1. Introduction
  2. Uranium
  3. Mineralogy of Uranium
  4. Radioactivity
  5. Nuclear Power
  6. Health Physics
  7. References


Uranium is now indelibly associated with atom bombs and nuclear power, but was a chemical curiosity of little importance for a hundred years after its discovery. Then, it led to the discovery of radioactivity and the elucidation of the structure of the atom. A half-century more saw its use in nuclear explosives and in power generation. Most of the importance of uranium has been in physics rather than in chemistry, unlike that of most chemical elements. In this article, as much as I can find on the chemistry of uranium will be outlined, but the physical subjects of radioactivity and nuclear power will be discussed at length. The elements of these subjects are not hard to understand, and can go a long ways towards a useful appreciation of subjects that are often in the news. In particular, it may help in making a meaningful and intelligent assessment of the risks and advantages in these matters.


Martin Klaproth, an early convert to Lavoisier's chemistry, discovered a new substance in pitchblende from Joachimsthal, Bohemia in 1789, and named it uranium after the ultimate planet in the solar system as then known, Uranus, which had been discovered by William Herschel a few years earlier, in 1781. Since Mendeleff's periodic system did not yet exist, Klaproth would not have realized that uranium, atomic number 92, would for many years be the ultimate element. When elements 93 and 94 were made many years later, the names neptunium and plutonium carried on the analogy with the solar system. Klaproth had actually found the oxide, not the metal, but at least he proved that pitchblende was not an ore of tungsten, iron or zinc. E. M. Péligot isolated the pure metal in 1842. Uranium was found to have the atomic weight of 238.07, the greatest of any element, and was assigned the symbol U.

A fresh surface of metallic uranium is white and lustrous, with a bluish tinge, but it soon tarnishes in air. There are three crystalline forms of metallic uranium; the usual form is orthorhombic, and is malleable and ductile. It can be cast, rolled and extruded, and is often produced in 1"-diameter rods. The metal oxidizes rapidly when heated to 200°C, and burns in air. This behavior is typical of all the heavy metals in this region of the periodic table, including radium and plutonium. When used as metals, they must be "cladded" with another metal, such as zirconium or aluminium, to protect them from moisture and oxygen. Uranium is soluble in mineral acids, and will displace hydrogen from water. Since the nitrates of uranium and its friends are soluble, chemistry is generally performed with them. A notable property of uranium is its high density, about 19 g/cc (sources give 18.9 and 19.04). This is about twice the density of lead (11 g/cc), and puts uranium up with the densest metals, such as platinum, osmium and iridium. Uranium melts at 1133°C with latent heat 2.5-3.0 kcal/mol, and boils at 3925°C with latent heat 110 kcal/mol. Its coefficient of thermal expansion is 15 x 10-6 per °C, and its electrical resistivity is 29 μΩ-cm.

The availability of metallic uranium as a by-product of the production of enriched uranium, and of plutonium for nuclear weapons ("spent uranium") has encouraged its use in armor-piercing projectiles, and perhaps as armor itself, taking advantage of its high density. The uranium in these devices escapes its cladding when the projectile melts on impact, releasing burning uranium which can do no one much good.

Uranium has a ground-state electron configuration of 5f36d7s2 outside a radon core. The f, d and s levels all have about the same energy, so it is not easy to guess what orbitals are used. Three of the valence electrons occupy the f orbitals that were left empty, moving from the d orbitals where they might be expected to reside. The first ionization potential of the atom is about 4 V, a relatively low value. Uranium is, therefore, an electron donor, easily oxidized, and exhibits valences of +3, +4, +5 and +6, with +4 (uranous) and +6 (uranyl) being the most prominent. Uranium is more liberal with valence states than most of its neighbors. The oxides are not soluble in water, so we do not know if they would give basic or acidic hydroxides, but basic would be more probable, so that uranium would definitely behave as a metal. The uranium atom is not unusually large, since the high nuclear charge sucks the electrons in closely. The ions are only about 0.1 nm in radius, a very typical value. Natural uranium is, on the average, an isotopic mixture of 99.27% U238, 0.72% U235 and 0.0055% U233.

UO2, UO3 and U3O8 are observed as oxides. Uranium exercises all its valences in combining with the halogens. The compound UF6, uranium hexafluoride, is typical. UF6 is a means of "gasifying" uranium, since it boils at 56.2°C at 65mm pressure, and is used in isotope-separation processes such as diffusion and centrifugation. Uranium does not seem to combine with hydrogen, but it forms the sulphides US2 and U2S3, which ignite rather readily. Thre is also a uranous sulphate, U(SO4)2·4H2O, which is nearly insoluble but dissolves in HCl. Uranous compounds are usually greenish or dark.

When the oxides are dissolved in acids, the stable uranyl cation, UO2++, is formed. This cation forms numerous soluble salts, such as UO2(NO2)2·6H2O. The yellow, orthorhombic crystals melt at 60.2°C, boil at 118°C and have a density of 2.807 g/cc. Deliquescent uranyl chloride, UO2Cl2, and uranyl iodate, UO2(IO3)2, are also known. Uranyl gives the color yellow in aqueous solution, and this color is typical of all hexavalent uranium.

