American Mineralogist, Volume 73, pages 313-323, 1988
Origin and significanceof blue coloration in quartz from Llano rhyolite (llanite), north-central Llano County, Texas
MrcHl.Br E. Zor,nNsxy Planetary Materials Branch, SN2/NASA Johnson SpaceCenter, Houston, Texas 77058, U.S.A. P.q.uLJ. Syr-vnsrBn* ELINASA Headquarters,Washington, D.C. 20546,U.S.A. Jtrms B. Pacns Geology Department, Michigan TechnologicalUniversity, Houghton, Michigan 49931, U.S.A.
Assrn,A.cr Blue quartz phenocrystsfrom the Llano rhyolite (llanite), Llano County, Texas, derive their coloration from Rayleigh scattering by abundant submicrometer-sized(type l) il- menite inclusions. Also included are larger, lessabundant ribbon-shaped(type 2) ilmenite inclusions lying on the rhombohedral growth surfacesof the host q\artz. These type 2 inclusions produce chatoyancein certain orientations; however, they are, in general, in- dividually too large (-0.1 by I by 20 pm) to contribute to the blue color by Rayleigh scattering.The total amount of ilmenite in the llanite blue quartz is calculatedto be -0.02 volo/0. Llanite blue quartz and groundmassexhibit distinct trace-elementcrystal/liquid parti- tion coefficientsthat deviate from the flat patterns characteristicof other quartzlrhyolite pairs. Partition coefrcientsfor Hf (0.335),Zr (0.38),Cr (0.10),and Lu (0.28)are signifi- cantlygreater than thosefor Rb (0.01),Ba (0.013),Th (0.004),La (0.003),and Tb (0.008), suggestingthat a majority of the ilmenite inclusions crystallizedfrom the llanite melt. This conclusion assumesthat quartz equally partitions all elements except Eu, which in the llanite blue quartz as well as some colorless quartz exhibits partition coefficientswith a positive anomaly (Eu/Eu* > 2.6).The trace-elementdata are compatible with either of the following scenarios:(1) both type I and type 2 ilmenite inclusions originated by crys- tallization from the llanite magma, or (2) the volumetrically major portion of the ilmenite inclusions (probably type l) was derived by crystallization from the magma, whereasthe remainder (probably only type 2) was formed by exsolution processes. If entrapment of early crystallizing ilmenite is a generallyapplicable model for the origin of blue quartz, two implications arise. First, calibrations of the temperature and oxygen fugacity of Fe-Ti oxide exsolution in blue quartz should not be made unlessan exsolution origin for the inclusions is assured.Second, the dominant occurrenceof blue quartz irr rocks of middle to late Proterozoic agemay reflect preferential conditions that promoted early ilmenite saturation during this time. These conditions remain largely undetermined but could involve particular physical parametersof magma equilibrium such as low tem- perature, high pressure,or high oxygen fugacity and/or processesof magma production resulting in Ti-, Fe-, alkali-, and REE-rich high-silica compositions.
INrnoluc.troN stricted in occurrenceto rocks ofPrecambrian, and par- ticularly middle to late Proterozoic age.The Proterozoic Quartz crystals exhibiting blue coloration are encoun- quartz tered variously in granites,granodiorites, rhyolites, char- occurrenceof blue as crystalsis in contrastto bluish gray which is nockites, as silicic segregationswithin anorthosites, and to cryptocrystalline silica, unconstrainedin with paper is in the regionally metamorphosed products of all these age,and which this unconcerned. quartz rocks (cf., Delong and Long, 1976;Herz and Force, 1984; The origin of blue coloration in crystalshas long Clarke, 1984;McConnell and Costello,1984). With few been known to result from the scatteringoflight. Several potential exceptions(Shaw and Flood, l98l), blue quartz is re- mechanisms have the to causethis scattering. Two less-citedmechanisms, strain deformation and Ray- proposed * Presentaddress: Department of Geoscience,New Mexico leigh scatteringfrom fluid inclusions,have been Instituteof Miningand Technology, Socorro, New Mexico 87801, as the agent for blue coloration for several occurrences U.S.A. (e.g.,Milford granite,Emerson and Perry, 1907).How- 0003-004x/88/0304-03I 3$02.00 3I 3 3t4 ZOLENSKY ET AL.: BLUE QUARTZ FROM LLANO RHYOLITE ever,in the majority of instances,blue coloration in quartz In this paper, we attempt to answer the first two of is believed to arise by Rayleigh scattering from submi- these questions for the notable doubly terminated blue crometer-sizedsolid mineral inclusions(e.g., Robertson, quartz phenocrysts from the Llano rhyolite, from Llano 1885; Iddings, 1904; Jayaraman,1939). The literature County, Texas, locally called "llanite" (Iddings, 1904). concerning the identity of the included phasesis enor- Llanite is a late Proterozoic, porphyritic, hypabyssal mous, and only the more conclusive (and contradictory) rhyolite intrusion within the Llano uplift of central Texas. studies will be summarized. The blue quartzes of the The igneous minerals of this rhyolite are well preserved, Roselanddistrict, Virginia, have been examined by Rob- although petrofabric data (Clabaugh and McGehee, 1963) ertson(1884, 1885)and, a century later, by Nord (cited and evidence of Rb-Sr redistribution (Delong and Long, in Herz and Force, 1984),who attributed the coloration I 976) suggest that the rock has been affected by low-grade to Rayleigh scatteringfrom submicrometer-sizedcrystals metamorphism. The blue quartz from this material ex- of an iron-titanium oxide (probably ilmenite). In the hibits a particularly beautiful blue coloration, which is course of an examination of the same material, Watson sky-blue in the center of each crystal and darker at the and Beard(1917) concluded that the colorationwas due margins (Iddings, 1904). In addition, in certain orienta- to