The Stability of Danalite, Feober(Sior3s Donero M. Bunr
Total Page:16
File Type:pdf, Size:1020Kb
American Mineralogist, Volume 65, pages 355-360, 1980 The stability of danalite, FeoBer(SiOr3S DoNero M. Bunr Department of Geology, Arizona State University Tempe,Arizona 85281 Abstract End-memberdanalite is compositionallyequivalent to a mixture of pyrrhotite, phenakite, and fayalite, but danalite's natural occurrenceswith pyrite, magnetite,and quartz suggest that it is stable under somewhat more sulfidizing conditions than pyrrhotite and more oxidiz- ing conditions than fayatte. Danalite tends not to occur with hematite. Thesefacts can be usedto constructa tentative log /",-log f o,diagtan for the stability of the end member. The diagram shows that danalite, if it is stable at all, has an extremely narrow stability field cen- tered near the pyrrhotite,/magnetitefield boundary. Outside this field, phenakite (or, at low temperatures,bertrandite) is stablewith magnetite,pyrrhotite (or pyrite), and quartz. A natu- ral example of theseequilibria apparently occurs at Iron Mountain, Bartlett, New Hamp- shire, where zoned danalite-helvite solid solutions occur as overgrowths on phenakite with magnetite,pyrite, and quartz. Danalite is the only end memberof the helvite group that is appreciablysensitive to hypo- gene oxidation. The Mn and Zn end,members helvite and genthelviteare instead sensitive only to S-O exchange(variation in log /",//",). Introduction 2 FeoBe,(SiOo)rS:2 FeS+ 3 Be,SiOo+ 3 FerSiOo Danalite, FeoBer(SiOo)rS,is a rare beryllium ore danalite: troilite * phenakite* fayalite mineral in iron-rich skarns,greisens, and related me- Troilite (Tr) is not a common terrestrial phase; for tasomaticrocks (Glasset al., 19441,Beus, 1966; Dunn, the purposesofthis discussionit can be replacedby 1976). lt typically contains considerable helvite, pyrrhotite, Fe,-,S. The non-stoichiometryof pyrrho- Mn Ber(SiOo)rS,or genthelvite,Zn Ber(SiOo)rS,i1 tite has been neglectedfor conveniencein balancing solid solution, and crystalsmay be distinctly zoned. reactions. The closestapproach to "danalite" in analyzednatu- The compositionalrelation expressedby the equa- ral specimensis 87 mole percentof the Fe end mem- tion is also shown in Figure l, a tentative depiction ber (Dunn, 1976).In a laboratory study (Mel?nikovet of mineral compatibilities in the reciprocal ternary al., 1968),danalite was the only end member of the system /Fe,Be/ /SiOo,S/ [or Fe"SiOo-2BeFe_,- helvite group that could not be synthesized. Sr(SiOr)-, in the "exchange operator" notation of The above data might suggestthat end-member Bvt, 19741.This systemis a subsystemof the four- danalite is unstable. This question cannot be an- component system FeO-BeO-SiOr-SO_r, itself a sweredwithout further experimentalwork; my objec- subsystemof the five-componentsystem Fe-Be-Si- tive is merely to examinethe reasonfor the rarity of O-S. Eight minerals in this system are listed in Ta- danalite-rich solutions in nature".The stabilities of ble l. helvite and genthelvite will be considered in later contributions. Incompatibilities The phase rule implies that not all eight of the Phases phasesin Table I can be compatibleat a given pres- Compositionally, end-memberdanalite is equiva- sure and temperature.Well-known incompatibilities lent to a mixture of troilite, phenakite, and fayalite, (unstable4ssemblages) in the Be.free subsystemFe- as shown by the following equation: Si-O-S include pyrrhotite-hematite, pyrite-fayalite, 0003-004x/80/0304-0355$02.00 356 BURT: STABILITY OF DANALITE 2FcS 2BeS and Tr ') ("ro 3Py+4Dan: l0Po+3Mt+6Phe +6Qtz (2) Hematite-danahteis not a well-documentedassem- tI blage, and the alternativeassemblage apparently oc- curs (Barton and Goldsmith, 1968,p. 59). Analogies 2S with other systemsalso suggestthat an FeO-rich 2S+5i04 mineral such as danalite will be incompatible with hematite.Reaction (l) hastherefore been assumedto go Foy to the right. Note, however, that Zubkov et al. terSiOO BerSiOO (1972,p.362-363) have reported a danalite with 5l atomic percent Fe end member Ssing altered by hematite, but it is not known if this representsan #----> equilibrium assemblage. Fig. |. Assumed mineral compatibilities in the reciprocal Reaction (2) is still more of a problem. The left as- ternary systemFe2SiOa-Be2SiOa-2FeS-2BeS (or Fe,Bel/SiOa,S). pyrite-danalite, The phases shown are probably compatible with quartz in the semblage, is relatively common, systemFeO-BeO-SiO2-SO_r. Abbreviations are given in Table I whereasthe right assemblageapparently has not yet and the text. beenreported. On the other hand, the danalite stabil- ity field in nature is undoubtedly extendedby solid and hematite-fayalite. Figure I implies that, if dan- solution of other components,and pyrrhotite-bearing alite is stable, pynhotite-fayalite-phenakite must assemblagesare rarely reported.Without experimen- also be an unstable assemblage.