Before the discovery of radioactivity, uranium was a curiosity with few uses. It gives a green color to ceramics and glass, as shown by the borax bead test. Uranyl nitrate is used in a reddish-brown photographic toner. Zinc uranyl acetate ("sodium reagent") is used to precipitate sodium in qualitative analysis, since pale yellow NaZn(UO2)3(C2H3O2)9 is one of the very few insoluble sodium compounds. Uranium metal was used for some gas-discharge tube electrodes, and compounds were used in making incandescent gas mantles. None of these uses is current, and it seems to be difficult even to procure uranyl compounds from chemical supply houses, probably more from lack of use than uranophobia. Uranium makes hard carbides like tungsten and molybdenum, which it resembles chemically. However, its use in hgh-speed tool alloys gave no advantage.

Uranium and its compounds are poisonous, and should not be taken internally, nor should their dusts be breathed. The chronic breathing of dusts is probably the most hazardous activity, but unless you work in the uranium-mining industry without a mask, you will not be exposed to this danger. Natural uranium minerals do not represent a hazard, since they are very insoluble and only very weakly radioactive. They can be collected, handled and studied with complete safety. Even uranium metal, natural or enriched, can be handled safely like any other metal. Granites usually contain a very small amount of widely-disseminated uranium, which generates the heavy inert gas radon. If this gas is trapped, in rare cases it may be present in such concentrations that breathing it is a hazard. The radon will decay in the lungs to a chain of radioactive substances that can do damage. This is an extremely feeble hazard that is completely avoidable by proper ventilation. The hazards of radioactivity will be discussed in more detail below.

Mineralogy of Uranium

The abundance of uranium in the earth's crust is about 4 x 10-4%, roughly the same as that of boron or tungsten. It is a little more common than gold or rubidium, but thorium and lead are 3 and 4 times more abundant, respectively. Boron and tungsten are found in common minerals, boron in Kernite and tungsten in Ferberite-Heubnerite, but no uranium mineral is common, and most are very rare. The availability of an element depends not only on its abundance, but on its concentration by geological processes into an economic ore. The demand for uranium as a weapons material led to extremely thorough prospecting, and its occurrence is now well-known. There are probably only some 2,000,000 tons of high-grade ore available, a very small amount for a material of wide use.

The most common uranium ore is pitchblende, which contains mainly uranous oxide, UO2, which as a pure mineral is uraninite. The "pitch-" came from its black, resinous appearance and botyroidal habit, which looks like pitch, and "-blende" because it cannot be smelted to any valuable industrial metal, like copper. It is a variable material, since the oxide is more or less partially oxidized to U2O8, which decreases the hardness from Mohs 5.5 down to 3-4, and its density from a very high 9.0-9.7 down to 6.5-8.5, which is still unusually heavy for a mineral. It is soluble in nitric, sulphuric and hydrofluoric acids, but not in hydrochloric.

Pitchblende was best-known from Joachimsthal, Bohemia in the Erzgebirge (now Jachymov)--this was the source of uranium for Klaproth, and later for the Curies. It is found also in Germany, Portugal, and notably in the Congo, at Lumumbashi in Katanga. It was known in Cornwall, found in Colorado at Central City, in Canada around the Great Bear Lake in the northwest, while large crystals come from Wilberforce, Ontario. Because pitchblende contains uranium, it also contains all the radioactive decay products of uranium, down to lead. The intermediate products are all present in constant proportions, while as uranium decays the lead builds up. This is true of any uranium mineral, not just pitchblende, of course. Because the half-life of uranium is so great, its daughters are present in very small concentrations.

The other economic ore of uranium is carnotite, with an average composition of K2O.2U2O3. V2O5.2H2O, a soft, yellow, powdery mineral that is much easier to work with than pitchblende, but rather rarer. A specimen is shown at the right (image © Amethyst Galleries). It's a powdery coating with the typical color. Its hardness is 2-4, its density 4.136 g/cc. Its monoclinic (or orthorhombic?) crystals are biaxial, with rather high indices 1.75 and 2.06. Normally, however, it is not found as euhedral crystals, but as crusts and cementing sandy sediments. Its solution in HCl is yellow, in HNO3 green, and in H2SO4 orange. It is a secondary mineral concentrating widely-disseminated uranium in granites. It is found in Utah and Colorado, notably in the Paradox Valley, and in Wyoming near Jeffrey City.

The mineral Coffinite, probably related to Carnotite, was discovered in 1955 in Colorado, but does not appear in common mineralogical references. The unusual violet Ianthinite is UO2(OH)2. Curite is said to be 3PbO.8UO3.4H2O. There is a number of very rare uranyl minerals, that are copper-uranyl or calcium-uranyl phosphates, vanadates or arsenates. Among the silicates are Kasolite, PbUO2SiO4.H2O, Uranophane, CaH2(SiO4)2UO2.5H2O, and Cuprosklodowskite, Cu(UO2)2Si2O6(OH) 2.5H2O. The structures of these rare minerals are not clear to me, so I simply report them. They do not appear in Dana, and probably do not represent practical sources of uranium. They do make very attractive colored crystals, however.

