Rare Earths


  1. Introduction
  2. Auer von Welsbach
  3. Mineralogy of the Rare Earths
  4. Uses of the Rare Earths
  5. Scandium
  6. Thorium
  7. References


The metals known as the "rare earths" comprise three members of Group IIIB of the Periodic Table, scandium (Sc, 21), yttrium (Y, 39), and lanthanum (La, 57), and the 14 lanthanides that are filling the 4f electron shell: cerium (Ce, 58), praseodymium (Pr, 59), neodymium (Nd, 60), promethium (Pm, 61), samarium (Sm, 62), europium (Eu, 63), gadolinium (Gd, 64), terbium (Tb, 65), dysprosium (Dy, 66), homium (Ho, 67), erbium (Er, 68), thulium (Tm, 69), ytterbium (Yb, 70) and lutetium (Lu, 71). For physical properties, electron configurations and such details, the reader is referred to the Periodic Table websites to which links are given on the Physics index page. In the sequel, I shall refer to the elements by their chemical symbols, saving a lot of typing. Also included in this page will be some information on thorium (Th, 90), which is not a rare earth, but is associated with them in several respects.

The rare earths are not, in fact, very rare. Ce is the 25th most abundant element in the earth's crust at 60 ppm, ahead of tin and lead, and only a little less abundant than zinc. Pm, however has a natural abundance of zero, because its longest-lived isotope, Pm147, has a half-life of only 2.6 days, so it has to be made freshly by nuclear reactions when any is desired. Nevertheless, it has been isolated as a metal. It was first prepared in 1963 at Oak Ridge. After Bohuslar Branner (1855-1935) predicted an element 61 in 1902, which was confirmed missing by H. B. G. Moseley in 1913-1914, Professor Luigi Rolla (1881-1960) and his student Fernandes reported its discovery in 1924 and named it Florentium, symbol Fr. Professor B. Smith Hopkins (1873-1952) of the University of Illinois, with his students Yntema and Harris, claimed its discovery in 1926 and named it illinium, Il, to fill the last hole in the periodic table. The discovery was not confirmed. There was an "Illinium Dinner" at the university that year. Of course, none of these elements existed in Florence, Illinois or anywhere else at the time, and American chemistry lost a claim to fame. Arguments over priority of discovery of elements are a disgusting reality, and the rare earths had their share.

The rare earths can be divided into two groups, the first containing Y, La, Ce, Pr, Nd and Sm (Yttrium group), and the second Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb and Lu (Dysprosium group), with Sc standing off at the side. All the elements are now available as pure metals. Before the 1950's, few if any of the dysprosium group had been actually isolated as metals, and only Ce was at all common, usually as an alloy with La, Nd and Pr called mischmetall. This word will be found spelled as "mischmetal" in English, and is a common item of commerce with various uses. Prices of the yttrium group currently range from $350 to $540 per kg, of the dysprosium group from $400 to $7500, while scandium is a breathtaking $18,000 per kg. For comparison, gold now is worth $12,110 per kg, so we see that any rare earth is cheaper than gold, except for scandium. Silver is worth $158 per kg, nickel $8.38 and chromium $7.50. The rare earths are available at quite reasonable prices, considering the difficulties of their processing and purification, and are slowly acquiring important industrial uses.

All the rare earths are typical transition metals. They are silver-white on a fresh surface, tarnish rapidly, are soft and usually ductile, and burn when heated. Ce can be cut with a knife, and can take fire if scratched. There are, of course, differences between them, but this is the general rule. None are of any use as structural metals. They are employed mainly in alloys and as compounds. Their chemical behavior is so similar that they are exceedingly difficult to separate. The sulphates of the yttrium group are soluble in water, while those of the dysprosium group are insoluble. The hydroxides of Sc, Y and La are the most basic, and the basicity decreases from Ce to Lu, as does the ionic radius. From Ce to Lu, the 4f electron shell is being filled. The 4f levels have about the same energy as the 6s levels, which are filled in the ground state. The often-heard explanation for the similarity of the rare earths based on the 4f levels being "hidden" is invalid. The ionic radius decreases from 0.106 nm for La+++ to 0.085 nm for Lu+++, and is responsible for the slight but steady variation of properties along the series. Gd, with the electron configuration Xe4f75d6s2, has eight electrons with parallel spins, and so is highly paramagnetic. It is used in cryogenics for adabatic cooling. The valence of all the rare earths is most commonly +3. Ce, Pr and Tb also exhibit +4, while Sm, Eu, Tm and Yb show +2, in addition. Gd and Lu put one electron in a 5d level, encouraged by the stability of 4f7 and 4f14 configurations, respectively.