crystalsof rutile. Indeed, most early workers concluded tions these crystals display an intense silver-blue chatoy- (assumed?)that rutile was the causeof blue coloration in ant flash. quartz (Robertson,1885; Postelmann, 1937; Jayaramary The results of our study of llanite may be applicable to 1939),and this view waspromulgated by later authorities the origin ofblue quartz in other felsic rocks, particularly (Goldschmidt, 1954;Frondel, 1962).Nevertheless, Par- those that have suffered only low-grade metamorphism. ker (1962)described inclusions of tourmalinewithin blue In this regard, we then attempt to answer the third out- quartz from the Wind River Range,Wyoming. standing question concerning the temporal association of An unusual controversy surrounds blue quartz from blue quartz, as posed above. Llano County, Texas. In the initial report ofthis occur- rence,Iddings (190a)concluded that the blue coloration in this material was probably due to apatite and also pos- ExpnnrrvrBNTAL PRocEDURE sibly ilmenite. Later workers, apparently misreading Id- Blue quartz phenocrysts were hand-picked from a partly ding's paper, stated that he reported rutile (e.g.,Jayara- weatheredoutcrop ofLlano rhyolite located in a roadcut on the man, 1939).Finally, in a recent re-examinationof the east side of Texas State Highway 16 (lat 3053'22" N, Iong Llano material,Barker and Burmester(1970) concluded 9839'23' W), approximately 17 km north of the Llano metro- that the blue coloration was due to zircon! politan area. Severalfresh whole-rock specimenswere also col- This lack of consensuson the included mineral oc- lected. The quartz phenocrystswere ground slightly in an alu- and ultrapure (>99o/o)blue quartz separate curred largely becausethe identification of microscopic mina shatter box; was extracted from this mixture by hand-picking under a bin- to submicroscopicinclusion grains was frequently made ocular microscope.Millimeter-sized chips of llanite microcline by visual observation, rather than by modern analytical phenocrystsand matrix material were separatedfrom the whole- techniques. However, it is also probable that all blue rock samplesby hand. The matrix chips were examined under quartzesdo not contain the same mineral inclusions. a binocular microscope, and those with quartz or microcline Another uncertainty is the mechanism by which the phenocrystadhesions were rejected from the separate.The mi- included phase in blue quartz was produced. Several crocline was ground and then purified by magnetic separation workers (Niggli and Thompson, 1979; G. L. Nord, Jr., and further hand-picking. A standard doubly polished petro- cited in Herz and Force, 1984,and pers. comm., 1986; graphic thin sectionwas also made from the whole-rock sample. quartz phenocrystswere mount- J. B. Thompson,pers. comm., 1986)have speculated that Severalthin chips of the blue Be grids for observationin a rsoI- toocxscanning the inclusions in blue quartzesfound in metamorphosed ed directly onto transmissionelectron microscope(srrrnl). Six other chips of blue feature, terranes may be a high-temperature exsolution quartz phenocrystswere set into epoxy and thin-sectionedusing quartzes potential geothermom- implying that blue are a a Porter Blum r'rr-z Ultra-Microtome and then subjectedto srur,r eter-oxygenbarometer for metamorphic conditions. Nig- analysis.Unfortunately, these sectionsaveraged approximately gli and Thompson (1979) also noted that blue quartz is 400 A thick, which was thicker than desired,because ofthe high highly strained in most occurrences,suggesting that the hardnessof quartz relative to that of the enclosing epoxy, but deformation of the quartz may promote the presumed were still serviceable.All rnineral identifications were verified exsolution process.A magmatic origin for the inclusions using electron diffraction (eo). The diffraction data were cali- in blue quartzesis also possible-i.e., the inclusionsmay brated by employing TlCl powder on the specimengrids and are quartz gralns be a liquidus mineral phase subsequentlyentrapped in consideredprecise to within 50/0.Additional were in lM HF and then heated;the residue was examined the growing quartz crystal. digested in a leor- :sc scanningelectron microscope (sru) and the srru. for most quartzes, there least three Thus, blue are at Energy-dispersivespectroscopy (ros) analyseswere obtained of outstandingquestions: (l) What mineral(s)forms the in- all material examined by electron microscopy using Princeton clusions?(2) Did this mineral crystallizefrom a magmatic Gamma Tech thin-window and windowless detectorsand ana- liquid or exsolveduring subsoliduscooling, metamorphic lyzers attachedto both the srrv and sBv. heating,or deformation?(3) And why is blue quartz largely Approximately 200 mg of the blue quartz separate,100 mg of restricted in occurrenceto Proterozoic and older rocks? the microcline separate,and 40 mg of the matrix separatewere ZOLENSKY ET AL.: BLUE QUARTZ FROM LLANO RHYOLITE 315 analyzedfor Fe, Na, and trace elementsby instrumental neutron than -0.3 ;rm in largestdimension (Nassau,1983), the activation analysis(rNae), following the proceduresof Jacobset type 2 inclusions are not blue coloring agents. al. (1977\. Most of theseelements have been determined to bet- ter than +50/o;the errorsare
Fig. l. A single llanite blue quartz crystal viewed down the c axis in reflectedlight. Chatoyancearises from arrays ofacicular inclusions aligned parallel to the facesofpositive and negative rhombohedra (and the arrow in the figure) and thus intersectingat anglesthat are multiples of 60', which is the angle betweenthe inclusion arrays illuminated in the two views shown in (a) and (b). Each field of view measures- 1 mm.
