Rare Earths From Supernova to Superconductor

7/27/2019 Rare Earths From Supernova to Superconductor

1/12

Pure & A p p l . C h e m. ,Vol. 65, No. 12, p p . 2453-2464, 1993.Printed in Great Britain.@ 1993 IUPAC

Rare earth s from supernova t o s uperconduc torDavid H CornellProfessor of Geology and Applied Geology, University of Natal, Durban 4001South Africa

Abs t rac tThe origin, geological evolution, mineral deposition and uses of rare earthelements (REE) are discussed. It i s shown how their nuclear, geochemical andchem ical properties n fluencetheir behav iour n geolog ical and industrial processes.Analysis i s critical and the appropriate method has to be selected carefully. Ageochem ical and chemical insight s required to achieve the analytical aims w hic hoften require concentrations from ppm to 50% to be determined. Significantinterferences exist. Various methods are reviewed.

I N T R O D U C T I O NThis paper aims to set the stage for several papers on the analys is of ra re earth elements at this confe rence.As these m ethods are m any and the de tailed papers m uch more com prehensive, I shall deal mainly wi thrare earth geochemistry, extraction and technology, to explain why so many scientists and engineersneed to a nalyze materials for the rare earths (ref. 1).More than for any other group o f elements, the imprint o f cosmic and geological processes i s indeliblystamped on rare earth minera ls and concentrates. I shall discuss these na tural imprints in an attem pt toremind the chemist and engineer that all our mineral resources spring ultimately from Mother Earthwhich in turn i s a mino r ch ild o f the solar system and the galaxy.

R A R E E A R T H S IN T H E P E R IO D IC T A B L E A N D T H E I R B O N D I N G E L E C T R ON SThe Rare Earth elements belong predom inantly to the Lanthanide Group of the pe riodic table, havingatomic number 58 to 71 (Table 1). Rare Earths i s a misnomer as these elements are neither rare, norearths, b ut the chemists wh o first iso lated them as oxides named them thus to distinguish them fromalkaline earth elements like M g, Ca and Sr. The lighter element yttrium has such similar properties tothe lanthanides hat it s usually included as a rare earth element. The other G roup I l lB element scandiumhas sufficiently distinct chemical and geochemical character that it has escaped from its 1968 formalinclusion in the rare earth elements by the International U nio n of Pure and Applied Chem istry.The lanthanides are usually depicted be low the m ain periodic table as they are a transition group, al lhaving two electrons in the outermost 6s orbital. As the atomic number increases from lanthanum tolutetium the 4 f orbitals inside the e lectron orbital structure are filled (Table 2 ) . Losing three electronsgives each rare earth a stable, x enon-like outer electron structure, whi ch explains wh y they a l l haveremarkably similar chem ical properties and norm ally form trivalent ions. O nl y two rare earth elementscan form different ions in natu re: Ce4+ nd Eu2+. he rare earths are litho ph ile (rock-loving) because theirions prefer to enter io nic silicate or oxide mineral structures rather than covalent sulphide or metalliccompounds.

2453


2/12

2454

Table 1 Th e rare earth elements an d Group / / / Ao f the perio dic table

rud I [Xe]4f46sZ 1[Xe14P6s2 [Xel4P

,tomic No

[Xe]4f3[Xe14f6

213957585 9606162636465666768697071

Lu I [Xe]4145d16s2

jyinbol

[XeI4f l4

scYLaCePrN dPmSn iEuC dTbDYHoErTmYbLu

CommentsJnlike lanthanidesj imi la r to ho lmiumThe first lanthanideCan form Ce4+Light REEs 57-62

Short lived radio-isotope

Can form Eu2+Heavv REEs 64-71

Th e last lanthanide

D. H. CORNELL

Table 2 Electron configuration of some rare earth element atoms

1 Neutra l atom I 2 + c a t i o n 1 3 + c a t i o n I 4 + cat ion Iand ions

L a CE Pr Nd S r n E u Gd Tb Dy Ho Er Tm YbFIGURE 1 . Ionic radii of the lanthanide group of rare

earth elements, illustrating the lanthanidecontraction (data from ref. 3)

In the crystal lattice si tes of niinerals the rare earths exhibit the lanthanide contraction, depicted in Figure 1.This decrease in size of the trivalent rare earth ions with increasing atomic number i s explained byshielding, whereby the electrons defining the radius of the ion are attracted more strongly by theincreasing number of protons in th e nucleus. The lighter element yttrium behaves identically toholmium, which has the same ionic radius.