Auer von Welsbach

The name of Carl Auer, Freiherr von Welsbach (1858-1921), usually called Auer von Welsbach, is closely associated with the rare earths. He was born in Vienna, educated at the Technische Universität there, and later at Heidelberg, where he was a student of Robert Bunsen's, and received his PhD. At the age of 27, in 1885, he resolved the rare earth substance then known as didymium into two elements by laborious fractional crystallization of the ammonim nitrates. The part that gave greenish salts he named praseodymium (praseos is the leek in Greek, a green thing), while the part that gave pink salts he called neodymium. Neodymium is about twice as abundant as praseodymium. He was ennobled by Kaiser Franz Josef for his accomplishments.

Ce had been discovered by J. J. Berzelius and Hisinger in 1804, three years after the discovery of the first asteroid, Ceres, in 1801. Carl Gustav Mosander (1797-1858), Berzelius' assistant at Uppsala, prepared pure Ce in 1827, and in 1839 separated La from it. Then he separated a substance with a rose-colored oxide from La and called it didymium, from the Greek "didymos" for twin, as it was a twin to La. He discovered Y, Tb and Er in 1843. Four rare earths bear names derived from the village of Ytterby, near Stockholm, where the minerals containing them were found. These are yttrium, ytterbium, terbium and erbium. For a long time, the rare earths were a Swedish monopoly. By 1875, Ce, La and didymium were being made by the electrolysis of molten halides. In 1879, Paul Émile Lecoq de Boisbaudran (1838-1912) found Sm, and Dy in 1886. He was a spectroscopist who labored for 15 years to find gallium in 1875. Per Theodore Cleve (1840-1905), Svante Arrhenius' research advisor, isolated the much rarer and harder-to-separate Ho and Tm in 1879. Ho was christened after the Latin name of Stockholm, Tm after Thule, the "ultimate." Tm did not remain the ultimate for long. The periodic table, as then known (it appeared in 1869), gave little hint as to how many rare earths were to be expected. It was not clear that there were 14 lanthanides until the electronic basis of the periodic table was elucidated early in the 20th century. P. T. Cleve is also known for his work in diatoms, including the classification of pseudonitzschia delicatissima, which secretes the powerful toxin domoic acid that causes the rare amnesiac shellfish poisoning (ASP).

Shortly after Auer von Welsbach's separation of Pr and Nd, Lecoq de Boisboudran announced Dy, and Marignac Gd, both relatively common members of the dysprosium group, in 1886. All these new elements came from samples of the previously-known rare earths, as separation methods were discovered. Eu was found by Demarcay in 1900, and purified by G. Urbain and Lacombe in 1904. Its salts are pink, like those of Nd. Most salts of the rare earths are colorless. Auer von Welsbach had meanwhile been working in spectroscopy, and suspected that the substance isolated earlier by Marignac called ytterbium contained two elements. He succeeded in separating the elements in 1905, calling them aldebaranium and cassiopeium. Unfortunately, he was a few weeks later than G. C. Urbain, who had also separated the elements, and because of his priority was allowed to name them ytterbium and lutetium. These were the last two rare earths to be discovered, although by 1920 the periodic table showed that there was still one missing, a hole that was soon, but erroneously, plugged in Illinois.