Commercial production of radium began in 1910. About 750 tons of raw pitchblende, or 12 tons of concentrated ore, must be processed to win one gram of radium. From 1913 to 1925 the chief source of uranium was Colorado carnotite. In 1925, discovery of pitchblende in the Congo diverted production to that source. In 1931, Canadian pitchblende from Great Bear Lake took over. The uranium boom of the 1950's saw uranium mills in Moab, Utah; Jeffrey City, Wyoming; Durango, Canon City and Uravan, Colorado; and many other places in the arid West, which processed mainly low-grade carnotite into oxide. These mills are now derelict, or scratch out an existence as best they can, beside huge piles of waste. These mines and mills produced only weakly radioactive concentrates ("yellowcake"), not metallic uranium or purified radionuclides.

Vanadium and uranium were concentrated at Durango, Colorado from 1942 to 1962, resulting in two large tailings piles at the southwestern edge of town. There are 23 other such sites in the West. These piles were unsightly, and leachings from them might be chemically obnoxious, but the radioactive hazard was extremely small. Any house built on them might be suspected for radon buildup, and the dust should not be chronically breathed (like so much nonradioactive dust), but they contained much less uranium than the original carnotite did, and presented even less of a danger. Indeed, if only vanadium had been extracted, which was actually the primary purpose, they would have attracted no interest, except by those who had to look at them. Uranophobia and ignorance, however, have led to expensive follies. At least they could get the government to pay for taking them out of sight. The heaps, I must assume, have now been moved to somewhere where they are also out of mind. Transportation accidents while moving the piles probably killed and injured more people than ever the dirt will do.

Thorium is usually found associated with the rare earths, especially Cerium. The principal source is Monazite sand, a Th-Ce-Y-La-Nd phosphate, which can contain up to 12% ThO2. Monazite is heavy and hard, so is easily separated from the lighter silicic gangue. A very rare mineral is Thorianite, (Th-U)O2, which contains uranium as well. Brannerite and Betafite are other rare minerals containing rare earths, thorium and uranium, but are not common enough to be ores.


Henri Becquerel (1852-1908) was studying the properties of fluorescent compounds in 1896, one year after Röntgen's amazing discovery of X-rays. He noticed that a uranyl compound, obtained from carnotite, fogged photographic emulsion that was wrapped to exclude light. He was, of course, delighted to have discovered a "radiation" himself, and set Marie S. Curie (1867-1934) to work on it. She was aided by Pierre Curie, her husband and a noted physicist, who quickly became interested as well. They observed that when the uranium was separated from pitchblende, the residue was very active. Curie named the phenomenon radioactivity. At length, the radiation was found to consist of positively charged heavy particles, negatively charged electrons (themselves only recently discovered), and electromagnetic radiation like X-rays but more penetrating. The three components were called α, β and γ radiations, respectively. Becquerel and the Curies received the Nobel Prize in Physics for 1903. Marie Curie received the Chemistry prize in 1911. There was even a Hollywood film of the 1940's about her discovery of radium.

Curie discovered that the pitchblende evolved small amounts of the gas helium, identified by its spectrum, which had been found in the sun some years earlier. This was the first terrestrial helium found. All helium in the atmosphere and in natural gas comes from radioactive decay, since it was soon demonstrated that the alpha particles became helium. This led Rutherford and Soddy to deduce that radioactivity was the result of the transformation of one atom into another in 1902, and from this to the concept of the nuclear atom. Radioactivity was the result of nuclear transformations. In 1913, Rutherford demonstrated the existence of nuclei, and Bohr showed how electrons moved. Alpha particles were simply the nuclei of helium atoms.

A nucleus can be characterized by its atomic number Z, the positive charge on its nucleus in units of the electronic charge, and its mass number A, the total number of nucleons in its nucleus. The nucleons are the proton, p, with mass 1.007596 amu and charge +1, and the neutron, with mass 1.008986 amu and charge 0. The atomic mass unit, amu, is just the reciprocal of Avogadro's number, and corresponds to 931 MeV in energy, according to Einstein's relation E = mc2. An MeV (mega-electron-volt) is 1.6 x 10-13 J. A nucleus (Z,A) contains Z protons and N = A - Z neutrons. Nuclei with the same Z, and therefore the same chemical properties, are called isotopes. The isotopes of uranium cover a range in mass number from 227 to 240. U227 is a 6.8-MeV alpha-emitter with T = 1.3 min, and U240 is a beta emitter with T = 14 hr.

The free neutron is heavier than a free proton, which favors beta decay of the neutron, n0 → p+ + e- + ν0. The ν is an antineutrino, necessary so that the fermion number (electron +1, antineutrino -1) is conserved (zero on both sides of the equation). The antineutrino carries off varying amounts of energy, so the maximum energy the electron can have corresponds to a zero antineutrino energy. Since the mass of the electron is 0.000548 amu (0.51 MeV), an amount of mass of 0.000842 amu has disappeared in this reaction, which must appear as the kinetic energy of the products. Most of this energy will be in the electron, since the proton is much heavier, and the amount will be 0.78 MeV.