N U C L E O S Y N T H E S I S A N D T H E O D D O - H A R K I N S E FF E CTOur sun and other stars emit light which on spectral analysis reveals that they consist mainly of hydrogenand helium, with minor amounts of al l the other elements. Helium i s being produced by the nuclearfusion of hydrogen inside them, but stars are not able to make elements more massive than carbon.Supernovae mplosion-explosion events are the only likely producers of heavy elements such as the rareearths. Such events took place in our corner of the galaxy from fifteen to five billion years ago. Theyproduced huge fluxes of neutrons, which were captured by atomic nuclei and subsequently decayedto proton plus electron, thus increasing both the atomic number and mass number of the nuclei by whichthey had been captured. The stamp of this pre-solar system process i s seen in the rare earth elementanalyses of all terrestrial, lunar and meteorite samples. Rare earth concentrations of two terrestrialmaterials are shown in Figure 2 , plotted against atomic number of the fourteen elements lanthanum tolutetium. The even atomic-numbered elements such as cerium, neodymium, and samarium are an orderof magnitude more abundant than the adjacent odd-numbered elements. This Oddo-Harkins effectresults from the even-numbered elements with lower neutron-capture cross-sections being less likelyduring neutron-flux events to capture another neutron and be transformed into the next-higher element.

1


3/12

Rare ear ths f rom supernova to superconductor

A Concentration T Chon drite norrnalised

2455

FIGURE 2 .Rare earth analysis of internationalrock standard BCR-1 compared w ithan estimate of the bulk earth com-position according to ref. 4

0 . 0 1 1 , I I I I , I I I I I I 1 1La Ce Pr Nd (Pm Sm Eu Cd Tb Dy Ho Er Tm Yb Lu

fulk Earth0.01

La Ce Pr Nd (Prn) Sm Eu Cd Tb Dy Ho Er Tm Yb LuFIGURE 3. Bulk earth compo sition plotted against logarithmic (parts per m illio n) andchond rite norm alised scales. Div idi ng the sample concentration by that of cho ndritemeteorites (refs 5, 6) yields a smooth chondrite-normalised rare earth element prof ile

Note the logarithmic abundance scale and that the light rare earth elements cerium and lanthanumconstitute the b ulk by mass of the rare earth group, while the heavy elements gadolinium to lutetiumhave very lo w concen trations.The Oddo-H arkins effect makes it diffic ult to work w ith rare earth element profiles as plotted in Figure2, so data i s usually plotted on Masuda-Coryell diagrams normalised using the concentrations inchondritic meteorite, which closely represents the primordial solar system material (refs 4 to 6) . Thusas shown by Figure 3, pre-so lar system processes are cancelled out, a llow ing us to concentrate o n thosewhic h have operated since the earth formed some 4650 m illio n years ago.

B A S A L T G E N ES I S B Y M E L T I N G O F T HE M A N T L ETrivalent rare earth ions wi th radii about one angstrom, are the same size as divalent ca lcium ions. Sothe rare earth elements fit into the crystal lattices of calcium-bearing rock forming minerals such aspyroxe ne CaMgSiO, and plag ioclase CaAI,Si,O,. But because of their high er charge and crystallineenergy, if such minerals are in eq uilib rium w ith a silicate m elt the rare earths are concentrated in theme lt and are thus called incom patible elements. Whe n a portion of the earths m antle, whic h i s close


4/12

2456 D. H. CORNELL

i IMantle Melting Model

*....*--..................................~..... X......~.....*.....***..L.*" '.....*- 1riginal mantle

