Chicago Pile-1: The First Controlled Nuclear Chain Reaction

At 3:25 PM on a cold Wednesday in December 1942, an Italian physicist standing on a balcony above a squash court in Chicago ordered a cadmium rod pulled out by 12 more inches. The counters clicking in the room below began to accelerate. The clicking became a roar. Enrico Fermi watched the rate climb on the recorder chart, let it run for 28 minutes, then said: "The reaction is self-sustaining. The experiment has succeeded." He then ordered the control rod reinserted. Humanity had produced the first controlled release of nuclear energy. The nuclear age had begun.

What a Nuclear Chain Reaction Is

Nuclear fission — the splitting of a heavy atomic nucleus (typically uranium-235 or plutonium-239) into two smaller nuclei when struck by a neutron, releasing energy (primarily as kinetic energy of the fragments) and 2–3 additional neutrons.

The chain reaction concept follows directly from the fission neutron yield. If each fission event produces, on average, more than one additional neutron, and if those neutrons go on to cause additional fissions, then the number of fission events per unit time grows exponentially. Like compound interest, but measured in microseconds.

The multiplication factor k — the average number of fission-causing neutrons produced per fission event, after accounting for all neutron losses; when k = 1, the chain reaction is self-sustaining (critical); when k > 1, it grows (supercritical); when k < 1, it dies out (subcritical).

In natural uranium ore or ordinary reactor fuel, the multiplication factor is below 1. Most of the neutrons produced by fission escape the material or are absorbed by non-fissile components before they can cause another fission. Achieving a self-sustaining chain reaction requires engineering the geometry, materials, and enrichment of the fuel to push k to exactly 1 in a controlled way.

Moderator — a material used to slow fast neutrons (which emerge from fission at high speeds) down to thermal energies, where they interact much more effectively with uranium-235; common moderators are graphite (carbon), heavy water (D₂O), and ordinary light water (H₂O); Fermi chose graphite.

Fermi's insight — building on work by himself, Leo Szilard, Herbert Anderson, Walter Zinn, and others — was that a carefully constructed arrangement of natural uranium and graphite could achieve criticality. The graphite would moderate the fast fission neutrons to thermal energies; the thermal neutrons would then trigger further fissions in uranium-235 (which constitutes 0.7% of natural uranium); and the geometry of the pile could be designed so that enough neutrons completed the cycle to keep k at or above 1.

Source: Fermi, E. (1952), "Experimental Production of a Divergent Chain Reaction," American Journal of Physics, 20(9), 536–558 — Fermi's own account of Chicago Pile-1, written after the experiment could be declassified; includes his description of the critical measurements and the December 2 run.

The Physics Behind the Design

The selection of graphite as a moderator was not obvious. Graphite is carbon, and carbon has a small but nonzero probability of absorbing neutrons — the "absorption cross-section." If the graphite contains even trace impurities of elements with higher absorption cross-sections (boron is particularly bad, absorbing neutrons voraciously), the pile would never reach criticality no matter how large it was built.

Fermi and his team needed graphite of exceptional purity. This was not commercially available in 1942. The team worked with graphite manufacturers — primarily the National Carbon Company — to develop production processes that reduced boron impurities to a few parts per million. The difference between "reactor-grade" graphite and commercial-grade graphite was, in this context, the difference between a possible reactor and an impossible one.

Source: Holl, J.M. (1997), Argonne National Laboratory 1946–96, University of Illinois Press — history of the Metallurgical Laboratory and its graphite procurement challenges; based on declassified Manhattan Project records.

Similarly, natural uranium contains 99.3% uranium-238 and only 0.7% uranium-235. Uranium-238 absorbs thermal neutrons without fissioning — it is a neutron absorber, which reduces k. The trick was to use a geometry — alternating lumps of uranium (or uranium oxide) with blocks of graphite — that maximized the probability of a fission neutron being moderated to thermal energy before encountering a U-238 nucleus. Detailed neutron physics calculations, done by Fermi's group at Columbia and then Chicago, showed that a "heterogeneous pile" — uranium lumps embedded in a graphite matrix — would be more efficient than a homogeneous mixture.