0.013(9)(2o) and a mean TiO, (wto/o)value of 0.030(16) apparent temperaturesand oxygen fugacity values, pre- (2o). These data suggestthat this blue quartz is signifi- clude great confidencein these calculations (as they in- cantly enrichedin Fe and Ti relative to colorlessquartzes, volve computational extrapolations,Rumble, 1976; Cza- which typically have FeO, and TiO, concentrations less manske et al., 1977); however, these low values clearly than 0.01 wt0/0,and often 0.005 wto/oor less (Frondel, do not representthe original magmatic conditions. 1962;Nash and Crecraft,1985). This conclusionis more Therefore thesedata support earlier suggestions(Delong definitive if the analytically superior (+50/0)rNaa Fe de- and Long, 1976) thaL the llanite rhyolite was affectedby terminations (describedbelow) of the blue quartz sepa- a very low grademetamorphic event (subgreenschist)sub- rate (0.022wto/o FeO) are considered.Blue quartz samples sequentto its crystallization. However, extensiveelement from other localities have reported FeO, concentrations redistribution at the whole-rock or mineral scale is not of 0.001 to 0.70 wto/oand TiO, concentrationsof 0.002 expectedto be significant at this low metamorphic grade. to 0.07 wo/o (Frondel, 1962). Microprobe analysisof groundmassilmenite and mag- rN.Lnanalyses netite yields compositions near end member FeTiO, and The concentrationsof FeO,, NarO, and trace elements FerOo,respectively. A relatively large pyrophanite in the blue quartz and microcline phenocrysts and (MnTiO.) component, up to 15 mol0/0,is present in groundmassof the llanite rhyolite are listed in Table l. groundmassilmenite. Strong partitioning of Mn into il- The chondrite-normalizedrare-earth-element (REE) con- menite relative to coexisting magnetite is a commonly centrations of these phasesare plotted in Figure 4. The observed phenomenon, particularly in high-silica rhyo- llanite groundmassis enrichedin Hf, Th, Zr, U, and REE, litic rocks (Czamanskeand Mihalik, 1972: Czamanskeet particularly LREE (La and Ce have concentrationsgreat- a1.,l98l) and in regionallymetamorphosed rocks (Rum- er than 300 times the chondritic value). A large negative ble, 1976).Using the techniquesof Spencerand Lindsley Eu anomaly (Eu/Eu* : 0.15) characterizesthe ground- (1981)with the modificationsof Andersenand Lindsley mass (Eu* is the Eu value derived by interpolating be- (1985),analyses from coexistinggroundmass oxide pha- tween the neighboring REE Sm and Gd). The microcline sesresult in consistentcalculated temperaturesand oxy- phenocrystshave 0.429 wto/oFeO, and 3.63 wto/oNarO genfugacitiesof396 t 125"Cand l0-32t5barsforthe and are enriched in Rb and Ba. The chondrite-normal- recalculationtechnique ofAnderson (1968),Lindsley and ized REE pattern of the microcline decreasesfrom La to Spencer(1982), and Stormer (1982).The high Mn con- Lu, with little or no Eu anomaly. tent of llanite groundmassilmenites, as well as their low The blue quartz phenocrysts are depleted in all ele- ZOLENSKY ET AL.: BLUE QUARTZ FROM LLANO RHYOLITE JLI
Frg.2. Transmission electron micrograph of type 1 rounded inclusions in the llanite blue quartz. Two inclusions are indicated by arrows. Inset shows a lower magnification micrograph of an ultramicrotomed llanite blue quartz grain, which is outlined by a dotted line. The sizeand distribution ofthe dark type I ilmenite inclusionsis clearly observed.The superimposedwavy interference is produced by chatter of the enclosingepoxy.
ments relative to llanite groundmass,but have relatively (0.0010/o)(Frondel, 1962;Nash and Crecraft,1985) con- high concentrationsof }{f,Zr, and HREE. A negative Eu centration of Fe, respectively. anomaly (Eu/Eux : 0.38) is present. Although several notable compositional differencesbetween blue and col- DrscussroN AND coNcLUSroNS orlessquartz are discussedlater, there are also somecom- Transmission electron microscopy of the llanite blue positionalsimilarities. The Ba concentrationof 3.5 ppm quartz revealedthe presenceofubiquitous, rounded type for the blue quartz is similar to that reported for colorless I inclusionsof ilmenite,averaging -0.06 pm in diameter. quarlz (0.45 to 13 ppm Ba, Ghiorso et a1.,1979; Four- Theseinclusions produce the blue color ofthis quartz by cadeand Alldgre, l98l). The blue quartz has equal con- the phenomenon of Rayleigh scattering.These type 1 in- centrationsof FeO,and NarO (0.022wto/o). This Na con- clusionsprobably make up most of the foreign inclusions centration falls within the range (2 to 683 ppm Na) within the quartz. The intensity of blue coloration in the reported for colorlessquartz (Frondel, 1962; Ghiorso et quartz varies directly with the number density of these al., 1979:.Nash and Crecraft, 1985).The Fe enrichment type I ilmenite inclusions. Apatite and zircon inclusions, of the blue quaftz relative to colorlessquartz was noted reportedfrom llanite blue quartz by Iddings (1904) and earlier. From the Fe concentration determined by rNe.o., Barker and Burmester(1970), respectively, and believed it may be calculated that the blue quartz contains either by them to be the causeofboth the blue color and cha- 0.014 or 0.024 volo/o(assumed stoichiometric, pure end- toyance,were not encounteredin the present study. member) ilmenite, dependingon whether the quartz lat- The type 2 ilmenite ribbons presentin the llanite blue tice itself contains either a relatively high (0.0080/o)or low quartz measure-0.1 by I by 20 pm, and are therefore 318 ZOLENSKY ET AL.: BLUE QUARTZ FROM LLANO RHYOLITE