--o-- 1% partial melt----.----esidue after melt0.1'a Ce Pr Nd Sm Eu Cd Tb D y Ho Er Tm Yb Lu FIGURE 4.A simplified model (ref. 7) showingthe calculated rare earth profilesresulting from the m elting ofchondritic mantle to produce 1OObasaltic partial m elt l iqu id and99 % residue

in com position to chon drite m eteorites, melts to form a basaltic liquid, the light rare earths are morestrongly con centrated in the basaltic melt, as shown by a mode l in Figure4, (ref.7). The ma ntle peridotitewhic h remains solid becom es depleted in light rare earth elements, wh ile the basaltic melts producedfrom it show both overall and light rare earth enrichment.Basalt makes up the crus t beneath the oceans. The continental crust was derived from it by a secondstage of me lting in su bduction zones. The rare earth elemen t profiles of a l l igneous rocks reflect (a) thesignature of their source rocks and (b) the signature of their me lting system and p roportion of meltproduced.

THE CONTINENTAL CRUSTThe crust on w hic h we live and work i s ma inly granite, covered by sedimentary rocks derived from itsweathering and erosion. Ma ntle peridotite contains about 40% silica, basalt 50 % and granite 60% andgranite isalsoenriched in alkal i elementsattheexpenseofMg Feand Ca. Models basedo ntraceelementpartition coefficients show that to derive granitic rocks from the m antle, tw o or mo re stages of m eltingare required, producing first basalt and then granite. The rare earths play an important role inconstraining models for the genesis of the continental c rust. A typical granitic rare earth element profilei s shown i n Figure 5 . The features w hic h it shows are: overall 40 times enrichment relative to chondrites(controlled by the degree of melting), a factor five enrichment of light relative to heavy rare earths(contro lled by the mineralogy of source rocks), and a negative europ ium anomaly (reflecting plagioclasefractiona tion). Plagioclase has a strong affinity for the reduced orm of Eu2+, epleting it in iqu id magmasfrom whic h it crystallises. This feature wh ich i s useful to geologists, i s not popular with min ing housesas w il l be shown below.Soils are derived from rocks by weathering, and they have rare earth element profiles wh ich reflect theirprovenance, whi le Ce and Eu anom alies may develop due to redox reactions during paedogenesis. InFigure 6 the rare earth profiles for some French wines, analyzed by direct asp iration ICPMS, show largedifferences in rare earth profile related to differences in soil compo sition. H owe ver one should bear inm ind that some fertilizers co ntain high concen trations of rare earths from carbonatite apatite!Inductive ly cou pled mass spectrometric (ICPMS) and therm ionic isotope dilu tion mass spectrometric(IDMS) analysis.To do ge ochem ical research on rocks and soils one needs an analytical technique whic h can attainroutine detection limits close to 0.01 parts per mil lio n in solid samples, and determ ine m ost of thefourteen lanthanides. Inductively-coupled plasma mass spectrometry i s one suitable technique as it


5/12

Rare earths from supernova to superconduc to r

10

2457

Europium anomaly

I I I I I I I I I I I I I

1000Oribi Gorge Granite

Light

.00001

EnrichmentP

Pomrnard

+ PorncrolA St EmilionI I I I I I I I I I I 1 -

FIGURE 5 . Chondrite normalised rare earth element profile o f a typical granitesample from O rib i Gorge, southern Natal, showing light rare earth enrichmentand a negative europium anomaly, Data from Thomas and Corne ll in prep.- ' - ' 3 I

,0011.-6s3c%

,00011Sanccrre


6/12

2458

DY ii+ErD.H. CORNELL

VG PQ2 ICPMS Spectrum

S a m d c NIM-G

162 164 166 168 170Mass (amu)FIGURE 7 . ICP mass spectrum s how ing some o f the isotopes of the h eavy rareearth elements dysprosium, erbium and thulium, for the rock standard sampleN IM -G b y induc tively coup led plasma mass spectrometry. The only significantspectral overlap i s between the isobars 16 4D y nd 164ErN E W M E T H O D S OF D A T I N G R OC K S