The chance that a neutron will decay in any time interval dt is proportional to dt. If N is the number of neutrons, this means that dN = -γNdt, since the number decaying in dt will be proportional to the number present. The constant γ is called the disintegration constant, and is independent of all physical influences. This is a remarkable law, by no means trivial or obvious, which says a lot about nuclear processes. This equation can be integrated to the familiar N = Noe-γt, which expresses the exponential decay of a radioactive nucleus. Half the original number of nuclei remain after a time T = ln 2/γ, which is called the half-life. The same answer is obtained no matter when we begin looking. The neutron has a half-life of about 13 minutes, so γ = 0.00128 per sec. If we have a thousand neutrons, then in the next second about one will transform into a proton. We speak of "disintegration" and "decay" but these are mere vivid terms--nothing disintegrates or decays in any meaningful sense. Half-lives vary from longer than can be measured, billions of billions of seconds, to less than can be measured, to mere nanoseconds. They are absolutely constant.

In the active residue from pitchblende, Curie isolated polonium in July 1898, and radium in September. Polonium (Z = 84) is chemically similar to bismuth, radium (Z = 88) to barium. All nuclei with atomic number 84 and greater are unstable, the ones up to about Z = 100 are radioactive, the remainder fission spontaneously. Only nuclei up to Z = 92 are found on earth, because any beyond this atomic number, if originally present, would have decayed by now to nuclei with Z < 84. Those with Z between 84 and 92 are found naturally only because they result from the decay of long-lived nuclei of uranium and thorium. U238 (4.67 x 109 y), U235 (7.13 x 108 y) and Th232 (1.39 x 1010 y) are the three patriarchs that give rise to three sequences of alpha and beta decay ending in Pb206, Pb207 and Pb208, respectively. The abundances of these isotopes of lead are 26%, 21% and 52%. Not all lead is of radioactive origin, but much is, and the atomic weight of lead varies depending on the source. The isotope Pb204, not the end of any radioactive series, has an abundance of only 1.3%.

The members of a radioactive series can be detected much more readily by radiation measurements than by chemical analysis, so the chemical nature of many of the daughter nuclei were not originally known. They were given designations like UZ (Pa234), RaF (Po210) or ThX (Ra224). The series often took alternative routes, with α,β or β,α alternatives. In equilibrium, every member of a radioactive series decays at the same rate it is created. Therefore, if N1 and N2 are the equilibrium number of two species with disintegration constants γ1 and γ2, N1γ1 = N2γ2, or the numbers are in inverse ratio to the disintegration constants. Those members with the longest half-lifes will be present in the greatest amounts. See if you can work out the ratio of the amount of radium to the amount of U238 in pitchblende. Answer: 0.34 g to 1 metric ton of uranium.

The ratio of uranium to lead in a rock can be used to measure the time since the rock was formed, if we assume all the lead came from radioactivity. If r is the ratio of the number of atoms of U238 to the number of atoms of Pb206 at time t, then the law of exponential decay can be used to find t = T [ln(1 + 1/r)/ln 2], if T is the half-life, 4.5 x 109 y for U238. If a particular rock gives r = 2, then t = 2.6 x 109 years. This method was suggested by Boltwood in 1905, soon after the discovery of radioactivity, and was brought to perfection by Arthur Holmes in the 1950's. Similar dating methods using the helium or argon trapped in igneous rocks can also be used. Thanks to uranium, we now know the absolute geological time scale, and can extend it back to Precambrian rocks. The oldest rocks found are a bit older than 3,000 million years old, while the age of the earth's crust is about 4,500 million years.

A fourth series would descend from U233, but its half-life is only 1.62 x 105 y, so it all would have been gone long ago. The end of its series is Bi209, whose isotopic abundance is 100%. However, this is not really the end, since Bi209 is estimated to have a half-life of some 3 x 1017 y against alpha decay to Tl205, which is stable and would be the end. Not nearly enough time has elapsed for this to have taken place, and there is still lots of bismuth. All isotopes of bismuth appear to be radioactive.

There are a few natural radioactive nuclides not included among the heavy elements and their radioactive series. The rare isotope K40, with abundance 0.0119%, is beta and gamma active with half-life 1.3 x 109 y. The energy of the gamma rays is 1.46 MeV. Rb87, abundance 27.8%, is also beta active with half-life 4.3 x 1010 y. These elements are valuable geologic clocks. Nd144 and Sm157 are alpha emitters, while Lu176 is a beta and gamma emitter. The radioactive nuclide C14 is produced by the N14(n,p)C14 reaction in the atmosphere due to cosmic rays, and is present in all living matter. Its half-life is 5600 y, and it is a beta emitter. One part in 1012 of atmospheric carbon is C14, and results in an activity of 15.3 dis/min per gram of carbon. After death, the carbon is no longer replaced by metabolism, and the activity decreases with the half-life of C14. In 1951, Willard Libby discovered how to use this effect for the dating of carbon-containing fossil matter. One of the first results was the dating of the last continental glaciation to 11,400 years before the present from the trees that were knocked down by the advancing ice. Comparison with other dating methods has shown that cosmic radiation intensity has been constant for at least 30,000 years. However, human activities have upset the C14 concentration. The carbon from fossil fuels has no C14, while C14 has been produced by nuclear weapons tests. C14 is present in such small amounts that these additions produce a notable effect, and must be allowed for, especially when current material is used as a standard. Radiocarbon dating can be quite accurate in determining the age of artifacts.

