Alteration of Spodumene, Montebrasite and Lithiophilite In

Home , Lithiophilite, Spodumene, White Picacho

American Mineralogist, Volume 67, pages 97-113, 1982

Alteration of spodumene,montebrasite and lithiophilite in pegmatites of the White PicachoDistrict, Arizona

Davrp Lor.rooxrnNo DoNer-uM. Bunr Department of Geology Arizona State University Tempe, Arizona 85281

Abstract

The crystallization sequence and metasomatic alteration of spodumene (LiAlSizOe), montebrasite(LiAIPO4(OH,F)), and lithiophilite (Li(Mn,Fe)PO+)are describedfor nine zoned lithium pegmatitesin the White Picacho district, Arizona. The observedcrystalliza- tion trends suggesta progressiveincrease in the activities of lithium species(spodumene follows microcline as the principal alkali aluminosilicate), as well as an increase in the activities of the acidic volatiles phosphorus and fluorine (montebrasite succeedsspodumene as the stableprimary lithium phase).Much of the lithiophilite occurs with columbite, apatite, beryl, zircon, and tourmaline in cleavelanditecomplexes that formed in part at the expenseof quartz-spodumenepegmatite. Fracture-controlledpseudomorphic alteration of the primary lithium minerals is widespread and apparently is the result of subsolidus reactionswith residualpegmatitic fluids. Spodumenehas been replacedby eucryptite, albite, and micas. Alteration products of montebrasite include low-fluorine secondary montebrasite,crandallite (tentative), hydroxylapatite, muscovite, brazilianite, augelite (tentative),scorzalite, kulanite, wyllieite, and carbonate-apatite.Secondary phases identified in altered lithiophilite include hureaulite, triploidite, eosphorite, robertsite, fillowite, wyllieite, dickinsonite,fairfieldite, Mn-chlorapatite, and rhodochrosite.Initial subsolidus metasomatismof the lithium minerals took place in an alkaline environment, as evidenced by albitization of spodumeneand calcium metasomatismof the phosphates.The formation of secondarymicas in spodumene,montebrasite, tourmaline, and much feldsparreflects a changefrom alkaline to relatively acidic postmagmaticfluids, as (K+H)-metasomatism producedgreisen-like or sericiticalteration. The abundanceof mineralscontaining Li, Be, Mn, Nb, Ta, and Bi indicate that these pegmatitesoriginated from a highly differentiated granitic source.These pegmatites were not fluorine-rich,as evidencedby the low fluorine contents of primary and secondarymontebrasite, by the formation of OH- and Cl-apatites, and by the absenceof topaz and the rarity of lepidolite, triplite, and fluorite.

Introduction also Precambrian(Laughlin, 1969),although a pre- The White Picachopegmatite district lies nearthe cise age has not been established. southeastend of the Arizona pegmatite belt (Jahns, Only nine lithium pegmatiteshave been identified 1952; see Fig. 1). The district, which is located in the district. Most of these were mapped and mostly on the Red Picacho7.5' topographicquad- described by R. H. Jahns (1952), and his work rangle map (U.S. Geological Survey, 1964),con- provideda foundationfor subsequentstudies by us tains severalhundred pegmatites. These pegmatites (Burt, London, and Smith, 19771'London,Bandy, intrudelow- to medium-gradePrecambrian schists, and Kealy, 1978;London and Burt, 1978;London, gneisses,and amphibolitesthat Jahns(1952) tenta- 1979).At present, most of Jahns' maps are still tively correlated with the Yavapai Series in the usable,inasmuch as only minor mining and devel- Jerome and Prescott areas (Jagger and Palache, opment have been carried out in the district over 1905;Anderson et al., l97l). The pegmatitesare the past thirty years. Additional maps of the region and of the pegmatites are available in London and rPresentaddress: Geophysical Laboratory, 2801Upton Street, Burt (1978)and in London (197il. N.W., Washington,D.C. 20008. In the White Picacho pegmatites,as at many m03-004)vE2l0r 02-0097$02.00 97 98 LONDON AND BURT: SPODUMENE, MONTEBRASITE AND LITHIOPHILITE

Moore, 1971,1972; 1973, 1974), the significanceof these replacements as indicators of pegmatite chemistry and evolution remains poorly under- ti::::i i- stood.In the White Picacholithium pegmatites,the t:::i: t i:i paragenetic tl sequenceof these secondaryminerals can be determined for each of the parent lithium minerals.Each paragenesisis interpretedas indicat- ing systematic changesin the geochemistry of the postmagmatic fluids. The central purpose of this paper is to characterize these pseudomorphic re- placementsand to evaluate their geochemical sig- nificance.Seventy-eight representative electron microprobe analyses and refined X-ray powder dif- fraction data for most of these phasesare listed in AppendicesI and2, respectively.ISome theoretical

rTo receivecopies of theseAppendices, order Document AM- 82-187(analyses, Appendix l) and AM-82-188 (X-ray data, Appendix 2) from the BusinessOffice, Mineralogical Society of

T nO America,2000 Florida Avenue,N.W., Washington,D.C. 20fi)9. Pleaseremit $1.00in advance for each microfiche.

ARIZONA SCALEOi MILES 0 20 40 60 t0 t00 Table l. Alteration products of spodumene,montebrasite, and lithiophilite in pegmatitesof the White Picacho district.

Ideallzed fornula*

Fig. 1. Map showingthe location of the White Picachodistrict in the Arizona pegmatitebelt (stippled area) as defined by Jahns sDodmene Lijl1S1206

(r95D. alblte I{aA1Si308

eucryPtlte r,tals104 K(Ll,A1) (S1'A1)4o10 (F'On) 3 2 xAr3sl30ro(oH) other localities, the lithium minerals have been 2 metasomatically altered to a large number of rare and mineralogically phases(Table rcntebraslte L1A1P04(0H'r) complex 1). This Iow-F nontebraaite LrA1P04 (oH'F) cleavage-controlledpseudomorphic replacement of crandallile csA13(PO4)(?o30H) (oH)6 the primary lithium mineralsis pervasiveat most of hydroxylepatlte ca, (PoO), (olt,F) nuscovlte KAl35130rO(OH) the pegmatitesand apparentlyis the result of sub- 2 brazillanlte N€al3(Po4) (oH) 2 4 solidusreactions with residual aqueouspegmatitic augellte A1, (Po/ ) (orl) r fluids. The sporadic distribution of alteration within scotzalite (rl2+,ile)er,ipo,) 2+, "

Abundances denoted by:

Geologyof the lithium pegmatites X = abundant O = comon Of the nine known lithium pegmatites in the R = rare A = not reported district, the Homestead, Independence,Midnight or observed

Owl, North Morning Star, and White Ridge were Pegmatite6 Spodumene I'lontebrasite Lithiophilite chosenfor detailedstudy, as thesefive pegmatites Homeslead 0 RX appearto encompassall variationsin generalgeolo- IndepeDdence 00 gy and in the occurrenceand alterationof primary Lone Giant OR lithium minerals (see Jahns, 1952, and London, Lower Junbo AA Midnlght ow1 x XO 1979,for index maps and detailed geologic maps I'lorning Star 0 AR and desqriptionsof the pegmatites). The lithium North Morning Star x XA pegmatitesdisplay highly irregularoutcrop patterns Picacho View 0 OA &trite Ridge x OR and usually occur as clustersof small steeply dipping pods or lenses rather than single, isolated *See Jahns (1952) and Londan (1979) for inder ncps md geo- bodies. Spodumene (LiAlSi206), montebrasite LogLe naps and deseiptims of indiuiduaL pegnatite.s. (LiAIPO4(OH,F)),and lithiophilite (Li(Mn,Fe)pOa) are the primary lithium minerals in these pegmatites. Micas generally are not abundant, and true lepidolitesoccur only as replacementsof primary occurs in the quartz cores and apparently was the lithium minerals.Green lithium tourmaline (verde- last primary lithium mineral to crystallize. Lithio- lite) occurs at one locality but dobs not figure philite also occurs with columbite, manganoantour- prominentlyin the paragenesisof the lithium miner- maline, zireon, and manganoanapatite in massesof als. relatively coarse-grainedalbite (cleavelandite)or All of the lithium pegmatitesare sharplyzoned by cleavelandite * mica. Although lithiophilite nod- grain size and by composition. With few excep- ules are abundantin these complexesat the White tions, the lithium mineralsare concentratedin the Ridge, Homestead,and Midnight Owl pegmatites, inner zonesand coresof essentiallygranitic pegma- the primary phosphateis usually highly altered. tite. The distributionof zoneswithin eachpegmatite In the White Picachopegmatites, cleavelandite or is somewhatirregular (nonconcentric), but a gener- cleavelandite* mica complexesusually are located al zonation sequencecan be established(Fig. 2). at the margins of quartz-spodumene pegmatite Basedon the assumptionthat crystallizationpro- zones;however, they crosscutall other zonesat the ceededsequentially from borders to core, spodu- North Morning Star and Homestead pegmatites, mene was the first lithium mineral to crystallize. and they are the only units that do so. At the Quartz-spodumenepegmatite appears to grade in- Independencepegmatite, a cleavelandite-lepidolite ward into quartz-montebrasitepegmatite, suggest- complexcontains rounded relicts of spodumenein a ing that montebrasitesucceeded spodumene as the textural relationship similar to that of the "spotted stableprimary lithium phase.Lithiophilite typically rock" zone at the Harding pegmatite,New Mexico (Jahnsand Ewing, 1976).At the Independence,the interior portions of the rounded spodumenerelicts have been replaced by fine-grained eucryptite + albite,whereas the rims of thesecrystal fragments, plagloclase-quartz-ntcrocllne-( tourultne)]]]- torder

