Download the Scanned
Total Page:16
File Type:pdf, Size:1020Kb
American Mineralogist, Volume 70, pages 379-387, 1985 Ma_nganesehumites and leucophoenicitesfrom Franklin and Sterling- Hill' NewJersev: 'i"? andimplications ;ifi':il1,;;lfiil11"r's' Perr J. Dullx Department of Mineral Sciences Smithsonian lnstitution, Washington, D. C. 20560 Abstract The manganesehumites, (alleghanyite, manganhumite, and sonolite),together with some Mn-bearing samplesof the Mg-humites,and the related phasesleucophoenicite and jerry- gibbsite,from the orebodiesat Franklin and SterlingHill, New Jersey,are describedtogether with analytical data. Solid solution betweenhumite and manganhumiteis at least partially continuous. Expected Mn/Mg solid solutions between alleghanyiteand chondrodite, and betweensonolite and clinohumite, are discontinuous; they are interrupted by apparently orderedphases. In all cases,the possibleorderings involve Zn as well as Mn and Mg. There are no Mn end-membersof the manganesehumites at this locality. Manganeseis apparently restricted in leucophoenicite(5.42-6.63 Mn per 7 octahedral cations) and in jerrygibbsite (7.79-8.02Mn per 9 octahedralcations). Calcium is common to both leucophoeniciteand jerrygibbsite,but among the Mn-humites,only sonoliteaccepts appreciable Ca (0.65Ca per 9 octahedralcations). There is a "threshold" level ofzinc in all studiedsamples; this "threshold" levelis a constantfor leucophoenicite1-9.3 Znper 3 Si)and alleghanyite(-O.2Zn per 2 Si). No samplesof leucophoeniciteor jerrygibbsite were found to be Zn-ftee,suggesting either that Zn is required for their stability, or that these two phasesmight not be stable as end-members.Fluorine is present in all the Mn-humites and is proportional to the Mg- content,but is absentin leucophoeniciteand jerrygibbsite. Introduction humite speciesoccur there; the Mg-humites occur in the The magnesiumhumite species(norbergite, chondrodite, host Franklin Marble for the most part, and the Mn- humite, and clinohumite) have been well-studiedand re- humites in the orebodiesthemselves. Franklin is also host cently summarizedby Ribbe (1982).Their manganeseana- to the two anomalous,yet related,species, jerrygibbsite and logues,however, have receivedless attention until recentlv. leucophoenicite,the latter occurringin relative abundance. They are, in general, less common, found in metamor- Second,Mn/Mg solid solution is common in the silicatesat phosed Mn-deposits, and may have lower p-T stability these deposits,and the activity of Fe is limited; in most rangesbecause of expansionof the structure by large Mn cases-it- has been preferentially absorbed by andradite, cations (Ribbe, 19g2j. Winter et al. (19g3) have rJcently which.also may serveto isolate silicate reactionsfrom the shownthat the Mn-humitesfrom Bald Knob. North carol_ ubiquitous franklinite. Third, the Mn-humite minerals are ina, require water-rich conditions in silica undersaturate4 moderately common at these deposits, occurring both rocks and that the formation of particular speciescan be within the primary ore and in secondaryvein assemblages. dependent upon complex phase relations involving Last, the Franklin and Sterling Hill assemblagesoffer a XIjI.2OI(XIH2O+ XCOri anA alSlOr;. In addition to thl y1q.". opportunity to study the behavior of zinc in the Mn isotypei of the humites, two other species, leu- Mn-humites, an area of investigationnot previously pur- cophoenicite and jerrygibbsite (Dunn et ut., tggi) a.e suedand which, as is shown below,is worthy of continued known, but the factors which govern their formation investigation, particularly from a structural standpoint. remain obscure.Crystal structureshave been determined The speciesdiscussed here are listed in Table 1, together for alleghanyite, magnesian alleghanyite, magnesian with related phases.The Mg-humites from the Franklin manganhumite,and sonolite,as cited below,but the deeree Marble, having already beenstudied by numerousinvesti- of solid solution betweenthe Mg- and Mn-humites is-in- gators, have beenexcluded from this study. Only samples completelydeflned, although Winter et al. (1983)provided with relationsto the ore mineralsare included. a partial plot of Mn/Mg miscibility, and Fukuoka (1981) Chemical composition provided compositionaldata for Japanesesamples. Franklin and SterlingHill provide an ideal "laboratory" The samples studied herein were chemically analyzed in which to investigatethese solid solution relations for a using an ARL-sEMeelectron microprobe utilizing an oper- number of reasons.First, all the known Mg- and Mn- ating voltage of 15 kV and a samplecurrent of 0.025pA, 0003-{04xl85/03O1-{379$02.00 379 3E0 DUNNI MANGANESE HUMITES AND LEUCOPHOENICITES Table l. The humite and leucophoenicitegroups overgrowthson the underlying alleghanyite,and as