Hydrothermal Phlogopite and Anhydrite from the SH2 Well, Sabatini Yolcanic District, Latium, Italy: Fluid Inclusionsand Mineral Chemistry H,Q.Nvnve

American Mineralogist, Volume 73, pages 775-797, 1988

Hydrothermal phlogopite and anhydrite from the SH2 well, Sabatini Yolcanic district, Latium, Italy: Fluid inclusionsand mineral chemistry H,q.nvnvE. Brr,xrN U.S. Geological Survey, 959 National Center, Reston, Virginia 22092,U.5.4. Grusnppn Clv.lRnrrr,c, Centro di Studio per la Geologia dell'Italia Centrale del C.N.R., % Dipartimento di Scienzedella Terra, Universitd degli Studi di Roma "La Sapienza,"00185, Roma, Italy BnNronrro Dn Vrvo Dipartimento di Geofisica e Vulcanologia, Largo S. Marcellino 10, 80138 Napoli, Italy Fn-c,NcnscA.Tpccn Centro di Studio per la Geologia dell'Italia Centrale del C.N.R., % Dipartimento di Scienzedella Terra, Universitd degli Studi di Roma "La Sapienza,"00185, Roma, Italy

AssrRAcr The SH2 well(2498.7 m) was drilled vertically in 1982-1983by an AGIP-ENEL Joint Venture as an exploratory hole to assessthe geothermal potential of the area north of Bracciano Lake, Latium, Italy, located in the Sabatini volcanic district. Drill-cutting samples from a thermometamorphic-metasomaticzone (1 140-2498.7 m) contain hydrothermal anhydrite + phlogopite (+ calcite + pyrite) and other authigenic volatile-rich phases. Microthermometry of primary and secondarytwo-phase [vapor (V) + liquid (L)] and multiphase (V + L * crystals)liquid-rich inclusions in anhydrite yields pressure-corrected temperatures of homogenization (trapping temperatures)that range from 144 to 304'C and that are generally coincident with measuredin-hole temperatures.The fluids have a variable salinity from 0.5 to 14.0 wto/oNaCl equivalent and also contain Ca2+ at least. Rare liquid COr-bearing aqueousinclusions have been verified by laser Raman spectroscopy. Also, rare liquid hydrocarbons(?)have beenobserved. Clathrates have been observed upon freezing, and crushing studies reveal noncondensablegas at P > I atm in some inclusions. Euhedral to subhedral phlogopite crystals (-0.5 to -2 mm) commonly are zoned and contain solid inclusions ofanhydrite and apatite and two-phase (V * L) and multiphase (V + L + crystals)fluid inclusions. Microthermometry of primary two-phase inclusions yields pressure-correctedtemperatures of homogenization (trapping temperatures) that range from 178 to 298 oC and are also generallycoincident with in-hole measured temperatures. Freezing studies show a variable fluid salinity (0.2-7.8 wto/oNaCl equiv.); the fluid contains Ca2+at least. If we assumethat the current hydrologic regime existed during anhydrite and phlogopite formation, the pressureof formation ranged from - 148to -220bars for phlogopite(1600--2500 m) and - 120to -220bars for anhydrite (1300--2500 m). Hence, the phlogopite and other authigenic phaseshave crystallized from a low P-2, volatile-rich, generallydilute, geothermal solution. Detailed microprobe analysesof fluid inclusion-bearing phlogopites indicate that they are nearly end member, Fer/(Fe, + Mg) : 0.02 to 0.1 and the t4tAlper I I oxygen equiv- alentsvaries from 1.0 to 1.35.F, BaO, and TiO, rangefrom 2.4 to 5.0 wt0/0,0 to 3.5 wt0/0, and 0.1 to 0.89 wt0/0,respectively, and an Fe2+-Favoidance is observed. Various cation substitutional schemesappear to be operative and are assessed.The dark zones are Fe- rich, but may have more or lessF, BaO, or TiO, than the light zones.xRF-KEvEx analysis of eight anhydrite samplesyield averagevalues (ppm) of Sr: 5500, Ba : 1900, Rb : 250' Ce: 140,and La: 100. The variation of fluid-inclusion salinities, the phlogopite zoning, and the chemical variation of the anhydrite and phlogopite suggestthat different fluids and/or episodic conditions were operative in this geothermal system.

0003404x/88/0708-0775$02.00 775 776 BELKIN ET AL.: HYDROTHERMAL PHLOGOPITE AND ANHYDRITE

basis of the various minerals recorded throughout the SH2 well, Cavarretta and Tecce (1987) have suggested that the ascendinghydrothermal solutions were rich in SO, + CO, plus other volatiles (B, HD and that perhaps a shallow Ca-rich fluid was present. They also have debated the hypothesisthat the brine was, in part, the result of remobilization of sedimentary (Triassic) sulfates in a low-pressureenvironment. They have concludedthat the hydrothermal solution and heat source was a magma of possible trachytic (or less differentiated) composition with a high volatile content that intruded in a subvolcanic en- BRACCIANOLAKE vironment in the northern Bracciano Lake area. The study of fluid inclusions in hydrothermal minerals may allow an evaluation of the fluid composition, tem- XA perature,and hydrology at the time of trapping, thus pro- CB viding information on the evolution of the geothermal systems(e.g., Browne et al., 1976; Taguchi and Hoyashi, {c 1983;Belkin et al., 1986).We have studiedthe fluid in- r Skm r /o clusions and mineral chemistry of the authigenic assem- blageof anhydrite and phlogopite. Anhydrite is becoming Fig. l. Location ofthe SH2 well and the generalfeatures of increasinglyrecognized important physicochemical the Sabatini volcanic district. A : scoria cones;B : explosive as an craters;C : caldera rims; D : fault. indicator in geothermal and mineralized hydrothermal systems(e.g., Cavarrettaetal., 1982;Queen and Motyka, 1984;Hattori et al., 1985).Authigenic phlogopite in geo- INrnolucrroN thermal or hydrothermal systems is rare; furthermore, The Sabatini volcanic district, located in northern La- fluid-inclusion studies of phlogopite are unknown to the tium, Italy, (Fig. l) has recently received considerable authors. Therefore, we have also undertaken extensive attention by the AGIP-ENEL Joint Venture (National Oil microprobe analysesof the phlogopite mineral chemistry and Electricity Boards, respectively; ENEL as operator) as a function of sample depth and crystal zoning. Phlog- for geothermalenergy exploration. The initial targetarea opite reflects the relative fugacities of HF, HCl, 02, H2, was the Cesano-Baccanocaldera that has the highest geo- and water during crystallization and hence monitors in- thermal gradient(l0G-150'C/km; Calamai et al., 1975) tensive variablesduring hydrothermal processes.The de- ofthe region.From 1975to date, l4 deepwells have been tailed characterization of the SH2 authigenic phlogopite drilled. Four ofthem produced a hot supersaturatedchlo- is especiallyimportant becausewe can define its environ- ride-sulfate brine (200 oC, reservoir conditions) charac- ment of formation in terms of pressure,temperature, and terized by unusually high values of total dissolved salts fluid composition. The trace-elementchemistry of the au- (up to 400 g/L, postflash; flashing steam estimated to be thigenic anhydrite was also determined in an attempt to 20-25o/oof total floq Calamaiet al., 1975).The difficul- provide additional information on the nature of the co- ties inherent with the use ofhot brines for electricity pro- existing solutions. duction have prompted the searchfor lower-salinity geothermal fluids outside the Cesano-Baccanocaldera. Gnor,ocrc sETTTNG The SH2 deep well (2498.7 m) wris drilled vertically in The Sabatini volcanic district is part ofthe perpotassic 1982-1983as an exploratoryhole to assessthe geother- Roman comagmatic province of Washington (1906) that mal potential of the area north of Bracciano Lake (Fig. borders the easternTyrrhenian Sea and is aligned along l). The bottom temperature was 290 oC,but no exploit- a NW-SE tectonic trend. Although the origin of the vol- able fluids were found (Calamai et al., 1985). Neverthe- canic activity is debated,the most current interpretation less,the pervasive hydrothermal alteration describedby defines the tectonic style as postcollisional and the asso- Cavarretta and Tecce (1987) is convincing evidencethat ciated magmatism as subduction related rather than re- an extensive hydrothermal system must have existed in sulting from intraplate rifting (e.g., Di Girolamo, 1978; the recent past. Extensive vein and pore-spacefillings by Rogerset al., 1985). hydrothermal minerals have effectively sealed the sys- The largest volcano-tectonic structure in the area, an tem. The lower 1630 m of the SH2 well is characterized almost square depression,is located near the Bracciano by the hydrothermal assemblageanhydrite + calcite + village and is now a lake (BraccianoLake). The Sacrofano pyrite plus, at various levels, the addition of K-feldspar, center, a stratovolcanic structure, is the main eruptive garnet, vesuvianite, apatite, phlogopite, spinel, pyroxene, center(De Rita et al., 1983).Numerous other minor cen- wollastonite, and the occurrenceof uncommon, volatile- ters (-20) give the areaa"crater-field" morphology.Their rich minerals such as wilkeite, cuspidine, harkerite, can- activity was mainly explosive, producing huge volumes crinite, and reyerite (Cavarrettaand Tecce, I 987). On the of pyroclastics;lava flows, very subordinate in volume, BELKIN ET AL.: HYDROTHERMAL PHLOGOPITE AND ANHYDRITE 777 are most abundant in the northern part of the Bracciano Lake area. The lake perimeter is characterizedby several recentminor eruptive centers,such as Trevignano, Vigna di Valle, Polline, and Acquerello. The oldest dated products of the Sabatini complex are the 0.6 Ma Morlupo center pyroclastics, whereas the youngestdated product is the 0.083 Ma Baccano pyro- A clastic flow. However, as defined by stratigraphy,the lat- est products were erupted from the Martignano, Le Case, and Stracciacappecenters (Di Filippo et al., 1984). The composition of products erupted within the Sa- o batini volcanic district is considered to be transitional o (to o between the mafic Alban Hills leucitites the south) E and the differentiated Vico and Vulsini suites (to the 20o north). The Sabatini lavas have been discussedin detail I F (1979) who showed that the lavas cluster in o- by Cundari ul B the tephritic leucite phonolite and tephritic leucitite fields o of the Streckeisen"APF" diagram. The compositional I related to crystalJiquid differ- I variations of the lavas are I entiation in a subvolcanic Qow-pressure)regime (Cun- I I dan, r979). I The volcanic rocks lie on a sedimenlary basementwhose I I I stratigraphyhas been reconstructedby Baldi et al. (1974) I ,ll G I I for the northern Roman Region and by Funiciello and I Parotto (1978), who provided additional data and ex- I I tended the reconstruction to southern Latium. From the I ll above studies the stratigraphy of the sedimentary base- 24 I I I ment of the area is as follows: (1) an upper allochthonous Itl D flyschoid complex (Late Cretaceousto Oligocene-Mio- CC AN PY KF MU GR VE PH BI CP SP WI CU HA cene) that varies from alternating calcarenites, marly major newly formed minerals as limestones, and clays to alternating calcarenites and Fig tion of the a function of depth in the SH2 well from Cavarrettaand Tecce quartzo-feldspathicsandstones (200 to 1000 m in thick- (1987).CC : calcite,AN: anhydrite,PY: pyrite,KF: K-feld- (oldest ness) and (2) a lower carbonate basal complex spar,MU : K-r'nicaand/or sericite, GR : garnet,VE: vesu- dated as Late Triassic) that consistsof marls, marly lime- vianite,PH : pglogopite,BI : biotite,CP : clinopyroxene,SP stones,and limestoneswith a marly limestone that con- : spinel,WI : wilkeite,CU : cuspidine,HA: harkerite.Also tains alternating dolomitic layers with minor anhydrite shown are the four zonesas distinguishedby Cavarrettaand layersin the lower section (total thicknessof complex can Tecce(1987). A: chlorite + zeolitezone; B : K-feldspara be 2000 m). garnet(grandite) + vesuvianite+ wilkeite + cuspidine+ har- keritezone; C: vesuvianite+ phlogopitezone; D : phlogopite + pyroxene+ spinelzone. Slvrpr,ns sruDIED The stratigraphy and hydrothermal mineral distribu- appearanceofgarnet. Figure 2 shows the distribution of tion of the SH2 well as reported by Cavarretta and Tecce the major authigenic phasesas a function of depth. (1987)can be summarizedas follows:Above 460 m, the Cavarretta and Tecce (1987; see Fig. 2) have defined well penetratedvarious volcanic products-lava and py- four main zones of hydrothermal minerals (calcite is roclastic flows (phonolites, leucitites, and leucitic teph- ubiquitous below 180 m, as are anhydrite and pyrite rites) and volcaniclastic breccias. Chlorite, calcite, and below 870 m): (A) a near-surfacechlorite + zeolite do- zeolites (mostly phillipsite) are presentas glassalteration main in the volcaniclasticunits, (B) a K-feldspl + garnet or newly formed minerals. Below 460 m the well entered (grandite) + vesuvianite + wilkeite + cuspidine + har- the allochthonous flyschoid complex, which was signifi- kerite + wollastonite + apatite domain in the top part of cantly recrystallized below 1070 m. Samples of the so- the "contact metasomaticcomplex," (C) a vesuvianite t called "Cicerchina," a faciesof the flyschoid complex, are phtogopite domain in the midpart of the same complex, distinctlypresent between 1450 and 1560m. Below 1560 and (D) a phlogopite + pyroxene + spinel + cancrinite m the original rock type is not recogttizable,and the pres- domain in the bottom Part. enceof the "carbonate basal complex" is assumed.From All the phlogopite samplesand all the anhydrite sam- 1070m to the bottom of the well at2498.7 m, authigenic ples except the core at 2168 m were selectedfrom drill- minerals are particularly abundant and mark the top of cuttings. The estimated positional uncertainty of the drill- the "contact metasomatic complex" defined by the first cutting samplesis +2 m. 778 BELKIN ET AL.: HYDROTHERMAL PHLOGOPITE AND ANHYDRITE