nlcrocline-quartz-plagloclase-(nuscovtte)f- wall andmedium-grained (5 cm alongthe c axis) second- Dicrocllne-quartz -( spodunene) ary spodumenecrystals in the complex, have been in!emedlate quartz-spodunene replaced by mica + albite. These features are quartz-montebrasite

quart z- ( 1 lthlophtlit e)------co re sufficientto indicate that the cleavelanditeunits crystallizedlate in the history of thesepegmatites, Fig. 2. General sequenceof primary crystallization for the and that they formed at least in part at the expense White Picacho lithium pegmatites. of pre-existingquartz-spodumene pegmatite. r00 LONDON AND BURT: SPODUMENE, MONTEBRASITE AND LITHIOPHILITE

Occurrencesof primary spodumene,montebrasite, Table 3. Fluorine contents of montebrasite from the White and lithiophilite Picacho pegmatites. Spodumeneis the only primary lithium alumino- Prlmary montebraslte silicatepresent in the district, and it occursat all of Sample 2e(131)* wt7" F the known lithium pegmatites(Table 2). It usually forms subhedral,prismatic MNO 4XX 52.64 6.8 crystals up to two me- WHR 408a 52.62 6.6 ters in maximum dimension; however, unaltered IND 4OO 52.6r 6.5 spodumeneis relatively rare. wHR 406 52.60 6.4 (o Montebrasiteoccurs at most of the pegmatitesas wHR 404 )2. )o wliR 400 52,55 5.8 blocky, subhedralcrystals or roundednodules up to MNO 400 52.55 5.8 a meterin dimension.The fluorinecontent of White MNO 419 52.55

Picachomontebrasite was determinedusing the X- WHR 4O8b 52.51 ray techniqueof eern6 et al. (1973).Representative HOr4 401 52.49 4.9 HOM 400 52.46 sampleswere analyzed by electron microprobe for MNO 425 4.3 Na, K, and Ca, to ensurethat no substitutionsother Secondary montebrasite than those involving F and OH affectedthe position Smple 2e(I31)* wrz F of the 2e(131)reflection (Cernr4 et a\.,1973);in all cases,the contents of Na. K. and Ca in White wHR408a 52.41 ?o wHR406 52.36 Picachomontebrasites were at or below the detec- r${s 400 52.35 3.2 tion limits for theseelements (less than 0.l0wt.Vo as MNo 407 52.35 3.2 oxides).Table 3 indicatesthat most primary materi- HOM400a 52.32 2.7 al examinedin this study containsfrom about 4.5- I1NO419 52.31 2.6 NMS401 )L. ZO 2.2 6.5 wt.VoF, correspondingto 30-50 mole Voof the MNO4&\ 52.27 amblygonitecomponent. Similar findings for blocky wHR408b >2. zo 2.0 white crystals are reported from the Varutrdsk noM 400b 52.22 r.4 pegmatite, (Quensel, MNO400 52.21 1.3 Sweden 1937)and from the }fi'lo 425 52.tL 0.1 Tanco pegmatite,Manitoba, Canada(Cernr4 et al., 1972). *X-rq data eoLlected on a Philips Lithiophilite occursat severalpegmatites as nod- goniometer, unfiLtered CuKq radia- tion, g"raphite monoehromaton, at ules or crude prismaticcrystals up to twenty centi- 40 kV, 20mA; scan rate: P z1/nrLn; metersin dimension.Microprobe analyses of repre- quottz intezmal standard. sentativesamples (Appendix 1)indicate that except for materialfrom the White Ridge pegmatite,most White Picacho lithiophilite is very near the als, commonly are themselves intimately intergrown. petrographic, LiMnPOaend memberof the lithiophilite-triphylite A combinationof X-ray dif- series. Such iron-free lithiophilite is not common fraction, and electron microprobe analyses was (c/. analysesof Palacheet al., 1951,p. 666),but requiredto identify the secondaryassemblages. A phases comparablematerial has been describedfrom a few few secondary still are not fully character- localities(e.9., Thomssenand Anthony, 1977). izedbecause of a lack of suitablematerial for study. Nevertheless,most of the secondaryminerals have beenpositively identified, and the crosscuttingrela- Determinationof secondary parageneses mineral tionshipsobserved among veinlets indicate the se- The metasomaticalteration of spodumene,mon- quenceof replacement. tebrasite, and lithiophilite occurred along cleavages, fractures, and crystal borders of the host primary phases.Veinlets are composedof one or Alteration of spodumene more fine-grainedreplacement minerals. In some The metasomatic alteration of spodumene in cases,individual veinlets contain two or more dis- these pegmatites has been discussed in several crete assemblagesthat are bilaterally symmetrical other publications(Burt et al., 1977;London and about the center of the veinlet. Individual mineral Burt,1979,1980,1981b), so only a briefdescription grainsaverage 25 micronsin dimension,and several is presentedhere. At the Independenceand Mid- assemblages,each containingtwo or three miner- night Owl pegmatites,much spodumenein contact LONDON AND BURT: SPODUMENE, MONTEBRASITE AND LITHIOPHILITE l0l

volved hydroxyl exchangefor fluorine, producing SPODUMENE-----'-+ * -'------> * +MUSCO\rITE along frac- A]JITE ALBITE low-fluorine, secondary montebrasite in the host primary phase(Fig. Fig. 3. General alteration sequencefor spodumene. tures and cleavages 6). Based on values of 20(131),the secondary montebrasitecontains about 1.5-3'5 wt.VoF (Table (Cetnd with primary quartz displays a thin reaction rim of 3). As at the Tanco pegmatite,Manitoba et fine-grainedalbite, whereas the interiors of the aL, 1972),replacement by secondarylow-fluorine spodumenecrystals are replacedto varying degrees montebrasiteis extensivein fresh-appearingmateri- by a very fine-grained,fibrous intergrowth of eu- al, but the primary and secondaryphases are virtu- cryptite * albite (Fig. 3). This initial replacement ally impossible to distinguish in hand specimen. produced bilaterally symmetrical veinlets of eu- Subsequentalteration of primary and secondary cryptite + albite whose long fiber axesare oriented montebrasiteproduced fine-grainedhydroxylapa- of a calcium perpendicularto the {110} cleavage of the host tite, muscovite, and minor amounts spodumene(Fig. a). In some samples,eucryptite aluminumphosphate that appearsto be crandallite nearveinlet centers and spodumenecrystal borders (Fig. 7). Positive identification of crandallite is has been subsequentlyaltered to muscovite or hamperedby the fact that the phase usually is lepidolite, resulting in the intergrowth of albite + disseminatedin and subordinate to apatite and mica called "cymatolite" by Brush and Dana montebrasite; thus it cannot be clearly distin- (1880).Finally, remnant spodumeneand secondary guishedin X-ray powder patterns. In addition, its (2 albite have been convertedto lithian muscoviteor fine grain size requires the use of a narrow lepidolite, producing soft, waxy pseudomorphsof micron) beam diameter for microprobe analyses; nearly pure mica after spodumene (referred to as this results in rapid volatilization and sample de- "killinite" by Julien, 1879).At all of the pegmatites, struction.Although thesethree secondaryminerals spodumenealso appearsto have been converted occur in grain contact with montebrasite,the cran- directly to albite * mica or to pure mica, but the dallite invariable occurs at montebrasite borders alterationsequence described above and illustrated and probably was the first mineral to crystallize.In in Figure 3 representsthe most completehistory of a typical specimen(Fig. 8), a thin reaction rim of metasomaticalteration of spodumeneat thesepeg- crandallite at montebrasiteborders gradesoutward matites. into veinlets of grayish-whitehydroxylapatite, followed by an intergrowth of apatite * muscovite, Alteration of montebrasite and finally fine-grained green muscovite at the At all of the pegmatites,montebrasite has been centersof replacementveinlets and near the crystal replacedby a numberof secondaryphases (Table 1) boundariesof the primary phase. in the general sequenceillustrated in Figure 5. Various combinationsof scorzalite,brazilianite, (Table Incipient alteration of primary montebrasite in- augelite(tentative), wyllieite, and kulanite 3) occur as scatteredcrystals in the apatite + muscovite replacementintergrowth and alsoform veinlets that clearly cut the apatite * muscovite assemblage.Thus theserare phosphatesformed late in the alterationhistory of montebrasite(Fig. 5)' Basedon availablesamples, augelite, wyllieite, and kulanite are very rare in montebrasitefrom all of the White Picachopegmatites, but scorzaliteand brazilianite are fairly common. Scorzalite occurs as anhedral disseminatedgrains that give specimensa light blue tint. Brazilianiteforms relativelylarge (2 mm-l cm) subhedralyellowish green crystals in specimens from severalpegmatites. A few colorlessgrains of an aluminum phosphatethat appearsto be augelite Fig. 4. Photomicrograph showing the cleavage-controlled have been found in two specimens;this phase is (Spd) to a fibrous intergrowth of alteration of spodumene and apatite, eucryptite (Ecr) + albite (Alb). The specimen is from the associatedwith scorzalite,muscovite, Midnight Owl pegmatite. Crossed polars. but positivg identificationof this aluminum phos- r02 LONDON AND BURT: SPODUMENE, MONTEBRASITE AND LITHIOPHILITE