Fig. 3. Type 2 ribbonlike inclusion crystal protruding from a thin chip of llanite blue quartz. Inset is an electron-difraction pattern produced by this crystal rn one orrentatron. in generaltoo large to causeblue coloration by Rayleigh The shape of the trace-element partition-coefficient scattering.In certain orientations, constructive reflection pattern for blue quartz formed by an exsolution process from these inclusions is observed to produce an intense would be similar to that of colorless quartz. The shapes white flash as a result of their alignment along (probably of these patterns are determined by the relative differ- rhombohedral) growth planes of the enclosing quartz ences between trace-element partition coefficients in a phenocrysts.In a section cut normal to the c axis of a given mineral. These relative differences(i.e., the shapes quartz phenocryst, six such chatoyant reflections, sepa- ofthe patterns) do not vary significantly as a function of rated by anglesof 60o,are observed. temperature, pressure,oxygen fugacity, and bulk liquid The trace-element data collected for the blue quartz composition, in contrast to the absolute values of the allows discrimination betweenan exsolution or magmat- partition coefficients(Philpotts, 1978). ic crystallization origin for the ilmenite inclusions. The Alternatively, if the ilmenite inclusions are early-crys- partitioning of a trace element between the llanite melt tallizing phasesthat have been trapped in the blue quartz and the crystallizing quartz phenocrystsis adequatelyde- crystalsgrowing in the llanite, the trace-elementpartition scribedby a simple Nernst partition coemcient(Mclntire, coefficientsof the blue quartz would reflect both colorless 1963).On the other hand,if the ilmenite formed by exso- quartz-melt equilibria and ilmenite-melt equilibria. In lution during either subsolidus cooling, metamorphic this case,the shapeof the trace-elementpartition-coeffi- heating,or deformation, the trace-elementpartition coef- cient pattern for blue quartz would reflect a combination ficients of the blue quartz would be governed solely by of the patterns for colorlessquartz and ilmenite. the initial colorlessquartz-melt equilibria. Trace-element Mineral/melt partition coefficientsfor the llanite blue partitioning between the blue quartz and the exsolving quartz and microcline phenocrystswere calculatedby di- ilmenite would not afect the original bulk composition viding their trace-elementconcentrations by those of the ofthe blue auartz. groundmass (Table l). This exercise assumesthat the ZOLENSKYET AL.: BLUE QUARTZ FROM LLANO RHYOLITE 319
U F oE o z - v U N r) T = z 0 z v f, m () U z (J o E L) =
Lo Ce Sm Eu Yb Lu Fig. 4. Chondrite-normalized (Haskin et al., 1968) REE concentrationsof blue quartz phenocrysts,microcline microperthite phenocrysts,and groundmassfrom the llanite rhyolite. quartz and microcline phenocrystscrystallized in equilib- The contrastingpartition coefficientpatterns of the two rium with a melt of the samecomposition as the ground- quartz types suggestthat the trace-elementcomposition mass.The microcline partition coefrcients are character- of the llanite blue quartz is not governed solely by col- ized by high values for Rb, Sr, and Ba, a preferencefor orlessquartz-melt equilibria. Consequently,it is doubtful LREE over HREE, and a positive Eu anomaly. Alkali that a majority of the ilmenite included within the blue feldspar/liquid partition coemcients reported elsewhere quartz formed by exsolution. (Mahood and Hildreth, 1983;Nash and Crecraft, 1985) In contrast, the partition-coefficient data are consistent are similar, but have lower absolute values for several with the hypothesis that the blue quartz includes a con- trace elementsincluding the REE. tribution from ilmenite inclusionscrystallized from a melt Mineral/melt partition coemcientsdetermined for the with the composition of the llanite groundmass.Mineral/ llanite blue quartz show that this quartz is enriched in melt partition coefficients determined for ilmenite in Cr, Co, Sc, HREE, Hf, Zr, and U relative to the LREE equilibrium with basaltic liquid (Fig. 58) generally in- (Fig. 5A). In contrast, it is widely assumed(cf. Cullers et creasewith atomic number for the REE, exceptfor a neg- al., 1981) that colorlessquartz does not preferentially ative Eu anomaly, and are enriched in Hf and Zr relalive partition any one trace element relative to another. How- to Lu, and in Sc and Co relative to La. lt is easily deter- ever, the requisite compositional data neededto confirm mined from a comparison of Figures 5,A'and 5B that this assumption have been collected only for colorless mixtures of ilmenite and colorlessquartz yield apparent quartz from rhyolitic lavas of the Twin Peaks volcanic partition coefficientswith relative values similar to those complex,Utah (Nashand Crecraft, 1985).The Twin Peaks of the llanite blue quartz. These data suggest(l) the ma- colorlessquartz displays relatively flat REE patterns and jority of the ilmenite inclusions (probably the more per- little or no Sc.Hf. and U enrichment relative to the LREE vasive type I inclusions) are early-fractionating phases (Fig. 5A). Both the llanite blue quartz and the Twin Peaks that have beentrapped within the quartz and (2) no other colorlessquartz exhibit large positive Eu anomalies. early-crystallizingphases are presentwithin the quartz. If 320 ZOLENSKY ET AL.: BLUE QUARTZ FROM LLANO RHYOLITE
Tnere 1. Llanite blue quartz, microcline microperthite,and oroundmassconcentrations and partitioncoeff icients