Ra226 has a half-life of 1620 y, corresponding to a disintegration constant of 1.354 x 10-11 per s. Alpha particles of 4.78 and 4.59 MeV are emitted (one or the other), and gamma rays of 0.187 MeV energy. It is the longest-lived member of its radioactive series, except for two uranium isotopes. One gram of radium contains 6.02 x 1023/226 nuclei, so the number decaying in a second will be 3.7 x 1010. This number is called a curie, a measurement of the number of disintegrations per second in any radioactive sample. An actual sample of radium that has been around a while is in equilibrium with all of its daughters, each of which contributes its curie. Radium was used in luminous watch dials in the ratio of one part to several thousand parts of the scintillator hexagonal ZnS, that emits a flash of light for every alpha particle that hits it. Zinc blende gives 7 times more light than the next most sensitive scintillator, Barium Platinocyanide. The method of making luminous paint is described in Glazebrook (see References). This was not a dangerous use of radium, because of the extremely small amounts used, but it is probably discouraged at present. The hazard to the user is totally negligible; that to the manufacturer can be avoided with care. Electroluminescent panels are a good substitute with no radioactivity hazard at all.

The result of the decay of Ra226 is Rn222, with half-life 3.825 days. If a gram of this gas were separated from the radium, it would have an initial activity of one curie. In radium therapy, this is what is done, and a little radon in a capsule is used instead of the actual radium, and its activity increases with time. The gamma rays emitted by the daughters of the radon provide the therapeutic effect, damaging cancer cells more than healthy cells. None of the alpha or beta radiation can penetrate the metal capsule.

Radioactivity is just a search for stability. For any Z, there is an optimum value of N that gives greatest stability. For small Z, Z = N is preferred, but as Z increases, N > Z is preferred. A nucleus that has too many neutrons can increase the Z/N ratio by emitting an electron in beta decay. This electron does not exist in the nucleus, but is created along with an antineutrino at the moment of decay. A nucleus with too few neutrons can get more by emitting a positive electron, or positron, along with a neutrino. The positron is the antiparticle of the electron, with fermion number -1, while the neutrino has fermion number +1. Nuclei of intermediate values of A, about A = 55 (in the vicinity of Fe), are more stable than those with smaller or larger mass. Heavy nuclei can move down the line of stability by emitting an alpha particle, reducing Z and N each by two. Since the line of stability curves upwards, this leaves the nucleus with too many neutrons, so generally a negative beta decay follows to set this right. The radioactive series are simply chains of these decays looking for a more stable state. Lighter nuclei, of course, have no alternative like inverse alpha decay that would make them heavier, and beta decays would get them nowhere.

Nuclear Power

Consider building up an atom of U235 by assembling 92 protons, 92 electrons and 143 neutrons. The total weight of these constituents is 237.0343 amu. If an atom of U235 is weighed, it turns out to weigh 235.1175 amu. What happened was that the parts attracted one another, and we were able to extract a large amount of energy in assembling them. From the mass defect, this energy turns out to be 1785 MeV, or and average of 7.59 MeV for each nucleon. Atoms with A about 55 show a maximum of 8.7 MeV per nucleon. Could we somehow split up the nucleus optimally into smaller parts, we would have (8.7 - 7.6) x 235 = 259 MeV to spare. If the nucleus just splits into two roughly equal parts, we get about 200 MeV. 1 MeV per atom corresponds to 2.3 x 107 kcal/mol, so 200 MeV gives us 4.6 x 109 kcal/mol, or about 2 x 107 kcal/g, enormously more than any chemical reaction. The heat of combustion of petroleum is about 10 kcal/g, for comparison. This process is called fission. Fission was first verified by Hahn and Strassman on 6 January 1939, though Fermi had been working on it since 1934.

At the other end of the mass scale, the mass of a helium atom is 4.003860 amu, while the mass of two hydrogen atoms and two neutrons is 4.035087 amu. This means that we will liberate about 29 MeV for each helium atom we make from protons and neutrons. The binding energy in the helium nucleus is about 7.27 MeV/nucleon, not much less than the maximum possible, so little is gained by further assembly. The energy output is 6.71 x 108 kcal/mol, or 1.7 x 108 kcal/g, even more than we obtained from fission. This process is called fusion. The stars get their energy from fusion of hydrogen into helium, so solar energy has its source in fusion. The mechanism involves several intermediate steps, so it is not quite as simple as it might appear. Fusion has been produced on earth by using a fission explosion to ignite a blanket of lithium deuteride, LiD, to fuse two deuterons, each consisting of a proton and a neutron, into a helium nucleus, in what is called a "hydrogen bomb." Attempts to make a fusion reactor began with Project Sherwood in 1948, and success has not yet been achieved. It is certainly possible, but we do not know how to do it.

The heaviest nuclei are on the verge of fission. They wobble like a large drop of mercury, and if given a little extra energy they can be pushed over the edge and separate into two parts. The extra energy can be obtained by the absorption of a neutron, which can approach and enter the nucleus easily because it is not charged, and will not be repelled as a proton would be. A few nuclei will fission if they absorb a neutron of any energy; these nuclei are called fissile. The only natural fissile nucleus is U235. When caused to fission by absorbing a slow neutron, it breaks into a light fragment with mass number 80-110, and a heavy fragment with mass number 125-155. The most probable masses are 95 and 139, respectively, and the full range is 72-160. These fission fragments possess about 80% of the 200 MeV released as kinetic energy, which they quickly distribute as heat. About 10% of the energy appears as gamma radiation, kinetic energy of emitted neutrons, and β-ray energy. The final 10% is carried away by neutrinos to an infinite distance.