HYDROXYLAPATITE HYDROXYLAPATITE MONTEBMSITE---+SECONDARYMONTEBRASITE----> + ------> + ----> MUSCOVITE->CARBONATE-A?ATITE CR3NDALLITE MUSCOVITE + BRAZILIA}IITE + AUGELITE + SCORZALITE + KULANITE + WYLLIEITE

Fig. 5. General alteration sequencefor montebrasite.

phateawaits a find of more material. Kulanite also intergrowthsof hureauliteand triploidite (Fig. 10, has been found in two specimensas forest green SchemeA, and Fig. l2). Along more extensively subhedralcrystals in associationwith apatite,mus- alteredveinlets, hureaulite becomes more birefrin- covite, quartz, and brazilianite. Its composition gentin thin section;this changein opticalproperties (analyses56-58, Appendix l) is similarto that of the appears to correlate with small but increasing type material (Mandarinoand Sturman, 1976),and amounts of calcium (analysis 12, Appendix 1). this representsthe secondreported occurrenceof Frequently, calcium-rich hureaulite is associated kulanite.Wyllieite was identifiedin thin sectionand by microprobeanalyses (analyses 69-72, Appendix 1) as a few grains that occur with scorzalite and brazilianitein one montebrasitespecimen. The latest-forming phase in the montebrasite alteration sequence is fine-grained carbonate-apatitethat formsbanded, colloform rinds that surroundmonte- brasitenodules and veinlets that crosscutall other replacementassemblages (Fig. 9).

Alteration of lithiophilite Lithiophilite displays two different replacement sequences,depending on whetherthe lithiophilite is embeddedin quartz or in albite (Fig. 10 and Table 3). The least-alteredcrystals occur in quartz (Fig. l1). In this association, the initial replacement involved hydration and removal of lithium to form

Fig. 7. Photomicrograph (top) showing the replacement of twinned primary montebrasite(Mbs) by a fine-grainedaggregate of hydroxylapatite (Apt) with minor crandallite (Crn). Notice in the lower right corner that the twinning presentin the host phase has been preserved in the pseudomorphic replacement. Fig. 6. Photomicrographshowing the replacementof primary Specimenis from the Midnight Owl pegmatite. Crossed polars. montebrasite(l' Mbs) by secondary, low-fluorine montebrasite Photomicrograph(bottom) showinga veinlet of muscovite (Mus) (2" Mbs). Both phasesare polysynthetically twinned. Specimen t apatite (Apt) that cuts twinned primary montebrasite (Mbs). is from the North Morning Star pegmatite. Crossedpolars. Sampleis from the Independencepegmatite. Crossedpolars. LONDON AND BURT: SPODUMENE, MONTEBRASITE AND LITHIOPHILITE 103

basisof optical propertiesand by X-ray and microprobe analysis. Olive-greendickinsonite and yel- lowish-brownwyllieite usually occur together and weredistinguished principally on the basisof microprobe analyses;both phasescommonly are intergrown with fillbwite. Fairfieldite commonly forms thin veinletsthat clearly cut the fillowite + dickinsonite+ wyllieite assemblage.As in the montebrasite alteration sequence,fibrous colloform apatite forms bandsthat surroundlithiophilite nodulesand veinletsthat cut all other secondaryassemblages. This apatite is Mn-rich chlorapatitethat contains almost 14 wt.% MnO; the presenceof abundant chlorine was determined qualitatively by energy- dispersivemicroprobe analyses. Along with Mn- chlorapatite,rhodochrosite forms massesand veinlets of pink crystals that representthe latest products of metasomatic alteration of lithiophilite Fig. 8. Arizona State University specimen #3288 from the crystals. Midnight Owl pegmatite,showing the typical alteration observed Spongymasses of Mn-oxides, purpurite, stewar- in montebrasitenodules. Phasespresent in this specimeninclude tite, and strengite(Jahns, 1952) also replacelithio- primary and secondary montebrasite (Mbs), crandallite (Crn), philite, but this assemblageof secondaryphases is hydroxylapatite (Apt), muscovite (Mus), and carbonate-apatite not presentin any freshly mined lithiophilite from (Cpt). assemblagesare outlined The approximate boundariesof excavations. These oxidized minerals by dotted lines. subsurface with eosphorite and fairfieldite; identification of eosphorite and fairfieldite was verified by micro- probeanalysis (numbers 15, 16and 3l-34, Appendix 1). Fairfieldite also forms thin veinlets in close associationwith dark reddish-brownrobertsite; ro- bertsitewas identifiedby its distinctive micaceous habit, by its intensepleochroism in plane-polarized light, and by X-ray and microprobeanalyses. This representsthe second reported occuffence of robertsite (Moore, 1974).Finally in this association, robertsite and all other earlier-formed replacement phaseshave been altered to fibrous Mn-rich chlorapatite (Fig. 13). Relict subhedral lithiophilite crystals in albite displaycrude prismatic[001] forms, but thesecrys- talsare almost completely replaced (Scheme B, Fig. 10). Cores of hureaulite + triploidite and minor lithiophilite are veined and replaced by fillowite, dickinsonite,and wyllieite (Fig. 14). Fillowite has beenreported previously only from the pegmatiteat Fig. 9. Photomicrograph showing a rim of carbonate-apatite Branchville,Connecticut (Brush and Dana, 1879a) (Cpt) surroundingprimary montebrasite(Mbs). The fine-grained and from the Buranga pegmatite, Rwanda (von assemblageabove the apatite is an intergrowth of muscovite (Mus) (Alb) As illustrated in this Knorring, 1963);in the White Picacho district, it + albite after spodumene. samplefrom the Midnight Owl pegmatite, wherever spodumene occurs abundantly as massive, fine-grainedbeige occurs with montebrasite(or near lithiophilite), the spodumene crystal aggregateswherever lithiophilite nodules is invariably more extensively altered than the phosphate. are embedded in albite and was identified on the Crossedpolars. LONDON AND BURT: SPODUMENE, MONTEBRASITE AND LITHIOPHILITE

HUREAULITE Ca-HIJREAIJI.ITE FAIRFIELDITE SCHEME A: LITIIIOPHILITE -+ * .------) * -----) * ------) }h-CHIORAPATITE TRIPLOIDITE EOSPEORITE ROBERTSITE