The calculations predicted the critical mass and geometry. But the calculations depended on nuclear constants — the fission cross-section of U-235, the absorption cross-sections of U-238 and graphite, the average number of neutrons per fission — that were themselves being measured in real time as the project progressed. Every material constant measurement fed back into the design calculations.

Source: Weinberg, A.M. & Wigner, E.P. (1958), The Physical Theory of Neutron Chain Reactors, University of Chicago Press — the standard technical reference for reactor physics, co-written by participants in the CP-1 program; covers the neutron physics underlying the pile design.

From Columbia to Chicago: The Metallurgical Laboratory

The work that culminated in Chicago Pile-1 had its origins at Columbia University in 1939–1941. After the January 1939 news of Otto Hahn and Fritz Strassmann's fission discovery, Fermi and Szilard immediately began thinking about chain reactions. By 1940, they had conducted preliminary exponential experiments at Columbia — measurements of neutron multiplication in small uranium-graphite assemblies that were not large enough to go critical, but which demonstrated that a sufficiently large pile of the right geometry would.

In December 1941 — the same week as the Pearl Harbor attack — the Manhattan Project's Army administration decided to consolidate the chain-reaction research under Arthur Compton at the University of Chicago. The program was given the deliberately vague name "Metallurgical Laboratory" (Met Lab) to avoid drawing attention. Fermi, Szilard, Herbert Anderson, Walter Zinn, and dozens of other physicists relocated to Chicago.

The original plan was to build the first pile at the Argonne Forest preserve outside Chicago, where remote location would reduce the risk to the public if something went wrong. But construction delays at Argonne, combined with mounting pressure to achieve the critical experiment before the end of 1942, led Compton to make a decision that would today be unthinkable: build the pile in the city, under the west stands of Stagg Field — the University of Chicago's sports stadium — in a squash court that was no longer used.

Source: Compton, A.H. (1956), Atomic Quest: A Personal Narrative, Oxford University Press — Compton's memoir, including his account of the decision to build CP-1 in Chicago rather than Argonne.

The decision was made without informing the University's president, Robert Maynard Hutchins, who learned of it afterward. The secrecy of the Manhattan Project made consultation impractical; the timeline pressure made delay unacceptable.

Building the Pile: November–December 1942

Construction began in mid-November 1942. The pile was assembled layer by layer, like a loaf of bread, in the 30-foot (9 m) high squash court. Workers — mostly physicists, technicians, and a team from the Met Lab — handled the graphite bricks and uranium slugs by hand, working in shifts around the clock.

The structure was an oblate sphere — roughly egg-shaped — about 20 feet (6 m) wide and 20 feet tall, composed of 385 tons of graphite bricks and 46 tons of uranium metal and uranium oxide. The uranium was contained in 22,000 "pseudospheres" (roughly cylindrical lumps) of uranium metal and 6 tons of uranium oxide, distributed throughout the graphite matrix in a precise pattern calculated by Fermi's neutron physics team.

Control rods — strips of cadmium metal that absorb neutrons strongly; inserting the rods into the pile absorbs enough neutrons to keep k below 1 (subcritical), preventing the chain reaction; withdrawing them allows k to approach 1.

Three sets of control rods were used. The primary safety rod was a heavy cadmium strip that could be released by breaking a rope — if all else failed, gravity would insert it and shut the pile down. A second rod was operated manually. A third rod — the one Fermi used to bring the pile to criticality — was withdrawn in measured increments, with neutron counter readings taken after each step.

Leona Woods (later Leona Marshall Libby), the only woman present on December 2, operated one of the boron trifluoride neutron detectors and was responsible for monitoring its output. She was 23 years old and had recently completed her PhD.

Source: Marshall Libby, L. (1979), The Uranium People, Crane Russak — memoir by Leona Marshall (Libby), one of the few first-person accounts from a non-Fermi participant in the CP-1 experiment.

