Melt Extraction during Formation of K-Feldspar + Sillimanite Migmatites, West of Revelstoke, British Columbia

An exposure of sillimanite-rich, strongly deformed, stromatic,K-feldspar-bearing migmatites in the Monashee Terrane west ofRevelstoke, British Columbia, has been examined to determinethe process of migmatization and to evaluate whether the systemwas open or closed during leucosome formation. An anatecticorigin for the migmatites is supported by: (1) the minimum meltcomposition of the leucosomes; (2) textures suggesting a fluidbehavior of the leucosomes and local pegmatitic textures; and(3) P–T estimates (720–820C; 75–9 kbar)above vaporabsent melting conditions of muscovitt + quartz. To establish whether melt was extracted or added during migmatization,measured volume percents of leucosome were compared with estimatesof melt production modeled by muscovite + quartz dehydrationmelting. Quantitative estimates of volume percent of leucosomeat present in the outcrop are between 20 and 30%. The amountof melt produced from the model muscovite dehydration meltingreaction is constrained by measured modal percent of sillimanite(15–25%) in the outcrop and is dependent on modal proportionof muscovite in the unmelted protolith and the melt water contentUsing a muscovite-rich protolith and a melt water content of4 wt%, complete dehydration melting of muscovite results ina production of 54 vol % melt and 25 vol % sillmanite, indicatinga melt loss of 29 vol %. A melt water content of 6 wt% resultsin production of 41 vol % melt and 23 vol % sillimanite, indicatinga melt loss of 16 vol %. Melt loss may have occurred by meltmovement along foliation planes during flattening, during formationof shear bands or locally along subvertical fractures. Spatialproximity of the outcrop to the Monashee dcollement suggeststhat thrusting was localized to zones of high melt production,which in turn facilitated melt migration. KEY WORDS: migmatites; British Columbia; Monashee Tarrane; anatexis; melt extraction ^*Corresponding author. Present address: Department of Earth and Planetary Sciences, The Univenity of New Mexico, Albuquerque, NM 817131, USA     

Temperatures of Granulite-facies Metamorphism: Constraints from Experimental Phase Equilibria and Thermobarometry Corrected for Retrograde Exchange

This study assesses temperatures of formation of common granulitesby combining experimental constraints on the P–T stabilityof granulite-facies mineral associations with a garnet–orthopyroxene(Grt–Opx) thermobarometry scheme based on Al-solubilityin Opx, corrected for late Fe–Mg exchange. We appliedthis scheme to 414 granulites of mafic, intermediate and aluminousbulk compositions. Our findings suggest that granulites aremuch hotter than traditionally assumed and that the P–Tconditions of the amphibolite–granulite transition portrayedin current petrology textbooks are significant underestimatesby over 100°C. For aluminous and intermediate granulites,mean corrected temperatures based on our method are 890 ±17 and 841 ± 11°C, respectively (uncertainties reportedas 95% confidence limits on the mean), consistent with minimumtemperatures for orthopyroxene production by fluid-absent partialmelting in these bulk compositions. In contrast, mean temperaturesbased on Grt–Opx Fe–Mg exchange equilibria, usingthe same thermodynamic data, are 732 ± 22 and 723 ±11°C, respectively, well below the minimum temperaturesfor Opx stability. For mafic granulites, the mean correctedtemperature using our method is 816 ± 12°C, similarto the mean temperature of 793 ± 13°C from Fe–Mgexchange. Reasons for the differences between the mafic granulitesand aluminous–intermediate granulites are unclear butmay be due to the lower Al concentrations in Opx in the maficrocks and possible deficiencies in the thermodynamic modellingof these low concentrations. We discuss a number of well-knowngranulite terrains in the context of our findings, includingthe Adirondacks, the Acadian granulites of New England, theincipient charnockites of southern India and Sri Lanka, andthe Kerala Khondalite Belt. Our findings carry implicationsfor thermotectonic models of granulite formation. A computerprogram to perform our thermobarometry calculations, RCLC, isavailable from the Journal of Petrology website at http://www.petrology.oupjournals.orgor from the authors at http://www.geo.ucalgary.ca/~pattison/drm_pattison-rclc.htm. KEY WORDS: granulite-facies metamorphism; thermobarometry; garnet; orthopyroxene     