FILLOI'ITE HTJREAI'LITE + Mn-CHLoRAPATITE SCHEMEB: LITHIOPHIIITE----> * +DICKINSONITE----+FAIRFIELDITE ----_-+ + TRIPLOIDITE + P.ITODOCHROSITE WY1LIEITE Fig. 10. Generalalteration sequencefor lithiophilite. SchemeA pertains to lithiophilite embeddedin massivequartz. SchemeB is for lithiophilite embeddedin coarse-grainedalbite (cleavelanditet appearto be the result of weathering rather than of (resulting in the conversion of eucryptite, then postmagmaticpegmatitic alteration. spodumeneand albite, to mica). Albitization of Jahns (1952) reported sicklerite as a common spodumenemay be describedby two reactions: alterationproduct of lithiophilite in these pegma- LiAlSirou + Sio2 + Na* : NaAlSi3o3 + Li+, tites, but most such material probably was robert- spodumene quartz albite site.Sicklerite does occur as the principalalteration productof lithiophiliteat the White Ridgepegmatite which explains the thin reaction rims of pure albite (which was not exposedduring Jahns' study). The separatingaltered spodumenefrom quartz, and White Ridge is the only lithium pegmatitein this 2LiAlSi2O6+ : + + district that did not undergo appreciablepostmag- Na+ LiAlSiOa NaAISi3Os Li+, spodumene eucryptite albite matic cation metasomatism.Spodumene and montebrasite from this pegmatite are relatively fresh which accounts for the formation of eucryptite * and unreplaced. albite intergrowths within spodumene crystals (London and Burt, 1980, 1981b).Whether spodu- Discussion menebreaks down to albiteor to eucryptite * albite Alteration of spodumene dependsprimarily on the presenceor absenceof quartz,and thus on the activity of silica within and The alterationsequence for spodumenefrom the around large spodumene crystals (London and White Picachopegmatites and from other localities Burt, 1980,198lb). (London and Burt, 1980,198lb) implies that initial The conversion of eucryptite to muscovite is replacementinvolved sodium-for-lithium ion ex- facilitatedby the fact that the Al:Si ratios are the change(as evidencedby the formation of albite), samein both minerals, and thus the formation of followed by potassium * hydrogen metasomatism pseudomorphicmica in spodumeneis enhancedby a local loweringof silicaactivity (London and Burt,

Fig. 11. Lithiophilite specimen from the Midnight Owl pegmatite, showing the alteration assemblagesobserved in specimensembedded in quartz (SchemeA, Fig. l0). The phases include lithiophilite (Lth), hureaulite (Hur), triploidire (Tpd), Fig. 12. Photomicrographshowing the incipient replacement eosphorite (Eos), robertsite (Rob), fairfieldite (Fld), Mn- of lithiophilite (Lth) by hureaulire (Hur) + triploidite (Tpd). chlorapatite(Mpt), and rhodochrosite (Rho). Specimenis from the Midnight Owl pegmatite. Crossedpolars. LONDON AND B(IRT: SPODUMENE, MONTEBRASITE AND LITHIOPHILITE IO5

Strafford,New Hampshire(see London and Burt, 1981b).Various replacementassemblages in the alteration sequencedescribed above (i.e., albite, eucryptite + albite, mica * albite, and pure mica) have been documented from many pegmatites, someof which are listed in Table 2 of London and Burt (198lb). These reported occurrencesindicate that the type of spodumenealteration observed in the White Picacholithium pegmatitesis commonin zoned lithium pegmatiteselsewhere, especially in those pegmatites that contain cleavelandite or cleavelandite-micacomplexes near spodumene- bearingzones.

Alteration of montebrasit e The formationof secondarymontebrasite by OH- for-F exchangein the primary phase presumably reflects an increase in the ratio asrslallp in the Fig. 13. Photomicrographshowing the late-stagereplacement residualfluid. Increasein this activity ratio implies of robertsite (Rob) by Mn-chlorapatite (Mpt). The veinlet cuts a depletionin fluorine (perhapsdue to crystalliza- hureaulite (Hur). Sample is from the Midnight Owl pegmatite. tion of apatite or of micas, or of the montebrasite Crossedpolars. itselfl or a relativeincrease in asr6 with decreasing

1980,1981b). The direct conversionof spodumene to muscoviteliberates silica: 3LiAlSi2O6+K*+2H- spodumene : KAlrSi:Oro(OH)z + 3SiO2+ 3Li+ muscovite qtartz but quartz usually is not present in muscovite pseudomorphsafter spodumenefrom the White Picachodistrict or from other localities (e.9., Gra- ham, 1975).Graham (1975)suggested that some, but not all, of the silica liberated by mica-forming reactionscould be incorporatedin secondarymicas with high Si:Al ratios. Another possible explana- tion lies in the fact that mica pseudomorphs after spodumenein the White Picachopegmatites (and at other localities) usually are embedded in nearly pure cleavelanditethat in some cases appearsto have replacedpre-existing pegmatite. The replacement of quartz-spodumene pegmatite by large massesof cleavelandite(with locally complete con- sumptionof quartz) may facilitate the conversion of remnant spodumene to mica without producing additional qtrartz (see Figs. 7 and 8, London and Burt. 1981b). The pseudomorphic replacementsof spodumene Fie. 14. Photomicrograph showing the replacement of in the White Picacho pegmatites are identical in hureaulite (Hur) + triploidite (Tpd) by dickinsonite (Dck) (top) mineralogyand texture to those observedin pegma- and followite (Fil) (bottom). Homestead specimen, in crossed tites at Branchville, Connecticut, and at Center polars. LONDON AND BURT: SPODUMENE, MONTEBRASITE AND LITHIOPHILITE temperature(Korzhinskii, 1957 Munoz and Eug- unstablethroughout most of pegmatitecrystalliza- ster. 1969). tion. Because montebrasiteoccurs near or with The initial cation metasomatismof primary and spodumene+ quartzin thesepegmatites, the alter- secondary montebrasite involved Ca-for-Li ex- ation of montebrasiteto brazilianiteprobably took change, producing hydroxyl-apatite and minor place only after spodumenehad been locally de- crandallite.Subsequent conversion of montebrasite stroyed(i.e., convertedto albite and/ormuscovite). to muscovitewas far more extensiveand reflectsa The alterationsequence of montebrasitefrom the high mobility of Li, K, Si, and P, but not Al. White Picachopegmatites has few known parallels In spite of their rarity in these pegmatites,the in other districts. Extensive replacementof the occurrencesof scorzalite,kulanite, wyllieite, auge- primary phaseby low-fluorinesecondary montebra- lite, and brazilianiteoffer additionalinsight into the site has been documented only from the Tanco chemistry of the postmagmaticfluids. Except for pegmatite,Manitoba (eernii et al., 1972),but this early formed schorl near the pegmatite borders, alterationprobably is more common than is pres- scorzalite,kulanite, and wyllieite are the only Fe- ently recognized. Products of Ca-for-Li ion ex- rich phasesin the White Picachopegmatites. The change (e.g., foggite, crandallite, and especially rarity and volumetric insignificanceof these three apatite)have been reported from numerouslocal- phosphatesemphasizes the extremedepletion in Fe ities (e.9., Ginzburg, 1950;Fisher and Runner, (and enrichmentin Mn) found in thesepegmatites. 1958;Gallagher, 19671' Moore et a|.,1975),but the The compositionof White Picachowyllieite is nota- extensivereplacement of White Picachomontebra- ble in that the weight ratio MnO:FeO is nearly 3:1 site by apatite * muscoviteand by pure muscovite (analyses69-:72, Appendix l), implying that the appearsto be unique. Late-stagerims of colloform M(1) and M(2a) sites are occupied predominantly gray carbonate-apatitesurrounding amblygonite- by Mn (c/. Moore and Ito, 1979).Kulanite was montebrasitemay be commonin pegmatitedistricts identifiedin alteredmontebrasite from the Midnight other than the White Picacho (Moore. 1973).but Owl pegmatiteby a comparisonof properties re- they are not well documented. ported by Mandarino and Sturman (1976)for kulanite and by Moore et al. (1973) for bjarebyite, Alteration of lithiophilite (Ba,Sr)(Mn2*,Fe2*,Mg;2A12(PO4)3(OH)3,from the The incipient alteration of lithiophilite coupled Palermo#1 pegmatite,New Hampshire. As in its hydration with removal of lithium to form hureau- type occurrence, kulanite from the White Picacho lite and triploidite. Thesesecondary phosphates are pegmatitescontains no Sr, and this fact may indi- abundant in most of the pegmatitesthat contain cate that the White Picacho pegmatites were rela- lithiophilite,and most subsequentalteration of lith- tively depletedin Sr (as comparedto the phospho- iophilite apparentlyinvolved replacementof these rus-rich Palermo #l pegmatite, which contains two phases.Following the formation of hureaulite bjarebyite, goyazite,and other Sr-bearingphases: and triploidite, alteration of lithiophilite nodules Kampf, 1981).The general absenceof aluminum essentiallyinvolved progressiveCa-for-Mn cation phosphatessuch as augelite, trolleite, etc., in the exchange(Fig. 15).Dickinsonite, fillowite, and wyl- White Picachopegmatites may be explainedby the lieite occur only in lithiophilite nodulesembedded fact that the postmagmatic fluids were relatively in albite. Formation of these three phasesreflects rich in K, Ca, and Si, thus favoring formation of an increasinSaguz*fo*rz* in the aqueousfluid, as apatite * muscoviteover augelite.Of the five rare well as a locally high activity of Na*. In the same secondaryphosphates noted above, brazilianite is way that brazilianite is chemically analogous to the most abundant.Its occurrencelate in the mon- lacroixite, dickinsonite,fillowite, and wyllieite can tebrasitealteration sequence suggests that its stabil- be consideredanalogous to natrophilite,NaMnPOa ity is analogousto that of lacroixite, NaAIPO+ (i.e., they are principally Na-Mn phosphates).Na- (F,OH). Lacroixite is incompatilewith lithium alu- trophilite, like lacroixite, is incompatiblewith spo- minosilicates in the presence of quartz (Burt and dumenein the presenceof quartz (Burt and Lon- London, 1978; London and Burt, 1981b).In the don. 1978:London and Burt. l98l) and is absentin White Picachopegmatites, the presenceof spodu- the White Picacho pegmatites.In its only known mene with albite and quartz probably buffered the occurrenceat Branchville,Connecticut (Brush and a14r*layi*ratio in the co-existing fluid at values Dana, 1890),natrophilite appearsto be the direct sufficiently low that brazilianite, like lacroixite, was product of Na-for-Li exchangein lithiophilite, and LONDON AND BURT: SPODUMENE, MONTEBRASITE AND LITHIOPHILITE