-2Oum

I n- 'l Y r-.,

-lOum B - loum 0": '"E -lg,g.s

-5um

-.. 2OOum M - -8 E!."*

Fig. 3. (A) Primary two-phase(V + L) inclusion in anhydrite brine (B) and CO, (confirmed by laserRaman spectroscopy)with from sample 2210 m, the third dimension is -35 pm. (B) Two- liquid CO. (L) preferentially wetting the vapor bubble (V). The phaseinclusion in anhydritefrom 2040m. The inclusionis necked- L + V CO, homogenizedin the vapor phaseat +28.3 'C. (F) A down, and further recrystallization would separatethe parts. A primary three-phaseinclusion in anhydrite from 2040 m. Two small single-phase(L) inclusion suggeststhat this processis a immiscible liquids are present;brine (B) and liquid hydrocarbon low-temperature one. The inclusions are flattened in the {010} (L) preferentially wetting the vapor bubble (V). The L and V plane. (C) An example of a curvilinear plane (X-Y) of secondary phaseshomogenize at + 32.9 "C in the vapor phase.(G) A primary (or pseudosecondary) two-phase inclusions in anhydrite from multiphase inclusion in anhydrite from 1600 m photographed 2150 m. The arrows indicate the common {0 I I } twin plane. (D) with partially crossedpolarizers. Four included crystals appear A primary three-phaseinclusion in anhydrite from 2040 m. The isotropic although some are very thin (-3 to 4 pm). Two are arrow points to an immiscible liquid phasethat is preferentially birefringent. (H) A primary multiphase inclusion in anhydrite wetting the vapor bubble. (E) A primary multiphase inclusion in from 2l 50 m. The bubble is preferentiallywetting a slightly pleo- anhydrite from 2168 m. Two immiscible liquids are present: chroic crystal (arrow) that may be a mica. (I) A zoned crystal of BELKIN ET AL.: HYDROTHERMAL PHLOGOPITE AND ANHYDRITE 779

The anhydrite and associatedcalcite, when observed Wejudge that samplepreparation probably would have decrep- in core, occupy common vein and pore-space fillings. itatedor partiallydecrepitated all fluid inclusionsin phlogopite. Phlogopite occursintergrown with anhydrite (Fig. 3K) in The crushing-stagetechnique (Roedder, 1970) was usedon a qualitativeestimate of the pres- both core and drill-cuttings. The evidencefor the hydro- anhydritein orderto obtain enceand compositionof noncondensablegases. The crushing thermal origin ofphlogopite is based on this textural re- mediawas immersion oil or an aqueousalkaline BaCl, solution. lationship and the fluid-inclusion and chemical data discussedbelow. of fluid inclusions in anhydrite Fr-urn TNCLUSToNS Description Three different inclusion types were distinguished. Type Fluid inclusions have been studied in hydrothermally A inclusions (Fig. 3,{) are two-phase (L + \4liquid-rich formed anhydrite and phlogopite throughout the SH2 deep aqueousinclusions with variable salinity. These are the well in order to better understand the physicochemical most common and were usedfor the heating and freezing characteristicsof the fluid at the time of trapping, its or- measurements.Type A inclusions are both primary and igin, and possible evolution. Although fluid inclusions secondary(pseudosecondary?). Primary type A inclusions were observedin other authigenic phases,only in phlog- (Fig. 3A,) typically are larger than secondary type A in- opite and especially anhydrite was their abundanceand clusions and have an equant shape. Type A secondary distribution sufficientfor a systematicstudy. In three cases (pseudosecondary) inclusions (Fig. 3C) commonly oc- the temperature of homogenization was determined for cupy curvilinear planes, are smaller, and tend to be flat- primary fluid inclusions in apatite crystals trapped in tened in the (010) plane. The correspondencebetween phlogopite crystals. microthermometric data (seebelow) of the primary and Fluid inclusions in micas have been rarely studied secondarytype A inclusions suggeststhat the latter may (Roedder, 1984). However, Pom6rleanu and Sabliovschi be pseudosecondary,but we have no supporting petro- (1980)reported primary fluid inclusionsin muscovitewith graphic evidence. temperaturesof homogenization (In) that vary ftom220 Type B inclusions are monophase,either liquid or va- to 600 "C. Trapping of fluid as a result of twinning on por, primary or secondary inclusions. Occasionally, we growth boundaries (twinning sutures)during fluorphlog- observe these inclusions to be the result of a necking- opite synthesis has also been noted by Kozlova et al. down process.Figure 38 shows that the necking-down ( l 984). processcan produce monophaseliquid inclusions. Neck- Methods ing can also produce monophasevapor inclusions. How- Microthermometrywas performedwith a USGSdesign and ever, we also observe isolated monophase vapor inclu- crurxvpca heatingand freezingstages at the Universityof Rome sions. Crushing experiments (see below) have revealed and the U.S. GeologicalSurvey, Reston. The stageswere inde- the presenceofnoncondensable gases at pressuresgreater pendentlycalibrated using organic and inorganic compounds as than 1 atm. Fracturing induced by the drilling process describedby Roedder(1984). The instrumentaluncertainty was could produce monophase vapor inclusions that are - +0.1 'C estimatedto be in the rangeof the final meltingpoint causedby the expansion of gasesduring decompression (7.) - +2.0 were of ice and "C in theheating mode. The samples on ascentwith the resultant expulsion of liquid. heatedonly onceto avoidspurious measurements that mightbe Type C inclusions are primary, multiphase, and liquid- dueto possiblestretching or decrepitation. and their Unpolishedbroken fragrnents (l-3 mm) of anhydritecrystals rich. Inclusions of this type are uncommon, andsmall (0.5-2 mm) euhedralto subhedralphlogopite crystals assemblagesare varied. The most common type C assem- wereused. This lackof samplepreparation (i.e., no doubly-pol- blage (Fig. 3G and 3H) comprisesvapor, liquid, plus one ishedplates) undoubtedly helped preserve fluid inclusionsthat or more crystal phases.The crystal phasesare birefringent otherwisemight havebeen removed or partiallydecrepitated. or nonbirefringent and euhedral to anhedral. In a few

phlogopite frorn 2210 m. The core contains numerous fluid in- and Fig. 3M. The details of the scallopeddissolution boundary clusions (one in white oval, shown in Fig. 3J) and is separated (arrow) are shown. (N) A zoned crystal of phlogopite from 2210 from a relatively inclusion-freezone by a sharp boundary. (J) A m. The upper left is a portion of the lighter-coloredcore. This is primary three-phaseinclusion from the core of a phlogopite crys- enclosedby a darker-colored zone (white oval) that is in turn tal (white oval, Fig. 3I). The included crystal is birefringent and surrounded by a faceted irregular rim (arrow). The darker zone appearsto be anhydrite. (K) A crystal of phlogopite, imaged in is inclusion-rich although the inner core also containsboth solid secondary-electronemission by a JEoLJMs-84osev, from 2210 m. and fluid inclusions.(O) A primary two-phaseinclusion from the A crystal of anhydrite (A) identified by snr"r energy-dispersive dark zone of Fig. 3N (white oval). (P) A primary multiphase analysis is partially enclosed by the phlogopite crystal. (L) A inclusion from the dark zone of Fig. 3N (white oval). (Q) A zoned crystal of phlogopite with a darker, inclusion-rich core primary two-phaseinclusion in a corroded apatite needle from separatedfrom the lighter outer portion by an irregularly scal- the core of Fig. 3N. The arrow points to the curved phlogopite- loped dissolution boundary. (M) An enlargementof an areafrom apatite boundary with the arrow on the phlogopite side. Fig. 3L; the solid triangle is in the samelocation in both Fig. 3L 780 BELKIN ET AL.: HYDROTHERMAL PHLOGOPITE AND ANHYDRITE caseswe can identify these crystal phases as daughter tions relative to the inclusion volume. Although these crystals(Roedder, 1984). crystals were only partially melted at vapor-liquid ho- An uncommon type C assemblage(Figs. 3D and 3F) mogenization, they are considered to be daughter crys- comprises two immiscible liquids plus vapor. Some of tals. Many large (>50 pm) inclusions contained crystals thesemultiphase inclusions were identified as containing but decrepitatedupon heating. In many casesthe crystals liquid COr, and others, liquid hydrocarbon(?).Figure 3D appeared to be cubic and hence looked like halite (Fig. shows a small globule preferentially wetting the vapor 3J), but all were birefringent and probably are anhydrite. bubble. This phenomenon was observed in a group of We did not observeany inclusions with two immiscible primary inclusions, which suggeststhat this is a primary liquids in phlogopite . Furthermore, we recognizethat the fluid component and not the result of later contamination population of inclusions on which we obtained heating (e.g., drilling fluids). Figure 3F shows a similar assem- and freezing data are biased. We observedthat many in- blage. This inclusion did not appear to absorb infrared, clusions in anhydrite, when crushed, yielded gas under and the vapor and adjacent liquid phasehomogenized at pressuregreater than I atm. We also noticed that many +32.9 "C in the vapor phase.This suggeststhat it is not fluid inclusions in phlogopite, even small ones, decrep- vapor and liquid COr. However, some type C inclusions itated along the weak (001) cleavagewhen heated. This contain immiscible fluids that appear to be liquid and biasesthe data toward those inclusions containing small vapor CO, (confirmed by laser Raman spectroscopy,see amounts of(or no) noncondensablegases. below). Figure 3E shows an inclusion where the vapor Small, usually corroded, elongatecrystals of apatite were and adjacent liquid homogenized at +28.3 oC and was commonly observedin the phlogopites.They were iden- an excellent absorber ofinfrared and hence is presumed tified as apatite by optical properties and by sEMenergy- to be COr. Figure 3E also shows the presenceof solid dispersiveanalysis on exposedcrystals. In three casesthe phasesthat are probably accidentally trapped crystals. apatite crystalscontained primary type A fluid inclusions Figure 3H showsa solid phase(daughter?) that is pref- (Fig. 3Q) that were amenable to vapor-liquid homoge- erentially wetted on its apparent basal surfaceby the va- nization measurement. por bubble. This phenomenonis similar to that described by Roedder(1984) and Belkin et al. (1985)for mica. MrcnorrmnMoMETRY RESULTS Description of fluid inclusions in phlogopite Pressurecorrection Two different types of fluid inclusions were observed A pressurecorrection (Potter, 1977) appropriatefor the in phlogopite . Type A inclusions (Fig. 30) are two-phase estimated trapping pressureand salinity was applied to (L + V), liquid-rich, and aqueous.Both primary and sec- all the Tn data subsequently discussed.The correction ondary (pseudosecondary?)type A inclusions were rec- was calculated with the following assumptions: (l) the ognized. The primary inclusions were most commonly presentwater table of - 50 m existed at the time of trap- found in the darker phlogopite zones (Fig. 3N) and were ping, (2) the pressurewas hydrostatic, and (3) the thermal usually isolated.We believe they are primary on the basis gradient at the time of trapping was similar to the one ofthe following criteria: (l) they are always concentrated now present in the well. A step-wise integration of the in the darker zones regardlessof the sequenceof zoning liquid densitiesfrom Keenan et al. (1969) and Potter and (e.g.,dark rim and light core or light rim and dark core); Brown (1977) was used with the water P-T data from we do not believe that the dark-zone preferencethat the Fisher (1976). The correction for both phlogopite and fluid inclusions show is the result of a difference in sol- anhydrite Tndata rangedfrom *8 to +14 "C. Thus, the ubility betweenthe light and dark zonesacting on a heal- data presentedon Figures 4 and 6 are trapping temper- ing fracture (Roedder, 1969),and (2) they sometimesoc- atures (I,). cur along with solid inclusions outlining growth planes, as distinguishedby color zoning. Anhydrite microthermometrydata Only the small (<35 pm) inclusions were usable for To observeany phasechanges upon warming, I 88 pri- heating and freezingmeasurements. Larger inclusions ob- mary and 28 secondary type A and C inclusions were viously leaked as also did some small ones. Secondary frozen (Fig. 5), and 276 pimary and 67 secondarytype (pseudosecondary?)type A inclusions were distributed A and C inclusions were heated to vapor-liquid homog- along healed fractures parallel to (001) that cut across enization (Fig. a). zonrng. On warming frozen inclusions, the eutectic tempera- Type B inclusions(Figs. 3J and 3P) are multiphase(L * ture (2") of the phase assemblagewas detectable;how- V + solid), liquid-rich, and aqueous with one or more ever, the small size or poor optical quality of the inclu- birefringent or nonbirefringent, euhedral to anhedral sol- sionsprevented this observationin most inclusions.When id phases. Most trapped crystals were isolated occur- observed,most f values were - -27 "C although a few rencesand did not melt during heating; hence they are ranged from -25 "C to -20'C. In two caseswe observed probably accidentally trapped crystals. In a few cases, 7. to be - -50 "C. Although the fluid composition prob- groups of type B inclusions were observed,all containing ably lies in the system containing Ca and Na with 7"": the same solid and fluid phaseassemblages and propor- -52"C (Crawford,l98l), apparentlythe Ca speciesis at BELKIN ET AL.: HYDROTHERMAL PHLOGOPITE AND ANHYDRITE 781

1300

? 2ooo o F o E

-o-22OO o t

2400

140 160 220 240 260 280 300 220 240 260 240 260 2AO 300 320 340 trapping temperature ("C)