P-T Conditions and the Influence of Graphite on Pelitic Phase Relations in the Ballachulish Aureole, Scotland

Pressure-temperature conditions of pelites in the Ballachulishaureole, Scotland, have been determined from a calibrated petrogeneticgrid and from published geothermometers and geobarometers. Tocalibrate the mineral reactions in the grid, thermodynamic datafor appropriate end members of Ms, Chi, Qtz, And, Sil, Ky, Crn,Crd, Kfs, and Bt were derived from experimental data. This approachwas hampered by the unknown compositions of many of the mineralsused in the experiments, and by apparent inconsistency betweenthe experiments. A best compromise grid that satisfies mostof the data was obtained, which is applicable to the Ballachulishand other contact aureoles. In this grid, the first developmentof sillimanite is constrained to lie between the Richardsonet al. (1969) and Holdaway (1971) andalusite-sillimanite boundaries. A pressure estimate of 3.0 + 0.5 kb is obtained from the calibratedgrid, within 0.3 kb of estimates from geobarometry and fromtwo other independent petrological studies. Temperatures rangedfrom 560?20?C at the first development of cordierite in theassemblage Ms+Qtz+Chl+Crd+Bt to 750–800?C in Grt+Crd+Hyassemblages in pelitic screens within the igneous complex. In graphitic slates, in contrast to non-graphitic pelites, anentire andalusite-bearing subzone is developed, and initialcordierite development occurs further from the igneous contacts.The presence of graphite lowered a_H2o in the slaters, expandingthe stability field of the andalusite-bearing assemblage And+Qtz+Bt+Ms+Crdrelative to the assemblage Kfs+Qtz+Bt+Ms+Crd developed in non-graphiticunits. Initial development of cordierite in the assemblage Ms+Qtz+Chl+Crd+Btwas also promoted by reduced a_H2o in graphitic slates. The regular sequence and spacing of mineral zones in the aureolesuggests that gross equilibrium was attained during contactmetamorphism, even though the thermal metamorphic pulse is estimatedto have been less than 0.2 Ma (Buntebarth, in press). Thereis no evidence for reaction overstepping in cordierite-producingreactions.     

Infiltration-driven dehydration and anatexis in granulite facies metagabbro, Grenville Province, Ontario, Canada

Dark hornblende + garnet-rich, quartz-absent metagabbro boudins from the Seguin subdomain, Ontario Grenville Province, are transected by anastomosing light-coloured veins rich in orthopyroxene, clinopyroxene, plagioclase and sometimes quartz. The veins vary in texture from fine-grained diffuse veins and patches that overprint the metagabbro, to coarse tonalitic leucosomes with sharp borders. The diffuse veins and patches are suggestive of channellized subsolidus dehydration of the metagabbro, while the tonalitic leucosomes are suggestive of local internally-derived anatexis. All vein types grade smoothly into each other, with the tonalitic leucosomes being the latest.
Relative to the host metagabbro, the veins have higher Si, Na, Ba & Sr, lower Fe, Mg, Ca & Ti, and similar Al. The coarser veins are enriched in K. Plagioclase becomes steadily enriched in Na in the transition from host metagabbro (An₄₇) to the veins (An₃₅), and in the coarsest veins it is antiperthitic. Differences in composition of the other minerals between host metagabbro and vein are minor. Pressure–temperature estimates are scattered, but indicate a minimum temperature during vein formation of 700°C at about 8 kbar.
Mass balance constraints indicate that the veins formed from the metagabbro in an open system. The transecting veins are interpreted to represent pathways of Si + Na + Ba + Sr ± K ± Al-enriched, low a _H2O fluids that metasomatized the host metagabbro to form the anhydrous veins. An initial period of localized solid-state dehydration of the metagabbro, represented by the diffuse veins, was followed by a transition to localized anatexis, represented by the tonalitic leucosomes. The change to anatexis may have been due to the addition of K to the infiltrating fluid. The source and delivery mechanism of the fluids is unknown.     