The fission fragments are excessively neutron-rich for their masses, and release several neutrons immediately, on the average, and a few more at a slightly later time. These neutrons are very important, because they can be used to cause further fissions. For a sustained chain reaction, only one neutron need survive to cause a fission from all those produced. Neutrons may leave the mass of uranium, or be absorbed in some non-fission process. The neutrons are emitted at high speeds as fast neutrons. They lose energy by elastic collisions with light nuclei, such as hydrogen, deuterium, lithium, or carbon, and are slowed down to become slow, or thermal neutrons if they escape absorption. Thermal neutrons are much more likely to cause fission in a fissile nucleus provided they are absorbed in the fissile nucleus. They are also much more likely to be absorbed without profit. The material used to slow down the neutrons is called the moderator. It must have a small nuclear mass, and a small cross-section (probability) for neutron absorption, which are very restrictive conditions. The best practical moderators are heavy water (deuterium oxide) and graphite (carbon). The absorption of fast neutrons is rather improbable, except for a critical velocity range for "resonance absorption" that most neutrons escape if they pass through it rapidly.

An assembly of uranium and moderator is said to be critical if just one neutron remains to cause fission from those released in any fission. Natural uranium is 99.3% U238 and only 0.7% U235, so it is a very delicate matter to achieve criticality. It can only be done with the uranium fuel in lumps, with a moderator of heavy water or graphite. A reactor assembly at Berlin using heavy water failed to achieve criticality in 1939. A group headed by Enrico Fermi succeeded in Chicago on 2 December 1942. This reactor consisted of 385 tons of graphite blocks, with 40 tons of natural uranium oxide in spherical lumps, carefully piled up. Cadmium has a very high cross section for neutron absorption, so cadmium rods were used for control. One rod could be released if an ionization chamber detected large amounts of radiation, another was tied with a cord that could be cut if necessary by an assistant on a balcony, and a third was intended for control by being moved in and out to change the reactivity. The "pile" went critical without incident, and operated at a power level of 2 kW, cooled by the natural convection of air. Control of the reactor was facilitated by the existence of the "delayed" neutrons that were essential for criticality, and by the time required to moderate the neutrons to thermal speeds.

It is impossible to construct a nuclear explosive with natural uranium, since a nuclear explosive must be "fast critical" and this requires more neutrons than can be produced by 0.7% of fissile material. The uranium must be "enriched" by increasing the proportion of U235, a very difficult matter. The method used was gaseous diffusion, relying on the more rapid diffusion of the lighter U235 relative to U238. The difference is tiny, so the most heroic measures are necessary, involving thousands of diffusion stages to get an even modest enrichment. Nevertheless, large quantities of enriched uranium were produced in case it was the only way to get fissile material.

One of the things that can happen to a neutron slowing down in a reactor is that it can be absorbed by U238. If the neutron is a fast one, it may cause fission in U238. Nuclei that will fission when they absorb a fast neutron are called fissionable. Not only U238 is fissionable, but also Th232, as determined by Nishina in 1939. Fast fission contributes to the reactivity of a thermal reactor, but only by a small amount. It is more likely that what is now U239 after eating the neutron will become Np239 by beta decay (T = 23.5 min), which then becomes Pu239 by a second beta decay (T = 2.33 days). Pu239 has a half-life of 24,300 years, so it will stay around for a while. By a stroke of luck, it happens that Pu239 is fissile, like U235!

When a reactor runs, fission fragments build up in the fuel elements. Some of these fragments have large neutron absorption cross-sections, and so are called reactor poisons, since they lower the reactivity. Very soon, the reactor would become subcritical and the fires would go out, unless the fuel elements were withdrawn and replaced by new ones. When this happens, only a few percent at most of the fuel has been consumed. This "spent fuel" is not by any means "nuclear waste," as so commonly described. The fission fragments must be cleaned out, and the fuel made into new fuel elements for further burnup. The first thing is to let the "hot" fuel elements cool down for 100 days or so, usually under water, until they can be handled. They are, however, still extremely radioactive and must be carefully managed.

The fuel elements are chopped up and dissolved in nitric acid. Then the solution is subjected to a countercurrent organic solvent extraction with n-tributyl phosphate diluted with kerosene. This solvent absorbs the +6 uranium and +4 plutonium components, leaving all the rest, fission fragments and all, in the nitric acid solution. The +4 plutonium is then reduced to +3 with ferrous sulphamate (the sulphamate removes the nitrite that was used to make sure the plutonium was in the +4 oxidation state) and back-extracted with water. This recovers all the plutonium as an aqueous solution. The uranium is left in the solvent and extracted with dilute nitric acid into an aqueous solution. The solvent is purified and recycled. Now we have three vats. One contains the highly radioactive fission fragments, another the uranium, and the third the plutonium. These two vats are hardly radioactive at all. The fission fragments are disposed of (easier said than done), the uranium precipitated and made into new fuel elements when combined with a suitable amount of fissile material, which can be some of the Pu239 also recovered. A thermal reactor burning U235 can often make as much Pu239 as it burns, becoming a pseudo-breeder. However, a fast reactor is usually necessary if an excess of Pu239 is required. A reactor that burns Pu239 and makes more than it burns is called a true breeder. The U238 that is turned into Pu239 is called a fertile species. Only one process, the "Purex" process, has been described here, but others are similar and have the same end results.