The pile grew layer by layer through November. After each layer, the pile's multiplication factor was measured by briefly withdrawing the control rods and measuring the neutron growth rate. Fermi could project, from these sub-critical measurements, exactly how many additional layers would be needed to achieve criticality. The calculations were extraordinarily precise: Fermi predicted criticality would occur at layer 57.

By the evening of December 1, the pile was at layer 57. Fermi ordered the control rods fully inserted for the night, declaring the experiment would proceed on December 2.

Component	Quantity
Graphite bricks (total)	385 tons (349,000 kg)
Uranium metal	~12,400 kg (approximately 12.4 tons)
Uranium oxide (UO₂)	~36,000 kg (approximately 36 tons)
Total uranium	~48,400 kg
Layers in the pile	57
Pile dimensions	~6 m wide × 6 m tall (oblate sphere)
Control rods	3 (1 gravity-drop safety, 2 manual)
Total building time	Approximately 18 days (mid-November to December 1)
Location	Squash court, West Stands, Stagg Field, University of Chicago

December 2, 1942: The Experiment

About 40 people were present in the squash court and on the balcony above it. The official count from participant memoirs and later interviews places the number at approximately 42. They included physicists, safety monitors, and a few project administrators. No journalists were present. Photographs were not taken at the moment of criticality — the work was classified.

Fermi arrived at about 9 AM. He spent the morning taking the pile through a series of careful approach-to-criticality measurements — pulling the main control rod out in small increments, measuring the neutron count rate at each position, plotting the inverse count rate (which should go to zero at criticality), and comparing with predictions.

Shortly after 11 AM, an automatic safety rod tripped — a pre-set neutron counter had reached its threshold. Fermi called a lunch break.

After lunch, the experiment resumed. At approximately 2 PM, Fermi ordered the control rod out to the 11.5-foot (3.5 m) mark. The counters climbed and leveled off. Not critical yet, but approaching. At 3:25 PM, he ordered it pulled to 13 feet (4 m). This time, the counters did not level off. They kept climbing.

Herbert Anderson, watching the chi-square recorder, said later: "the reaction was not increasing. It was rising and it was rising fast. It was like watching a clock."

Fermi held the pile critical — with k slightly above 1 — for 28 minutes, during which the neutron flux doubled approximately every two minutes and the power output reached a maximum of about 200 watts. He then ordered the control rod reinserted. The clicking of the counters dropped back to background.

Source: Allison, S.K. (1965), "The First Pile," Bulletin of the Atomic Scientists, 21(10), 4–7 — account of the December 2 experiment by a Met Lab physicist who was present; one of several first-person accounts gathered in the 1960s.

200 watts is a deliberately underwhelming number. A standard incandescent light bulb uses 60–100 watts. The world's first nuclear reactor, in its first run, produced enough power to light two light bulbs. This was correct: the goal was not to generate power but to achieve and demonstrate a self-sustaining reaction at a measurable, controlled rate. Running at higher power would have produced significant radiation and heat — neither of which the squash court was equipped to handle.

Source: Fermi, E. (1952), American Journal of Physics, 20(9), 536 — Fermi states the maximum power as "about half a watt" in his first postwar account, later revised to approximately 200 watts based on dosimetry measurements; the Atomic Energy Commission historical records use 200 watts.

"The Italian Navigator"

After the pile was shut down, Arthur Compton telephoned James Conant, the chairman of the National Defense Research Committee, at Harvard. The call was in plain language on an ordinary telephone line — presumably someone judged that German intelligence would be unlikely to intercept a conversational exchange between academic administrators.

The exchange, as both men later recorded it:

Compton: "The Italian navigator has just landed in the New World."

Conant: "How were the natives?"

Compton: "Very friendly."

"The Italian navigator" was Fermi. "The New World" was the territory of controlled fission. The analogy to Columbus's 1492 landing — Italian navigator, new world — was both pointed and appropriately coded.

Compton later recalled that he then opened a bottle of Chianti (which had been brought by Szilard in anticipation of the event) and everyone present signed the straw-covered bottle's label. The bottle and its signed label are preserved at Argonne National Laboratory.