Auer von Welsbach did not work only in the rare earths, though they were a favorite of his. His most widely known invention, the incandescent mantle, was patented in 1885, and called the "Auerlicht." In German, the incandescent mantle is known charmingly as the "Glühstrumpf" or "glow-stocking." A knitted sleeve was soaked in a solution of nitrate salts and dried. To use it, it was first mounted on the lamp burner above the frame, and then lighted with a match. The organic parts burned, leaving fragile network of oxides in the path of the flame. When the lamp flame was lighted, and adjusted to be blue and nonluminous (the Argand burner was ideal for this) the mantle was heated to incandescence, giving a bright white light. The original mantles used Mg, La and Y. Pr, whose oxide is light brown, was used for the label printed on the mantle, which could be seen when the mantle was in use. They were very fragile, and so had a short life, and gave a greenish light. A factory was established in Atzgersdorf in 1887 to make mantles, but it soon closed because of the deficiencies of the product. Auer von Welsbach worked assiduously to improve the mantle, and in 1890 discovered that thorium was better then magnesium, and in 1891 found the combination of 99% Th and 1% Ce that gave a long-lasting mantle with a brilliant white light. The factory reopened, and a new, larger factory was established in Althofen, in Carinthia (Kärnten), where he made his home. The Wesbach Mantle soon spread world-wide, and revolutionized gas lighting. Illuminating gas no longer had to be carburetted to make luminous flames, and mantles worked excellently even with natural gas. Welsbach mantles are still used on camp lanterns and similar illuminating devices, though Y has replaced Th because of ignorant radiophobia.

The Welsbach and General Gas Mantle Company made mantles in Camden, NJ from the 1890's until 1941. More than 50 years later, this factory, then converted to other uses, was identified as a radiation hazard by the EPA. It is not possible to determine the facts of the case from published EPA documents, since they never give any survey figures or even say what their standards are, but it is very probable that this was an extremely minor hazard, giving gamma radiation levels scarcely higher than the cosmic-ray background. It is reasonable that some thorium contamination was left in the area, but thorium is extremely feebly radioactive, and thoron contamination would have been equally negligible, easily counteracted by ventilation. The fear of traces of radioactivity has also driven thorium out of a reasonable use as an alloying agent for magnesium in aerospace alloys, where it was an equally small hazard. Fortunately, Y substitutes for Th in mantles, and a mixture of Y, Zr and other lanthanides is a substitute in Mg alloys. However, I have not seen any credible evidence that Th has ever been a health hazard because of its feeble radioactivity. There are far greater hazards in other places, and ignorant campaigns like this interfere with public protection rather than benefiting it by absorbing valuable resources.

Gas mantles give a light in every way competitive with electric light, and even cheaper. However, the bother of dealing with gas and mantles has made gas lighting disappear even in street lighting, where mantles lasted until late in the 20th century. Of course, the competition of electricity was realized from the first, since Edison's carbon-filament lamps were already widespread when the mantle was invented. Auer von Welsbach developed metal-filament lamps, patenting the Auer-Oslicht in 1902, which had an osmium filament made with powder metallurgy, in which Auer von Welsbach was a pioneer. This was a very good lamp, more economical than the Edison lamp, and a good deal brighter. It led in a few years to the replacement of the carbon filament with metallic filaments, of osmium, tantalum and finally tungsten. The "os" syllable remains in commercial names such as "Osram."

The observation that Ce easily caught fire, and burning sparks were easily made by striking it with a knife, led Auer von Welsbach to investigate "funkengebenden metalle" as an easy way to ignite a gas flame. In 1903 he patented an alloy of 70% Ce and 30% Fe which gave copious sparks when scratched by steel. In 1907, he founded a factory at Treibach for the manufacture of this Cer-Eisen to be used in lighters. The small pieces of alloy are called "flints" in English, "Zündsteine" in German, and "piedras de encendedor" in Spanish, and are still widely used. A small cylinder of the alloy is held by spring pressure against a serrated steel wheel. When the wheel is rotated by the thumb, a shower of sparks is produced. Pure Ce is not required. The cheaper alloy called "mischmetall" is just as good. This alloy is often incorrectly but commonly called "pyrophoric," a word that should be used for substances that take fire spontaneously, but there appears to be no existing term for funkengebende substances.

China is today the largest supplier of mischmetall. Two typical alloys are Ce 51, La 26, Nd 17, Pr 6 and Ce 53, Yb 24, Nd 16, Pr 5, Gd 2. The alloys do not have to be specially made. They result from the electrolysis of whatever combination happens to be in the natural ore, saving the cost of purification, which is totally unnecessary. When only the general properties of the rare earths is required, mischmetall is an economical alternative.