After the CP-1 reactor proved the principles valid, it was decided to build large, water-cooled graphite-moderated reactors at Hanford, Washington to be used as pseudo-breeders to make plutonium. This was much, much easier than isotope separation to obtain fissile material, since a far smaller degree of enrichment could be used in the fuel, and the product was, in effect, 100% enriched. A test reactor was built at Oak Ridge, Tennessee that went critical in November 1943. This reactor was forced-air-cooled, operating initially at 1 MW, later at 4 MW, and used to test the materials for the Hanford reactors, which went into operation in September 1944. The story of the production of the atomic bomb is a fascinating one, for which the reader is referred to the references.

There is also insufficient space here to give a proper account of the use of reactors for power production. The reactor serves only to generate the heat to make steam (or boil mercury); once this is done, the power plant is exactly like one burning coal. Much of the world concentrated on reactors using natural uranium, like the water-cooled, heavy-water moderated Canadian reactors, or the CO2-cooled, graphite moderated British reactors. The United States was crawling with enriched uranium, so American reactors could use light-water moderators and other luxuries, with plenty of excess reactivity. The first reactor to enter commercial service was the pressurized-water, highly-enriched uranium reactor at Shippingport, PA, on 2 December 1957. It resembled the smaller reactors used in nuclear submarines that were developed at Arco, Idaho. From 1954 until 1971, enriched uranium was leased to power plants, apparently so that it could be returned to extract the plutonium for weapons. Without this considerable subsidy, nuclear power plants using enriched uranium would be economically impossible. By 1971, the need to enlarge the huge plutonium stockpile seemed unnecessary, and public resistance to nuclear power had rendered its future dim. Many efforts, like the Advanced Gas Cooled Reactor at Ft. St. Vrain, Colorado, were utter failures because of engineering deficiencies, though the utilities originally envisioned money flowing like water from the government. Since 1971, uranium has been available for purchase by utilities. It is not sold at anything near the cost of its production.

One very significant change is that the processing of spent fuel has ceased in the United States, and is not encouraged anywhere in the world. Instead of the horribly radioactive waste tip of reprocessing, we now have the horribly radioactive pile of spent fuel under water, which, as we have pointed out, is certainly not waste, just fuel of which 2% has so far been used. The danger, more than that of the radioactive waste, seems to be the danger of using the plutonium that results from reprocessing for unauthorized nuclear explosives. Without reprocessing, the limited reserves of fissile material will soon be used up, and then nuclear power is gone for good. Reactors themselves are safe, clean and nonpolluting. They compete mainly with coal as a fuel, and several centuries of cheap coal are still available, even at the current rates of consumption. They do nothing to reduce dependence on cheap petroleum, however, and that will be exausted much sooner. The greatest problem is the huge amount of radioactive fission products that will accumulate, and proper management of this resource seems well beyond the capacity of modern politics and politicians. Even the casual reader of this article now knows more about nuclear power than the whole lot of them.

Health Physics

"Health Physics" was the wholesome and cheery name given to everything concerned with the protection of workers in the nuclear industry. Exposure to radiation can be chronic or acute, and the effects are different. It is first necessary to be able to estimate radiation "doses" numerically, and this is not an exact science, since there are so many variables. So far, we have had only a measure of the number of disintegrations per second, the curie, or 3.7 x 1010 dis/sec. The rutherford, 106 dis/sec, is also used. But these are not really what we want. The biological effects of radiation depend on its ionizing power, since this represents molecular disruption, which is the root effect. Alpha and beta particles are charged, and when moving rapidly through matter, dislodge electrons. Alpha particles may knock other light nuclei into motion in collisions, and these knock-ons may also ionized. In air, every ion pair, electron plus ion, requires about 34 eV of energy. Alpha rays have ranges of a few centimeters in air, beta rays of a few metres. A sheet of paper stops both. Gamma rays may eject electrons from atoms, disappearing in the process, or may collide with electrons, knocking them on while being deflected (Compton effect), or if they are more energetic than 1.05 MeV, may produce an electron-positron pair. The fast electrons then cause ionization. The roentgen is the amount of X or gamma radiation that, in air, produces 1 esu of charge of either sign per cubic centimeter. (The electronic charge is 4.80 x 10-10 esu.) This is the basis of dose measurement.

For the effect of radiation in soft tissue, the rad, which is the amount of radiation depositing 100 erg per gram of tissue, was defined. It happens that this is just about equal to 1 roentgen. The relative biological efficiency, RBE, of a radiation is the ratio of the effect that a dose of X-rays of 200 kV energy would have divided by the dose of a particular form of radiation in rads. This expresses the fact that the same deposit of energy will do different amounts of damage in different tissues. Finally, the dose in rem (roentgen-equivalent-man) is equal to the dose in rads times the RBE. This procedure at least allows an estimate of the biological effects of various forms of radiation. Doses in rem are additive. Neutrons and gamma rays are the radiations of major concern in health physics. Fast neutrons have an RBE of 10, thermal neutrons 2.5, and gamma rays 1.0. Alpha particles, with an RBE of 10, would be very dangerous, but they are easily stopped, and it is almost impossible to get an alpha-ray dose unless the radioactive substance is inside the body.