Concentrations* Partition coefficients Llanite ground- Micro- Species Bluequartz Microcline mass Quartz cline FeO, 0 022 0.429 z-ov Naro 0.022 3.63 397 Bb 0.76 300 144 0.01 2.1 Sr 90 40 2 Ba 3 5 1760 270 0 013 6.5 TLMENTTE/BASALT B 0.006 0.36 0.69 0.01 0.52 0.420 47 6 117.9 0.003 0.404 Ce 0.91 111.6 285 0.003 0 391 0.108 8.73 z+.cJ 0.004 0.359 F Sm z 2.54 1.21 0.013 2 10 u EU 0.016 o Tb 0 029 1,37 3.80 0.008 0 360 E .ro Yb 0.332 2.00 14.0 0024 0.143 U LU 0.064 0.287 2.27 0.028 0.126 $ .os z Sc 0 048 0.679 5.49 0 009 0j24 9 Co 0 027 0.26 120 0.019 0 19 tr Cl 0 26 0.81 2.5 0.10 0 32 E .or HI 5.50 1 44 16.4 0.335 0.088 c .m5 Th 0 089 1.21 22.0 0.004 0.055 Zr 160 100 420 0.38 0.24 U 0 09 0.40 c.o 0.02 0 09 .m1 - In ppm, exceptfor FeOrand Na,O,which are in wt7o. 60
small amounts (0.01 to 0.02 volo/o)of early-crystallized ilmenite are also trapped within the microcline pheno- crysts,their compositional contribution is maskedby the relatively greatertrace-element concentrations of the mi- crocline. Unfortunately, blue quartz tace-element data cannot completely discriminate between the two different types of ilmenite inclusions. The chemical data presentedhere are compatible with a crystallization mode of origin for both types of inclusions. If this scenariois correct, then Cr Sc Ca Nd SmGd Dy Er Yb Hf U the two types of ilmenite found within the blue quartz Co La Pr Pm Eu Tb Ho Tm Lu Zr Th probably nucleatedat separatelocations or times and were Fig. 5. Mineral/melt partition coefficientsfor quartz and il- then incorporated together into the growing quartz crys- menite. (A) Blue quartzlllanite rhyolite (X; this study) and col- tals. It seemsimprobable that two ilmenite types with orlessquartzlTwin Peaksrhyolite (.; Nash and Crecraft, 1985). such dissimilar morphologies could have nucleated to- (B) Ilmenite/synthetic high-Ti mare basalt (O, Nakamura et al., gether. In this instance,it is also possiblethat the type 2 1986;a, McKay et al., 1986;v, McKay and Weill, 1976),il- ilmenite crystals nucleated onto clusters of smaller, pre- menite/plagioclase-poikiliticilmenite lunar basalt 70135 (!, existing type I ilmenite crystals,although this possibility Haskin and Korotev, 1977) and ilmenite/Skaergaardmafic lay- was not investigatedin the present study. ered series (O, Paster et al., 1974). (C) Ilmenite/Twin Peaks It is also possible, however, that some part of either rhyolite (1, Nash and Crecraft, 1985) and ilmenite/Sierra La Primavera rhyolite (O, Mahood and Hildreth, 1983),with REE type of ilmenite inclusion, or one type in entirety, formed partition coefficients recalculated- for illustrative purposesonly- by exsolution processes.This alternate scenario is per- assumingarbitrary amounts of apatite and allanite contamina- mitted becausethe exsolution processwould have no ef- tion (n, Twin Peaks, 40/oapatrte,0.22o/o allanite; O, La Prima- fect on the blue quartz bulk trace-elementcomposition. vera, 0.040/oapatite, 0.050/oallanite), using allanite and apatite However, since it is probable from the presenttrace-ele- partition data from Mahood and Hildreth (1983) and Nagasawa ment study that most of the ilmenite in the llanite blue (1970), respectively.Recalculation assuming contamination by quartz crystallized from the rhyolite magma before or chevkinite and titanite [which have REE partition coefrcients during the formation of the quartz phenocrysts,if one broadly similar to allanite (Cameron and Cameron, 1986) and type of ilmenite inclusion formed in entirety by exsolu- apatite (Green and Pearson, 1986b), respectivelyl also yields a tion, it was most likely the volumetrically inferior type 2 HREE-enriched pattern. Limited knowledge of partitioning be- inclusions. If the type 2 ilmenite crystals indeed formed havior of other trace elementsin these accessoryminerals pre- possible cludes a more complete recalculation of the observedilmenite by exsolution,it remains that they nucleatedonto data. pre-existing type I ilmenite crystals. ZOLENSKY ET AL.: BLUE QUARTZ FROM LLANO RHYOLITE 321
The shapesof partition-coemcient patterns for ilmenite (Dennen,1966; Ghiorso et al., 1979),thanto the equilib- in equilibrium with rhyolitic liquid (Fig. 5C), which is the rium melt. The consequentpreferential partitioning of the more appropriatecompositional system forthe discussion complexesinto liquidus quartz would result in Eu2* en- of llanite crystallization, are claimed to be significantly richment. different from ilmenite/basalt partition coefficients(Ma- Our conclusion that the llanite blue quartz contains hood and Hildreth, 1983;Nash and Crecraft,1985). Cam- ilmenite inclusions, of which the majority (at least) are eronand Cameron(1986) have noted, however, that such liquidus phases,suggests that the same result may be ob- "anomalous" mineral/rhyolite liquid partition coeffi- tained for blue quartz from other locales,ifthey are sub- cients may reflect trace-element-cnrichedaccessory-min- jected to a thorough examination using modern analytical eral inclusions (e.g.,allanite, chevkinite, apatite, or tita- techniques.It also follows that extremecaution should be nite) trapped within host minerals. Detection of these exercisedby workers wishing to develop blue quartz as a inclusionsin opaqueminerals such as ilmenite is most geothermometer-oxygenbarometer. For example, Nord difficult and allows the possibility that available ilmenite/ has speculatedthat iron-titanium oxide grains (probably rhyolite partition coefficientsmeasure accessory mineral ilmenite) in blue quartz from a leucocharnockite-anor- contamination. This contention is supported by the con- thosite in the Roseland