Fig. 4. The temperatureof trapping (7,) data for primary and 2168, and 22IO m have been offsetto the right to avoid overlap. secondary fluid inclusions in SH2 anhydrite as a function of The in-hole temperatureswere measuredin stabilizedconditions sample depth (m). The Tndata have been pressurecorrected as to minimize the effect of external interferences.The uncertainty + describedin the text. Also shown are the measuredin-hole tem- in the in-hole temperaturemeasurement is l-2 "C in the range peratures,and the boiling-point curve (BPC)for pure water (Haas, 140-200'C and t2-5'C in the range200-300'C. 1971)adjusted for a water table of -50 m. Samplesfrom 1990, such a low concentration that the Z" melting at low tem- enized to a fluid by the disappearanceofthe vapor phase perature (- - 50 'C) was insufrcient to observeroutinely. upon heating. The 7, values of primary and secondary Upon further warming, the temperature of the last type A inclusions were similar except for a sample from melting of ice (Z* ice) was generally observable.Values 2410 m where the secondary inclusions had Z, values for I- ice (vapor-present)for primary inclusions ranged -20 "C lessthan the averagefor primary inclusions. The - from -0.3 to -10.0'C with most between-0.3 and f, values for primary type A inclusions ranged from 140 -3.0 .C. We have recastthe Z- ice data into weight per- "c at I 300 m to - 320 "C at 2490 m. cent NaCl equivalent assuming the dominance of NaCl Heating type C inclusions revealedthat some included as the dissolved species.Figure 5 shows the intra- and crystalsdissolve below vapor-liquid homogenization and intersample variability of the salinity data. The salinity some remained unaffected. ranges from -0.5 to 14 wto/oNaCl equivalent. Upon Figure 4 shows that the mode of all the pressure-cor- freezing, a few two-phase (L + V) inclusions yielded a rected ?"ndata for eachsample except for those from 2410 solid that persisted above 0.0 "C. The melting point of m falls within + l0 "C of the measuredin-hole tempera- thesepresumed gas clathrates ranged from + l. 55 to + 20.0 ture. This agreementis very good considering the limi- 'C "C. Since CO, clathrates are unstable above -+10 tations and assumptionsof the data. (Collins, 1979), the higher I- ice values indicate that Phlogopite microthermometrydata other gases(hydrocarbons) must be present. Clathrate formation was recognizedfrom all sample depths. How- Ninety primary type A and B inclusions were heated ever, CO, as a solute can depressthe freezing point by a to vaporJiquid homogenization(Fig. 6), and 79 were fro- maximum of - 1.48 'C without forming clathrate (Hed- zen to observeany phasechanges upon warming (Fig. 7). enquist and Henley, 1985).Hence, we cannot distinguish After freezing, the inclusions were slowly warmed in an -45 .C betweenthe effectsof dissolved saltsor CO, in some low- attempt to determinate 2". The 7" was ! 5 but salinity inclusions in anhydrite and phlogopite. was difficult to observeat best. The determination of Z- A small number of secondaryinclusions were frozen ice was easierbecause of better resolution of the remain- and yielded Z- ice that ranged from -0.4 to -3.9 "C. ing ice crystals.In only one inclusion from sample 2480 : 'C) The salinity for these inclusions ranged from 0.7 to 6.3 m was evidence of clathrate ice (2. +9 observed. wto/oNaCl equivalent. Figure 7 showsthe T^ice data recastinto weight percent All primary and secondarytype A inclusions homog- NaCl equivalent. The salinity rangedfrom 0.2 to 7.8 wto/o 782 BELKIN ET AL.: HYDROTHERMAL PHLOGOPITE AND ANHYDRITE

anhydrite 1300 ru o primary

. secondary

o H- . H F-- n- o o 2000 E o 2100 o E ct o o E !t -c 2200 o. o !t

2400

160 180 200 220 240 260 2AO 300 trapping temperature ("G) Fig.6. Thetemperature of trapping(I".) data for primaryfluid o4812 inclusionsin SH2phlogopite as a functionof sampledepth (m). wt% NaGlequivalent The 7n data havebeen pressure corrected as describedin the Fig. 5. The salinity (wto/oNaCl equivalent)data for primary text. Also shownare the measuredin-hole temperatures.The andsecondary fluid inclusionsin SH2anhydrite as a functionof samplefrom 2210m hasbeen ofset to avoid overlap. sampledepth (m). observed in both primary and secondary inclusions. A few crushing runs in an alkaline BaCl, solution revealed NaCl equivalent with a considerablescatter and was sim- the faint presenceof a white precipitate (BaCOr) after ilar to that shown by the anhydrite data. sudden bubble collapse,confirming the presenceof CO, Type A and type B inclusions were heated, and all ho- as a volatile componenl. mogenization to fluid was by the disappearanceof the vapor phase. Over 500/oof the heated inclusions decrep- Llsrn R.LMlN spEcrRoscopy itated before vapor-liquid homogenization.This suggests Limited laser Raman spectroscopicanalyses were per- the presenceof dissolvedgases. No included crystalswere formed on one multiphase (Fig. 3E) and six two-phase seen to melt before vapor-liquid homogenization, al- inclusions in anhydrite. The technique used was similar though one group becamenoticeably rounded. to that described by Dhamelincourt et al. (1979). The f,, rangedfrom 178"C at 1600m to 298 .C at 2480 m. analysesconfirm the presenceof CO, in the multiphase Figure 6 shows that although the data are limited, there inclusion and indicate the absence of noncondensable is a definite increaseof Z, with depth. Three two-phase gases(CO2, H2S, CO, and Nr) in the six two-phaseinclu- primary fluid inclusions in apatite (Fig. 3Q) homogenized 'C sions. This latter result supports the observations,made to liquid at 250 from sample 2210 m. This tempera- by crushing and petrography, of a variable presenceof ture is similar to that of the inclusions in the host phlog- noncondensablegases in the fluid inclusions of the SH2 opite . anhydrites. CnusnrNc n,q.Ta Prrr,ocoprrr MINERAL CHEMISTRy Selected anhydrite crystals from all samples were Analyticalmethod crushed in oil. The good cleavageof anhydrite and the Small(0.5-2 mm) phlogopite mode of occurrenceof the fluid inclusions in groupsmade euhedralto subhedral crystals werehandpicked, where present, from the drill-cuttings.Only this technique difficult. Evolution of gas, at a pressure thosecrystals that containedprimary fluid inclusionswere se- greater than atmospheric, was noted in all samples but lected.A standardpetrographic slide was frostedon one side only in some inclusions. The vapor bubbles expanded with 800grit AlrO3abrasive and then cleaned and dried. A very various amounts from just filling the inclusion volume to thin layerof clearepoxy (e.g., Devcon 2-Ton) was applied with evolving bubbles into the crushing medium. This was the edgeof a razorblade. Under a low-powerbinocular micro- BELKIN ET AL.: HYDROTHERMAL PHLOGOPITE AND ANHYDRITE 783

X 16OOm PhlosoPite O 2 lgom a 221Om 1600 H n 15

1.4 x Oi i r.s o ta 1800 ci t" o {^' r...tiJt^ : ? ,., ls", o 1.O d! o o.o2 0.04 0.06 0.08 6 2O0O ^'ld:3:':?;t=""'", E Fel(Fe+Ms) + ANNTTE .c CL o HF Fig. 8. Variation of the phlogopitecomposition from the SH2 E -H - H substitutions (basedon an I I - EEFH H well as a function of two coupled 2200 oxygen-equivalentnormalization). The abscissashows the substitution Mg'* = Fe'?+,phlogopite to annite, as expressedby Fel (Fe + Mg), Fe : total Fe by microprobe. The ordinate shows the substitutiotr tetlMr*) 1 t+t(gir+)= t6l(A13+)+ I4r(Al3+)as expressedby the variation in totAl per formula unit (pfu). All the 2400 phlogopite data presentedin Figs. 8-10 and 12-16 representthe completepopulation of microprobe analyses(Table l). The light and dark zones in single crystals are plotted as separatepoints. Only in Fig. 9 are these points paired and marked with respect 048 to their relative color. wt% NaCl equivalent Fig. 7. The salinity (wto/oNaCl equivalent) data for primary phlogopite function fluid inclusions in SH2 as a of sampledepth determinedby the precessiontechnique, and the resultsare given (m). in Appendix 1.

Norrnalization procedure Microprobe analysiscannot distinguishbewteen Fe2+ and Fe3+, crystalswere arrangedin a recordedpattern, scope,the separated and the HrO in hydrous silicates is not determined. We have (l h). The must be then set aside for the epoxy to cure epoxy usedthe program and normalization schemeofFlohr (1983) that gently exceedinglythin in order to allow the single crystalsto be involves normalization to I I oxygen equivalents, i.e., total cat- pushed againstthe slide without the epoxy flowing up and over ion charge : 22. We have assumed that no Fer+ is present. crys- their top surfaces.After curing, cleaning,and C-coating,the Hence, all Fe discussedsubsequently is total Fe expressedas with tals were examined reflected-light optics to distinguish those FeO. Table I gives the formula proportions using the above (001) were perfectly flat, areason the surfacethat smooth, and scheme.The OH was calculatedby the method of Flohr (1983), for suitable electron-microprobeanalysis. assuminga full OH-site occupancy.We have also assumedthat phlogopite zoned when viewed The crystals are commonly the substitution of O'- for (OH) does not occur. The cations perpendicular (001). from in to Any analysis any other direction are assignedto crystallographicsites for the stoichiometry (lt'?E, a zoned crystal would give valuesthat may or may not be mean- Na, Ca,K Ba) (Fe,Mg, Mn,161Al,Ti,16r !)3 (4rAl, Si)oO'o(OH,F,Cl), ingful in terms of the systematicelement distribution of the zon- (! : vacancy). The vacancies are then calculated by difference' 1ng. An anr--serraqelectron microprobe with four fixed and four Results driven spectrometerswas used for wavelength-dispersiveanal- phlogopite nearly end-member com- ysis.Operating conditions were 15-kV voltageand a 0.l-pA beam The SH2 have : current. Standardizationwas done before eachanalytical session positions, with average Fe/(Fe + Mg) 0.05, and vary on silicates of known composition, similar structure, and ele- in composition by two major cation substitution schemes. ment abundancessimilar to thoseanticipated in the phlogopites. First, all the analyses are enriched in A1rO., save two, Selectedrepresentative analyses are given in Table l. Table I compared to phlogopite-annite. The dominant mecha- also gives the standard deviation (lo) and the detection limit nism causing Al enrichment in biotite involves the Al- associatedwith each element. The count data were reduced on- Tschermak's substitution line by a Bence-Albeescheme as adaptedby McGee (1983). The backgroundinterpolation method of M. Mangan (pers. comm., r61(M2+)4 r+r($ia+)= 16r(413+)1tot1Al3*). I 985) was used.Each analysisreported in Table I representsthe Figure 8 shows the variance sf tet(lvlz+) utt6 tollAl3*). averageof 3 to 9 individual analyseson different l0-pm beam from 1.0 to 1.35 t4rAl per ll oxygens. spots. Phlogopite analysesof crystals from any given sample The values range + were always run during different analytical sessionsto avoid any The other cation substitution involves Mg'* Fe2+; the systematicerror associatedwith any one analytical sesslon. phlogopite component ranges from 90 to 980/0.Figure 9 The crystallographyof a phlogopite crystal from 2210 m was shows the variance of Mg with Fe, and the data show a 784 BELKIN ET AL.: HYDROTHERMAL PHLOGOPITE AND ANHYDRITE

TnaLe1. Representativemicroprobe analyses of zoned and unzonedphlogopites from the SH2 well

Depth (m): 'I Sample: 1234 2 3 +co Description: zoned unzoned unzoned unzoned unzoneo unzoneo zoneo unzoned unzoned unzoned lightcore light core Numberof anaryses: o o sio, 40.10 36 13 41.44 38.06 40.73 40.78 40.76 41.34 38.93 39.62 Alros 1593 18.57 13.79 17.00 13.31 14.59 13.88 13.37 13.60 14.20 FeO 2.08 1.73 2.09 1.44 1.21 1.25 1.19 0.99 0.99 1.17 Mgo 26.1I 23.89 25.s9 25.12 26.54 26.57 26.75 26.82 26.79 27.61 n.o. n.d. n.d. n.d. n.d. 0.03 n.d. n.d. n.d. n.d. Naro 0.10 0.10 0.10 0.10 0.09 0.08 0.09 0.07 0.09 0.08 KrO 10.36 9.23 10.10 9.66 10.12 9.92 10.04 10.04 9.64 10.26 Tio, 008 0.36 0.16 0.17 0.15 0.16 0.20 0.16 0.15 0.15 MnO 010 0.08 0.09 0.07 0.04 0.04 0.05 0.04 n.d. 0.04 BaO 114 3.81 1.02 2.52 1.38 1.46 1.01 1.52 2.01 2.11 cl 0.26 n.d n.d. n d. n.d. n.d. 0.19 0.19 n.d. n.d. F 3.35 2.95 3.46 3.32 4.21 3.90 3.90 J.CO 3.69 3.87 Sum 99.69 96.85 97 84 97.82 97.78 98.78 98.06 97.91 95.89 99.11 -Cl=O 0.07 0.00 0.00 0.00 0.00 0.00 0.05 0.05 0.00 0.00 -F=O 1.41 1.24 1.46 1.40 1.77 1.64 1.64 1.50 1.55 1.63 Sum 98.21 95.61 96.38 96.06 96.00 97.14 96.37 vo.rtc 94.34 97.48 H,O (calc)- 261 2.70 2.57 2.58 2j7 2.38 2.30 2.46 2.34 2.37 Sum 100.82 98.30 98.95 98.64 98.17 99.52 98.67 98.82 96.67 99.85 Cations calculatedon the basis of 11 oxygen equivalents,total Fe as FeO Si 2.818 2.645 2.952 2.747 2.930 2.889 2.908 2.951 2.856 2.826 r4tAl 1.182 1355 1.048 1.253 1.070 1.1 11 1.092 1.049 1.144 1.174 T site 4.000 4 000 4.000 4.000 4.000 4.000 4.000 4.000 4.000 4.000 r6jAl 0.138 0.248 0.110 0.194 0.059 0.107 0.076 0.059 0.032 0.021 Fer* 0.122 0.106 0.125 0.087 0.073 0.074 0.071 0.059 0.061 0.070 Mg 2.743 2.607 2.717 2.702 2.845 2.805 2.844 2.853 2.929 2.935 Ti 0.004 0.020 0.009 0.009 0.008 0.009 0.011 0.009 0.008 0.008 Mn 0.006 0.005 0.005 0.004 0.002 0.002 0.003 0.002 0.000 0.002 M site 3 014 2.985 2.965 2.996 2.988 2.997 3.005 2s82 3.030 3.036 0.000 0.000 0.000 0.000 0.000 0.002 0.000 0.000 0.000 0 000 NA 0 014 0.014 0.014 0.014 0.013 0.011 0.012 0.010 0.013 0.011 K 0.929 0 862 0.918 0 890 0.929 0.896 0.914 0.914 0.902 0.934 Ba 0.031 0.109 0 028 0.071 0.039 0.041 0.028 0.043 0.058 0.059 A site 0.974 0.985 0.961 0.976 0.980 0.950 0.954 0.966 0.974 1.004 > cations 7.988 7.971 7.925 7.970 7.968 7.947 7.959 7.948 8.002 8.040 F 0.745 0.683 0.780 0.758 0.958 0.874 0.880 0.804 0.856 0.873 cl 0.031 0.000 0.000 0.000 0.000 0.000 0.023 0.023 0.000 0.000 OH (calc). 1.224 1.317 1.220 1.242 1.042 1.126 1.097 1.173 1.144 1.127 Sum 2.000 2.000 2.000 2.000 2.000 2.000 2.000 2.000 2.000 2.000 Fel(Fe + Mg) 0.043 0.039 0.044 0.031 0.025 0.026 0.024 0.020 0.020 0.023 F(F + OH) 0.373 0.343 0.390 0.379 0.479 0.437 0.440 0.402 0.428 0.436 Notei The standarddeviations (1o, wt7") for 3 to 9 individualanalyses are SiO, : 0.52, Al,Oo: 0.43, FeO : 0.20, MgO = 0.47, QaO: 0.01, NazO: 001,K,O:0.19,TiO,:0.04,MnO:0.02,BaO:0.15,andF:02l.Theaveragedetectionlimits(wt%)forthemicroprobeconfigurationusedare SiO,: 0.05,ALO.: 0.03,FeO: 0.06,MgO: 0.04,CaO :0.03, Na"O: 0.02,K,O: 0.04,TiO,:0.04, MnO: 0.04,BaO:0.11, Cl :0.10, and F : 0.08. n.d. : not detected. . H2Ocalculation is basedon the assumptionof full OH site occupancy(OH : 2.00 - F - Cl) and is only usedas a checkon analysissums.