Mineralogy of the Rare Earths

Minerals containing rare earths are not especially rare, and the recent demand for rare earths has led to the discovery of new ores. There is little danger of a rare earth shortage at the present. The combination of potassium, rare earths, and phosphorus seems to be the last fraction that crystallizes from a magma, and is called KREEP by geologists. All the rare earths occur together in varying amounts, since nature is no better than chemists in separating them.

The principal ore of the rare earths has long been monazite, a dark, hard, heavy mineral that is found mainly in placer deposits, though vein deposits are known. It is found in commercial amounts in the U.S., Madagascar, Brazil, India, Sri Lanka and Australia. Monazite is principally CePO4, cerous phosphate, but almost always contains a considerable amount of Th, together with La and other rare earths. Most Th, in fact, has been obtained as a by-product of the extraction of rare earths from monazite. Xenotime, YPO4, found near Ytterby and the historical source of many rare earths, is similar. Euxenite is a mixed oxide of Y, Ce, Er and Nb, while Allanite is an aluminium-iron silicate with Y, Ce and Ca. Of these, only monazite is a commercial ore.

The improved techniques for production of rare earths that were developed after 1950, and the increasing demand, led to the discovery of a large resource of Bastnäsite at Mountain Pass, California. Bastnäsite is mainly Ce2(CO3)3 or CeF3 (?), a mixed fluoride-carbonate. The mineral is long known from Sweden, where its name comes from the village of Bastnäs. This deposit has made the U.S. one of the principal producers of rare earths, next to China and Russia.

Rare earths are now separated by ion exchange resins, which is much less tedious than the earlier fractional crystallization methods.

Uses of Rare Earths

Recently, the largest use of rare earths has been in glass polishing and ceramics, 39%, followed closely by automotive catalytic converters, 22%. 16% is used in permanent magnets, 12% for petroleum refining catalysts, 9% metallurgical uses, 1% in cathode-ray tube phosphors, and 1% in miscellaneous uses, including lighter flints. As an example of the metallurgical uses, mischmetall was used in the steel alloy for the Alaska crude oil pipline. The cathode ray tube phosphor is Eu-doped YVO4 (yttrium vanadate), giving the red color. Other phosphors also contain rare earths.

Didymium, the mixture of Pr and Nd that Auer von Welsbach separated in 1885, is used in glass for welders' and glass-workers' goggles. The violet-tinted glass is 80% transparent in the visible, but cuts off abruptly at about 340 nm for UV protection, and, remarkably, over a short range near 590 nm that blocks the Na glare. Neodymium-yttrium-garnet (NdYAG) and gadolinium-yttrium garnet (GdYAG) are laser materials, alternatives to ruby.

Rare earths are used in the permanent magnet alloys samarium-cobalt and neodymium-iron-boron. Sm-Co appeared in the late 1960's, Nd-Fe-B in 1983. Sm-Co magnets have the composition SmCo5 or Sm2Co17, and Nd-Fe-B magnets are Nd2Fe14B. The 2-17 alloy gives better temperature characteristics than the 1-5. These are very hard and difficult materials that must be shaped by sintering or compression bonded with a bonding agent, and worked with diamond tools. A typical Sm-Co material has a remanent flux of 10,700 gauss and a coercive force of 10,300 oersted, giving a BH product of 28 MG-oe. A Grade 30 La-Fe-B powder has a remanent flux of 11,400 gauss and a coercive force of 10,400 oersted, for BH = 30 MG-oe. There are many different materials, but this shows typical high performance. 1-5 Sm-Co has a Curie temperature of 750°C, and 2-17 825°C, but the maximum service temperature is about 350°C. La-Fe-B has a lower Curie temperature, 310°C, and a maximum service temperature of 150°C. The corrosion resistance of La-Fe-B is low, but that of Sm-Co is high. The mechanical strength of brittle Sm-Co is low, while that of La-Fe-B is moderate. The density of La-Fe-B is about 7.5 g/cc, and that of Sm-Co is 8.3 g/cc. Sm-Co is expensive because of the cobalt. These materials require a very high magnetizing field, which is usually applied after assembly. These crystals are actually antiferromagnetic, with the lanthanide and cobalt or iron moments opposite. However, they have very high magnetic anisotropy, which makes them good permanent magnets, with a very high internal coercive force.