district, Virginia (cited in Herz sistency between the relative REE partition coefrcients andForce, 1984; G. L. Nord, Jr.,pers. comm., 1986) may for (l) the recalculatedilmenite-rhyolite system, assuming be a high-temperature exsolution feature that could be small amounts of allanite and apatite contamination (Fig. calibrated in T-J6, space.However, INAA data indicate 5C), (2) the ilmenite-rhyolite system inferred from the that the majority of the ilmenite inclusions in llanite blue blue quartz data, and (3) the ilmenite-basalt system, de- qvartz are not an exsolution feature. smu imaging indi- termined experimentally. catesthat theseinclusions are probably the type I ilmen- The one characteristic of the trace-element partition ites, which causethe blue coloration in the llanite blue coefficientsshared by colorlessand blue quartz-a posi- qvaftz. The less abundant type 2 ilmenite inclusions are tive Eu anomaly-is remarkable in that it is commonly possibly of exsolution origin. However, theselatter inclu- assumed(e.g., Nash and Crecraft,1985) that quartz does sions are unrelated to the blue coloration of the quartz not preferentiallyincorporate Eu relative to the other rare- phenocrysts.If sample selection for the blue quartz geo- earth elements.Since the colorlessquartz from the Twtn thermometer-oxygen barometer is made on the basis of Peaksrhyolite has not been affectedby low-grade meta- inclusion-producedcolor, workers should ensurethat the morphism and does not include ilmenite, it is unlikely color-producing inclusions are genetically related to the that the Eu anomaly is a metamorphic phenomenon or event they are attempting to characterize. the result of ilmenite contamination. The anomaly is also The entrapment of liquidus ilmenite in the llanite blue unlikely to be the result of feldspar contamination of the quartz phenocrystsalso suggeststhat the llanite melt was blue quartz separate,because inclusions in or adhesions capableof crystallizing ilmenite very early in its crystal- on a transparent mineral such as qvarlz arc easily iden- lization history. Early ilmenite saturation may occur pri- tifled during microscopic examination. In order to pro- marily in melts extraordinarily rich in Ti and Fe. It has ducethis Eu anomaly by feldsparcontamination, it would also been established,however, that ilmenite saturation be necessaryfor 330/oofthe blue quartz separate(partition is promoted by decreasingtemperature. increasing pres- coefrcient Eu/Eux : 2.60) to consistofthe analyzedllanite sure, increasing oxygen fugacity, and increasing concen- microcline microperthite (partition coefficient Eu/Eu* : trations of silica, alkalis, and REE in equilibrium liquids 5.83).Besides being a very evidentproportionofthe hand- (Greenand Pearson,1986a). The insetin Figure6 shows picked sample, this amount of contamination would re- a significant lowering of the TiO, content required for quire the quartz separateto exhibit FeO,(0. l4 wto/o),NarO ilmenite saturation by these additional conditions. The (l .2wto/o\,and trace-element concentrations that aremuch llanite rhyolite certainly exhibits a high-silica, alkali-rich higher than thoseobserved. (75 wto/oSiOr, 8.0 wto/oNa, O + KrO; Goldich, 1941; There are at leasttwo crystal-chemicalexplanations for Barkerand Burmester, 1970), REE-enriched composition, the Eu anomaly, both involving the reduction of Eu3*to but in the absenceof quantitative compositional data for Eu2t. The first explanation is that Eu2* may enter into the ilmenite inclusions, estimates of their temperature, sixfold coordination with oxygen and thus form a substi- pressure,and oxygen fugacity of crystallization are pre- tutional solid solution in quartz, analogousto the sup- cluded. posedbehavior of Mg'?*(Frondel, 1962).Ditrerences in Delong and Long (1976) suggestedthat Fe- and Ti- the ionic radii of Sio*(0.40 A) and Eu2"(1.31 A) (Shannon rich melarhyolite dikes of the Llano uplift may be related and Prewitt, 1969)preclude direct substitution. geneticallyto the llanite rhyolite. The melarhyolite con- A secondexplanation involves the formation of com- tains lower concentrationsof silica and higher concentra- plexes of Eu2* with aluminosilicate (AlrSirO! ) in mag- tions of Fe and Ti relative to the llanite rhyolite (Fig. 6). matic liquids (Moller and Muecke, 1984). These com- Someof the melar\olite containsminor groundmassblue plexesmay be more similar to the structure of crystalline quarlz.In comparison to the massivegranites of the Llano quartz, which is typically characterizedby substitution of uplift, the melarhyolite exhibits greaterFe and somewhat Si by Al and a valence compensation ion such as Na greater Ti concentrations,whereas the llanite rhyolite is 322 ZOLENSKY ET AL.: BLUE QUARTZ FROM LLANO RHYOLITE
at <900 "C, 1.5 kbar, and the F' of the hematite-mag- l2 netite (HM) buffer (Fig. 6). If this model is correct, the restriction of blue quartz largelyto middle to late Proterozoicand someolder rocks l0 may reflect preferential production of the required high Hypolheticol Ti-rich Ti-Fe felsic magmacompositions and their crystallization phose solubility curve - 08 . olkolinei under low-temperature, high-pressure,or high fo2 con' o F ditions during this time. It is well known that a distinctive .R \--F-tp i=.=Sl'.n Fe- and Ti-rich suite ofA-type (rapakivi) granites,massif- E _ ou \\ type anorthosites,and mangeritic rocks is largely confined q o--- to the middle to late Proterozoic(Emslie , 1918:,Anderson, quartz anotherman- o4 1983).In some cases,blue may be ifestation of this type of magmatism.