good agreementfor the substitution Mgz+ -. Fe2+.Figure phlogopite is one-twentieth that of biotites. The dimen- l0 shows the variation of Fe/(Fe + Mg) with sample sion of the (OH,CI) site, hence its occupancy,is mainly depth. From 1600 m to 2190 m there is a slight decrease controlledby the rotation angle(cy) ofthe tetrahedra,which in Fe/(Fe + Mg), then the averageratio increases.This is a direct function of the XF" in the biotite. Four phlog- picture is complicatedby thefaclthalzoning alsobecomes opite crystalswere analyzedfor CrrOr. The CrrO, content more prevalentat depthsbelow 2210 m. was found to be at or below the detectionlimit (0.03 wto/o). Other elementsanalyzed for are Mn, Ti, K, Ba, Ca, N4 Ti, K, Ba, and F variations are discussedbelow. Cl, and F. The MnO contents vary from below the de- Phlogopitecolor and zoning.The SH2 phlogopites,when tection limit to 0.26 wto/o.However, it clearly varies by viewed perpendicularto (001), display variation in color the substitution2 Mgz+ + Fe2+* Mn2+ (e.g.,Fig. I l). from colorlessto very dark green,which is in part a func- CaO was usually below the detection limit. NarO was tion of crystal thickness.Very striking, especiallyin small always minor and varies from near the detection limit to (-2 mm) euhedral to subhedral crystals, is a simple but 0. 19 wt0/0.Cl was always at or below the detection limit. sharp light and dark zoning (Fig. 3I). As either the core Volfinger et al. (1985) indicated that rhe Cl content of or the rim can be light or dark, the color is probably a BELKIN ET AL.: HYDROTHERMAL PHLOGOPITE AND ANHYDRITE 785

TABLE1-Continued 2210 112 23 45 5 6 7 zoned zoned zoned zoned zoned zoned zoned zoneo zoned zoned zoneo zoneo lightrim dark core dark rim lightcore darkcore lightrim lightrim dark core lightrim dark core lightrim dark core

3 o 3 3 e 39.99 40.00 40.47 41.51 40.40 39.65 38.80 40.96 39.84 38.04 38.64 39.16 14.60 1504 13.93 12.95 13.88 15.00 14.43 11.83 13.73 15.02 15.41 15.22 1.07 3.52 2.92 2 53 3.02 1.56 1.36 3.38 1.47 4.74 1.85 3.53 28.42 26.38 27.50 28.25 24 74 25.85 27.41 26.78 27.29 24.18 26.58 24.35 n.d. n.d n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. 0.05 0.04 0.04 0.05 0.13 0.07 0.08 0.08 0.07 0.07 0.07 0.07 9.99 10.17 10.30 10.45 10.02 9.85 10.43 11 .01 10.70 10.45 10.72 10.53 0.13 0.28 0.35 0.'t7 0.25 0.18 0.23 0.15 0.19 0.53 0.19 0.26 n.d. 0.11 0.09 0.04 0.13 0.05 0.06 0.26 0.05 0.10 0.06 0.18 1.45 0.27 n.d n.d. 0.48 1.O4 1.67 0.13 1.13 1.12 1.18 0.58 n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. O.22 4.42 3.73 4.40 4.57 3.52 4.68 3.33 3.58 4.22 3.33 3.53 3.41 100.12 99.54 100.00 100.52 96.57 97.93 97.80 98.16 98.69 97.58 98.23 97.51 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.06 186 1.57 1.85 1.92 1.48 1.97 1.40 1.51 1.78 1.40 1.49 1.44 9826 97.97 98.15 98.60 95.09 95.96 9640 96.65 96.99 96.18 96.74 96.02 2.16 2.48 2.17 2.12 2.47 1.94 2.59 2.48 2.16 2.53 2.51 2.47 100.42 100.45 100.32 100.71 97.56 97.90 98.99 9913 99.16 98.71 99.25 98.49 Cationscalculated on the basisof 11 oxygenequivalents, total Fe as FeO 2.816 2.820 2.849 2.906 2.925 2.856 2.791 2.941 2.855 2.777 2.771 2.833 1.184 1.180 1.151 1.069 1.075 1.144 1.209 1.001 1.145 1.223 1.229 1.167 4.000 4.000 4.000 3.974 4.OOO 4.000 4.000 3.943 4.000 4.000 4.000 4.000 o 027 0.071 0.006 o.0oo 0.109 0.129 0 015 0.000 0.014 0'069 0.074 0.132 0.063 0.208 0.172 0.148 0.183 0 094 0.082 0.203 0.088 0'289 0.111 0.214 2.982 2.772 2.886 2.947 2.669 2.775 2.939 2.866 2.914 2'630 2.841 2.626 0.007 0.015 0.019 o.oo9 0.014 0.010 0.012 0.008 0.010 0.029 0.010 0.014 0.000 0.007 o.oos 0.002 o.oo8 0.003 0.004 0.016 0.003 0.006 0.004 0.011 3.079 3.072 3.087 3.107 2.983 3.011 3.052 3.093 3.030 3-024 3.039 2.996 0.000 0.000 o.ooo o.0oo 0.000 0.000 0.000 0.000 0.000 0'000 0.000 0.000 0 007 0.005 o.oo5 o.oo7 o.olI 0.010 0.011 0.011 0.010 0.010 0.010 0.010 0.897 0.915 0.925 0.933 0.925 0.905 0.957 1.009 0.978 0.973 0.981 0.972 0.040 0.007 o.ooo o.oo0 0.014 0.029 0.047 0.004 0.032 0.032 0.033 0.016 0.944 0.928 0.930 0.940 0.957 0.944 1.015 1.023 1.020 1.015 1.024 0.998 LO24 8.000 8.017 8.021 7.939 7.955 8.067 8.059 8.049 8.039 8.063 7.994 0 984 0.832 0.980 1.012 0.806 1.066 0.758 0.813 0.956 0.769 0.801 0.780 0.000 0.000 o.ooo o.ooo o.ooo 0.000 0.000 0.000 0.000 0.010 0.000 0.027 1.016 1 168 1.020 0.988 1.194 0.934 1.242 1.187 1.044 1.23'l 1.199 1.193 2.000 2.000 2.OOO 2.OOO 2.OOO 2.000 2.000 2.000 2.000 2.000 2.000 2.000 0.021 0.070 o.os6 0.048 0.064 0.033 0.027 0.066 0.029 0.099 0.038 0.075 0.492 0.416 0.490 0.506 0.403 0.533 0.379 0.407 0.478 0'385 0.401 0.390

function of time of nucleation. In a few casesthe phlog- Phlogopite zoning is not common, and although there opites displayedmore than this simple paired zoning. Fig- is considerablespeculation as to the exact origin of the ures 3N and I I show euhedral crystals with three major zoning, there is agreementthat it is causedby some major zones.Superimposed on these major zonesare very thin perturbation in one or more intensive parameters(Rim- and compositionally subtle oscillatory zones.Detailed in- saite,1969; Rashkova, 1981). Various mechanismshave vestigation ofthese oscillatory zoneswas not undertaken, been considered:(l) a changein the bulk composition of but it appearsthat they result from changesin Fe2+/Mg2+ . the precipitating fluid; (2) a change in pressure and/or Also observedin a small number of phlogopite crystalsis temperature; (3) a preferential loss of H, during some a dissolution texture always associatedwith the darker event by which the solution becomesmore oxidized, the zones (Figs. 3L and 3M). In every case the dissolution activity of Fe2+ is lowered, and, hence, the phlogopites event was followed by the crystallization of a lighter (Mg- becomeenriched in Mg; and (4) that the Fe content varies rich) zone. The dissolution has occurred primarily on the as a function ofgrowth rate. Baronnet and Velde (1977) edgesof the phlogopite crystalsand not on the (001) sur- have experimentallyinvestigated the growth of phlogopite face. for constant chemical and physical parameters.They found 786 BELKIN ET AL.: HYDROTHERMAL PHLOGOPITE AND ANHYDRITE

f ABLE1-Continued

Depth(m): 2410 Sample: 8 8 I 911 2233 Description: zoneo zoneo zoneo zoned zoned zoned zoned zoned zoned zoned lightcore darkrim lightrim dark core light rim dark core lightrim darkcore darkcore lightrim Numberof anaryses: e 2 333 3333 sio, 40.53 39.25 40.34 37.81 40.18 39.69 41.79 41.39 39.64 39.27 Alr03 11.95 13.87 11.67 14.74 14.23 15.50 12.82 12.76 14.32 14.54 FeO 2 37 3.7't 2.28 3.53 1.58 1.74 1.20 2.08 4.03 3.47 Mgo 2715 25.06 27.53 25.96 28.47 27.43 29.15 28.25 26.11 26.94 CaO n.d. n d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. Na.O 0.08 0 08 0.08 0.07 0.05 0.06 0.04 0.05 0.04 0.05 KrO 11.02 10.80 10.84 10.75 10.26 10.00 10.57 10.33 10.11 9.80 Tio, 0.20 0.30 0.17 0.24 0.16 0.16 0.15 0.11 0.40 0.41 MnO 0.08 0 12 0.08 0.16 n.o. n.d. n.d. 0.10 0.14 0.10 BaO 0.44 0 65 0.27 0.84 0.40 o.74 n.d. 0.37 n.d. 0.19 cl n.d. n d. n.d. 0.16 n.o. n.d. n.d. n.d. n.d. n.d. F 3.57 3.24 5.00 3.79 4.29 4.06 4.58 4.56 4.14 4.05 Sum 97.39 97.08 98.26 98.05 99.62 9938 100.30 100.00 98.93 98.82 -Cl=O 0.00 0.00 0.00 0.04 0.00 0.00 0.00 0.00 0.00 0.00 -F=O 1.50 1.36 2.11 1.60 1.81 1.71 1.93 1.92 1 74 1.71 Sum 95.89 95.72 96.16 96.41 97.81 VT.OT 98.37 98.08 97.21 97.12 H,O (calc). 2.46 2.59 1.76 2.28 2.23 2.33 2.'t3 2.10 2.24 2.29 Sum 9835 98.30 97.92 98.69 100.04 100.00 100.50 100.18 99.44 99.40 Cations calculatedon the basis of 11 oxygen equivalents,total Fe as FeO Si 2.928 2.854 2.926 2.753 2.828 2.797 2 915 2.914 2.829 2798 ratAl 1.018 1 146 0.998 1.247 1j72 1.203 1.054 1.059 1.171 1.202 T site 3.945 4 000 3.923 4.00 4.000 4.000 3.969 3.974 4.000 4.000 16lAl 0.000 0.043 0 000 0.018 0.009 0.085 0.000 0.000 0.034 0.019 Fe2* 0.143 0.226 0.138 0 215 0.093 0.103 0.070 0.122 0.241 0.207 Mg 2.923 2716 2.976 2.817 2.987 2.881 3.030 2.965 2.777 2.861 Ti 0.011 0.016 0 009 0.013 0.008 0.008 0.008 0.006 0.021 0.022 Mn 0.005 0.007 0.00s 0.010 0 000 0.000 0.000 0.006 0.008 0.006 M site 3.082 3.009 3.128 3.073 3.097 3.076 3.108 3.099 3.082 3.115 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 NA 0.011 0.011 0.011 0.010 0.007 0.008 0.005 0.007 0.006 0.007 K 1.016 1.O02 1.003 0.999 0.921 0.899 0.941 0.928 0.921 0.891 Ba 0.012 0.019 0.008 0.024 0.011 0.020 0.000 0.010 0.000 0.005 A site 1039 1.032 1.022 1.032 0.939 0.928 0.946 0.945 0.927 0.903 2 cations 8.066 8.040 8.073 8.105 8.036 8.003 8.023 8.018 8.009 8.018 F 0.816 0.745 1.147 0.873 0.955 0.905 1010 1.015 0.935 0.913 0.000 0 000 0.000 0.020 0.000 0.000 0.000 0.000 0.000 0.000 OH (calc.)- 1.'184 1 255 0.853 1.108 1.045 1.095 0.990 0.985 1.065 1.087 Sum 2.000 2.000 2.000 2.000 2.000 2.000 2.000 2.000 2.000 2.000 Fel(Fe + Mg) 0.047 0.077 0.044 0.071 0.030 0.034 0.023 0.040 0.080 0.067 F(F + OH) 0.408 0.375 0.574 0.436 0.478 0.453 0.505 0.507 0.467 0.456

that increasedgrowth rate increasesthe Fe and to a lesser with K* and the position ofthe two substitutionalschemes. extent the Ti content of the mica. The abundanceof solid The most statistically significant variation of Ba with K and fluid inclusions observedin the dark zonescompared appearsto behave according to scheme Ba2+* Al3* + to the light zones also suggeststhat the dark zones grew K* + Si4+.However, the displacementsof the 1600-m more rapidly (Roedder,1984). and one of the 2210-m regressioncuryes may indicate Ba in SH2 phlogopites.The BaO content of SH2 phlog- additional substitution into the interlayer site by other opite rangesfrom 0 to 3.8 wto/oand can be enriched or elementsor vacancies. depletedin similarly colored zonesof the sameor different The BaO content in relation to the obvious color zoning crystals.Although the Ba content of SH2 phlogopite has is complex. Figure I I shows a zoned crystal where BaO scatter in any given sample, there appearsto be a mini- in one zone can be either enriched or depletedrelative to mum in the averageBaO content at 2410 m with increas- the other zone. Generally the darker (Fe-rich) zoneshave ing values at lesserand greaterdepths (Fig. l2). A similar intermediateamounts whereas adjacent lighterzones have variation in the Ba content ofcoexisting anhydrite is also BaO contents that range from 0 to > 3 wto/0.The ratio of seen(Fig. l2). Ba-enriched to Ba-depleted adjacent light zones is ap- Ba2+is nearly identical in size to K*, and various sub- proximately one.Although the data are limited, there does stitutional schemeshave beenproposed: (l) Ba'z*+ tr2lE= not appear to be any systematic difference between the 2K* (Shmakin, 1984)and (2) Ba2+* Al3* + K+ + Si4+ Ba contents of light-colored centersor light-colored rims. (Wentlandt, 1977).Figure l3 showsthe variation of Ba2+ Ti in SH2 phlogopites. Ti in phlogopite is of interest BELKIN ET AL.: HYDROTHERMAL PHLOGOPITE AND ANHYDRITE 78',1