Accessory phase petrogenesis in relation to major phase assemblages in pelites from the Nelson contact aureole, southern British Columbia

Monazite petrogenesis in the Nelson contact aureole is the result of allanite breakdown close to, but downgrade and therefore independent of, major phase isograds involving cordierite, andalusite and staurolite. The development of garnet downgrade of the staurolite and andalusite isograds does not appear to affect the onset of the allanite-to-monazite reaction but does affect the textural development of monazite. In lower pressure, garnet-absent rocks, allanite breakdown results in localized monazite growth as pseudomorphous clusters. In higher pressure, garnet-bearing rocks, allanite breakdown produces randomly distributed, lone grains of monazite with no textural relationship to the original reaction site. Fluids liberated from hydrous phases (chlorite, muscovite) during garnet formation may have acted as a flux to distribute light rare earth elements more widely within the rock upon allanite breakdown, preventing the localized formation of monazite pseudomorphs. Despite these textural differences, both types of monazite have very similar chemistry and an indistinguishable age by electron microprobe chemical dating (157 ± 6.4 Ma). This age range is within error of isotopic ages determined by others for the Nelson Batholith. Garnet from the garnet, staurolite and andalusite zones shows euhedral Y zoning typified by a high-Y core, low-Y collar and moderate-Y annulus, the latter ascribed to allanite breakdown during garnet growth in the garnet zone. The cause of the transition from high-Y core to low-Y collar, traditionally interpreted to be due to xenotime consumption, is unclear because of the ubiquitous presence of xenotime. Accessory phase geothermometry involving monazite, xenotime and garnet returns inconsistent results, suggesting calibration problems or a lack of equilibration between phases.     

Metamorphic evolution of granulites in the Minto Block, northern Québec: extraction of peak P-T conditions taking account of late Fe-Mg exchange

In the central Minto Block of northern Québec, the Lake Minto and Goudalie domains are dominated, respectively, by orthopyroxene-bearing plutonic suites (granite-granodiorite and diatexite) and a tonalitic gneiss complex, both of which contain scattered remnant paragneisses. Two main granulite-grade mineral assemblages are observed in the paragneiss: garnet (Grt)-orthopyroxene (Opx)-plagioclase-quartz (GOPQ) and garnet (Grt)-cordierite (Crd)-sillimanite-plagioclase-quartz (GCSPQ). These show distinct lithological associations, with the GCSPQ assemblages occurring exclusively within the diatexite in the Lake Minto domain. Petrogenetic grid considerations demonstrate that the GOPQ rocks are higher grade than the GCSPQ rocks. Maximum temperatures for GOPQ rocks, obtained from equilibria based on Al solubility in orthopyroxene in equilibrium with garnet, range from 950 to 1000d? C, significantly higher than garnet-orthopyroxene Fe-Mg exchange temperatures of 700 ± 50d? C, the latter probably representing a closure temperature below peak conditions. The Al temperatures were corrected for late cation exchange by adjusting the Fe/(Fe + Mg) ratios in garnet and orthopyroxene, to achieve internal consistency between the GOPQ thermometers and barometers. Grt-Crd thermometry records temperatures of 750±50d? C. Peak P-T conditions range from 5-6 kbar and 750-800d? C in the Goudalie and eastern Lake Minto domains, to 7-10 kbar and 950-1000d? C in the western and central Lake Minto domain. This variability contrasts with the uniform crustal pressures of 5 ± 1 kbar recorded by the GCSPQ assemblages in the diatexites and the hornblende granodiorites (c. 4-5 kbar) across the same area. The GOPQ rocks are inferred to record earlier P-T conditions that prevailed before the formation of GCSPQ assemblages and the intrusion of the granodiorites. Partial P-T paths in GOPQ rocks from both domains, based on net transfer equilibria corrected for Fe-Mg resetting, document cooling of 100-250d? C from thermal-peak conditions, concomitant with a modest pressure decrease of 2-3 kbar. Although textures diagnostic of isobaric cooling are not developed, the paths are consistent with a tectonic model in which granulite metamorphism and crustal thickening in the Minto Block were consequences of magmatic underplating. The progression from higher P-T conditions recorded by GOPQ assemblages to lower P-T conditions recorded by GCSPQ assemblages is attributed to variable amounts of synmagmatic uplift and cooling in a single, continuous thermal event in the Minto crust, associated with protracted crustal magmatism. In the Goudalie and eastern Lake Minto domains, where GOPQ and GCSPQ rocks and Hbl granodiorites have similar P-T conditions of equilibration, the crust may not have been thickened as much as further west, where GOPQ P-T conditions are significantly higher than those of the hornblende granodiorites and the GCSPQ rocks.     