The search for a light, cheap rechargable battery with high energy density continues, but the lead-acid cell is hard to beat. A recent contender has been the Ni-Cd rechargeable battery, and more recently the Ni-metal hydride (Ni/MH) battery, which avoids the environmental restrictions on cadmium. These batteries have used the expensive cobalt, however, and have a limited life due to expansion and contraction of the plates during the charge-discharge cycle. An alloy has been developed at Brookhaven National Laboratories that uses La, Ni and Sn, so that it is free of expensive Co and toxic Cd. It was found that putting slightly more Ni/Sn in than the usual 1 La to 5 Ni/Sn prevented the stressing of the plates on charge and discharge, making a much more durable battery.

Ce2O3 is the catalyst used in self-cleaning ovens, and possibly also in automotive catalytic converters that break down unburned hydrocarbons in engine exhaust.


Scandium, a light metal between calcium and titanium in the first long period of the periodic table, acts much more like a rare earth than either of these, or of boron, which it might also resemble. It was discovered by Lars Fredrik Nilson in 1876 (or 1878, or 1879, according to different sources) in scandinavian euxinite, and was recognized by Cleve as the "eka-boron" predicted by Mendeleev in his periodic table of a few years before. It is light, like aluminium, but has a much higher melting point, which suggests aerospace uses. However, its cost would seem to put it out of the competition, though it would be very similar to titanium in this application. Scandium reacts readily with water to displace H2, which would seem to eliminate it as a structural material. Scandium salts are colorless, and it is not paramagnetic.

The USSR was the main producer of scandium, but ceased export of it in 1968, after which its price went sky-high, to some $75,000 per kg until other sources were found. It seems to have been important in some laser application, but just what is not evident, and information is lacking. Scandium is used in aluminium alloys for baseball bats, bicycle frames and handguns. It is also used in high-intensity metal halide lamps, as the oxide scandia, Sc2O3. ScI3 is added to mercury vapor lamps to make the light more like sunlight. Sc46, a beta-emitter with a half-life of 85 days, is used as a tracer in oil refineries. Natural scandium is 100% Sc45, with a rather large cross-section for thermal neutron absorption. It is expected that scandium may be used in fuel cells.

Scandium is believed to be the trace ion that gives the blue color to beryl in the gem aquamarine. It is relatively much more common in the Sun than on earth, where it is 50th in order of abundance.


Thorium, atomic number 90, was first isolated by J. J. Berzelius in 1828 or 1829. The heavy (d = 11.72 g/cc), shiny white metal soon tarnishes in air, gradually becoming black from a layer of oxide. The pure oxide, thoria, ThO2, however, forms clear cubic crystals with the high index of refraction of 2.2 and specifc gravity 10.03. It is usually encountered as a white powder. The mineral thorianite, (Th,U)O2, is black, with hardness Mohs 6.5 and specific gravity 8.0-9.7. The metal is face-centered cubic, soft and ductile. Thoria melts at 3300°C, nearly as high as tungsten does, which makes it an excellent refractory, used for laboratory crucibles. Thoria, up to 30%, also makes optical glass of a high index of refraction. Thorium hardens magnesium without interfering with its passivity, giving a strong, light alloy with aerospace uses. We have already noted its use as the principal component of incandescent mantles, although Y is a replacement. Thoria also is a catalyst in the oxidation of NH3 to HNO3, and other reactions.

Powdered thorium is pyrophoric (in the proper sense of the word), like many powdered metals (Al, Mg for example). It is attacked only slowly by water, but dissolves readily in hydrochloric acid. Only hot sulphuric acid attacks thoria. Its hydroxide, Th(OH)4, makes a gelatinous precipitate, and is amphoteric, dissolving in alkalis. Although thorium occurs in minerals where it is the principal component, as in thorianite, ThO2, it is commonly associated with the rare earths, and has been obtained mainly as a by-product in their extraction, and has benefited from the improvements in rare-earth metallurgy. Thorium occupies the same place at the head of the actinide series in the periodic table that cerium occupies in the lanthanide series.