N o less than bi x of the rare earth elements have one or more ra dioactive isotopes, but as their ha lf-livesrange from 1O ' O to 1Ol6years, i t has not been feasible to use their radioa ctive decay as geolog ical clocks,because they p rod uce such small am ounts of daughter isotopes in geological tim e spans betwee n 1O6and lo9years. However as more stable thermionic mass spectrometers were developed, three newdating methods have been proven: samarium-neod ymium, lutetium-ha fnium and lanthanum -cerium.Sm/Nd has been the most successful. The Sm/Nd techn ique has no w been wid ely used in dating someof the worlds old eit rocks, no tably in southern Africa the Ancient Gneiss Comp lex o f Swa ziland (ref. 10)at 341 7 -34 Ma, and the volcanic rocks of the Onverw acht Grou p in Barberton Mo un tain Land at 3540- 0M a (ref. 11 .This me thod also provides informa tion about the source from w hic h igneous rocks werederived wh ich can be used in models describing the structure and origin of the m antle and crust of theearth.

R A R E E A R T H OR E M I N E R A L SIn perid otite and basa lt, rare earth elements usually total less than 100 parts per mi lli on spread evenlythroughout most of the m ajor m inerals feldspar and pyroxene. They o ccupy a small proportion of thecrystal latt iceposit ionsofmajorelements i kecalc ium. In granit ic melts rareearthsencounter new m inorminerals not fou nd i n basalts, such as apatite Ca,(PO,),(OH), and zircon ZrSiO,, w h ic h have crystallattice positions m ore invitin g to rare earths than those of com mo n mine rals. As a result these traceminerals take up larger propo rtions of the available rare earth elements. In the trace-element ric h fluidswh ich remain after most of the magmatic l iqu id has solidified, rare earth minerals l ike m onazite mayform in coarse-grained pegmatites. This im portan t ore mineral, a cerium -tho rium phosphate, containsall the other rare earths, decreasing wit h increasing atomic num ber as show n in Figure 8. Monaziteoccurs in small amounts in most granitic rocks, and when they weather and are carried to the sea inrivers, the mo naz ite grains mo ve wi th them . As mon azite grains are denser than silicate minerals, theyare conce ntrated w it h other heavy minerals lik e ilme nite FeTiO, and zirco n on beaches and dunes.W he n such placer mineral deposits are min ed for their titaniu m content, for exa mple i n Australia andNatal, South Africa, the mo nazite i s also extracted and sold as a by-product.


7/12

Rare ear ths f rom supernova to superconductor 2459

1000004 1-

\ I01 A Palabora concentrateN Dune monazite0 Bastnasite

La Ce Pr Nd (Pm) m Eu Cd Tb Dy Ho E r Tm Yb LuFIGURE 8. Ch ond rite normalised rare earth profiles for a typical du nedepos it mon azite, compa red w it h M ou nta in Pass bastnasite (data fro m

ref. 15) an d a rare earth concentrate from the Palabora carbonatite(data from FEDMIS, pers. com m.) show significant differences in eu ropiu mconcentrationMo naz ite was originally m ine d for thorium coatings on incandescent gas lam p mantles. The rare earthcontent i s n o w more valuable, bu t mona zite has tw o disadvantages as a rare earth ore. Its radio activ ityrequires preca utions in h and ling and extraction. Secondly, because monazites form late in the igneouscrystallization sequence, they contain little europiu m as shown in Figure 8, because it has already beentaken up by plagioclase. Europium i s the most im portant rare earth elem ent in today's market,contributing to the development of square-eyes in our children.The other major rare earth ore mineral i s the carbonate bastnasite, which i s found m ainly in onecarbonatite intrusive plug at Mo unt ain Pass, California. This i s the major producer of rare earth minerals,about 20 000 tons per annum and contributed to rare earth technology in an unusual wa y. The comp anywh ich owned it had no market for bastnasite, so invested heavily in research. This pa id off wh en theydiscovered that europium cou ld form a highly eff icient red phosp hor fortele vision screens;every c olou rTV today contains about one rands wo rth of europium, mu ch of i t com ing from Mo unt ain Pass. Themu ch higher Eu content of M ou nt ain Pass bastnasite show n i n Figure 8 makes it a m ore valuable orethan monazite, and it also has lo w thorium .Apatite i s a com mo n calciu m phosphate mineral also found in carbonatites wh ich can yield a rareearthconcentrate. Th e analysis shown in Figure 8 i s from Palabora, a carbon atite pipe n the eastern Transvaal,South Africa. Althou gh rare earths are only min or constituents of this apatite, the co mp any w hi ch minesit for p hosphate fer tilizer separates the rare earths as a by -pro duc t. This p rod uc t also has a favourableeuro pium concen tration as plagioclase i s absent from the carb onatite system.