(Strictly speaking, tal data, then, the stable assemblagein this reaction due to the non-stoichiometryof pyrrhotite, this con- cannot be determined. clusion doesnot follow directly from Figure l. Nev- Stability diagrams ertheless,it can easily be shown that non-stoichio- Given the above data and assumptions,danalite metric pyrrhotite plus fayalite plus phenakite is has five possiblebreakdown reactionsinvolving Or, compositionally equivalent to the presumably more stable assemblagemagnetite-danalite-quartz.) Sr, and the phasesin Table l. These reactions are listed in Table 2 and are depicted on Figure 2, The phasesin Table 1, and the incompatibilities a schematic isobaric, isothermal log dia- mentioned above,can be used as input for computer /",-log .fo, gram. In view of the lack of thermochemicaland ex- program REACTTON(Finger and Burt, 1972),h order perimental data on the Be-bearingphases, the dia- to determineif there are other incompatibilities.Two gram is drawn at an arbitrary (moderate or low) reactionsthat are printed out are: pressureand temperature.The slopesof the lines on 6 Hem * 2Dan: Py + 6 Mt + 3 phe + etz (l) this diagram are fixed by compositionalrestrictions, and the topology should remain the same over a range of pressuresand temperatures (in Table l. Eight phasesin the systemFe-Be-Si-O-S the vicinity of 0.2-2.0kbar, 300-500'C). The reactionsinvolving the Be-freesubsystem Fe- Name Abbreviation Composition Si-O-S on Figure 2 are too well known to require Pyrrhotite Po cornment(Holland, 1965).Superimposed on theseis r-x a hypothetical danalite stability field, outlined with Pyrite Py FeS2 dark lines. The danalite stability field is seento be re- Magnetite Mt Ft3o4 stricted to the vicinity of the pyrrhotite/magnetite Hematlte ftn F"203 field boundary. The assemblagedanalite-quartz is stableover the samearea as danalite alone,except at Fayallte Fay FerSiOO the lower left corner of the diagram. Danalite's com- Phenaki.te Phe BerSiOO patibility with pyrite is, as mentioned,uncertain, and the danalite stability field could be considerably Danalite Dan FeOBe,(SiOO) rS smaller than Figure 2 implies. (As mentioned in the Quart z Qtz sio2 introduction, end-member danalite could be un- stable.) BURT: STABILITY OF DANALITE 357 6(0 cl .: 3 o u (9 Fe O2 Fig. 2. Hypothetical log .fs,-log 1o, diagram for danalite Figure 2, but in the presenceof a stability in the system Fe-Be-Si-O-S. The inset shows mineral Fig. 3. Diagram similar to fugacity of CO2, and showing hypothetical (schematic) compositions projected onto the composition plane Fe-Be-Si. The constant limits for danalite solid solutionsof 75 and 50 atomic Vo thicker lines outline the danalite stability field; the reactions stability assumes that the end-member danalite field involving danalite are listed in Table 2. Mineral abbreviations are Fe. This diagram (Dan,*) indicated on Figure 2. given in Table l. is more restricted than that Siderite stability is not considered. Danalite-rich solid solutions are common enough in nature, however,and their stabilitiescan similarly 2Dan+ 6 CO, : 6 Sid + 3 Phe+ 3 Qtz * 2 Po be depicted (Fig. 3). Figure 3 is analogousto Figure 2, exc,eptthat a small fugacity of CO, allows the In this regard, Palache(1907) reports the alteration graphite line to eclipse the fayatte stability field (cl of danalite in a granite cavity from Cape Ann, Mas- Holland, 1965).By analogywith other systems,solid sachusettsto a mixture of siderite,phenakite, quartz, solution of, e.g., helvite is shown as allowing the end- pyrite, and sphalerite, all coated by later kaolinite member danalite field (here shown as not touching and hematite. pynte) to expandconsiderably. Again due to the lack of data on danalite,the fields and numbersshown on Table 2. Five breakdown reactions of danaliter this diagram are purely hypothetical. Mhor sub- stitution of Mn in magnetitehas been neglected. The superpositionof a stability field for siderite (Sid) on Figure 3 has also been neglected.At low 1. 6 Dan * 4 Or= 2 Mt + 9 Fay + 9 Phe + 3 S, temperatures,with increasingCO, fugacity, siderite t 2 Dan * Qtz + 02 = 4 Fay * 3 Phe + S, would gradually replacethe magnetitefield (and part 5 Dan * 7 Or= 8Ut + 9 Phe + 9 Qtz + 3 52 of the pyrrhotite field). The lines of equal danalite content in helvite solid solutionswould be deflected 4. 2 Dan * 8 S, = 3 Py + 3 Phe + 6 Qtz + 4 Oz upwards as they enteredthe siderite field (to a slope ( 2 Dan * 4 Sr= 8 Po + 3 Phe + 6 qtz + 4 Oz of +1, instead of +7/3 as in the magnetite field). Eventually, increasingCO' fugacitieswould lead to the destruction of danalite, by reactionssuch as the * Abbreviatlons are es ln Table 1. following: 358 BURT: STABILITY OF DANALITE Table 3. Selectedassociations ofdanalite Locallty Association References Cape Ann, Massachusetts CavlEl-es ln granlte. Cooke, 1866 Danalite altered to slderlte, phenakite, Palaehe,1907 qttartz, pyrite, sphalerite. Pough,1936 Iron Mountaln, Bartlett Metasomatlcally replaced granlte. Wadsworth,1880 New Hampshlre Danallte-helvlte solLd solutLons overgrowing Glass et aL., 7944 phenakite, wlth nagnetite, pyrite, Warner et a1., 1959 sphalerlte, quartz, galena, fluorite, Barton and Goldsmith hematlfe, and pyroxene. 1968 Dunn, 1976 Authorts vlslts, 1976, 1977 North End Peak, Iron Mountaln "Ribbon rock'r banded skarn replacing class et al., L944 New Mexlco limestone.