tions suffers from their extremely limited occur- apatlre (I3) rences. The subsequentformation of fairfieldite, robertsite, and apatite implies that ag^z+laMnz*irl- creasedthroughout the alteration episode. Chlorine-rich apatites produced at this stage of alteration further reflect that postmagmaticfluids in alrfleldlte (5) these pegmatiteswere F-deficient. Although pri-

lte (3) mary apatite containing 5-7 wt.% MnO is common in theseand other pegmatites,the secondaryapatite in this associationcontains almost 14 wt'%oMnO and is the most Mn-rich apatite known to us. The flllowtte (4) of these apatites can reasonably /oplckineonrte {e) high Mn contents be attributed to the locally high activity of Mn2* lploidite (6) (liberated of hureaulite, triploi- aod eosphoitte (4) by the breakdown dite, etc.). The highly oxidized phases, such as Fig. 15. Compositionsof lithiophilite replacementminerals (in wt.Vo) on the ternary CaO{MnO + Mn2O3)-(FeO + FezOr). purpurite, stewartite, strengite, etc., reported by Numbers in parentheses indicate the number of specimens Jahns (1952) appear to be weathering products. analyzed(by electron microprobe). Representativeanalyses are Sicklerite at the White Ridge pegmatite also may be given in Appendix l. the result of weathering or of near-sutfacereactions with meteoric water, inasmuchas spodumeneand the structuralsimilarities between these two miner- montebrasitefrom this pegmatitedo not exhibit the als seems to confirm this (Moore, 1972).In the metasomaticalteration observed in the other peg- White Picacho pegmatites,it appearsthat the pres- matites. ence of spodumene with albite and quartz kept The metasomaticreplacement assemblages iden- values of a5i,*/cs;* in the fluid sufficiently low that tified in the White Picacho lithiophilite are most lithiophilite was not converted to natrophilite. By similarto thosedescribed by Brush and Dana (1878, the time the ratio ayuJayy had increasedsufficient- 1879a,1879b,1890) from the Branchvillepegmatite, ly to promote the growth of Na-rich phosphates Connecticut; however, the sequenceof replace- (i.e., afterspodumene had beendestroyed by albiti- mentsin Branchville lithiophilite has not been clear- zation), most lithiophilite probably had been con- ly established,largely because the replacement verted to hureaulite + triploidite, and the activity of products are relatively coarse grained and occur as Caz* may have increasedin the fluid. This circum- irregularly distributed patches in nodules rather stanceapparently led to the formation of dickinson- than in the discrete veinlets often observedin the ite, fillowite, and wyllieite from hureaulite and White Picacho material. These suites of Mn-rich triploidite. secondaryphosphates reflect the high MnO/(MnO Fillowite, dickinsonite and wyllieite are usually +FeO) ratios of the primary lithiophilite at these intimately intergrown, and this fact suggeststhat localities.They are similar to the productsof meta- they formed contemporaneously.In the White Pica- somaticallyaltered lithiophilite from other localities cho material, these three phases possess sharp (e.g., Matsubaraand Kato, 1980),except that the crystal boundariesand appearto be optically homo- higher Fe content of primary lithiophilite in most geneous;however in somerespects, the distinction other pegmatitesleads to a somewhatdifferent suite between phases is unclear. For example, some of alteration minerals that contain essential Fe3* materialthat possessesthe optical properties and (seeMoore, 1973, for a list of secondaryphosphates composition(analyses 2l-24, Appendix 1) of dick- containingFe'*). insonite apparently has the unit cell (both symmetry Na- and Ca-rich phosphates similar to those and dimensions) of wyllieite. Fisher (1965) has identified in the White Picacho district are common demonstratedthat a complex paragenetic relation- products of metasomaticalteration of lithiophilite- ship exists among these minerals that depends on triphylite. Moore (1971)has identified large masses temperature and asr6. There is little doubt that the of alluaudite that replace triphylite in several peg- crystal-chemicalrelations among these three phases matitesof the Black Hills district, SouthDakota. In are very complex (Moore and Ito, 1979;Moore, an occurrence similar to that described from the 1981),and an understandingof their stability rela- White Picacho district, Moore (1973)also reported LONDON AND BURT: SPODUMENE. MO NTEBRA SITE AND LITH I OPH I LITE