Source: Compton, A.H. (1956), Atomic Quest, p. 144 — Compton's account of the phone call; corroborated by Conant's own recollection in: Conant, J.B. (1970), My Several Lives: Memoirs of a Social Philosopher, Harper & Row.

What Made CP-1 Different from What Came Before

The months of exponential experiments at Columbia and the sub-critical measurements during pile construction were all preparatory. What was new on December 2 was the demonstration that k could be held at exactly 1 in a stable, controllable way.

The theoretical concern was whether criticality would be stable or runaway. If withdrawing the control rod by a small amount caused a self-sustaining chain reaction, would it also cause an uncontrolled acceleration? Or could it be arrested cleanly by reinserting the rod?

The answer hinged on a phenomenon Fermi had calculated carefully: delayed neutrons. In a fission chain reaction, most neutrons are emitted instantaneously at the moment of fission — the "prompt neutrons." But a small fraction (about 0.65% for uranium-235) are emitted seconds to minutes later, from the radioactive decay of fission fragments. These "delayed neutrons" are responsible for the relatively slow response time of a reactor to control rod movements.

Delayed neutrons — neutrons emitted by excited fission product nuclei with a characteristic delay of 0.2 to 55 seconds after fission, constituting about 0.65% of all neutrons in U-235 fission; their delayed emission makes reactor control possible, because without them the response to control rod movements would be too fast for mechanical control systems to manage.

If delayed neutrons are included in the multiplication factor, the effective k for purposes of control is close to 1 even when the "prompt k" (based only on prompt neutrons) is slightly below 1. This means that the chain reaction grows on the timescale of delayed neutron emission — seconds — rather than the microsecond timescale of prompt neutrons. Control rods moving at human speed are fast enough to regulate the reaction.

Without delayed neutrons, a nuclear reactor would be mechanically uncontrollable. Their existence was known theoretically but had not been confirmed in a true chain reaction before CP-1. Their effect was exactly as Fermi had calculated. The pile responded to the control rod on a timescale of seconds, not microseconds, and the reaction could be arrested in less than a second by reinserting the rod.

Source: Hughes, D.J. & Harvey, J.A. (1955), Neutron Cross Sections, BNL-325 (Brookhaven National Laboratory) — the standard compilation of neutron cross-section data underlying reactor physics calculations; reflects measurements taken during and after the Manhattan Project period.

The Timeline: From Fission to Criticality

Date	Event
December 1938	Otto Hahn and Fritz Strassmann split uranium in Berlin; Lise Meitner and Otto Frisch identify the reaction as fission, January 1939
January–March 1939	News of fission spreads through the physics community; Fermi, Szilard, and others immediately consider chain reaction possibility
March 1939	Szilard and Fermi confirm that fission releases multiple neutrons — necessary for a chain reaction
August 2, 1939	Einstein-Szilard letter to Roosevelt, warning of potential German bomb program
1939–1941	Exponential pile experiments at Columbia University; refinement of graphite purity requirements
December 6, 1941	Manhattan Project reorganized under Army Corps of Engineers; Compton consolidates chain-reaction work at Chicago
December 7, 1941	Pearl Harbor attack; US enters World War II
January–November 1942	Met Lab established at University of Chicago; graphite and uranium procurement; design calculations
November 16, 1942	Construction of Chicago Pile-1 begins in Stagg Field squash court
December 1, 1942	Pile reaches layer 57; Fermi predicts criticality; experiment paused for the night
December 2, 1942, 3:25 PM	First self-sustaining chain reaction achieved; power ~200 W; run for 28 minutes
December 2, 1942, ~5 PM	Compton calls Conant: "The Italian navigator has just landed in the New World"
February 1943	CP-1 dismantled and rebuilt as Chicago Pile-2 at Argonne, with radiation shielding
September 1944	Hanford B Reactor — the first full-scale production reactor — achieves criticality
July 16, 1945	Trinity test — first nuclear weapon detonation
August 6 & 9, 1945	Hiroshima and Nagasaki
December 1951	EBR-I at Idaho produces the first electricity from nuclear fission

Who Was in the Room

The approximately 42 people present on December 2 included several who would shape nuclear physics for the next half-century:

Enrico Fermi — directed the experiment; Nobel Prize in Physics 1938 (for neutron-induced radioactivity); widely considered the last physicist equally at home in theory and experiment.