appar- ently randomly oriented euhedra,perhaps of a subsequent has beenfound on a l,r2*,si Humite grorJp fh-humitegroup Leucophoenicite growth period. Although alleghanyite few Franklin specimensfrom other assemblages(e.g., r.uNtt 3:l norbergj te unknownin nature unknown C6S84),most of the preservedpink-colored material from However,it is not clear wheth- 5i2 chondrodi te al 1eghanyi te unknown Franklin is leucophoenicite. er this is due to selectiveretention of the more attractive 7t3 humite mnganhumite I eucophoenici te bright pink leucophoeniciteby miners' casualcollecting or 9t4 cl inohumite sonol i te jerryg i bbsi te if this apparent predominanceof leucophoeniciteis due to geochemicalconditions, a viewpoint hereadopted. At Sterling Hill, alleghanyiteoccurs as euhedralcrystals measured on brass. The standards used were synthetic and thin seamswhich crosscutthe ore. Thesecrystals fre- tephroite(Mn,Si), synthetic ZnO (Zn), hornblende(MgFe), quently accompanyarsenate species in hydrothermal vein- and fluorapatite(F) for Mn-rich samples.Forsterite (Si,Mg) lets and formed in apparent equilibrium with rare species and rhodonite (Mn,Zn) were employed for Mg-dominant such as kolicite, holdenite,magnussonite, adelite, kraisslite, samples.The samplesare homogeneousat the microprobe Aside from theseuncommon level. The resultant analysesare presentedin Tables 2-6. chlorophoenicite,and others. with alleghanyite are All sampleswere individually verifiedby X-ray powder dif- arsenates,other speciesassociated and carbonates,all of second- fraction techniquesprior to analysis. franklinite, willemite,barite, ary recrystallization.The observedspecies are all formed Alleghanyite on willemitefranklinite ore which contains abundant cal- cite. The best sampleshave beenfound in recentyears. The Introduction ubiquitous twinning observedby Winter et al.' (1983)in Alleghanyite was first describedfrom the Bald Knob end-memberalleghanyite from Bald Knob was not ob- manganeseprospect, Alleghany County, North Carolina, servedeither in these magnesiancrystals, or in the more by Ross and Kerr (1932)who noted that it was an anhy- Mn-rich material from both Franklin and SterlingHill. drous manganesesilicate related to tephroite.A subsequent study of Bald Knob materialby Rogers(1935) found the Chemical composition mineral to contain HrO, redefined the species as Microprobe analysesof alleghanltes are presentedin Mnr(SiOn)r(OH)r, and noted the apparent relationship to Table 2. Examination of the data revealsseveral features. chondrodite. This isostructural relationship was further First is the absenceof material with end-membercompo- supportedby Campbell Smith et al. (1944)in their descrip- sition. Although near end-membermaterial is known from tion of alleghanyitefrom Rhiw, Wales. The crystal struc- Bald Knob (Simmonset al., 1981;$y'inter et al., 1983),it is ture of alleghanyitewas determinedby Rentzeperis(1970) unknown at Franklin and SterlingHill; most material from usingBald Knob material. there is highly magnesian.Second, samples from Franklin Alleghanyite,ideally Mnr(SiO4)2(OH)2,was flrst noted are markedly lower in Mg than those from Sterling Hill, from Franklin and Sterling Hill by Cook (1969) who reflectingthe higher generalconc€ntration of Mg in Ster- clearedup part of the confusionarising from earlier mor- ling Hill silicates.Third, consideringthe first 6 analyses, phological studies.White and Hyde (1982)examined sev- there appearsto be partial solid solution betweenthe most eral Franklin samplesusing TEM techniques.Although Mn-rich material and that with an Mn: Mg ratio of 4: I they found several phases associatedwith alleghanyite (the next flve analysesin the table),but there is an evident (chiefly leucophoeniciteand sonolite),these minerals were break in solid solution betweenmaterial with MnoMgt and presentas fragments,and not as intergrowths.Recently, a the rest of the samples.The possibleordering of somesam- high-Mg alleghanyitefrom Sterling Hill was describedby ples with a Mn:(Mg,Zn) ratio of 4:1 (8:2) might be per- Petersenet al. (1984),and a crystal structure refinementof missableinasmuch as alleghanyitehas spacegtoup PLtlb, Sterling Hill material of similar composition by Francis which has equipoint ranks of 4 and 2. Such material,how- (1985a)showed a high degreeofcation ordering.Extensive ever,has not yet beenrefined by crystalstructure analysis' analytical studies(present study) indicate that such order- The next 12 analysesin Table 2 representthe compo- ing may be very commonin samplesfrom SterlingHill. sitions of magnesianalleghanyite crystals from varied sec- ondary fissuresat Sterling Hill. In one samplewhich con- Description tained both massivematerial as a vein-filling,and euhedral Alleghanyiteoccurs at both Franklin and Sterling Hill. crystals