Interplay between equilibrium and kinetics in prograde metamorphism of pelites: an example from the Nelson aureole, British Columbia

The distribution of metapelitic mineral assemblages in the Nelson aureole, British Columbia, generally conforms to what is predicted from phase equilibria. However, in detail, the sequence and spacing of isograds, mineral textures and mineral compositions and mineral chemical zoning do not. Two of the main disequilibrium features in the aureole are: (i) delay in the onset and progress of several reactions, i.e. overstepping in temperature; and (ii) unreactivity of staurolite and especially garnet porphyroblasts when they are reactants in prograde reactions. The thermal overstepping is ascribed to difficulty of nucleation of the product porphyroblasts and sluggishness of dissolution of porphyroblasts when they are reactants. The extent to which these kinetic barriers delay the onset of reaction is related to the reaction affinity of each reaction, defined herein as the Gibbs free‐energy difference between the thermodynamically stable, but not‐yet‐crystallized, products and the metastable reactants. For oversteps in temperature (ΔT), reaction affinity is, in turn, related to the difference in entropy (ΔS) between these two states through the relation A = ΔT * ΔS. Mineral reactions which release large quantities of H₂O, such as chlorite‐consuming reactions, have a higher entropy change per unit of temperature overstep, and therefore a higher reaction affinity, than those which release little or no H₂O, such as the chlorite‐free staurolite‐consuming reaction. Thermal overstepping is consequently expected to be less for the former than for the latter, as was estimated in the aureole where 0 to 30 °C overstepping was required for garnet, staurolite and andalusite growth from a muscovite + chlorite‐bearing precursor rock and ～70 °C overstepping was required for the growth of Al₂SiO₅ from a staurolite‐bearing, chlorite‐free precursor. In all cases, reaction progress was strongly influenced by the presence or absence of fluid, with presence of fluid lowering kinetic barriers to nucleation and growth and therefore the degree of thermal overstepping. Textural features of rocks from the nearly coincident garnet, staurolite and andalusite isograds are suggestive of a fluid‐catalysed ‘cascade effect’ in which reaction took place in a narrow temperature interval; several competing muscovite + chlorite‐consuming reactions, some metastable, appear to have occurred in parallel. Metamorphic reaction, fluid release and possibly fluid presence in general in the aureole were episodic rather than continuous, and in several cases well removed from equilibrium conditions. The extent to which these findings apply to regional metamorphism depends on several factors, a major one being enhanced deformation, which is expected to lower kinetic barriers to nucleation and growth.     

Metasomatism in the Generation of Granulite Veins: Mass Balance, Mass Transfer, and Reference Frames