Leo Szilard — had conceived the nuclear chain reaction concept in 1933 (patenting it in 1934); co-wrote the Einstein letter to Roosevelt; was deeply ambivalent about the military application of what they were building.

Herbert Anderson — Fermi's primary experimentalist; operated the main neutron monitoring equipment; later built on CP-1 work for reactor design.

Walter Zinn — managed the pile's construction logistics; later became the first director of Argonne National Laboratory and a pioneer in breeder reactor design.

Leona Woods (Marshall) — the only woman present; operated a boron trifluoride detector; one of the few participants who wrote a first-person memoir of the event.

George Weil — operated the final control rod that Fermi directed on December 2; his hand on the rod when criticality was achieved.

Source: Rhodes, R. (1986), The Making of the Atomic Bomb, Simon & Schuster — Pulitzer Prize-winning history of the Manhattan Project, with detailed coverage of CP-1; the standard narrative history of the event.

Szilard, as the others filed out, reportedly took Fermi's hand and said: "This day will go down as a black day in the history of mankind." He was the only one present who seemed fully focused, in the moment, on the implications of what they had done.

What CP-1 Did Not Do — and Why It Mattered Anyway

Chicago Pile-1 was not a power plant. It produced 200 watts — enough to light a lamp, not enough to power a building. It had no cooling system: the graphite absorbed the small amount of heat generated, and the power was deliberately kept low enough that the temperature rise was negligible. It had no radiation shielding: at 200 watts, the dose rates were low enough that the people in the room were not at significant risk during the 28-minute run, though chronic occupancy of the space over subsequent months would have been another matter.

What CP-1 proved was not a power-generating concept. It proved three things:

1. A self-sustaining chain reaction is physically achievable with natural uranium and graphite — no enrichment required, no exotic materials.
2. The reaction can be controlled by mechanical means (control rods) on human timescales, because delayed neutrons make the system respond slowly enough.
3. The neutron physics calculations — the cross-sections, the multiplication factor, the geometry requirements — were correct. The theorists had gotten the physics right.

Points 1 and 2 were necessary demonstrations before the Manhattan Project would commit to building multi-hundred-megawatt plutonium production reactors at Hanford, Washington. A failure at Chicago would have set the production reactor program back by months or years — potentially affecting the outcome of the war.

Point 3 mattered for a different reason: it validated the entire theoretical framework on which reactor design, bomb physics, and ultimately nuclear power engineering would be built.

Source: Hewlett, R.G. & Anderson, O.E. (1962), The New World: 1939–1946 (A History of the United States Atomic Energy Commission, Vol. I), Penn State University Press — official history of the Manhattan Project period, with detailed coverage of the CP-1 results and their role in accelerating the production reactor program.

What People Get Wrong About Chicago Pile-1

"CP-1 was the first nuclear reactor." In a strict sense, yes — but the word "reactor" was not used at the time. The device was called a "pile" (from Fermi's Italian term pila, meaning "pile" or "battery"). The term "nuclear reactor" came into common use only after the war. The pile had no heat extraction, no cooling, no shielding, and no electricity generation — it was an experiment, not a generator.

"It could have caused a nuclear explosion." No. A graphite-moderated natural uranium pile cannot produce a nuclear explosion. The physics do not allow it. An uncontrolled runaway in CP-1 would have produced a large release of heat and radiation — a "criticality accident," potentially lethal to those present — but not a nuclear detonation. Nuclear weapons require highly enriched material and an implosion or gun-type geometry that compresses the fuel to supercriticality faster than the material can blow itself apart; neither condition was present at CP-1.

"The experiment was kept secret from everyone." The University of Chicago's president was not informed; the local government was not informed; no press was present. But approximately 42 people witnessed the event, and the Met Lab employed hundreds of scientists and support staff who had varying degrees of knowledge about the project's goals. The event was classified, but not invisible within the project.