an occurrenceoflarge, partially replacedcrystals of volatiles P and to a lesser extent F increasedas triphylitefrom the PleasantValley pegmatite,South crystallization proceeded, and that spodumene Dakota, in which relict cores of triphylite were eventuallybecame unstable with respectto monte- surroundedby a thick massof secondaryalluaudite, brasite * quartz: which is finally rimmed by secondaryapatite. LiAlsi2o6 + + H2o + HF] Sickleriteis a common product of oxidation and [P2o5 spodumene componentsof fluid alkali leachingin Fe-rich lithiophilite and triphylite, : LiAIPO4(OH,F) + 2SiO2 as at the Stewartpegmatite, California (Shigley and montebrasite quartz Brown, 1980),and the Varutriisk pegmatite,Swe- den (Quensel,1957). Sicklerite occurs sparinglyat The composition of primary montebrasiteindi- the White Ridge pegmatite in the White Picacho catesthat the White Picachopegmatites were not F- district; lithiophilite from the White Ridge also is nch (cf. Loh and Wise, 1976). Although Jahns muchmore Fe-richthan at the other localitiesin the (1952)reported F-rich minerals such as lepidolite, district (analysis 3, Appendix l), a factor that triplite, and fluorite from the White Picachodistrict, undoubtedlyfacilitated its oxidationand conversion lepidolite is not abundant, and triplite and fluorite to sicklerite.As at most other lithiophilite-triphylite must be very rare (the latter two phases were localities,highly oxidized and alkalileachedphases soughtbut not found during this study). Topaz was such as purpurite, stewartite, and strengite are not observedby Jahnsor by us. common in White Picacho lithiophilite (Jahns, The abundanceof minerals containing Li, Be, 1952);however, as noted above, these phosphates Mn, Nb, Ta, and Bi in thesepegmatites implies that and sicklerite appear to have been formed by they originated from a highly differentiated source weathering. magma.The very high MnO/(MnO + FeO) ratio in lithiophilite and other Fe-Mn minerals (columbite Geochemicaltrends and tourmaline)with which it occurs indicatesthat the White Picacho pegmatiteswere highly differen- Crystallization of the primary lithium minerals tiated with respect to Mn ys. Fe when they were The primary textural and zonal features observed emplaced. Foord (1976) similarly observed that in the White Picacho pegmatites are typical of garnets and tourmalines from the Mesa Grande steeply dipping, zoned lithium pegmatites else- district, California, show increasingenrichment in where. Nevertheless,a few details are worthy of Mn vs. Fe at successivelylater stagesof pegmatite emphasis.Primary montebrasite(or amblygonite) development.The sudden and late appearanceof and lithiophilite (or triphylite) are commonly em- Mn-rich minerals in the White Picacho pegmatites bedded in massive quartz in the White Picacho remainsunexplained. pegmatites,as at other well-known, P-rich pegma- Relatively coarse-grained cleavelandite units tites such as the Palermo #l pegmatite, New containing lithiophilite, columbite, beryl, apatite, Hampshire,and the Newry pegmatite,Maine (Cam- zircon, tourmaline, and some mica are present in eronet al., 1954).Inthe White Picachopegmatites, most of the White Picacho lithium pegmatites. the primary lithium phosphates appear to have Cleavelandite units with similar mineralogy and crystallized after spodumeneand primary feldspars, texture have been described from at least fiftv but this relationshiphas not been demonstratedin pegmatitesin the UnitedStates alone (e.g., by Page pegmatites from most other districts (c/. zonal et al., 1953;Cameron et al., 1954;Norton et al., assemblages6 through 1l in the generalizedcrystal- 1962;Jahns and Ewing, 1976),and summariesof lization sequenceproposed by Cameron et al., pegmatite occuffences elsewhere (e.g., Vlasov, 1949).These observationsimply that successive, 196l; Beus, 1960) indicate that these units are fluid-saturated magmas in the White Picacho lithi- present in highly differentiated lithium pegmatites um pegmatitesbecame increasinglyrich in phos- throughout the world. The origin of cleavelandite phorus,and phosphates,rather than silicates,were bodieshas been a subjectof considerablespecula- the last primary minerals to crystallize with qtrartz. tion. For example, their formation may be ex- Thus montebrasite and lithiophilite apparently suc- plained by the separation of a volatile- and silica- ceeded spodumene as the stable primary lithium rich fluid from primary magma, in a model of phases in the White Picacho pegmatites. This se- differentiation similar to that proposedby Jahnsand quence implies that the activities of the acidic Tuttle (1963)and by Osieki and Jahns(1980) for the LONDON AND BT]RT: SPODUMENE, MONTEBRASITE AND LITHIOPHILITE 109 segregationof large K-feldspar massesfrom finer- Ewing, 1976),textural evidence usually does not grained albite. In this context, cleavelandite units provide sufficientproof of an origin by replacement. may be analogousto sodic aplite from which a K-, In the White Picacho pegmatites' however, textural Li-, and Si-rich fluid separatedbut did not escape features (described above) indicate that at least the closed pegmatitesystem. These cleavelandite some parts of cleavelandite or cleavelandite-mica units may have crystallizedfrom essentiallyalbitic complexesreplaced quartz-spodumenepegmatite' melts that also might be enriched in less volatile, It is not clear if spodumenereacted with residual residualelements such as Ca and Mn; such melts magmaor aqueousfluids, but the formation of these also would be enriched in the elementsF, P, Zr, complexes seems to represent the last stage of Nb, and LREE that form especially stablecomplex- coarse-grained crystallization that preceded the es with alkalis and alkaline earths (e.g., see Ko- fine-grained,postmagmatic pseudomorphism of the garkoet al.,1968; Fleischerand Altschuler,1969; primary lithium minerals. Watson, 1979).Alternatively, cleavelanditecom- Postmagmatic plexesmay be producedby reactionof solid phases fluids (especiallyspodumene * quartz) with residual mag- The initial subsolidus metasomatism of lithium ma (magmatic replacement) or with residual fluids minerals in these pegmatitestook place in an alka- (postmagmaticreplacement). Although albite com- line environment(Fig. 16),as evidencedby Na-for- plexes clearly replaced pre-existing pegmatite at a Li exchange(albitization) in spodumeneand Ca-for- few localities(e.g., Norton et al',1962; Jahnsand Li and Ca-for-Mn exchangein the phosphateg:This

orDEsr TrME Yo1'NGEST SEQUENCE --+ ---) MONTEBMSITE_>LCI^I-FLUORINE .-> CMNDALLITE--> HYDRO)ffLAPATITE MUSCO1IITE CARBONATE-APATITE MONTEBRASITE + MUSCOVITE + BRAZILIANITE

--) LITITIOPHILITE -->TRIPLOIDITE -+ Ca-HUREAULITE --)FAIRFIEIDITE -----> Ih-CHLOMIATITE RITODOCITROSITE ++t HUREAULITE DICKINSONITE ROBERTSITE

FILLq,'ITE

SPODIJMENE EUCRY?TITE MUSCOVITE MUSCOVITE + AIAITE A]AITE

PROCESS

OII-F EXCHANGE -----)

Lt-LEACI{ING

Ca-METASOMATISM

Na-METASOMATISM K + H METASOMATISM

REPLACEMENT OF PIIOSPIIATE BY CARBONATE

A].KALINE METASOMATISM ACIDIC METASOMATISM (ALBITIZATION) (WEAKGRXISENIZATION OR SERICITIZATION)

processes Fig. 16. Generalizedreplacement sequences for spodumene,montebrasite, and lithiophilite, and the inferred chemical that give rise to the observed alteration assemblages.The replacement sequencesfor montebrasite and lithiophilite have been abbreviatedto show the essentialmineralogy. The assemblagesare laid out along the time axis in the order and relative timing in which they seem to occur. The beginning of acidic metasomatismis marked by the appearanceof muscovite in spodumeneand montebrasite. ll0 LONDON AND BURT: SPODUMENE, MONTEBRASITE AND LITHIOPHILITE