X - R A Y F L U O R E S C EN C E A N A L Y S I S A N D G E O C H E M IC A L P R O S P EC T IN GScanning electron m icroscope X -ray spectra of two rare earth phosphate mine rals from a dune sandmine, m onazite and xenotime are shown in Figure 9. W hi le m onazite shows large peaks for the lightrare earths, xenotim e has a smaller lattice site whic h accommodates on ly yttriu m and the heavy rareearths. The X-ray tech niqu e clea rly distinguishes these minerals, bu t overlaps of the four fold L-spectrallines of the in dividu al elements makes quantitative analysis difficu lt. The four most abunda nt ligh t rareearth elements La, Ce, Nd and Sm can be determined with high-resolution wavelength-dispersivespectrometers at levels above ten pp m in rock and ore samples, but even w ith group separation andconcen tration methods (ref. 12) elements like Eu rema in hidde n in the backgroun d.


8/12

2460 D. H.CORNELL

60

50 -.- 4 0 -v101-.E 30-rrx 20-

10-

CF Cc p1 Nd DYXenotirne

FIGURE 9.Scanning electron microscope energy-dispersive X-ray spectra of monazite andxenotime, tw o rare earth phosphate

Er

J0 1 I I I I I

4 5 6 7 8 9Energy, KeV

X-ray m ethods are thus not used to de termine rare earth profiles, bu t because of their lo w cost and rap idturnround, are ideal for prospecting i f only La, Ce and Y (representing he heavy rare earths) are required .Stream sediments and soil samples can be analyzed in large numbers, revealing anomam ous targetregions in whic h prospecting may be concentrated (ref. 13). Once a deposit has been discovered anddelineated, m ore comprehensive and expensive analytical methods come in to play.G R O U P S EP A R A T I O N A N D L O W - T E C H N O L O G Y U S E S

Rare earth minerals are norma lly sold as concentrates containing more than 90% of the ore mineral. Thefirst step in extraction aims to p roduce an undifferentiated rare earth product conta ining m ostly ceriumand lanthanum w ith min or amounts of all the others according to their proportions in the ore mineral.The group extraction i s achieved by reaction with either caustic soda orwith sulphuric acid. If monazitei s he feedstock to this process, radioactive thorium becomes an effluent and representsan environmentalproblem (ref. 14). Such combined rare earth products can be used in applications where any one of therare earth elements has the desired effect, for example the production of steel alloys, in catalysts or asan abrasive for glass polish ing. The term "misch metal" i s used forthe hard, silver rareearth metal, whic hhas the in timid ating property of spontaneous combustion in air. It i s used in lighter flints as a spark-inducer, as we ll as i n alloy me tallurgy (ref. 15). once had an exciting experience trying to weigh ceriummetal powder to make a standard solution, whilst it, once free of the argon it was stored in, was benton becoming an oxide as quic kly as i t could!

T H E W O RL D M A R K E T FO R R A RE E A R T H S A N D E S T A B L I S H ED U S E SAbout 2.2 m il lio n tons of rare earths have been produced in the past century (Figure 10); present annua lproduction is about 50 kilotons. Steenkampskraal monazite mine at Van Rynsdorp in Namaqualand,South Africa was a major producer during the second world war bu t was qu ickly m ined out. W ith therapid growth in demand for rare earths geological exploration has also increased and the officia l wor ldreservesof50 mi llion tons wo uld last a thousand years at the present rate of consumption. Specific rareearth elements are mu ch more expensive than the g roup concentrates. For example "cerium H polishingpowder, conta ining all the other rare earths costs 0.31 US cents/gram, wh ile 99.9% pure cerium oxideis 2.1 dg and 99.9% pure metal i s 19 dg (ref. 16, standard container loads). Prices quoted by JohnsonMatthey Ltd. for one gram o f the individ ua l rare earth metals range from US $5 (lanthanum) to $140(europium) and $3 50 for lutetium (p robably because there i s little demand and the separation s costly).These prices are comparable w ith those of gold, p latinum and silver, so in this respect the rare earthsare precious metals! As w ith precious metals, the price depends on demand, and ove rproduction w il lprobably lead to massive price reductions.Some interesting aspects of present-day use are: Rare-earth zeolite catalysts are used for "cracking"crude oil to produce sho rter-chain molecules i n petrol refining. The white blebs on the sides of self-clean ing ovens are rare earth oxides wh ich prom ote the oxidation o f fat. Spectacles, telescope mirrorsand television screens are polished wit h "ceria" polish ing powder. In glass manufacture the unwanted