02 AcxNowr-nncMENTS We thank G. Rossman,G. L. Nord, Jr., J. B Thompson,and D Gust for helpful commentsand calculations We appreciatethe helpful reviews 0 of this manuscript by G. McKay, D Gust, and G L. Nord., Jr D S. McKay, and D Blanchardkindly provided accessto the elec- 8 McKay, G tron-beam and rNa.A.facilities ofthe Solar SystemExploration Division at gtr Melorhyolite trtr dikes the NASA/Johnson SpaceCenter. B Bansalgraciously performed the HF s digestion. T Cobliegh is acknowledgedfor enlighteningus as to the sig- 6 tr nificanceofthis problem and ensuringthat we remained sensitiveto the particularneeds ofthe llanite o RnrnnnNcrs crrpo = Andersen,DJ., and Lindsley,D.H. (1985)New (and final!) modelsfor O Llonite geothermometer v rnyollre the Ti-magnetite/ilmenite and oxygenbarometer. EOS Mossive 66,416. Llono gronites Anderson,A T ( I 968) Oxidation ofthe LaBlancheLake titaniferous mag- ) netitedeposit, Quebec. Journal ofGeology, 76,528-541 Anderson,J L. (l 983)Proterozoic anorogenic granite plutonism ofNorth 70 72 America.Geological Society of AmericaMemoir l6l, 133-154 Barker,D.S, and Burmester,RF. (1970)Leaching ofquartz from Pre- cambrian hypabyssalrhyolite poryhyry, Llano County, Texas.Contri- Fig. 6. Whole-rock compositions of the massivegranites (.), butionsto Mineralogyand Petrology,28, l-8. melarhyolite dikes (!), and llanite rhyolite dike (O, whole rock; Barnes,V.E., Dawson,R F, and Parkinson,G A (1947)Building stones O, groundmass)from the Llano uplift (Goldich, 1941;Barnes et ofcentral Texas.Texas University Publication 4246,198 p al., 1947;Barker and Burmester,1970; Delong and Long, 1976). Cameron,K.L , and Cameron,M. ( I 986)Whole-rock/groundmass differ- A hypothetical Ti-rich phasesolubility curve for thesesupposed entiation trends of rare earth elements in high-stlica rhyolites. Geo- liquid or near-liquid compositions is shown by the dashedline. chimicaet CosmochimicaActa, 50, 759-770. Insert: Experimentallydetermined solubility curvesin TiOr-SiO, Clabaugh,S.E., and McGehee,R V. (1963)Precambrian rocks of the Llano region,Texas University Bureau of EconomicGeology Guidebook 5, spacefor the Ti-rich accessoryphases at 7.5 kbar in subalkaline 63-78. (solidcurve, 1000'C; long-dashedcurve, 900'C and REE-nch; .C) Clarke,J.W. (1984)The coreofthe BlueRidge anticlinorium in northern dot-dashedcurve, 900 and alkaline (short-dashedcurve, 1000 Virginia.Geological Society ofAmenca SpecialPaper 194,155-160. 'C) compositions(Green and Pearson,1986a). All experiments Cullers,R L, Koch, R.J , and Bickford,M E (1981)Chemical evolution conducted under magnetite-wtistite-buffered oxygen-fugacity of magmas in the Proterozoic terrane of the St. Francis Mountains, conditions except for dot-dashed curve, which was under con- southemMissouri; 2. Traceelement data. Journal ofGeophysical Re- ditions of the hematite-masnetitebufer. search,86, 10388-10401. Czamanske,G.K, and Mihalik, P (1972) Oxidation during magmatic differentiation,Finnmarka complex, Oslo area,Nomay; I The opaque oxides JournalofPetrology, 13, 493-509. clearly enriched in both Fe and Ti at the 75 wto/o SiO, Czamanske,G.K., Wones,D R , andEichelberger, J.C (1977)Mineralogy level (Fig. 6). Blue quartz has not been identified in the and petrologyof the intrusive complex of the Pliny Range,New Hamp- massive Llano granites. shire.American Journal of Science,2'7 7, 1073-1 123 K, Ishihara,S., and Atkin, S.A (1981)Chemistry ofrock- It is possible that some combination of low-tempera- Czamanske,G forming minerals of the Cretaceous-Paleocenebatholith in southwest- ture, high-pressure, or high-oxygen fugacity conditions em Japanand implicationsfor magmagenesis. Journal ofGeophysical allowed early ilmenite saturation in the Ti-, Fe-, REE-, Research,86, 10431-10469 alkali-rich, high-silica llanite rhyolite melt. A hypothetical Delong, S E, and Long,L.E (1976)Petrology and Rb-Srage of Precam- Ti-rich phase solubility curve drawn in TiOr-SiO, space brian rhyolitic dikes, Llano County, Texas Geological Society of AmericaBulletin. 87 . 275-280. ilmenite crystallization in the llanite rhyolite to allow early Dennen,w H (1966)Stoichiometric substitution in naturalquartz. Geo- and in some of the melarhyolites is compatible with ex- chimicaet CosmochimicaActa, 30, 1235-1241 perimentally determined curves for similar compositions Emerson,B K., and Perry,J.H (1907)The greenschists and associaled ZOLENSKY ET AL.: BLUE QUARTZ FROM LLANO RHYOLITE 523