"Fermi was taking enormous risks." Fermi was methodical to an unusual degree. The sub-critical measurements throughout November gave him confidence in the predicted critical mass. The control rod system was multiple-redundant. The operating power (200 W) was chosen specifically to be well below any danger threshold. Fermi's "suicide squad" — a team of men stationed above the pile ready to flood it with cadmium solution — was a final safety backup. The risk was real but carefully bounded.

FAQs

Q: What exactly happened at 3:25 PM on December 2, 1942?
A: Enrico Fermi ordered George Weil to withdraw the main control rod to the 13-foot (4 m) mark, leaving about 6 inches of rod inside the pile. The neutron count rate, instead of leveling off as it had at each previous position, began to climb continuously. This indicated that the multiplication factor k had exceeded 1 — the chain reaction was self-sustaining. Fermi allowed the pile to run for 28 minutes before ordering the control rod reinserted, bringing the reaction to a stop.

Q: Why was the pile built under a sports stadium?
A: The original plan was to build at the Argonne Forest site outside Chicago, but construction delays and time pressure (the Manhattan Project needed a proof of concept before committing to full-scale production reactors) led Arthur Compton to authorize building at Stagg Field. The squash court there was unused, it was physically large enough, and secrecy could be maintained within the classified wartime research program. The decision was made without informing the University president.

Q: Was anyone in danger during the experiment?
A: At 200 watts of power, the radiation dose rates during the 28-minute run were low and did not pose an acute danger to the ~42 people present. The main safety risk would have been a runaway reaction caused by a control system failure, which could have produced a criticality accident with lethal radiation doses in the room. The multiple control rod systems and the "suicide squad" with cadmium solution were in place to prevent this. No one was injured.

Q: How does CP-1 relate to modern nuclear power plants?
A: Every nuclear power plant in operation today is based on the same fundamental physics demonstrated by CP-1: a moderated, controlled fission chain reaction with k maintained at approximately 1 by mechanical control systems. The moderators (usually light or heavy water rather than graphite), fuels (enriched uranium or mixed oxide), and safety systems have evolved enormously, but the principle is unchanged. CP-1's demonstration that the delayed neutron effect makes mechanical control possible was essential to the entire subsequent history of nuclear power.

Q: What happened to Chicago Pile-1 after December 2, 1942?
A: CP-1 was operated a few more times through December 1942 and early 1943, then dismantled and rebuilt with radiation shielding at the Palos Hills forest preserve outside Chicago as Chicago Pile-2 (CP-2). The original Stagg Field squash court was eventually demolished in 1957. A memorial plaque marks the site today, near the Henry Moore sculpture "Nuclear Energy," installed in 1967 on the 25th anniversary of the experiment.

Q: What did "The Italian navigator has just landed in the New World" mean?
A: "The Italian navigator" was Enrico Fermi, an Italian-American who had emigrated after receiving his Nobel Prize in Stockholm in 1938 rather than returning to Mussolini's Italy. "The New World" referred to the new territory of controlled nuclear energy — with the secondary meaning of America, where the discovery was made. "Very friendly natives" meant the experiment had gone smoothly without incident. The coded language was used because the call was made on an unsecured telephone line.

The Takeaway

Chicago Pile-1 ran for 28 minutes and produced enough power to light two light bulbs. That is its entire physical output. Its actual legacy is the proof of concept for every nuclear reactor and every nuclear weapon that has existed since December 2, 1942. Fermi's meticulous buildup through 57 layers of graphite and uranium, his sub-critical approach-to-criticality measurements, his multiple-redundant safety systems, and his precise prediction of the critical configuration demonstrated that nuclear fission can be controlled — that the most energetic reaction known to physics can be brought to heel by a rod of cadmium. What followed that demonstration — the plutonium production reactors, the bomb, and eventually the commercial nuclear power industry — was the application of what Fermi proved on a cold Wednesday afternoon in Chicago.

---

Read next:

• The Rutherford Nuclear Atom
• The Chadwick Neutron Discovery
• The Millikan Oil Drop Experiment