The Making of the Atomic Bomb

neutron chain reaction in one or more of the materials known to show nuclear fission.” Serber said one kilogram of U235 was approximately equal to 20,000 tons of TNT and noted that nature had almost located that conversion beyond human meddling: “Since only the last few generations [of the chain reaction] will release enough energy to produce much expansion [of the critical mass], it is just possible for the reaction to occur to an interesting extent before it is stopped by the spreading of the active material.” If fission had proceeded more energetically the bombs would have slept forever in the dark beds of their ores.

Serber discussed fission cross sections, the energy spectrum of secondary neutrons, the average number of secondary neutrons per fission (measured by then to be about 2.2), the neutron capture process in U238 that led to plutonium and why ordinary uranium is safe (it would have to be enriched to at least 7 percent U235, the young theoretician pointed out, “to make an explosive reaction possible”). He was already calling the bomb “the gadget,” its nickname thereafter on the Hill, a bravado metonymy that Oppenheimer probably coined. The calculations Serber reported indicated a critical mass for metallic U235 tamped with a thick shell of ordinary uranium of 15 kilograms: 33 pounds. For plutonium similarly tamped the critical mass might be 5 kilograms: 11 pounds. The heart of their atomic bomb would then be a cantaloupe of U235 or an orange of Pu239 surrounded by a watermelon of ordinary uranium tamper, the combined diameter of the two nested spheres about 18 inches. Shaped of such heavy metal the tamper would weigh about a ton. The critical masses would eventually have to be determined by actual test, Serber said.

He went on to speak of damage. Out to a radius of a thousand yards around the point of explosion the area would be drenched with neutrons, enough to produce “severe pathological effects.” That would render the area uninhabitable for a time. It was clear by now — it had not been clear before — that a nuclear explosion would be no less damaging than an equivalent chemical explosion. “Since the one factor that determines the damage is the energy release, our aim is simply to get as much energy from the explosion as we can. And since the materials we use are very precious, we are constrained to do this with as high an efficiency as is possible.”

Efficiency appeared to be a serious problem. “The reaction will not go to completion in an actual gadget.” Untamped, a bomb core even as large as twice the critical mass would completely fission less than 1 percent of its nuclear material before it expanded enough to stop the chain reaction from proceeding. An equally disadvantageous secondary effect also tended to stop the reaction: “as the pressure builds up it begins to blow off material at the outer edge of the [core].” Tamper always increased efficiency; it reflected neutrons back into the core and its inertia — not its tensile strength, which was inconsequential at the pressures a chain reaction would generate — slowed the core's expansion and helped keep the core surface from blowing away. But even with a good tamper they would need more than one critical mass per bomb for reasonable efficiency.

Detonation was equally a problem. To detonate their bombs they would have to rearrange the core material so that its effective neutron number, which corresponded to Fermi's k, changed from less than 1 to more than 1. But however they rearranged the material — firing one subcritical piece into another subcritical piece inside the barrel of a cannon seemed to be the simplest option — they would have no slow, smooth transition as Fermi had with CP-1. If they fired one piece into another at the high velocity of 3,000 feet per second it would take the pieces about a thousandth of a second to assemble themselves. But since more than one critical mass was necessary for an efficient explosion the pieces would be supercritical before they had completely mated. If a stray neutron then started a chain reaction, the resulting inefficient explosion would proceed from beginning to end in a few millionths of a second. “An explosion started by a premature neutron will be all finished before there is time for the pieces to move an appreciable distance.” Which meant that the neutron background — spontaneous- fission neutrons from the tamper, neutrons knocked from light-element impurities, neutrons from cosmic rays — would have to be kept as low as possible and the rearrangement of the core material managed as fast as possible. On the other hand, they did not have to worry that a fizzle would drop an intact bomb into enemy hands; even a fizzle would release energy equivalent to at least sixty tons of TNT.