Thorium is so rare that its chemical toxicity is not an issue, and no specific biological hazards are known. It is feebly radioactive, and widely distributed in igneous rocks. Basalt shows a thorium activity of 270-405 pCi/kg, and granite 1890 pCi/kg. There is little thorium in sedimentary rocks. In crustal abundance, it is between Pb and Be, the 34th in abundance, but has not been concentrated as well as these elements into economic ores. The Nuclear Regulatory Commission exempts items containing less than 2g or 0.25% of thorium from control, as well as optical glass containing less than 30% Th by weight.

The only naturally-occurring isotope of Th is Th232, an alpha-emitter with a half-life of 1.39 x 1010 years. Tiny amounts of radioactive isotopes of Th occur in the decay chains of U238 and U235. By the absorption of a neutron, Th232 becomes the fissile isotope Th233, which can be used in nuclear reactors. Advantage has not been taken of this possibility, however. The end of the thorium radioactive series is Pb208, so Th232/Pb208 ratios can be used for geological dating. One member of the decay chain is the gas Rn220, also called thoron, which is a hazard similar to the radon that occurs in the uranium series. However, it is very rare, since it has the short half-life of 54.5s, compared to 3.825 days for radon, and can scarcely be a hazard under normal circumstances (for equal thorium and uranium activities, thoron will be 0.00016 times less concentrated than radon).

It's good to remember that all nuclides with Z > 83 are radioactive, and even Bi209, the only naturally-occurring isotope of bismuth, has a half-life of about 3 x 1017 years, so, strictly speaking, it is not stable either. Only the nuclides with half-lives of about 109 years or more have survived since the origin of the earth, and there are only three of these, Th232, U235 and U238. All the other elements in this region of the periodic table are either contained in the radioactive chains headed by these nuclides, or are artificially made. Of these, only Bi210 and three isotopes of uranium have long half-lives, which means that they will exist in reasonable amounts. The difficulty in isolating radium, with a half-life of 1620 years, is evidence of their rarity.

K40, a rare isotope of potassium that is radioactive, is present in the earth's crust at 0.0029%, while the abundance of Th232 is 0.00115%. Even allowing for the radioactive daughters of the thorium, which potassium does not have, and the 10-times longer half-life of the thorium, natural gamma radioactivity due to potassium is larger than that due to thorium. I hope the EPA does not take fright at this and try to remove potassium from the environment.

Thermionic cathodes used in vacuum tubes emit electrons when heated. A tungsten filament heated to 2500K (white heat) is a useful emitter, but not a copious one. It was found in 1904 that barium and strontium oxide surfaces were copious emitters when heated only to 1000K (red heat), but were easily destroyed by ion emission when used at higher plate voltages. In 1914, Langmuir and Rogers found that a good emitter could be made by adding 1-2% ThO2 to the tungsten. The filament had only to be heated to 1900K (orange heat) in this case. Before use, the filament had to be heated briefly to 2100K to activate it. The thorium is adsorbed as a monoatomic layer at the surface of the tungsten, reducing the work function from 4.52V to 2.63V, making it easier for the electrons to get out. As the thorium boils off, more diffuses from the bulk of the filament to replace it. These "thoriated tungsten" filaments are still used in thermionic emitters, and were widely used in transmitting tubes for plate voltages up to 5000V.

Thorium is also found as a coating for tungsten welding rods, and in electric arc carbon cores.


The Auer von Welsbach museum website is at Welsbach Museum.

R. Duda and L. Rejl, Minerals of the World (New York: Arch Cape Press, 1989).

Current information on the rare earths, including prices, production and consumption, is available from USGS. Search for "rare earths."

EPA Region 2 Superfund, ID#NJD986620995, Welsbach and General Gas Mantle Contamination.

Good thorium information is at Thorium Waste. This firm specializes in chemical and radioactive waste cleanup, including Se, Tc(!), Th, U and H3.

Didymium goggles can be seen (and bought) at Art Glass.

For a good discussion of Sm-Co and Nd-Fe-B permanent magnet materials, see S. R. Trout.

Return to Physics Index

Composed by J. B. Calvert
Created 9 February 2003
Last revised 11 February 2003