TABLE1-Continued

2410 2450 2480 4 1 123 45667 zoned zoneo zoned zoned zoned zoned zoned zoned zoned zoned zoned zoned (very (very dark (very (very dark dark dark dark light dark dark slightly) slightly) core slightly) slightly) core core core core core nm core 2 e 3 o eee 33333 41.09 40 17 37.94 38.82 39.42 39.48 39.52 37.80 36.55 35.79 40.33 3926 13.40 13.32 17.42 16.23 15.53 16.21 1494 17.37 18.48 17.48 13.19 13.59 205 3.47 3j2 2.88 3.08 2.20 2.81 3.38 2.31 2.72 3.81 4.82 2607 24.61 23.87 23 90 24.47 24.24 24.60 23.09 23.94 24.07 24.42 24.56 n.o. 0.06 n.d. n.d. 0.05 n.d. n.d. n.d. n.d. n.d. n.d. n.d. 0.08 0.19 0.10 0.16 0.11 0.08 0.08 0.08 0.09 0.08 0.06 0.06 9.83 10.25 9.78 9.65 10.67 10.46 10.55 10.19 10.10 9.68 10.79 10.71 0.30 0.49 0.22 0.39 o.25 0 20 0.29 0.23 0.27 0.28 0.89 0.86 0.08 0.14 0 07 0.07 0.09 0.05 0.10 0.11 0.05 0.08 0.11 0.13 093 0.45 1.96 1.28 0.84 0.98 0.54 1.84 2.30 2.97 0.34 0.27 no n.d. n.d. 0.10 n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. 374 3.34 2.53 2.68 3.19 2.98 4.12 2.80 3.13 2.43 3.10 2.77 97.57 96.49 97.01 96.16 97.70 96.88 97.55 96.89 97.22 95.58 97.04 97.03 0.00 0.00 0.00 0.03 0.00 0 00 0.00 0.00 0.00 0.00 0.00 0.00 1.57 141 1.O7 1.13 1.34 1.25 1.73 1.18 1.32 1.02 1.31 1.17 9600 95 08 95.94 95.01 96.35 95.63 95.82 95.72 95.90 94.56 95.73 9586 2.42 2.54 2.95 2.84 2.66 2.76 2j9 2.79 2.64 2.90 2.68 2.82 9841 97.62 98.90 97 86 99.01 98.38 98.01 98.50 98.54 97.45 98.41 98.68 Cationscalculated on the basisof 11 oxygenequivalents, total Fe as FeO 2.941 2.921 2.739 2.812 2.832 2.837 2.860 2.748 2.657 2.650 2.917 2.849 1.059 1.079 1261 1.188 1.168 1.163 1.140 1.252 1.343 1.350 1.083 1.151 4.000 4.000 4.000 4.000 4.000 4.000 4.000 4.000 4.000 4.000 4.000 4.000 0.o72 0.062 0222 0.198 0.148 0.210 0.134 0.237 0.242 0.176 0.042 0.011 0 123 0.211 0.188 0.174 0.185 0.132 0.170 0.206 0.140 0.168 0.230 0.293 2.781 2.667 2.568 2.580 2.620 2.596 2.653 2.502 2.594 2.656 2.632 2.656 0.016 0.027 o.o12 0.021 0.014 0.011 0.016 0.013 0.015 0.016 0.048 0.047 0.005 0.009 0.004 0.004 0.005 0.003 0.006 0.007 0.003 0.005 0.007 0.008 2.996 2.975 2.994 2.979 2.972 2.952 2.979 2.964 2.994 3.021 2.960 3.015 0.000 o 005 0.000 0.000 0 004 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.011 oo27 0.014 0.022 0.015 0.011 0 011 0.011 0.013 0.011 0.008 0.008 0.898 0.951 0.901 0.892 0.978 0.959 0.974 0.945 0.937 0.914 0.996 0.991 0.026 0.013 0.0ss 0.036 0.024 0.028 0.015 0.052 0.066 0.086 0.010 0.008 n oa( 0.995 0.970 0.9s0 1.021 0 997 1.001 1.009 1.015 1.012 1.014 1.007 7.931 7.970 7 964 7.929 7.993 7 950 7.980 7.972 8.009 8.032 7.973 8.022 0 847 0.768 0 578 0.614 0.725 0.677 0.943 0.644 0.720 0.569 0.709 0.636 0 000 0.000 0.000 0.012 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 t.tcJ 1.232 1.422 1.374 1.275 1.323 1.057 1.356 1.280 1.431 1.291 1.364 2.000 2.000 2.000 2.000 2.000 2.000 2.000 2.000 2.000 2.000 2.000 2.000 0.042 0.073 0.068 0.063 0.066 0.048 0.060 0.076 0.051 0.060 0.081 0.099 0.424 0.384 0.289 0.307 0.363 0.338 0.472 0.322 0.360 0.285 0.3s5 0.318

especially from high-pressureand high-temperature re- provide substantialevidence that F may have been a sig- gimes (e.9.,Arima and Edgar, l98l) as a potential P-Z nificant component in certain hydrothermal systems.Hy- indicator, although Ti substitutions in phlogopite are not droxyl-bearing silicates, such as phlogopite, efficiently well understood (Dymek, 1983). Various schemeshave scavengeF from aqueous fluids, and F also greatly en- beenproposed: (1) a Ti-Tschermak'send member, 2torsi+ hances the thermal stability of phlogopite (Munoz and r61Mg+ 2r4rAl+ r5rTi(Robert, 1976),(2)lror![g ;- rorfi { Eugster,1969; Foley et a1., 1986). Furthermore,recent I6rtr(Bohlen et al., 1980),and (3) t6lMg+ 2(OHf = t6rTi advancesin curved-crystalspectrometer design allow for + 2O2- (Bohlen et al., 1980).The TiO, content of the routine F analysisas emphasizedby Petersenet al. ( I 982). SH2 phlogopites is low and rangesfrom 0.1 to 0.89 wto/o The fluoride-hydroxyl exchangein biotite has been in- with most values below 0.4 wto/o.Figure 14 shows that vestigatedexperimentally by Munoz and Ludington (1974) the mechanism of Ti substitution does not appear to rig- in terms of the fluid composition and other intensive pa- orously follow any of the above schemes.Nevertheless, rameters. Hence, we analyzed for F in SH2 phlogopites there is a weak negative correlation with Mg. Ti is gen- in an attempt to discernany correlation with temperature, erally higher in the darker zones (Fig. I l) of zoned crys- pressure,or bulk chemistry and to possibly place limits tals and generallyhigher in deeper samples. on the fluid composition. Although the geothermal so- F in SH2 phlogopites.Although F is a minor constituent lutions are Cl-bearing, the Cl content of SH2 phlogopite of many silicic-alkalic magma series(e.g., Bailey, 1977), was usually at or below the detection limit. recent studies (e.9.,Gunow et al., 1980; Plimer, 1984) The F/(F + OH) ratio for all SH2 phlogopites ranges 788 BELKIN ET AL.: HYDROTHERMAL PHLOGOPITE AND ANHYDRITE

8b..-. ...

Fe / (Fe+Mg) -? Fig. 10. Variationof Fe/(Fe+ Mg) in SH2 phlogopitesas a ct functionof samplelocation. All Fe assumedto be Fe2*. .8)o.re

and emphasizedby, e.9.,Munoz and Ludington (1974) and Valley et al. (1982).The SH2 phlogopiteshave a low Fe content [Fe/(Fe + MC) < 0.1], which may in part account for their high X.. Figure 16 shows a positive correlation of F with Mg2+. Such a good correlation is surprising consideringthe limited range of Mg'z+content o.12 and the higher uncertainty associatedwith F analysis. F in phlogopitezoning. The darker zonesofSH2 phlogopite (Fig. I 1) generallyhave lessF, as the Fe2+-F avoidance principle would demand. However, the distribution of F in phlogopites may be more complicated. For example, although most Mg-rich (light-colored) zoneshave more F than do the adjacent dark zones, in three cases they had less.Furthermore, a traverse acrossa light-colored zone (Fig. I l) showsthat the distribution is not constant. 2.4 2.6 2.9 3.O 3.2 Mg to.r.u.l ANuvnnrrn rRAcE-ELEMENTcHEMrsrRY The partition coefficient of minor elements between Fig. 9. The variationin SH2 phlogopitesof Mg and Fe pfu authigenicminerals and geothermalsolutions can be used on the basisof 1l oxygenequivalents. The line *-* showsthe parameters geo- positionof thesubstitution Mgt* = Fe2+.The circled data points to estimate some of the chemical of the indicatedark phlogopiteor dark phlogopitezones. The other thermal fluid and perhaps to resolve the precipitation datapoints are light phlogopitesor light phlogopitezones. The mechanismsresponsible for various authigenic minerals tielinesconnect dark (circled) and light zonesof thesame crystal. (e.g.,Holland, 1956; and Shikazonoet al., 1983).As a All Feassumed 1o be Fe2*. preliminary study of the trace elementsin SH2 anhydrite, crystal fragmentscarefully handpicked from various SH2 drill-cutting sampleshave been analyzedby a rrvnx en- from 0.28 to 0.57 (Table l). Using Figure 5 of Munoz ergy-dispersivesystem. Table 2 showsthe data as a func- and Ludington (1974) and the temperature data derived tion of depth and givesthe estimateduncertainties for Rb, from primary fluid inclusions in phlogopite, the logfsro/ Sr, Ba, La, and Ce. There are three major limitations and 1[. values in equilibrium with the phlogopite depositing assumptions of this study. First, the value determined solutionsare - >6. representsan averagefrom anhydrites that perhaps were Figure 15 shows the distribution of F with respect to deposited by different fluids. The salinity data (Fig. 5) samplelocation. Although the data are limited, there is a show that within the same sample, fluid inclusions have definite maximum at22l0 m with generallylower values different values. Second, although the anhydrites were at greaterand lesserdepths. carefully examined prior to crushing, we cannot preclude Fe2+-Favoidance. Although F substitutesfreely for hy- the presenceof contaminating phases. Third, we have droxyl, the principle of Fe2*-F avoidance-that an Fe- assumed equilibrium crystallization between the anhy- rich silicate requires higherfrrto stabilize a given X. than drite and precipitating solutions. the equivalent Mg-rich silicate-has beenstudied recently There is very little published information on the rare- BELKIN ET AL.: HYDROTHERMAL PHLOGOPITE AND ANHYDRITE 789

SOOpm

taaaaaa tt ataaaa a MnO'. aaraa {:" Fig. 12. The variationof BaO(wto/o) of SH2 phlogopitesas a functionof samplelocation (depth in meters).The Ba (ppm) {:' trace-elementdata in anhydriteare also shown.

Although the above data are limited, they corroborate our conclusions that the anhydrite and phlogopite were FeOr a precipitated from the same solutions and that the factors that causedthe elementalvariation in both minerals were similar. The generallyenriched trace-elementvalues support the thesis that the fluid is, in part, derived from a

ttt' a t . tt 1.O2 F a a o'

' BaO' 0.98 a. aaaa aoooo'a t at l a' j ..i Fig. 11. An electron-microprobetraverse (offset black line) o of a euhedral,zoned phlogopite crystal from the SH2 sample Yo.s+ 2210m. Thevariation of BaO,F, FeO(total Fe), TiOr, andMnO (wto/o)across the light-dark-lightsequence is shown.The beam sizewas -10 pm in diameter,and the standarddeviation (1o) associatedwith standardizationis shown. earth-elementcontent of anhydrite; however,Morgan and Wandless (1980) presenteddata on two samples of hydrothermal anhydrite. Their REE values (in ppm; sample PBB-2-78,La: 9l andCe : 180;sample PBB-3-78, La : 52 and Ce : 102) are similar to those found in the SH2 anhydrites. 0.86 o o.o4 0.08 0.12 The most notable and consistentaspect of the data for Ba(p.r.u.) all elementsis a minimum at 2040 m with generally in- creasingvalues to lesserand greaterdepths. If we compare Fig. 13. The variation of K and Ba (ptu) of the SH2 phlog- the Ba data ofanhydrite with the BaO content ofcoex- opites on the basis of I I oxygen equivalents. The substitution = = isting phlogopite (Fig. l2), we see a similar pattern, but schemesBa2+ + Al3+ K+ + Si4+ and Ba2+ * E 2K* are *-*. for from displaced toward slightly greater depth. Furthermore, a shown as the lines Regressioncurves the data samplesfrom 1600, 2210, and2489 m are shown. The data for comparison of F in coexistingphlogopite (Fig. I 5) shows 2210 m have been divided into two groups; thus two regression a definite pattern antithetic to the anhydrite trace-element curvesare shown for 2210m. 2210m (uppercurve), r : -0.818; and phlogopite BaO data; the F maximum is in a similar 2210 m (lower curve),r : -0.919;2480 m, r : -0.970; 1600 position to their minima. m, r : -0.989. 790 BELKIN ET AL.: HYDROTHERMAL PHLOGOPITE AND ANHYDRITE

o.o5

2190m 221Om t ro.o o

A A A

3'o a'o 5'o F(*,r) Fig. 15. The variation of F (wto/o)in SH2 phlogopitds as function of sample location (depth in meters).

pressures, limited ? various salinities, and temperaturesare 1983). o (e.g.,Kushnir,1982b; Shikazonoand Holland, The distribution of Sr and Ca between two phasesis I governed by the equation o.02 : K"* (a..roo/ a"^"o o)/ (a",r* / aqur*), where K"* is the distribution coefficient, c5,560arrd a"^"oo are the activities of the SrSOoand CaSOocomponents of anhydrite, and a.,r* and a"^r*are the activities ofSr2* and Ca2* in solution. K"" is a Berthelot-Nernst-type distri- o.o1 bution coefficient(e.g., Mclntire, 1963). We have calculatedthe Sr/Ca ratio in the precipitating brine for four samples(Table 3). Kushnir (1982a)studied the partitioning of Sr betweenanhydrite and brine during gypsum to anhydrite transformation at temperaturesbe- low 150 "C. He found a linear decreaseof Ki' with temperature. Shikazono and Holland (1983) found that Ki' dependsless on the NaCl concentration,temperature, and precipitation rate than on the degreeof supersaturation 2.4 2.6 2.8 3.O of the solutions with respectto anhydrite and/or the mor- Mgtn.t.u.t phology of the precipitated anhydrite crystals. They distinguished acicular and rectangular anhydrite. The rect- Fig. 14. The variation in SH2 phlogopitesof Ti and Mg (pfu) angular anhydrite morphology is similar to the SH2 on the basis of I I oxygen equivalents.The substitution schemes anhydrite. The Shikazono and Holland (1983) K$ values 2Mgt* - Ti4+ + tr and Mgr* + 2Si4++ Ti4+ + 2Al3* are shown between solution and rectangular anhydrite at 150, 200, by the lines *-*. and250 "C are0.35,0.24,0.27 (+9.95;, respectively. Note that they show a minimum at 200'C. Thus, using these experimental K!'values, the Sr/Ca ratio in the precipi- highJevel fractionating magma (Cavarretta and Tecce, tating solution rangesfrom 0.060 to 0.10. However, the l 987). experimental K!' values do not agree with the higher K!' values of natural hydrothermal anhydrites from the Sr in SH2 anhydrite East Pacific Rise (Shikazonoand Holland, 1983) or from Natural anhydrite contains minor amounts of Sr, which vein anhydrites from the Salton Sea geothermal system substitutesfor Ca in the anhydrite structure (Purkayastha (M. A. McKibben, pers.comm., 1987).The Sr/Ca ratio and Chatterjee, 1966).The coprecipitation of Sr with cal- of seawater (-0.021) is significantly ditrerent from those cium sulfate minerals has been studied to try to clarify values (Table 3) calculated with the experimental K:' the depositional environments and especially the brine data. If the higher, natural K!' values are used, the SH2 composition(e.g., Herrmann, 1961;Butler, 1973;Kush- Sr/Ca ratio would be closer to the sea-waterratio. Thus, nir, 1982a).However, experimental studies of the parti- we cannot preclude the presenceof seawater as a com- tioning ofSr betweenanhydrite and aqueoussolutions at ponent in the SH2 geothermal fluid. BELKIN ET AL.: HYDROTHERMAL PHLOGOPITE AND ANHYDRITE 791