granitesand porphyriesofRhode Island U S GeologicalSurvey Bul- ccedingsofthe SeventhLunar ScienceConference, p.2427-2447 Per- letin3l 1.45-46 gamonPress, New York Emslie, R.F. (1978) Anorthosite massifs,rapakivi granites,and Late Pro- McKay, G, Wagstaf,J., and Yang, S.-R (1986)Zirconium, hafnium, terozoic rifting of North America PrecambrianResearch, 7 , 6l-98 and rare earth elementpartition coefrcientsfor ilmenite and other min- Fourcade,F , and Alldgre,C.J. (l 98I ) Traceelement behaviour in granite eralsin high-Ti lunar mare basalts:An experimentalstudy. In Pro- genesis:A casestudy. The calc-alkalineplutonic associationfrom the ceedingsofthe SixteenthLunar and PlanetaryScience Conference, Part Quengut Complex (Pyrenees,France). Contributions to Mineralogy and 2 Journal of GeophysicalResearch, 91, D229-D237. Petrology,7 6, l'77-195. lVlollcr.P , and Muecke,G.K. (1984)Significance of europiumanomalies Frondel,C (1962)The systemofmineralogy: Vol III. Silica minerals, in silicatemelts and crystal-meltequilibria: A re-evaluationContri- 334 p Wiley, New York butionsto Mineralogyand Pelrology,87,242-250 Ghiorso,M S, Carmichael,I S E, and Moret, L.K. (1979)lnverredhigh- Nagasawa,H (1970) Rare earth concentrationsin zircons and apatites temperaturequartz Unit cell parametersand properties ofthe alpha- and their host dacitesand granites.Earth and PlanetaryScience Letters, beta inversion Contributionsto Mineralogyand Petrology,68, 307- 9, 359-364 J/J. Nakamura,Y , Hirokazu, F, Nakamura,N, Tatsumoto,M., McKay, Goldich, S.S.(1941) Evolution ofthe centralTexas granites Joumal of G A , and Wagstaff,J (1986) Hf, Zr, and REE partition coefficients Geology,49,697-720. between ilmenite and liquid: Implications for lunar petrogenesis In Goldschmidt,V M. (1954)Geochemistry, 365 p. ClarendonPress, Ox- Proceedingsofthe SixteenthLunar and PlanetaryScience Conference, ford Part 2. JournalofGeophysical Research, 91,D239-D250. Green,T H , and Pearson,N.J. (1986a)Ti-rich accessoryphase saturation Nash, W P., and Crecrafl, H.R (1985) Partition coefficientsfor trace ele- in hydrous mafic-felsiccompositions at high P, Z Chemical Geology menls in silicic magmas.Geochimica et Cosmochimica4c1a,49,2309- 54,185-201. 2322 -(1986b) Rare-earthelement partitioning between sphene and co- Nassau,K (l 983)The physicsand chemistryof color,p. 233-249 Wiley, existing silicateliquid at high pressureand temperature.Chemical Ge- New York. ology,55, 105-l19. Niggli, E, and Thompson,J.B. (1979)Petrogenetic significance ofblue, Haskin,L A , and Korotev,R L. (1977)Test ofa model for traceelement opalescentquartz in metamorphic rocks Memorie degli Istituti di Geo- partition during closed-systemsolidifrcation of a silicateliquid Geo- logiae Mineralogiadell' Iniversitddi Padova,33, 258. chimicaet CosmochimicaActa, 41, 921-939. Parker,R B. (1962)Blue quartzfrom the Wind River Range,Wyoming. Haskin,L A., Haskin,M A , Frey,F.A., and Wildeman,T.R. (1968)Rel- AmericanMineralogist, 47, 120l-1202 ative and absolute terrestrial abundancesof the rare earths In L.H Paster,T.P, Schauwecker,D.S., and Haskin, LA (1974)The behavior Ahrens,Ed, Origin and distributionofthe elements,p 889-912 Per- of some trace elementsduring solidification of the Skaergaardlayered gamon,New York. seriesGeochimica et CosmochimicaActa, 38, 1549-1571. Herz, N., and Force,E. (1984)Rock suitesin Grenvillianterrane ofthe Philports,J.A (1978)The law ofconstantrejection. Geochimica et Cos- Roselanddistnct, Virginia Part I Lithologic relations GeologicalSo- mochimica Acta, 42, 909-920. ciety ofAmerica SpecialPaper 194, 187-200 Postelmann, A (1937) Die ursache der blaufarbung gesteinsbildender Iddings, J P (1904) Quartz-feldspar-porphyry(graniphyro liparose-alas- quarze NeuesJahrbuch li.ir Mineralogie, Geologie und Palaontologie, kose)from Llano, Texas.Journal ofGeology, 12,225-231 12.401-440. Jacobs,J.W., Korotev,R L., Blanchard,D P , and Haskin,L A (1977)A Robertson,R (1884)Examination ofblue quartzfrom NelsonCo., Va. well-testedprocedure for instrumental neutron activation analysis of ChemicalNews, 50, no 1301,207. silicaterocks and minerals.Joumal of RadioanalyticalChemistry, 40, -(1885) An examinationof blue quartzfrom NelsonCounty, Vir- 93-l | 4 ginia The Virginias,6,2-3. Jayaraman,N (1939)The causeof colour of the blue quartzesof the Rumble,D (1976)Oxide mineralsin metamorphicrocks. Mineralogical charnockitesof South India and the Champion Gneissand other re- Societyof America ShortCourse Notes, 3, Rl-R24 latedrocks ofMysore. Proceedingsofthe Indian AcademyofSciences, Shannon,R.D , and Prewitt,C T (1969)Effective ionic radii in oxides 9. 265-285. and fluorides.Acta Crystallographica,425, 925-946 Lindsley,D H , and Spencer,K.J. (1982)Fe-Ti oxide georhermometry: Shaw,S.E, and Flood, R.H (1981)The New Englandbatholith, eastem Reducing analysesof coexistingTi-magnetite (M0 and ilmenite (Ilm) Australia: Geochemicalvariations in time and space.Joumal of Geo- (abs)EOS, 63,471 physicalResearch, 86, 10530-10544 Mahood,G., and Hildreth,W (l 983)Large partition coemcienrs for trace Spencer,K J , and Lindsley,D.H. (1981)A solutionmodel for coexisting elementsin high-silicarhyolites Geochimicaet CosmochimicaActa, iron-titaniumoxides. American Mineralogist, 66, I 189-1201 47, t-30. Stormer,J.C. (1982)The recalculationof multicomponentFe-Ti oxide McConnell,K.I., and Costello,J O. ( I 984)Basement-cover rock relation- analysesfor geothermometry:A quasi-thermodynamicmodel (abs.). ships along the westernedge ofthe Blue Ridge thrust sheetin Georgia. EOS,63,471 GeologicalSociety ofAmenca SpecialPaper 194, 263-280. Watson,T.L, and Beard,R.E (1917)The color of amethyst,rose, and Mclntire, W L. (1963)Trace elemenl partition coefficients-A reviewof bluevarieties ofquartz. Proceedingsofthe U S. NationalMuseum, 53, theory and applicationsto geology Geochimicaet CosmochimicaActa, ffJ-)bJ 27. 1209-1264 M.rxuscrrrr REcETVEDFesnuenv 16, 198'7 McKay, G.A, and Weill, D. F (1976)Petrogenesis of KREEP. In Pro- Mnrrrscnryr AccEprEDNovprrrsrn 20. 1987