observation implies that the fluid chemistry the exocontactsof these pegmatites.The lithium changedsignificantly from the later primary stage, amphiboleholmquistite, common in some districts when acidic P- and F-rich fluids promoted the (reviewed by Heinrich , 1965),appears to be absent formation of montebrasite + quartz, to subsolidus in the White Picacho district. conditions sufficiently alkaline to stabilize eucryptite + albite (cf. the stabilities of eucryptite, spodu- Summary and conclusions mene, and amblygonitein Fig. 14 of London and As in most other pegmatite districts, the vast Burt, 1981b).The sodium phosphatesbrazilianite, majority of pegmatites exposed in the White pica- fillowite, wyllieite, and dickinsonite occur late in cho district are granitic in composition, poorly the alkaline stage of metasomatic alteration, and zoned, and do not contain lithium minerals. The these phases probably formed only after spodu- nine lithium pegmatites that have been recognized menehad been convertedto albite and/or micas. in the district are complexly zoned and contain The widespread formation of secondary mica in mineralsof the elementsLi, Be, P, Nb, Ta, Bi, and spodumene,montebrasite, tourmaline, and much of probably other rare elements.Spodumene, monte- the feldspar (especially in oligoclase of the granitic brasite,lithiophilite, lithian muscovite, and lithian zones) signifies a gradual change from alkaline to tourmaline are the primary lithium minerals found relatively acidic postmagmaticfluids, as (K+H)- in thesepegmatites; however, only spodumeneand metasomatismproduced alteration characteristic of montebrasiteoccur in abundance (Table 2). ln sericitization or weak greisenization. Late-stage general, the lithium minerals are restricted to the fluids were not F-rich, however, as evidencedby inner zonesand coresof essentiallygranitic pegma- the low F contentsof secondarymontebrasite, by tite. the formation of OH- and Cl-apatites, and by the If crystallization proceeded sequentially from absenceof F-rich phases. borders to core, then the primary crystallization Fluid compositionprobably varied with location sequencein these pegmatites reflects the usual (as well as time) in these pegmatites.The large, increase in activities of Li species (spodumene sharply bounded pegmatitic zones are essentially follows perthitic microcline as the stable alkali mono- or bi-mineralic, and with respect to the fluid aluminosilicate),as well as an increasein the activi- phase, each zone may have constituted its own tieS of the acidic volatiles P and F (montebrasite chemicalsystem. As a result of this isolation,much succeedsspodumene as the stable primary lithium of the observedpseudomorphic alteration may have phase). Coarse-grainedcleavelandite units, com- been produced as (upwardly) migrating fluids at- monly containingmajor or accessorylithiophilite, tempted to re-equilibrate with different solid-phase apatite,columbite, beryl, and zircon,occur in all of assemblagesfrom zone to zone. The movementof the lithium pegmatites and are located near the fluids through the crystallized pegmatite appearsto marginsof quartz-spodumenezones. These appear have been controlled by fracturing, with the result to be the latest-crystallizingcoarse-grained units, that fluids almost certainly moved across zonal and they formed at least in part by replacementof boundariesas channelwaysopened up. As an exam- quartz-spodumenepegmatite. ple, the albitizationof spodumeneobserved in these Subsolidus metasomatic alteration of spodu- and other pegmatitesmay have been produced as mene, montebrasite,and lithiophilite was pseudo- Na-rich fluids escapedalong fractures from cleave- morphic in nature and produced a large number of landite complexes into (overlying) quartz-spodu- fine-grainedsecondary phasesand replacement as- menepegmatite. The sporadicdistribution and the semblages.As summarizedin Figure 16,the subsol- variations in observed alteration assemblagesare idus replacementof the primary lithium minerals in further confirmation that the fluid composition was thesepegmatites can be divided into two stages:an non-uniform with location (and with time). alkaline stage (principally involving albitization), Lithium was removedfrom the pegmatitesduring followed by an acidic stage (with alteration charac- the alterationepisode. Although someLi was incor- teristic of greisenizationor sericitization).The in- porated in the fine-grained secondary micas, spec- cipient alteration of these primary lithium minerals troscopic evidence indicates that much Li (perhaps involved cation exchangewith a fluid rich in Na and liberated by the pseudomorphic replacementof the Ca relative to Li. This initial stage of alteration primary lithium minerals) was trapped in large seemsto indicatethat in someinterval betweenlate massesof metasomaticmicas and tourmalinesat primary crystallization and the onset of postmag- LONDON AND BURT: SPODUMENE, MONTEBRASITE AND LITHIOPHILITE matic alteration, the pegmatitic fluid possibly Acknowledgments evolved from a relatively acidic phase (rich in We are especially grateful to Sidney B. Anderson of Mesa, volatile P- and F-species)to an F-deficient but Arizona, for unlimited access to his pegmatite claims in the Daniel E' presumably saline fluid dominated (chemically) by White Picacho district. We would like to thank EugeneJarosewich, Joseph A. Nelen, and John S' and alkaline earths. Albitization of spodu- Appleman, alkalis White, Jr. for the use of microprobefacilities in the Division of mene may have been contemporaneouswith Ca- Mineralogyof the SmithsonianInstitution, Washington,D' C' metasomatismof the phosphates,but the formation We also thank CharlesF. Lewis, Centerfor MeteoriteStudies, of the Na-rich phosphatesbrazilianite, fillowite, Arizona State University, for determining the carbon content of R. Holloway, dickinsonite,and wyllieite probably occurred only carbonate-apatitesamples. John M. Ferry, John PaulB. Moore, and David R. Veblenare to be spodumenehad been converted to JohnW. Larimer, after nearby thanked for their reviews of this manuscript' Finally, we thank albite or micas. The subsequentreplacement of Kathryn Gundersen,who typed the manuscrtpt. eucryptite, spodumene,and montebrasiteby mus- This work was funded in part by a National ScienceFounda- covite records the developmentof a weakly acidic tion Grant, #EAR-7814785,to Arizona StateUniversity' Addi- to the senior K- and H-rich fluid. This sericitic or greisen-like tional support in the form of academicscholarships were provided by Inspiration ConsolidatedCopper Com- is widespreadin tourmaline and in author alterationalso pany, Cities Service Corporation, and Phelps Dodge Corpora- much feldspar (mostly in plagioclaseof the outer, tion. granitic zones). Finally a more COz-rich fluid de- positedcarbonates (mostly rhodochrositein lithio- References philite) and carbonate-rich apatite as rims on mon- Albee, A. L. and Ray, L. (1970)Correction factors for electron tebrasitenodulqs. Similar rims of apatite surround- probe microanalysis of silicates, oxides, carbonates, phos- ing altered lithiophilite are very manganese-and phates,and sulfates.Analytical Chemistry,42 (12)' 1408-1414' chlorine-rich. Anderson,C. A., Blacet,P. M., Silver,L. T., andStern, T' W' The extremely high MnO/(MnO + FeO) ratios (1971)Revision of the Precambrian stratigraphy in the Pres- area, Yavapai County, Arizona. U.S. Geological observedin White Picacholithiophilite, columbite, cott-Jerome Survey Bulletin lf24-C, Cl-C16. e/c. suggestthat thesepegmatites originated from a Appleman, D. E. and Evans, H. T., Jr. (1973)Indexing and highly differentiated granitic source. The pegma- least-squaresrefinement of powder difraction datei' U'S' tites introduced Li, K, and B to the metamorphic Geological Survey Computer Contribution 20 (Available file no' rocks, producingmasses and veins of metasomatic through National Technical Information Service, usGs-GD-73-003). micasand tourmalineat the pegmatiteexocontacts. Beus, A. A. (1960)Geochemistry of Beryllium and Genetic All of the mineralsand most of the assemblages Typesof BerylliumDeposits. Akademii Nauk SSSR'Moscow identifiedin the White Picachopegmatites occur in (transl.W. H. Freeman,San Francisco,1967). complex lithium pegmatites elsewhere. In fact, Brush, G. J. and Dana, E. S. (1878)On a new and remarkable with a many of the textures and assemblagesin these mineral locality in Fairfield County, Connecticut; several new speciesoccurring there. American pegmatitesare characteristic of lithium pegmatites. description of Journal of Science, 116,3346 and ll4-123- However, we know of no other pegmatite or group Brush, G. J. and Dana, E. S. (1879a)On the minerallocality in of pegmatiteswhere the history of subsolidusmeta- Fairfield County, Connecticut. Secondpaper. American Jour- somatisminvolving cation exchangebetween crys- nal of Science,l17, 359-368. in tals and fluid is so complex, yet decipherable'To Brush,G. J. and Dana, E. S. (1879b)On the minerallocality County, Connecticut. Third paper. American Journal encouragefurther studiesof these and other pegma- Fairfield of Science,l18. 45-50. tites, we have made representativesamples from Brush, G. J. and Dana, E. S. (1880)On the minerallocality at the White Picacho pegmatitesavailable for study Branchville, Connecticut. Fourth paper. Spodumeneand the from the museumslisted below: results of its alteration. American Journal of Science, ll8, Departmentof Geology,Arizona StateUniversi- 257-285. Dana, E. S. (1890)On the minerallocality at Arizona 85281; collection numbers Brush, G. J. and ty, Tempe, Branchville, Connecticut. Fifth paper. Analyses of several 3288-3303. manganesianphosphates. American Journal of Science, 139' U.S. National Museum, Division of Mineralogy, 20lJl6. SmithsonianInstitution, Washington, District of Burt, D. M., London, D., and Smith, M. R' (1977)Eucryptite phase diagram' Columbia 20560; collection numbers 145123- from Arizona, and the lithium aluminosilicate (abstr.) Geological Society of America, Abstracts with Pro- 145130. grams,9(7),917. MineralogicalMuseum, Harvard University, 24 Cameron,E. N., Jahns,R. H., McNair, A' H.' and Page,L' R" Oxford Street, Cambridge, Massachusetts02138; (1949) Internal structure of granitic pegmatites. Economrc collectionnumbers ll7 233-117244. Geology MonograPh2, ll5 P. l12 LONDON AND BURT: SPODUMENE, MONTEBRASITE AND LITHIOPHILITE