TneLe2. xRF-KEVEXanalyses (in ppm) of anhydriteseparates from ther SH2 well

Depthof drillcuttingsample (m) 1300 1600 1990 2040 2150 2190 2410 2480 Rb 190 320 100 40 110 130 450 670 Sr 6100 5700 6200 2000 5100 8400 5600 5400 '| oa Ba 1600 3800 1300 10 730 280 3600 3900 A 400 70 30 <2 14 110 130 4 500 80 20 16 50 30 210 180 ooo Note.John R Evans,analyst. The uncertaintyis +5% of the givenvalue a'o exceptfor values<20, wherethe uncertaintvis +10%. llo a DrscussroN Deposition of anhydrite Quartz, calcite, and anhydrite are perhaps the three most common primary gangueminerals in hydrothermal ore deposits(Samartsev et al., 1980; Shikazonoet al., 1983; Drummond and Ohmoto, 1985).The conditions that deposit anhydrite are of interest to help model the geothermal and ore-forming environments. Geothermal solutions at the SH2 well site have deposited anhydrite Fig. I 6. Thevariation of Mg (pfu)(on the basis of I I oxygen as the result ofperhaps various processes. equivalents)and F (wto/o)in SH2phlogopites. Anhydrite solubility in pure water at geothermal temperaturesis retrograde;however, the retrogradesolubility reachesa minimum with increasingNaCl content at con- Kuroko deposits. Fluid inclusions in these Kuroko an- stant pressure(Blount and Dickson, 1969).Blount and hydrites have In values that range from 240 to 340 "C Dickson (1969) also found that a pressureincrease raises with salinity values from 0 to 5 wto/oNaCl equivalent. the solubility of anhydrite over the entire range of their They concluded on the basis ofthese data, Sr concentra- 875r/865r experimental conditions. However, the retrograde effect tions, and ratios that anhydrite deposition was of increasingtemperature far outweighsthe positive effect the result of a subsurfacemixing of fluids. Although we of increasing pressure except in NaCl solutions greater find some evidence for perhaps multiple fluids at SH2, than 2M. The range of salinity from our fluid-inclusion our data are insufficient to discuss deposition by this studiesin SH2 anhydrite is 0.5 to 14.0 wt0/oNaCl equiv- mechanism conclusively. alent,which correspondsto -0 to -2MNaCL Thus, the Phlogopite and biotite in geothermalsystems effect of a minimum in anhydrite solubility as a function of salinity would be minimized in SH2 solutions. How- Biotite is a common primary constituent of the host ever,the presenceofthe other dissolvedcomponents (Ba2+, rock in many geothermal systemsbut is rare as a hydro- Sr'z*)and volatiles (COr, B, HF, hydrocarbons)may have thermal mineral (Browne, 1984) except in the deeper, significantly enhancedthis minimum solubility effect. higher-temperatureportions of somegeothermal systems. Drummond and Ohmoto (1985)have modeledboiling Biotite appearsbetween the 325 and 360 "C interval in hydrothermal systemsand concludedthat anhydrite, and Borehole Elmore I, Salton Sea geothermal field (Mc- sulfatesin general,will not be common depositionalprod- Dowell and Elders, 1980),in the hotter (>325 "C) parts ucts of boiling. However, in relatively dilute and slightly of the Cerro Prieto field (Elders et al., 1979), at Ngawha, acidic solutions of about 250 "C, the decreasein proton New Zealand, at temperatures above 220 "C (Browne, concentration associatedwith boiling can cause sulfate 1984),and in wellsMF2, MF5, and SV3 in the Phlegraean minerals (anhydrite) to precipitate (Sasada, 1986). Queen and Motyka (1984) found that in water-dominated zones TneLE3. Calculationof the Sr/Caratio in thegeothermal fluid of the Makushin geothermal system,anhydrite is a com- of theSH2 well mon authigenic mineral, whereas it is absent from the Calculated vapor-dominated zones. We find no petrographic evi- Sample In-holef Sr ppm ftuid dence for boiling in SH2, and the measured down-hole depth (m) ('C) (Table2) K:" sr/ca temperatures and the similar homogenization tempera- 1300 143 6100 035 0.060 tures lie far to the left of the depth with boiling-point 1600 190 5700 o.24 0.082 2150 -250 5100 0.27 0.064 curve calculatedfor this environment. 2190 -250 8400 0.27 0.10 Shikazonoet al. (1983) have examined the mode of - Shikazonoand Holland(1983). anhydrite occurrenceand its depositional mechanismsin 792 BELKIN ET AL.: HYDROTHERMAL PHLOGOPITE AND ANHYDRITE

Fields at temperaturesabove 300 "C (Belkin et al., 1986). .SH2 PHLOGOPITES Cavarretta et al. (1985) reported biotite as a contact metasomatic mineral at or above 300 "C in the Latera geothermal field, Italy. Biotite is also reported in the SH2 well as an uncommon phase below 1600 m and above 200 "C (Cavarretta and Tecce, 1987). Another phyllosil- icate not generallyfound in geothermal systemsis lepid- 6 olite. However, from Yellowstone National Park bore- ae holes, Bargar et al. (1973) reported lepidolite forming F =4 between130 and 140'C. Phlogopiteis also a rare authigenic geothermalfields. Cavarrettaet al. (1985) IL mineral in reported phlogopite from the Lalera geothermal field, It- 2 aly, in the contact metasomaticzone above 200 "C. Phlog- opite has also been reported in paleohydrothermal sys- o tems. Zhang (1986) reported the assemblagephlogopite + anhydrite (with pyrite) from a volcanic area in the 8 Yangtze River valley. Fluid-inclusion studies in the anhydrite indicate the temperature to be between 300 and 500'c. n6 - The common occurrence of authigenic phlogopite in = the SH2 well indicates that equilibrium crystallization of 4 phlogopite can occur at temperaturesfrom - 200 to - 300 d\r 'C F and pressuresfrom 140 to 220 bars from a F-bearing, volatile-rich, generallydilute geothermal solution. 2 Cornparisonwith phlogopitesfrom other environments Phlogopiteis often the only volatile-bearingconstituent of mantle-derived rocks and potassic mafic and ultra- mamfic rocks. Becauseof its petrogenetic importance, geothermometersand geobarometersinvolving various elementalvariations have beenproposed (e.g., Luhr et al., 1984;Tronnes et al., I 985).Figure 17 shows a comparison a of F, TiOr, and BaO vs. FeO as total Fe betweenthe SH2 F phlogopites and phlogopites or phlogopitic biotites from = different metamorphic and igneousenvironments. This is o o not a comprehensivesurvey and was further limited by d} the paucity ofstudies that provided data on all three elements-F, Ba, and Ti. Minor-element abundancesin any mineral are a complex interrelationship of intensive and extensive parameters. SH2 phlogopites represent compositions formed from a low-temperature, low-pressure, aqueous,volatile-rich environment and might be used to constrain the lower-pressure,lower-temperature stability FeO WT% TOTAL IRON field of phlogopite compositions. Fig. 17. A comparison of the F, TiOr, and BaO vs. FeO A generally negative correlation of F with FeO is ap- variations for SH2 and other phlogopites. ESS : Exley et al. parent (Fig. 17) and would be consistentwith the crystal- (1982), host : spinel therzolite (F was not detectedin micas with chemical rule of Fe2+-Favoidance. Two fields of phlog- FeO > 3 wt9o). M : Matson et al. (1986), host : peridotite opites (M, ESS, Fig. 17) from kimberlites and spinel xenoliths (primary and secondary)and mica-rich xenoliths (glim- lherzolites have a low F content with respect to other merite) in kimberlite. BS1,BS2 : Bachinski and Simpson(1984), phlogopites with similar FeO content and probably rep- : : : host minettes, BSI group I phenocrysts,BS2 group II resentthe influence of a low-F bulk composition. A pos- : groundmass.JSl, JS2 Jonesand Smith (1985),host: minettes, itive correlation of TiO, with FeO (Fig. 17) suggeststhat, : Hannaborough locality, : Loxbeare Farm JSI Quarry JS2 in general,the substitution Mg2+ + f'sz+ controls the Ti locality. Vl, V2 : Valley et al. (1982), host : marble, Vl : content and/or a bulk composition richer in Fe is also sampleGOV50 -2, V2 : sampleBMT- I (BaO for GOV50-2 was not determined). W : Wendlandt (1977), host : monticellite peridotite. GW : Gasparand Wyllie (1982),host : carbonatites. Y : Yau et al. (1984), host : granulite-facies(metasedimentary) lamprophyres.C : Cross(1897), host : wyomingite. AC : Allan retrogradecalcsilicate rock. SSN : Schulzeet al. (1985), host : and Carmichael (1984), host : lamprophyric lavas. BELKIN ET AL.: HYDROTHERMAL PHLOGOPITE AND ANHYDRITE 793 richer in Ti. The position of the F and Ti SH2 phlogopite meability by mineral deposition, and then repeated hy- compositional fields with respect to other fields suggests drofracturing. that neither pressurenor temperature alone control Ti or F variation. Evolution of the geothermal system The pattern of BaO distribution (Fig. l7), however, A comparisonofthe well-fluid compositionand meaured suggeststhat the major controlling factor in the Ba content in-hole temperatureswith the fluid inclusion Zn and sa- of phlogopitesis the bulk composition. In general,phlog- linity data would allow an assessmentof geothermalevo- opite has BaO values of I wto/oor less.Three fields (SH2 lution (e.g.,Browne et al., I 976;Belkin et al., I 986)'How- phlogopites,GW, W) are major exceptions.The replace- ever, becauseno fluid was recoveredfrom the SH2 well, ment of K+ by Ba2+in the phlogopite interlayer site re- our analysisis limited. quires chargecompensation that can be achievedby var- The correspondenceofthe 7i data (Fig. 4) for anhydrite ious substitution schemes (see above). Higuchi and with in-hole temperaturesis quite good. The scatterof the Nagasawa(1969) and Jensen(1973) reported partition ?",values and especiallythe low ?] values from 2410 m coefficientsin a dacite-Ku" for biotite/matrix of - l0 and could result from one or a combination of factors: (l) the K" for biotite/matrix of -2. Thus, phlogopite would act hydrologic regime (water table) was different at the time as a very effectivescavenger ofBa. The variability ofBaO oftrapping than the regime that exists today, and hence at relatively constantFeO content in the SH2 phlogopites the pressure correction is in error; (2) the drill-cutting suggestssignificant Ba variations in the geothermal fluid. samples may have significant vertical-and hence temperature-scatter; (3) inclusions that had necked-down (4) the sample may representseveral Hydrocarbons in the inclusion fluid were measured;and different times, temperatures,or pressuresof formation. The occurrenceofliquid hydrocarbonsand a gas-clath- The L data (Fig. 6) of phlogopite show a similar corre- with greater rate a melting temperature significantly than spondenceto the in-hole temperature. The data are less l0'C are not common in the fluid inclusions in the SH2 numerous than the T, data for anhydrite, and this un- presence in geothermal anhydrites. The ofhydrocarbons doubtedly adds to th€ scatter. Nevertheless,the T, dala (e.g., and hydrothermal environments is not unusual for anhydrite and phlogopiteindicate that theseauthigenic Roedder,1984; Hodge and Norman, 1985),especially if phasesare relatively recent and/or the thermal flux ofthe passed such environments are located in, or if the fluid water-dominated geothermal systemhas remained stable presence through, a sedimentary sequence.Thus, the of during the latter stagesof its evolution. The salinity data hydrocarbons in the SH2 inclusion fluid suggestssome for anhydrite and phlogopite (Figs. 4 and 5) suggestthe interaction with or fluid derivation from the local sedi- existenceof multiple fluids that resulted in the trapping mentary environment. of a rangeof salinities (-0 to -14wto/o NaCl equivalent). However, the data are too limited to define any system- CO, in the inclusion fluid atics between sample location and host. The evidence for CO, as a volatile constituent in fluid inclusions in SH2 anhydrite is widespread but not uni- An episodic and multiple-fluid environment versal. Crushing and freezing data suggestthat during some The large-scale(light-dark, etc.) zoning in some SH2 anhydrite-depositional periods, the percipitating fluid phlogopites and their element variations indicate an ep- contained COr. Becausethe drill-cutting samplemay rep- isodic and/or multiple-fluid environment during the evo- resent the entire depositional time range, each sample lution of the SH2 geothermal system. The major phlog- would tend to show the complete range of fluid compo- opite color-zone difference is associatedwith the Mg-Fe sitions. The minimum CO, concentration necessaryto substitution. The dark zones are enriched in Fe but de- produce CO, clathrate in liquid-rich inclusions trapped pleted in F relative to the light zones.A plausible expla- in the temperaturerange of 200 to 300'C is 1.5 molo/o nation for such a distribution would be the episodic in- (Sasada,1985). troduction of magmatic fluid into the geothermalregime. The rare occurrenceofliquid-CO, fluid inclusions from A fluid, perhaps derived from a fractionating trachytic geothermal fields has been discussedby Sasada(1985). magma (Cavarretta and Tecce, 1987),would be enriched He has arguedthat pressuresgreater than hydrostatic are in HF and other volatile species(COr, etc.). This would necessaryto form inclusions with liquid CO, (at room reduce the activity of HrO and hence the activity of Fe. temperature)in the geothermalenvironment. A vein sys- Phlogopite crystallizing at this time would be depletedin tem, sealedby mineral deposition, could representsuch Fe but enriched in F. The common sharp, color-zone a circumstance.The self-sealedvein systemcould still be boundaries suggestthat these pulsesproduced relatively subject to greater than lithostatic pressure from the as- rapid changesin fluid conditions. Cavarretta and Tecce cending fluid, especially if some or all of the ascending (1987) have reported similar zoning in vesuvianite and liquids and volatiles evolved from crystallizing magma. suggestedthat it has resulted from the processofepisodic In the SH2 well, the abundant filled and cross-cutting hydrofracturing. Uncommon phlogopite dissolution tex- veins suggestthat such overpressureresulted in common tures also suggestthat during some periods the fluid be- hydrofracturing, subsequentself-sealing of fracture per- came undersaturatedwith respectto phlogopite. 794 BELKIN ET AL.: HYDROTHERMAL PHLOGOPITE AND ANHYDRITE

