made, as might have been the fact that it was a “performance” with an audience, done on probably a tighter schedule than it might normally have been.

      A perhaps-relevant passage from Charles Perrow’s Normal Accidents:

      A frequent “component” failure is the personnel. We do not know the extent to which personnel are trapped into “forced errors” through production schedule demands or long shifts below the ground (eleven-hour shifts are mentioned), or, on the other hand, the extent to which there is a “macho” culture that provides psychic rewards for risk taking. I am sure that the first exceeds the second; a risk-taking, macho culture has probably developed to make sense out of what is a fairly inhuman activity, a view of this world that makes it conceivable that one could function in it. But whatever the cause, operator error, forced or unforced, is common.

  3. Howard says:

    Fascinating read. I’ve been interested in the “Rufus” story since first seeing “Fat Man and Little Boy” many years ago. Probably one of the worst ways to die I can think of. The statistical odds of the same core killing two scientists in almost the same way much be pretty astronomical. Thanks so much for doing the research on its final fate and sharing this with us.

  4. And after this second accident, as I recall, Bradbury made a hard rule that all criticality tests would be run by remote control. Which they were.