Cameron,E. N., and others (1954)Pegmatite investigations, Korzhinskii, D. S. (1957)Physicochemical Basis of the Analysis 1942-45,New England. U.S. Geological Survey professional of the Paragenesisof Minerals. Izvestiya Akademii Nauk Paper 255. SSSR, Moscow (transl. Consultants Bureau, New York, Cern6, I., Cernf, P., and Ferguson,R. B. (1972)The Tanco r9s9). pegmatiteat Bernic Lake, Manitoba. III. Amblygonite-monte- Laughlin, A. W. (1969)Excess Radiogenic Argon in Pegmatite brasite. CanadianMineralogist, ll (3), 643-659. Minerals.Ph.D. thesis,University of Arizona,Tucson, Arizo- Cern6,I., Cernf, P., and Ferguson,R. B. (1973)The fluorine na. cont€nt and some physical properties of the amblygonite- Loh, S. E., and Wise, W. S. (1976)Synthesis and fluorine- montebrasiteminerals. American Mineralogist, 58, 291-301. hydroxyl exchangein the amblygoniteseries. Canadian Miner- Fisher, D. J. (1965) Dickinsonites, fillowite. and alluaudites. alogist,14,357-363. American Mineralogist, 50, lU7-1669. London, D. (1979)Occurrence and Alteration of Lithium Miner- Fisher, D. J. and Runner, J. J. (1958)Morinite from the Black als, White Picacho Pegmatites,Arizona. Masters thesis, Ari- Hills. American Mineralogist, 54, 585-594. zona State University, Tempe, Aizona. Fleischer, M. (1980)1980 Glossary of Mineral Species.Mineral- London,D., Bandy,G. L., and Kealy, S. L. (1978)Metasomatic ogicalRecord, Inc., P.O. Box 35565,Tucson, Arizona. alteration of primary lithium phosphates from the White Fleischer, M. and Altschuler, Z. S. (1969)The relationship of Picacho pegmatite district, Maricopa and Yavapai Counties, rare-earth composition of minerals to geologic environment. Arizona. (abstr.) Geological Society of America, Abstracts Geochimicaet CosmochimicaActa, 33 (6),725J32. with Programs,l0 (7), 446. Foord, E. E. (1976) Mineralogy and petrogenesis of layered London, D. and Burt, D. M. (1978)Lithium pegmatitesof the pegmatite-aplitedikes in the Mesa Grande district, San Diego White Picacho district, Maricopa and Yavapai Counties, Ari- County, California.Ph.D. thesis,Stanford University, Stan- zona.In Burt, D. M. andP6w€, T. L., Eds., Guidebookto the ford. California. Geofogy ofCentral Arizona. Aizona Bureau ofGeology and Gallagher,M. J. (1967)Phosphates and other mineralsin pegma- MineralTechnology, Special Paper 2,6l-72. tites of Rhodesia and Uganda. Mineralogical Magazine, 36, London, D. and Burt, D. M. (1979)Processes of formation and 50-59. destructionof eucryptite in lithium pegmarites.(abstr.) Trans- Ginzburg, A. I. (1950)Montebrasite and its replacement reac- actionsof the AmericanGeophysical Union, 60 (18),417. tions [in Russian]. Trudy MineralogicheskogoMuzeya, Aka- London, D. and Burt, D. M. (1980) Local lowering of silica demii Nauk SSSR,2, 72-85 (not seen; extracted from Chemi- activity during albitization of spodumene-its relation to the cal Abstracts, 49, #12207, 1955). formation ofmicas and eucryptite. (abstr.) Transactionsofthe Graham, J. (1975)Some notes on o-spodumene,LiAlSi2Ou. AmericanGeophysical Union, 61 (17),404. American Mineralogist, 60, 919-923. London, D. and Burt, D. M. (l98la) Lithium-aluminosilicate Heinrich, E. W. (1965) Holmquistite and pegmatitic lithium occurrences in pegmatites, and the lithium aluminosilicate exomorphism.Indian Mineralogist,6, l-13. phasediagram. American Mineralogist, in press. Heinrich, K. F. J., Yakowitz, H., and Vieth, D. L. (1972)The London, D. and Burt, D. M. (l98lb) Chemical models for correction for absorption of primary X-rays. Proceedingsof lithium aluminosilicate stabilities in pegmatitesand granites. the 7th Annual Meeting of the Electron Microprobe Analysts American Mineralogist, in press. Societyof America, San Francisco,3A-3F. Mandarino,J. A. and Sturman, B. D. (1976)Kulanite. a new Jagger,T. A. and Palache, C. (1905) Descriprion of the Brad- barium iron aluminum phosphatefrom the Yukon Territory, shawMountains quadrangle,Arizona. U.S. GeologicalSurvey Canada.Canadian Mineralogist, 14, 127-131. Geologic Atlas, Folio 126. Matsubara, S. and Kato, A. (1980)Pegmatite phosphates from picacho Jahns, R. H. (1952) Pegmatite deposits of the White Yukiiri, Ibaraki Prefecture,Japan [in Japanese].Kobutsugaku district, Maricopa and Yavapai Counties, Arizona. Arizona Zasshi, 14 (4),269-286 (not seen; extracted from Chemical Bureau of Mines Bulletin 162. Abstracts, 94, #33833t, l98l). Jahns,R. H. and Ewing, R; C. (1976)The Harding Mine, Taos Moore, P. B. (1971)Crystal chemistry of the alluaudite srnrcture County, New Mexico. New Mexico GeologicalSociety Guide- type; contribution to the paragenesisof pegmatite phosphate book, 27th Field Conference,Vermejo Park,263-276. giant crystals. American Mineralogist, 56, 1955-1975. Jahns,R. H. and Tuttle, O. F. ( 1963)Layered pegmatite-aplite Moore, P. B. (1972) Natrophilite, NaMnPOa, has ordered cat- intrusives. Mineralogical Society of America SpecialPaper l, ions. American Mineralogist, 57, 1333-1344. 78-92. Moore, P. B. (1973)Pegmatite phosphates: descriptive mineral- Julien, A. A. (1879) On spodumene and its alrerarions from ogy and crystalchemistry. Mineralogical Record,4, 103-130. granite-veinsof Hampshire County, Massachusetts.Annals of Moore, P. B. (1974)I. Jahnsite, segelerite,and robertsite, three the New York Academyof Science,l, 318-359. new transition metal phosphate species. II. Redefinition of Kampf, A. R. (1981)The phosphatemineralogy of the palermo overite, an isotype of segelerite. III. Isotopy of robertsite, pegmatite.(abstr.) Fourth Annual Meeting of the Mineralogi- mitridatite, and arseniosiderite. American Mineralogist, 59, cal Societyof America-Friends of Mineralogy Symposiumon 48-59 and 640. the Mineralogy of Pegmatites,February 15-16, 1981,Tucson, Moore, P. B. (l9El) The primary pegmatitephosphate minerals. Aizona, 19-20. (abstr.)Fourth Annual Meeting of the Mineralogical Society of Kogarko, L. N., Krigman, L. D., and Sharudilo,N. S. (1963) America-Friends of Mineralogy Symposium on the Mineral- Experimental investigations of the afect of alkalinity of sili- ogy of Pegmatites,February 15-16, 1981, Tucson, Aizona, cate melts on the separation of fluorine into the gas phase. l7-18. Geokhimiya, (E),948-956 (transl. GeochemistryInternational, Moore, P. B., Irving, A. J., and Kampf, A. R. (1975)Foggite, s,782J%, 1968\. CaAI(OH)2(HrO)[PO+]; goedkenite (Sr,Ca)zAl(OH)[pO+]z; LONDON AND BT]RT:SPODUMENE. MONTEBRASITE AND LITHIOPHILITE |l3

and samuelsonite, (Ca,Ba)Fe3+Mn?+CasAlz(OH)zlPO+lo: Shigley, J. E. and Brown, G. E., Jr. (1980)Phosphate mineral- three new species from the Palermo No.l pegmatite, North ogy of the Stewart pegmatite,Pala district, SanDiego County, Groton, New Hampshire. American Mineralogist, ffi, 957- California. (abstr.) Geological Society of America, Abstracts 9&, with Programs,l2 (7), 521. Moore, P. B. and Ito, J. (1979) Alluaudites, wyllieites, and Thomssen,R. W. and Anthony, J. W. (1977)Lithiophilite crys- arrojadites: crystal chemistry and nomenclature. Mineralogi- tafs from the Foote mine. Mineralogical Record, E (2),95-97. cal Magazine, 4t, 227-235. Vlasov, K. A. (1961)Principles of classifying granite pegmatites Moore, P. B., Lund, D. H., and Keester, K. L. (1973)Bjare- and their textural-paragenetictypes. Izvestiya Akademii Nauk byite, (Ba,Sr)(Mn,Fe,Mg)zAl2(OH)3eO4L,a new species. SSSR, Seriya Geologicheskaya,l, 5-20 (transl. Izvestiya of Mineralogical Record, November-December, 282-2E5. the Academy of Sciences,USSR, GeologicSeries, 1,5-20, Munoz, J. L. and Eugster, H. P. (1969)Experimental control of l%l). fluorine reaction$in hydrothermal systems. American Miner- von Knorring, O. (1963)VI(c) Report on mineralogicalresearch. alogist, 54, 943-959. 7th Annual Report (1961-1962)Research Institute of African Osieki, R. A. and Jahns, R. H. (1980)Feldspar zonation within Geology, University of Leeds, 33-37. pegmatite masses of the Milford-Mason, New Hampshire Mineralogischeund PetrographischeMitteilungen, 55, 9-l 8. granite.(abstr.) GeologicalSociety of America, Abstracts with von Knorring, O. and Fransolet, A. M. (1975)An occurrenceof Programs,12 (2),76. bjarebyite in the Buranga pegmatite,Rwanda. Schweizerische Page,L. R. and others (1953)Pegmatite investigations, 194245, mineralogischeund petrographischeMitteilungen, 55' 9-18. New England, U.S. GeologicalSurvey ProfessionalPaper 247. Watson, E. B, (1979)Zircon saturation in felsic liquids: experi- Palache,C., Berman,H., and Frondel,C. (1951)The Systemof mentalresults and applicationsto trace elementgeochemistry. Mineralogy, Volume 2, 7th Edition. Wiley, New York. Contributions to Mineralogy and Petrology, 70 (4), 407420. Quensel, P. (1937)Minerals of the Varutriisk pegmatite. VIII. The amblygonite group. Geologiska F