9/12

Rare ear ths f rom supernova to sup erconductor 2461

Brazil18Metallurgy

30

2514FIGURE 10. Glob al rare earth element produc tion(mt) of the major producing countries 1885- 1980,acco rding to ref. 15. Total produc tion was 1.7 Mt.Mo st of South African produc tion was from thenow-d efunct Steenkampskraal m onazite m ine

FIGURE 11. Rare ear th element usage by mass,updated from ref. 15. The 5% used in high-techapplications i s the fastest-growing and highest-value sector

blue colour of ferric Fe3+ro n i s removed b y ox idiz ing it to the ferrous Fe3+ tate using ceric ox ide CeO,(ref. 15). The europium-based red phosphor for television screens allow ed manufacturers to increasethe relative brightness of the other colours b y 40%. The development o f all these products required amix of ap plied chemistry and crystallography, with a sprinkling of physics, backed u p as always byanalytical chemistry.Although industrial consumption should increase exponentially, the inexo rably high demand foreuropium for TV screens seems sure in he m edium term to be the major factor in he supply and demandmech anism wh ich determines the price of rare earths.

E L E M E N T S E P A R A T I O N FOR H I G H - T E C H P R O D U C T SThe presen t uses of rare earths are sum marised in Figure 10.The high tech sector represents a rapid lygrowing new market, which requires individual rare earth elements at purities ranging from 60 to99.999% (ref. 15). Upg rading rare earth ores to high-purity metals adds orders of m agnitude to theirvalue, as shown by Figure 12. Analytical control of the purity of pure rare earth products requires

Potential value of elements in monazite

La Cc Pr Nd Sm Eu Cd Tb Dy Ho Er Tm Yb LuFIGURE 12 . The potential value of each element proportion in one gramof m onazite if refined to 99.9 % pure metal, in US cents


10/12

2462 D. H. CORNELL

techniques wh ich can detect low impurity levels in essentially 100/~ f a single oxide or metal. Op ticalor mass spectrometry i s most commonly used, the solid sample being vaporised by eithe r spark, DC arcor laser abla tion. The laser ablation ICPMS can determine impurities at the sub-ppb (ng/g) level.To obta in such products from the group-separated material, the subtle differences between the elementshave to be exploited. Europium and cerium ar e usually converted to the divalen t or tetravalent ionsrespectively, to yield insoluble comp ounds w hic h precipitate wh ile the trivalent rare earths are stillsoluble. The other rare earths ar e usua lly purifie d by extrac tion in organic solvents, wh ich process hasthe advantage of being amen able to con tinuous-flow operations (ref. 17). These techniques are w ell-kn ow n and used in the first-world economy, n otably in the USA, EC and japan. Althou gh South Africano w produces bo th monazite concentrates from dune m ines and a rare earth group conce ntrate fromPalabora apatite, the ma jor value-a dding process of separation into pure oxides and metals i s still donein true first wo rld countries. There seems little prospect of rare-earth refining industries developing closeto the m ines unless a major cost-saving breakthrough i s made by MINTEK or other minera l technologygroups.Analyt ical meth ods us ing chrom atograph ic separa t ion : Ion chromatography and ICP atomic emiss ionspectrographyIn ion chromatography the operator passes a sample in solution though an ion-exchange column inwhic h the rare earths are separated because of their s lightly diffe ring pa rtition coefficients. The elementsare detected as conductivity peaks on emerging from the column. This method uses relativelyinexpensive equipm ent and has p roved successful for rock analysis (ref. 18 ). How ever it cannot easilyhandle the large differences in concentration between adjacent elements produced in industrialrefining. Another m ethod wh ich depends on a similar separation for rare earth analysis i s inductivelycoup led plasma atom ic e miss ion spectrography (ICPAES). The major elements such as iron present inmost samples cause severe overlap interferences with many of the rare earths in the optical spectrum.Grou p separation using ion exchange columns i s the salvation of this m ethod wh ich has dyna mic rangeand sensitivity similar to that of the more expensive ICPMS (ref. 19 ). However the skilled m anpowerneeded for separation i s a major cost factor. N ote also that one m ay easily produce negative ceriumanomalies during separation by allo win g the ion exchange resin to dry out in the presenceof nitric acid;the resin retains Ce4+ rom subsequent next samples.