In the Seguin subdomain, Grenville Province, dark hornblende–garnet-richquartz-absent metagabbro is transected by anastomosing light-colouredveins rich in orthopyroxene, clinopyroxene, plagioclase, and,in some cases, quartz. Three types of veins form a smoothlytransitional series from fine-grained diffuse veins and patchesto a more defined intermediate vein type, and finally to coarsegrainedtonalitic veins with abrupt contacts to the metagabbro host.The textures and contact relations of diffuse and intermediateveins are suggestive of channelized subsolidus dehydration ofthe metagabbro. Mass-balance calculations show that dehydrationwas accompanied by metasomatism; all reasonable solutions requirethe addition of SiO₂, Na₂O, and large ion lithophile elements(LILE) to the host rock to produce the veins. Reference framesfor the evaluation of mass transfers can be represented by twocases: (1) constant volume, and (2) constant FeO, MgO, V, andZr requiring volume increases by factors of 13—33. Thesimplest model (minimizing the number of mobile constituentsand giving progressively larger volume increases for more leucocraticveins) is based on constant FeO, MgO, V, and Zr. In this case,model mineral reactions and mass transfers for two types ofveins, distinguished on the outcrop as (1) diffuse and (2) intermediate,are: These mass balances show that SiO₂, Na₂O, and Al₂O₃ were added,and H₂O lost, in producing the vein assemblages. Minimum fluid/rockvolume ratios of 3–20 imply flow-through. The fluid didnot induce melting of hornblende+plagioclase, and hence it musthave had relatively low a_H2O. The open-system transformation of hornblende–garnet-richmetagabbro to pyroxene-rich veins was driven primarily by theintroduction of silica and feldspar components in a low-a_H2Osupercritical fluid at granulite-facies P–T conditions.The application of quantitative mass balance, taking into accountpropagated errors and a critical appraisal of the possible referenceframes, enhances our ability to resolve open-system processes.     

Genesis of the Kapuskasing (Ontario) migmatitic mafic granulites by dehydration melting of amphibolite: the importance of quartz to reaction progress

Migmatitic, granulite-grade mafic gneisses make up a significant part of the Kapuskasing Structural Zone (KSZ), Ontario. Although they contain a common mineral assemblage [hornblende (Hbl)+plagioclase (Pl)+diopside (Di)±garnet (Grt)+quartz (Qtz)±titanite (Ttn)], the mafic gneisses show wide variations in modal mineralogy from hornblende-rich to diopside+garnet-rich varieties and all gradations between. Up to 25 vol.% segregated plagioclase+quartz-rich (trondhjemitic) leucosome (Tdh) is intimately associated with the mafic gneiss, occurring in a continuum of patches, veins and transecting dykes at scales ranging from decimetres to micrometres. The texture and composition of the leucosome, combined with P-T estimates for the host rocks above the solidus, suggest it represents crystallized trondhjemitic melt. Quartz is mainly restricted to the segregated leucosomes but more rarely occurs in a variety of interstitial textures in the mafic gneiss, suggesting that it crystallized from a melt phase rather than having been present as a solid phase at peak metamorphic conditions. Modal and textural data indicate a reaction relationship of the form: Hbl+Pl(+Qtz?)=Grt+Di+Ttn+leucosome (Tdh), implying that the granulite-forming process involved dehydration melting of an amphibolite protolith. Pressure-temperature estimates from Grt+Di+Pl+Qtz geothermobarometry are 9 kbar and 685-735 °C; however, based on experimental studies of dehydration melting of amphibolite, we estimate that peak conditions were closer to 11 kbar, 850 °C. Mass balance analysis, using the technique of singular value decomposition, and reaction space analysis were used to quantify the reaction and to determine the controls on reaction progress. The following mass balance provides a model for the natural reaction:1.00 Hbl+0.92 Pl+3.76 Qtz=1.14 Grt+1.54 Di+0.21 Ttn+1.49 Tdh+0.14 ‘pg’+0.39 Fe_?1Mg+0.33 NaSiCa_?1Al_?1where ‘pg’ is a pargasite-like exchange. In all model mass balances tested, quartz is a reactant with a large coefficient. We argue that the abundance of quartz in the amphibolite protolith was the primary control on the differing extents of reaction observed. Mineral compositional variation exerted a secondary control on reaction progress, with Fe-richer layers containing An-richer plagioclase and more actinolitic amphibole reacting earliest (i.e. at lowest temperatures). Comparison of the calculated amount of melt produced in the gneisses with that now observed implies expulsion of 5–30% of the melt. These volumes are similar to those predicted from REE modelling of Archaean tonalities and trondhjemites from a garnet amphibolite source, suggesting that the KSZ mafic gneisses may be representative of partially depleted source rocks for trondhjemite-tonalite generation.