Predetonation would reduce the bomb's efficiency, Serber repeated; so also might postdetonation. “When the pieces reach their best position we want to be very sure that a neutron starts the reaction before the pieces have a chance to separate and break.” So there might be a third basic component to their atomic bomb besides nuclear core and confining tamper: an initiator — a Ra + Be source or, better, a Po + Be source, with the radium or polonium attached perhaps to one piece of the core and the beryllium to the other, to smash together and spray neutrons when the parts mated to start the chain reaction.

Firing the pieces of core together, the Berkeley theoretician continued, “is the part of the job about which we know least at present.” The summer-study group had examined several ingenious designs. The most favorable fired a cylindrical male plug of core and tamper into a mated female sphere of tamper and core, illustrated here in cross section from the Los Alamos Primer:

The target sphere could be simply welded to the muzzle of a cannon; then the cylinder, which might weigh about a hundred pounds, could be fired up the barrel like a shell:

The highest muzzle velocity available in U.S. Army guns is one whose bore is 4.7 inches and whose barrel is 21 feet long. This gives a 50 lb. projectile a muzzle velocity of 3150 ft/sec. The gun weighs 5 tons. It appears that the ratio of projectile mass to gun mass is about constant for different guns so a 100 lb. projectile would require a gun weighing about 10 tons.

For a mechanism eight times lighter or with double the effective muzzle velocity they could weld two guns together at their muzzles and fire two projectiles into each other. Synchronization would be a problem with such a design and efficiency might require four critical masses instead of two, a demand which would significantly delay delivering a usable bomb.

Serber also described more speculative arrangements: sliced ellipsoidal core-tamper assemblies like halves of hard-boiled eggs that slid together; wedge-shaped quarters of core/tamper like sections of a quartered apple mounted on a ring. That was an odd and striking design, sketched in the mimeographed Primer as probably on a blackboard before, and it did not go unnoticed. “If explosive material were distributed around the ring and fired the pieces would be blown inward to form a sphere”:

Autocatalytic bombs — bombs in which the chain reaction itself, as it proceeded, increased the neutron number for a time — looked less promising. The cleverest notion incorporated “bubbles” of boron-coated paraffin into the U235 core; as the core expanded it would compress the neutron-absorbing boron and render it less efficient, freeing more neutrons for fission chains. But: “All autocatalytic schemes that have been thought of so far require large amounts of active material, are low in efficiency unless very large amounts are used, and are dangerous to handle. Some bright ideas are needed.”

Their immediate work of experiment, Serber concluded, would be measuring the neutron properties of various materials and mastering the ordnance problem — the problem, that is, of assembling a critical mass and firing the bomb. They would also have to devise a way to measure a critical mass for fast fission with subcritical amounts of U235 and Pu239. They had a deadline: workable bombs ready when enough uranium and plutonium was ready. That probably gave them two years.

The Japanese physics colloquium in Tokyo had decided in March 1943 that an atomic bomb was possible but not practically attainable by any of the belligerents in time to be of use in the present war. Robert Serber's lectures at Los Alamos in early April asserted to the contrary that for the United States an atomic bomb was both possible and probably attainable within two years. The Japanese assessment was essentially technological. Like Bohr's assessment in 1939, it overestimated the difficulty of isotope separation and underestimated U.S. industrial capacity. It also, as the Japanese government had before Pearl Harbor, underestimated American dedication. Collective dedication was a pattern of Japanese culture more than of American. But Americans could summon it when challenged, and couple it with resources of talent and capital unmatched anywhere else in the world.

The Europeans at Los Alamos complained of the barbed wire. With the exception, apparently, only of Edward Condon, who found security so oppressive he quit the project within weeks of his arrival and went back to Westinghouse, the Americans accepted the fences around their work and their lives as a necessity of war. The war was a manifestation of nationalism, not of science, and such did their duty on the Hill appear at first to be. There was “relatively little nuclear physics” at Los Alamos, Bethe says, mostly cross-section calculations. They thought they were assembled to engineer a “practical military weapon.” That was first of all a