An acute whole-body dose of 25 rem and under has no clinical effect. This once led to the belief that small doses were not injurious, but this is not true for chronic exposure. Greater doses may produce blood changes, but not until the dose exceeds 100 rem will there be a chance of vomiting and fatigue, with recovery in a few weeks. Over 300 rem the effects are serious, with hemorrhage, infection and hair loss, and recovery taking up to a year. Above 600 rem, the dose is usually fatal within a month or so, but a few may survive with serious effects.

Chronic low-level doses do not produce these dramatic responses, but may result in cancer, genetic damage, and similar effects. For people occupationally exposed to radiation, the chronic dose should not exceed 3 rem in 13 weeks, or 12 rem per year. The total dose should not exceed 5(N - 18) rem, where N is the age of the person in years. These were the levels accepted in the 1960's--they may well be lower now. Note how very much smaller they are than the acute doses with clinical effects. For people not occupationally exposed, the guideline is 0.5 rem/year, and no more than 1.5 rem up to age 30. These levels are comparable to the levels of natural radiation, from cosmic rays, uranium in igneous rocks, and other sources. The cosmic radiation dose is greatest at high altitudes, and decreases rapidly approaching the surface. At high altitudes, it seems to be less than 25 mr (approximately rem) per day, so a maximum yearly dose would be about 9 rem. At the surface, the dose would be much smaller, perhaps close to the recommended 0.5 rem/year.

Radiation can be detected in very small amounts. The dosimeter carried by workers in the field is a pencil-like capacitor that is charged to a high voltage. If you look in the end, an electroscope can be seen showing that the central wire is charged. A rather small amount of ionizing radiation passing through the device will produce enough charge to discharge the capacitor, and the amount of discharge can be seen by observing the calibrated electroscope. There are, of course, more sophisticated devices that are charged and measured by a machine. The Geiger-Müller tube responds to individual events, usually gamma rays, though beta rays can enter if a thin window is provided. It is good for low levels of radiation. Any dangerous level of radiation will completely block a G-M counter, so if you hear it ticking, you are quite safe.

Because radiation is silent, and can be detected in minuscule amounts, people who have trouble in estimating risks are excessively frightened by it, causing uranophobia and other mental illnesses. It is extremely difficult to receive a dangerous acute dose as long as you stay away from unshielded high-level sources, such as are used for metallography or sterilization. The irradiation of food leaves no radiation in the food whatsoever, and makes fewer chemical changes than ordinary exposure to the environment (but often decreases palatability), while eliminating some actual, dangerous risks from pathogens. Exposure to dental X-rays, for example, carries little risk to the patient. A lead apron is used to drape the patient, when it is actually the operator that is in most need of protection. This is almost completely for patient reassurance, not need. In the 1940's, fluoroscopes were used in shoe stores to show the bones of the foot in footwear. To get a bright image, rather strong X-rays were necessary, although the part of the body exposed was rather insensitive to radiation, and stray X-rays were probably well blocked by shielding. Of course, an end has been put to this entertainment, but one wonders if there was any real hazard. Whatever the answer, a salesman exposed many times daily was more at risk than any customer who visited once a year.

While on the subject of risk assessment, it might be good to mention the threat of a radioactive "dirty bomb" here. As a weapon, this is probably one of the worst ideas to appear in a long time. It would be even less productive than chemical agents, which are more effective, but still prove to be very poor weapons. The radioactive material would have to be collected inside the device (a nuclear weapon makes its own just when needed), and to produce sufficient unpleasantness the amount would probably kill whoever transported the weapon to its place of use. Proper shielding would be very heavy and cumbersome, rather difficult to conceal. Whatever is in the device would then be distributed by an explosion unevenly over a rather small area, which could be quickly evacuated before more than a small acute dose was received by anyone. A dose of 50 rem would require a very large amount of any kind of radioactive material that is available. Anyone messing with fission fragment waste would become very dead very rapidly. The contamination could easily be detected, and warnings posted. Dirty bombs are of the same order of effectiveness as the Italian terrorists who thought they would poison the water supply with potassium ferrocyanide. It sounds like cyanide, but it is not poisonous, and wouldn't poison anyone. Give the emergency personnel Geiger counters, and forget about dirty bombs.


B. Jaffe, Crucibles: The Story of Chemistry, 4th ed. (New York: Dover, 1976).

The website of the World Nuclear Association is a treasure chest of information on uranium and nuclear power, covering all aspects.

Image of carnotite kindly furnished by Amethyst Galleries, Inc.. This is an excellent website and specimens can be purchased online. This is probably the best mineral website, and the company should be supported for making it available.

R. M. Kirkham, R. W. Blair and W. R. Junge, in Geology of the Paradox Basin (Denver: Rocky Mountain Association of Geologists 1981 Field Conference), pp. 227-232.

A. Holmes, Principles of Physical Geology, 2nd ed. (New York: Ronald Press, 1965). Chapter XIII. Radioactive dating.

R. Glazebrook, A Dictionary of Applied Physics (London: Macmillan, 1923), Vol. IV, pp 194-196; preparation of luminous radium paint.

Return to Physics Index

Composed by J. B. Calvert
Created 20 January 2003
Last revised 9 November 2003