The BaO distribution in the phlogopites, where, for possible magmatic source for the geothermal COr. Cor- example, light zones can be enriched or depleted with tecci et al. (1980) reported that the 63aSvalues for sulfate respectto the dark zones, suggeststhat the episodic con- from the CesanoC-l well brine and anhydrites from the ditions are not just the simple alternation of two uniform CesanoRC-l well show similar valuesof about * 157o, fluid compositions or physicochemical conditions. The but quite different6t8O values of +4.8Vmand *l4.9Vm + Ba distribution indicates that each fluid pulse was some- 0.37m,respectively. The low d'8Ovalue of +4.87@suggests what different with respect to some components. This that the origin of the Cesanosulfates may not be just the arguesfor an evolving magmatic fluid or varying degrees simple dissolution and reprecipitation of Upper Triassic of interaction (mixing) with another Ba-rich fluid. It is sedimentaryevaporites. interesting to note that the first Ba-rich igneous biotite, A detailed S- and Sr-isotope study of the authigenic containing 7.32 wto/oBaO, was reported by Thompson mineral assemblagein the SH2 well would provide valu- (1977) ftom a leucitite from the Alban Hills, Italy, a sim- able information on the oriein and evolution of the SH2 ilar volcanic district in the perpotassicRoman province. geothermal fluid. Subsequentreported Ba-rich biotites from similar feld- spathoid-bearingvolcanic rocks are interpreted as late- crystallizing phases. CoNcrusroNrs Although the element distribution can be zoned within 1. Phlogopite and anhydrite (plus pyrite and calcite) a single phlogopite crystal, the element distribution with occur as authigenicand hydrothermal minerals in fracture respectto sample location (depth) in the phlogopites and and pore-spacefillings in the SH2 geothermalwell, located anhydrites also indicates zoning. The stratigraphic inter- in the Sabatini volcanic district, Latium, Italy. Also pres- val from approximately 2000 to 2200 m showsa general ent in the well, at various levels, are K-feldspar, garnet, depletion of BaO in phlogopite and trace elementsin an- vesuvianite, apatite, spinel, pyroxene, wollastonite, wilk- hydrite and an enrichment of F in phlogopite. This sug- eite, cuspidine, harkerite, cancrinite, and reyerite. gestsa stratigraphic zone where the presenceof a fluid 2. The primary fluid inclusions in phlogopite and the rich in Ba, Sr, and other trace elementswas limited. This primary and secondaryfluid inclusions in anhydrite have implies lithologic and/or permeability differencesor per- vapor-liquid homogenization temperaturessimilar to the haps a fluid stratification (Oakes, 1985). measuredin-hole temperature of the SH2 well. This in- The origin and evolution of various S speciesin geo- dicatesthat the minerals are relatively recent and/or that thermal systemsrelated to volcanic districts is the subject the thermal regime has remained stable during the past. of continual investigation and debate (e.g., Kiyosu and The salinity data show a considerablescatter; this and the Kurahashi, 1983).Furthermore, the presenceof SO, and rare occurrence of hydrocarbons and daughter crystals stableigneous sulfate minerals (e.g.,anhydrite) in magmas indicate the presenceof fluids with different compositions. and their products has received considerablerenewed in- Crushing data, clathrate formation, and the rare occur- terest,especially after the eruption of El Chich6n in 1982 rence of liquid CO, also indicate that CO, was a common (Luhr et al., 1984).Rye et al. (1984)argued on the basis volatile species. of stable isotopesthat a major source of the high El Chi- 3. Detailed microprobe analysesof unzoned and zoned ch6n S content (pre-eruption, -2.6 wto/oSO.) must be, in phlogopite crystals that contain primary fluid inclusions part, magmatic and thus the S was not entirely remobi- reveal that they have nearly end-member compositions, lized from nearby evaporites. The releaseof SO, as part with Fe/(Fe + Mg) : 0.02 to 0.l, and have variablecon- of the volatile magma fraction degassingfrom the upper tents of F, BaO, and TiOr. This variability is directly portions of a fractionating magma will eventually allow related to the phlogopite zoning and samplelocation. The the S to become part of a hot ascendingaqueous fluid. crystal zoning indicates episodic changesin growth con- TheseS and other volcanic exhalationswill participate in ditions whereas the variation with depth also suggests reactions with magmatic, connate, and/or meteoric some kind of lithologic (permeability?)control on growth aqueousfluids. Kiyosu and Kurahashi (1983) suggested conditions.e.g.. fluid mixing. that the sulfate-rich waters in volcanic areasresult from 4. Energy-dispersivetrace-element analysis of anhydrite a disproportionation reaction of SO, to yield the more crystalsshows a relatively high abundanceof Sr, Ba, Rb, oxidized sulfate species.HrS is also produced and would Ce, and La. The overall variation of anhydrite composi- tend to react with available Fe to form, for example, py- tion with samplelocation is similar to that of the variation rite-a ubiquitous accessoryin the SH2 well. Thus, it is of phlogopitechemistry with samplelocation and suggests reasonableto assumethat the S present in the SH2 well a similar and related control in growth conditions. could be partially magmatic and that appeal to the re- 5. The fluid-inclusion salinity data,the presenceofzoned mobilization of Mesozoic evaporites as the sole S source phlogopites,and the compositional variation among dif- is unnecessary. Bertramiet al. (1986)compared the DrsO ferent samplesofphlogopite and anhydrite reveala picture and d'3Cin calcitesfrom various Latium geothermalwells of episodic and/or changingfluid and/or mineral growth and their host Mesozoic carbonates.The D'8Oof SH2 conditions during the evolution of the SH2 geothermal calcites is significantly diferent from the Mesozoic car- field. The presenceof hot, mineral-rich, ascendingsolu- bonates, being more than 20Vmlighter. This suggestsa tions has effectively reducedthe permeability of the host BELKIN ET AL.: HYDROTHERMAL PHLOGOPITE AND ANHYDRITE 795 rock by successiveepisodes ofdeposition in pore spaces Calamai,A., Cataldi, R , Dall'Aglio, M., and Ferrara,G.C. (1975)Prelim- and fractures. inary report on the Cesanohot brine deposit (Northem Latium, Italy). United Nations SymposiumofGeothermal Energy,San Francisco, Cal- AcrNowr,BocMENTS ifomia,1,305-313 Calamai,A., Fiordelisi,A., Pandeli,E., and Valenti, P. (1985)First ex- The AGIP-ENEL Joint Ventureis gratefullyacknowledged for supplying ploratory well in the Sabatiniarea (northern Latium). In A.S. Strub and the samplesand drillingdata of the SH2 well. Dr. JeanDubessy of CREGU P. Ungemach, Eds., European geothermal update, p 433439. Pro- (Nancy, France)kindly allowed accessand instruction (F. T.) for the laser ceedingsofthe Third International Seminar on the ResultsofEC Geo- Raman probe. Financial support to this project for G. C., B. D. V., and thermal EnergyResearch, Munich, 1983. F T. was provided by C.N.R., Gruppo Nazionale di Vulcanologia. We Cavarretta, G., and Tecce, F (1987) Contact metasomatic and hydro- thank Malcom Ross (USGS) for determining the phlogopite crystallog- thermal minerals in the SH2 deep well, Sabatini volcanic district, La- raphy and John R. Evans (USGS) for the KrvEx analyses.We also thank 1ium,Italy, Geothermics,16, 127-145. Elaine S. McGee (USGS) and Marta J. K Flohr (USGS) for stimulating Cavarretta,G., Gianelli, G., and Puxeddu, M. (1982) Formation of au- discussionsconceming phlogopite chemistry. The manuscript benefited thigenic minerals and their use as indicators of the physicochemical from the careful reviews of E. Roedder. M.J K Flohr. and M. A Mc- parametersofthe fluid in the Larderello-Travalegeothermal field. Eco- Kibben. nomic Geology, 77, 107l-1084. Cavarretta,G , Gianelli, G , Scandiffio,G., and Tecce,F. (1985)Evolution RnrnnnNcns crroo of the Latera geotherrnalsystem II: Metamorphic, hydrothermal mineral assemblagesand fluid chemistry. Journal ofVolcanology and Geo- Allan, J.F, and Carmichael,I.S.E. (1984) Lamprophyric lavas in the Co- thermalResearch. 26. 337-364. lima graben,SW Mexico. Contributions to Mineralogy and Petrology, Collins, P.L.F. (1979) Gas hydrates in COabeanng fluid inclusions and 88.203-216. the use of freezingdata for estimation of salinity Economic Geology, Arima, M., and Edgar,A.D. (1981)Substitution mechanisms and solu- 74.1435-1444. bility oftitanium in phlogopitesfrom rocks ofprobable mantle origrn Cortecci,G., Bertrami, R., and Ceccarelli,A (1980) Circulation patterns Contributions to Mineralogy and Petrology, 77,288-295. and geothermometryofsome Italian spring systemsby sulfateisotopes Bacbinski,S.W., and Simpson,E.L. (1984)Ti-phlogopites of the Shaw's Third International Symposium on Water-Rock Interaction, Edmon- Coveminette: A comparisonwith micasofother lamprophyres,potassic ton, ExtendedAbstracts, I l5-l 16. rocks, kimberlites, and mantle xenoliths. American Mineralogist, 69, Crawford, M.L. (1981) Phaseequilibria in aqueous fluid inclusions. In 4r-56. L.S. Hollister and M.L. Crauford, Eds., Fluid inclusions:Applications Bailey,J.C. (1977)Fluorine in granitic rocksand melts, a review. Chemical to petrology, p 75-100. Mineralogical Association of Canada Short Geology,19,142. CourseHandbook. Baldi, P., Decandia,F.A.,LazzarcIto, A, and Calamai,A. (1974)Studio Cross,W (1897) Igneousrocks ofthe l€ucite Hills and Pilot Butte, Wy- geologicodel substratodella copertura vulcanica laziale nella zona dei oming. American Journal of Science,fourth series,4, I I 5-14 I laghi di Bolsena,Vico e Bracciano Memorie della SocietdGeologrca Cundari, A (1979) Petrogenesisof leucite-bearinglavas in the Roman Italiana.13. 575-606. volcanic region,Italy: The Sabatini lavas.Contributions to Mineralogy Bargar, K.E., Beeson,M H., Fournier, R O., and Muffier, L.J.P (1973) and Petrology,70, 9-21. Present-daydeposition oflepidolite from thermal watersin Yellowstone De Rita, D., Funiciello, R., Rossi, U., and Sposato,A. (1983) Structure National Park. American Mineralogist, 58, 901-904. and evolution ofthe Sacrofano-Baccanocaldera, Sabatini volcanic com- Baronnet, A., and Velde, B. (1977) Iron content of synthetic phlogopite plex, Rome Journal ofVolcanology and GeothermalResearch, l7 ,219- as a function ofgrowth rate Earth and Planetary ScienceLetters, 37, z)o. l 50-l 53. Dhamelincourt,P., Beny,J.M., Dubessy,J., and Poty,B (1979)Analyses Belkin, HE, De Vivo,8., Roedder,E., and Cortini, M. (1985) Fluid d'inclusions fluides i la microsonde MOLE a effet Raman. Bulletin de inclusion geobarometry from ejected Mt. Somma-Yesuviusnodules. Min6ralogie,102, 600-610 American Mineralogst, 70, 288-303. Di Filippo, M, Funiciello,R, and Parotto,M., Eds.(1984) Geological Belkin, H.E., Chelini, W., De Vivo, 8., and Lattanzi, P. (1986) Fluid map of the Sabatini area.Conisiglio Nazionale delle RicercheProgetto inclusionsin hydrothermal minerals from Mofete 2, Mofete 5 and San F inalizzato Geodinamica Vito 3 geothermalwells, PhlegreanFields, Campania, Italy. Interna- Di Girolamo, P. (1978) Geotectonic setting of Miocene-Quatemaryvol- tional Volcanological Congress,Proceedings of Symposium 5: Volca- canism in and around easternTyrrhenian Seaborder (Italy) as deduced nism, Hydrothermal Systemsand Related Mineralization, 7-12. by major elementgeochemistry. Bulletin Yolcanologiqte, 4243, 229- Bertrami,R., Friedman,I., Gleason,J., Lombardi, S., and Pandeli,E. 2s0. (1986) Isotopic composition of carbon and oxygen in the carbonate Drummond, S. E., and Ohmoto, H. (1985)Chemical evolution and min- materialspresent in Italian geothermalwells: On theirgenesisandgeoth- eral deposition in boiling hydrothermal systems.Economic Geology, ermometric application.Fifth Intemational Symposiumon Water-Rock 80, 126-147. Interaction, Iceland, ExtendedAbstracts, 56-59. Dymek, R. F. (1983) Titanium, aluminum and interlayer cation substi- Blount, C.W , and Dickson,F W. (1969)The solubility of anhydrite(CaSO") tutions in biotite from high-gradegneisses, West Greenland.American in NaCl-HrO from 100 to 450 .C and I to 1000 bars Geochimica et Mineralogist,68, 880-899. CosmochimicaActa, 33, 227-245. Elders,W.A., Hoagland,J R., McDowell,S.D., and Cobo R, J.M (1979) Bohlen,S.R, Peacor,D.R., and Essene,E.J. (1980) Crystal chemistry of Hydrothermal mineral zonesin the geothermalreservoir of Cerro Pne- metamorphic biotite and its significancein water barometry. American to In W.A. Elders,Ed., Geologyand geothermicsof the Salton Trough, Mineralogist,65,5542 Field Trip no. 7, p. 3643 GeologicalSociety of America 92nd Annual Browne, P.R.L. (1984)Lectures on geothermalgeology and petrology.The Meeting,November 1979. United Nations University, Geothermal Training Programme,Iceland, Exley,R. A., Sills,J. D, and Smith, J. v. (1982)Geochemistry of micas Report 1984-2,92 p from the Finero spinel-lherzolite,Italian Alps. Contributions to Min- Browne, P.R.L., Roedder, E., and Wodzicki, A. (1976) Comparison of eralogyand Petrology,81, 59-63. past and present geothermalwaters, from a study of fluid inclusions, Fisher, J.R. (1976) The volumetric properties ofHrO-A gtaphical por- Broadlandsfield, New Zealand. In J Cadek, and T. Paces,Eds , Pro- trayal U.S GeologicalSurvey Journal ofResearch, 4, 189-193. ceedings,Intemational symposium on water-rock interaction, Czecho- Flohr, M J.K. (1983) BASIC programs for calculation of cation site oc- slovakia, 1974, GeologicalSurvey, Prague, 140-149 cupanciesand plotting ofmineral data with applicationsfor evaluation Butler, G.P (1973) Strontium geochemistryof modem and ancient cal- of metamorphic mineral assemblages.U S. Geological Survey, Open- cium sulphateminerals. In B.H. Purser,Ed., The PersianGulf, p. 423- File Report83-905, 65 p. 452 Spnnger,Berlin. Foley,S.F, Taylor, W.R., and Green,D.H. (1986)The effectof fluorine 796 BELKIN ET AL.: HYDROTHERMAL PHLOGOPITE AND ANHYDRITE