R A R E E A R T H S A S A F U T U R E R E S O U R CEMan y new applications of rare earth elements are currently being developed. One w hic h i s already faradvanced i s the produ ction of perm anent magnets based on samarium or neodymium , whic h are morethan five times stronger than conventional alnico magnets. These are used in small motors andappliances, particularly where weight savings are important in car and aircraft manufacture. Productionin 1988 was about 300 tons per annum and i s rapidly increasing (ref 19). Materia l scientists are usingrare earths to stabilize crystal structures l ike cubic zirconia, and to make new compounds such asyttrium-aluminium garnet (YAG)a laser generator and yttrium -iron garnet (YIG)for microw ave systems.An exciting chapter in rare-earth ech nolog yfollow ed he discovery by Bednorz and Mu ller (ref. 20) andsubsequent w orkers o f high-temp erature superconductorscomp rising lanthanum barium cop per oxidessuch as La,+BaxCu04 (Figure 13). These have no resistance to electric current at 90K or -1 83-C, easilyattainable using cheap -1 95-C liq uid nitrogen, while current superconductors need expensive liqu idhelium at -269-C. This discovery led to an explos ion of research and hasty papers in Na ture, w ithmassive funding being channelled into new mu ltidisciplinary research groups. Crystallography suddenlybecame fashionable: the unusual prospect arose of Nobe l prizes for m aterials scientists, the prom ise ofsuperconducting pow er transmission cables wh ich wou ld save 50% of po wer losses in national grids,and other advances such as liquid-nitrogen-based superconducting magnets for medical body-scannersand nuclear research accelerators.


11/12

Rare earths from supernova to superconductor

Table 3 Comparsion of some analy tical m ethods for rare earth elements.O ne US ollar equals approximately 3 SA Rands

0.5-0.005I

2463

Number ofelementsdetermined

1 2

3peratortimeincludingsamplereparationan dilculat ion:hours

Total

hours

cost ofnstrument1992 inin SouthAfrica

R1OO 000

rechn ique

Units

Cost peranalysisin SouthAfrica1992; Rands

Routinedetectionl imi t insol id

samplePPm(ridkg1

Neutronactivationanalysis 5-0.1 8-1 0 2 2.5 15 0

15 0

100

150

Inductivelycoupledplasmamassspectro-metry 0.1-0.0 1 15 2 10Inductivelycoupledplasmaatomicemissionspectro-metry 32-1 5 6Ionchroma-trography 0.1-0.01 1 2 2

Isotopedi lut ion(thermio-nic ) massspectro-metry 1-0.01 8-1 0 8 50 0

50

40

10

X-rayfluores-cencespectro-metry

I

I

l 10 5 2Two facts have emerged as the dust settles: copper, not lanthanum , forms the basisof the superconductor,the other cations which may include other rare earths, bismuth and thallium, are needed simply toprovide the right framework in the perovskite structure for copper oxide planes to carry the currentwithout resistance. Secondly, there are technical problems in making long, superconducting cableswhich can transmit significant amounts of electrical power. This i s because of the difficulty ofma intaining superconductivity across grain-bounda ries, and because these ceram ic m aterials are brittleand fracture easily. In this area of endeavour analytical methods are proving crucial, to introduce andmonitor the desired levels of impurities and in determining the crystal structures using electronmicroscop es and X-Ray or electron diffraction. It remains to be seen whether rare earths form part ofthe final product o f this unexpected technological breakthrough, wh ich i s a good exam ple of a spin-offfrom "pure" research.