on phaserelationships in the systemKAlSiOr-MgSiOn-SiO, at 28 kbar Mclntire, W.L. (1963) Trace element partition coefrcients-A review of and the solution mechanismoffluorine in silicate melts. Contributions theory and application to geology.Geochimica et CosmochimicaActa, to Mineralogy and Petrology, 93,46-55. 27, 1209-1264. Funiciello, R., and Parotto, M. (1978) Il substratosedimentario nell'area Morgan, J.W., and Wandless,G.A. (1980)Rare earth elementdistribution dei Colli Albani: Considerazionigeodinamiche e paleogeografichedel in some hydrothermal minerals: Evidence for crystallographiccontrol. marginetirrenico dell'Appennino centrale.Geologica Romana, l7 ,233- Geochimica et Cosmochimica Acta. 44.973-980. 287. Munoz, J.L., and Eugster,H.P. (1969) Experimental control offluorine Gaspar, J.C., and Wyllie, PJ. (1982) Barium phlogopite from the Jacu- reactions in hydrothermal systems.American Mineralogist, 54, 943- piranga carbonatite,Brazll. American Mineralogist, 67, 997-l0OO. 9s9. Gunow, A.J, Ludington,S., and Munoz, J.L. (1980)Fluorine in mica Munoz, JL., and Ludington, S. (1974) Fluoride-hydroxyl exchangein from the Henderson molybdenite deposit, Colorado. Economic Geol- biotite. American Journal of Science,27 4, 396413 ogy,75,ll27-1137 Oakes,C.S. (l 985) Strong salinity gradientsin the northeasternpart ofthe Haas, J.L., Jr., (1971) The effect of salinity on the maximum thermal Salton Seageothermal field, CA: A fluid inclusion and brine chemistry gradient of a hydrothermal system at hydrostatic pressure.Economic study. GeologicalSociety ofAmerica Abstractswith Prograrns,I 7, 679. Geology,66,940-946. Petersen,E.U., Essene, E.J , Peacor,D.R., and Valley, J.W. (1982) Fluorine Hattori, K., Cameron, E.M., and Smith, R (1985) Occurrenceand geo- end-membermicas and amphiboles.American Mineralogist, 67, 538- chemistry of anhydrite in the Hemlo Basin. GeologicalAssociation of 544 Canadaand Mineralogical AssociationofCanada, Joint Annual Meet- Plimer, I.R. (1984) The role of fluorine in submarine exhalative systems ing 1985, Program with Absfacts, 10, A25. with specialreference to Broken Hill, Australia Mineralium Deposita, Hedenquist,J.W., and Henley, R.W. (1985)Effect ofCO, on freezingpoint 19, 19-25 depressionmeasurements offluidinclusions-Evidencefrom active sys- PomArleanu,V., and Sabliovschi,V. (1980)Contribution to the geochem- tems and implications for epithermal ore deposition. Economic Geol- istry of muscovite from Romania 26th Intemational GeologicalCon- ogy,80, 1379-1406. gress,Paris, Abstracts, l, 139. Herrmann, A.G. (1961) Zur Geochimie des Strontiums in den salinaren Potter, R.W., L (1977) Pressurecorrections for fluid-inclusion homoge- Zechsteinabbgerunngun der Stassfurtserie die SubharzbezirkesChemie nization temperaturesbased on the volumetric propertiesofthe system der Erde.21. 137-194. NaCl-HrO. U.S. GeologicalSurvey Journal ofResearch, 5, 603-607. Higuchi, H., and Nagasawa,H. (1969) Partition of trace elementsbetween Potter, R.W., II, and Brown, D.L. (1977) The volumetric properties of rock-forming mineralsand the host volcanic rocks Eanh and Planetary aqueoussodium chloride solutions from 0'to 500'C at pressuresup to ScienceLetters. 7. 281-287 2000 bars basedon a regressionofavailable data in the literature. U.S. Hodge, G., and Norman, D.L (1985) Organic compounds in epithermal GeologicalSurvey Bulletin l42l-C,36 p. ore-depositingfluids(abs). In Epithermal Depositsin New Mexico. New Purkayastha,B.C., and Chatterjee,A. (1966) On the uptake ofstrontium Mexico Bureau of Mines and Mineral ResourcesCircular 199, 53. tracer by different forms of calcium sulphate. Joumal of the Indian Holland, H.D. (1956)The chemicalcomposition of vein mineralsand the Chemical Sociery,43, 687-.693. nature of ore forming fluids Economic Geology, 5 l, 7 81-797. Queen,L.D., and Motyka, R.J. (1984) Changesin the Makushin geother- Jensen,B.B. (1973) Patternsoftrace elementpartitioning Geochimica et mal system: Evidence from alteration mineralogy. Geological Society CosmochimicaActa, 37, 2227-2242. of America Abstracts with Programs, 16, 329. Jones,A.P, and Smith, J.V. (1985)Phlogopite and associatedminerals Rashkova,D.M. (1981)Zonal phlogopitecrystals from the Rossencopper- from Permian minettes in Devon, South England.Bulletin of the Geo- molybdenum deposit, district ofBourgas Comptes Rendus de I'Acad- logical Society of Finland, 57, part l-2, 89-102. 6mie Bulgarede Sciences,34, 1529-1532. Keenan,J.H., Keyes,F.G., Hill, P.G., and Moore, J.G (1969) Steam Rim9ite, J. (1969) Evolution ofzoned micas and associatedsilicates in tables:Thermodynamic propertiesofwater including yapor, liquid and rhe Oka carbonatite. Contributions to Mineralogy and Petrology, 23, solid phases,162 p. Wiley, New York. 340-360. Kiyosu, Y., and Kurahashi, M (1983) Origin of sulfur speciesin acid Robert, J.L. (1976) Titanium solubility in synthetic phlogopite solid so- sulfate-chloride thermal waters, northeastern Japan. Geochimica et lutions.Chemical Geology, 17, 213-227 CosmochimicaAcra. 47. 1237-1245. Roedder,E. (1969)Varvelike banding ofpossible annual origin in celestite Kozlova, O.G., Stepanova,T.V., and Egorov-Tismenko, Yu.K. (1984) crystalsfrom Clay Center,Ohio, and in other minerals.Amencan Min- The mica law of twinning and the formation of primary inclusions. eralogist,5 4, 79 6-8 10. Vestnik MoskovskogoUniversiteta. Geologiya,39, no. 3, 52-57 (trans- -(1970) Application of an improved crushing microscope stageto lated in Moscow University Geology Bulletin, 1984,39, no. 3, 5,t-60) studiesofthe gasesin fluid inclusions. SchweizerischeMineralogische Kushnir, J. (1982a)The partitioning ofseawater cations during the trans- und PetrographischeMitteilungen, 50, part 1, 41-58. formation of gypsumto anhydrite. Geochimicaet CosmochimicaActa, - (1984)Fluid inclusions.Mineralogical Societyof America Reviews 46,433446. in Mineralogy,12,644 p. - (l 982b) The composition and origin ofbrines during the Messinian Rogers,N.W., Hawkesworth, C.J., Parker, R.J. and Marsh, J.S. (1985) desiccationevent in the Mediterraneanbasin as deducedfrom concen- The geochemistryofpotassic lavas from Vulsini, central Italy and im- trations of ions coprecipitatedwith gypsum and anhydrite. Chemical plications for mantle enrichment processesbeneath the Roman region Geology,35, 333-350. Conrributions to Mineralogy and Petrology, 90,244-257. Luhr, J.F Carmichael,I.S.E., and Varekamp,J.C. (1984) The 1982erup- Rye, R O, Luhr, J.F., and Wasserman,M.D. (1984)Sulfur and oxygen tions of El Chich6n volcano, Chiapas,Mexico: Mineralogy and petrol- isotopic systematicsofthe 1982eruptions ofEl Chich6n volcano, Chia- ogy ofthe anhydrite-bearingpumices. Joumal ofVolcanology and Geo- pas, Mexico. Joumal of Volcanology and Geotherrnal Research,23, thermal Research,23, 69-108. r09-t23. Matson, D.W., Muenow, D.W., and Garcia,M.O. (1986)Volatile contents Samartsev,I.T., Ovsyannikov, I.I , and Inshina, V M. (1980) Gypsum- of phlogopite micas from South African kimberlite. Contributions to anhydrite mineralization in somegold-ore deposits ofthe CentralUrals. Mineralogy and Petrology,93, 399408. Izvestiya Akademii Nauk SSSR:Seriya geologicheskaya, no. 2,70-74. McDowell, S.D., and Elders, W.A. (1980) Authigenic layer silicate min- Sasada,M. (1985) COrbearing fluid inclusions from geothermal fluids. erals in Borehole Elmore l, Salton Sea geothermal field, California, Geothermal ResourcesCouncil, Transactions,9, part l, 351-356. USA. Contributions to Mineralogy and Petrology,74,293-310. -(1986) Anhydrite precipitation by low pressureboiling at DW-5 McGee, J.J (1983) $ANBA, a rapid, combined data acquisition and cor- drill hole, Hohi geothermalarea, Japan: Fluid inclusion evidenceand rection program for the SEMQ electron microprobe U.S. Geological geologicalsignificance. Geothermal ResourcesCouncil, Transactions, SurveyOpen-File Report 83-817,45 p r0,149-r54. BELKIN ET AL.: HYDROTHERMAL PHLOGOPITE AND ANHYDRITE 797

Schulze,D.J., Smith,J.V., and Ncmec,D. (1985)Mica chemistryof lam- wood Mountains, Montana. CarnegieInstitution of Washington Year prophyresfrom the Bohemian massif,Czechoslovakia. Neues Jahrbuch Book 76, 53+539. filr Mineralogie Abhandlungen, | 52, 32 | -33 4. Yau, Y.-C , Anovitz, L.M., Essene,E.J, and Peacor,D.R. (1984)Phlog- physical conditions during ret- Shikazono,N , and Holland, H D. (1983) The partitioning of strontium opite-chlorite reaction mechanismsand betweenanhydrite and aqueoussolutions from 150'to 250'C. Economic rogradereactions in the Marble Formation, Franklin, New Jersey.Con- Geology,Monograph 5, 320-328. tributions to Mineralogy and Petrology,88, 299-306. (1954) and stability Shikazono,N., Holland, H D., and Quirk, R F (1983) Anhydrite in Ku- Yoder, H S., and Eugster,H.P. Phlogopite synthesis roko deposits:Mode ofoccurrence and depositional mechanisms Eco- range Geochimicaet CosmochimicaActa, 6, 157-185' nomic Geology, Monograph 5, 329-344. Zhang, R. (1986) Sulfur isotopesand pyrite-anhydrite equilibria in a vol- Shmakin,B.M. (l 984)Causes and consequencesofhigh contentsofbanum canic basin hydrothermal systemof the middle to lower YangtzeRiver in sheetmuscovite and phlogopite. In Non-metallic mineral ores. Pro- Valley EconomicGeology, 81,3245. ceedingsofthe 27th InternationalGeological Congress, 15,261-271. Taguchi, S., and Hoyashi, M (1983) Fluid inclusion study in some geo- MANUscRrprRECEIvED Jr-ms 26, 1987 thermal fields of Kyushu, Japan. Fourth International Symposium on Meruscnrpr ACcEPTEDMmcn 17, 1988 Water-rock Interaction, Misasa, Japan, ExtendedAbstracts, 459462. Thompson, R.N. (1977) Primary basaltsand magma genesisIII. Alban Hills, Roman comagmaticprovince, central Italy Contributions to Min- eralogyand Petrology,60, 9l-108. ApprNlrx 1. Pnr-ocoplTE cRYSTALLocRAPHY Tronnes,R.G, Edgar,A D., and Arima, M (1985)Titanium solubility phlogopite from 2210 m con- in phlogopite:An experimentalstudy on its potential as a P-T indicator A handpicked euhedral crystal for upper mantle-lower crustalrocks. Geological Association ofCanada taining primary fluid inclusions but relatively free of trapped and Mineralogical Association of CanadaJoint Annual Meeting, Pro- solids was examined by the single-crystalprecession technique. gram with Abstracts,10, A63. The unit-cell data (Malcolm Ross, pers. comm., 1986) for the Valley,J.W., Petersen, E.U., Essene, E.J., and Bowman,J.R (1982)Fluor- 2210-m crystal based on Buergerprecession single-crystal pho- phlogopite and fluortremolite in Adirondack marbles and calculated tographsof the h\L h3h1,h3h\lkl hhl, afld hhl nets are 4 : Mineralogist, 67, 545-557. C-O-H-F fluid composition. American 5.318,D :9.228,c: 10.070A,B: l}1.2,spacegroup: C2/m. Volfinger,M Robert, Vielzeuf, D., and Nevia, A.M.R. (1985)Struc- , J.-L., The crystal examined is twinned by spiral +60'rotations of tural control of the chlorine content of OH-bearing silicates(mica and the individual layers(twinning about 10] and 10]).The space amphiboles).Geochimica et CosmochimicaActa, 49,37-48 [l [1 form. The /M form occurs Washington,H S (1906) The Roman comagmaticregion CamegieInsti- grotrp,C2/m, correspondsto the 1M tution Publication no 57, 199 p throughout the entire range ofgeologic, physical, and chemical Wentlandt, R F (1977) Barium-phlogopite from Haystack Butte, High- conditions (Yoder and Eugster,1954).