12/12

2464 D. H. CORNELL

Helium NitrogenLa Ba Cu oxide

Y Ba Ag oxide

Th Ba Pb(Zr)oxide

Y Ba Cu oxide

B i Sr Ca Cu oxide

FIGURE 13.Transition temperatures of somehigh-temperature superconductors.Three of these materials are super-conducting when immersed inliquid nitrogen, whic h i s muchcheaper than liq uid helium

I I I I I-250 -200 -150 -100 -50 0Superconductivity region C

W H I C H M E T H O D T O U S E ?In Table 3 I have summ arised some im portan t features of several techniques used for rare earth elementanalys is. As I have pointed out d uring above each method has its applications to one or other area ofendeavour. For many applications it i s probably better to pay an already expert laboratory to do theanalysis than to embark on a long and expensive learning curve. If one i s contemplating in-houseanalysis of rare earths, the cost and r elia bility of equipmen t are m ajor factors, howeve r sophisticatedequipm ent cannot overcome the need for an expert analyst wh o has the time to ro ll up his sleeves, dobattle with suppliers and get his test-tubes wet!

REFERENCES1. D. H. CORNELLed.) Symposium on Rare Earth Elements as a Future Resource, Dep t Geology, Un iv.Stellenbosch (1988).2. P. HENDERSON,h 7 in P. Henderson, (ed.)Rare Earth Element Geochem istry. Elsevier (1984) 51Opp.3 . R. D. SHANNON. Acta Crystallogr., Sect A, 32: 751-767 (1976).4. R. GANAPATHYnd E. ANDERS.roc. 5th Lunar Sci. Conf. 1181 1 206 (1974).5 . A. MASUDA.eochim. Cosmochim. Acta 37: 239-248 (1973)6. N.M. EVENSEN,P. J. HAM~LTONnd R. K.ONIONSeochim. Cosmochim. Acta 42: 1 199-121 2 (1978)7. L. A. HASKIN.h 4 in Henderson, P (ed) Rare Earth Element Geochem istry. Elsevier (1984) 51 Opp.8. K. E. JARVIS. J. Anal.At. Spectrom. 4: 563-570 (1989).9. P. HENDERSONnd R . 1. PANKHURST.h 73 in Henderson, P (ed) Rare Earth Element Geochemistry.Elsevier (1 984 ) 510pp.10. R. W. CARLSON,.R. HUNTERnd F. BARKER.ature 305, 701 -702 (1983).11 . P. J. HAMILTON,. M. EVENSEN,R. K. ONIONS,H. S. SMITH and A. J. ERLANK. Nature 279, 98-300(1979).12 . G . N. EBY.Anal. Chem., 44: 2137-2143 (1972).13 . D. H. CORNELL,. SCHADEand R. SCHEEPERS. .Afr.J. Sci. 89, 54-658 (1989).14 . D. G . GOULD,n Cornell, D H (ed.) 24-25 (1 988).15 . C. R. NEARYnd D.E. HIGHLY.,h 72 in Henderson, P(ed.) Rareearth element geochem istry. Elsevier

(1984) 51 Opp.16. Com mercial p rices for Lanthanide oxides and minerals are given in the weekly journal AmericanMetal Market.17 . R. L. PAULand J. S. PRESTON,n Cornell, D H (ed.) 44-45 (1988).18 . A. P. LE Roux and R. T. WATKINShem. Geol. 88,151 -1 62 (1990).19. J.N.WALSH,. BUCKLEYnd J. BARKER.hem. Geol. 33: 141-153 (1981).20 . J. G. EDNORZ nd K. A. MULLER.Phys. B64, 189-193 (1986).

Rare Earths From Supernova to Superconductor

Documents

Transcript of Rare Earths From Supernova to Superconductor