Library

Notes: Abstract We report the result of H2O-undersaturated melting experiments on charges consisting of a layer of powdered sillimanite-bearing metapelite (HQ36) and a layer of powdered tonalitic gneiss (AGC150). Experiments were conducted at 10 kbar at 900°, 925° and 950°C. When run alone, the pelite yielded ∼40 vol% strongly peraluminous granitic melt at 900°C while the tonalite produced only ∼5 vol% weakly peraluminous granitic melt. At 950°C, the pelite and the tonalite yielded ∼50 vol% and ∼7 vol% granitic melt, respectively. When run side by side, the abundance of melt in the tonalite was ∼10 times higher at all temperatures than when it was run alone. In the pelite, the melt abundance increased by ∼25 vol%. When run alone, biotite dehydration-melting in the tonalite yielded orthopyroxene and garnet in addition to granitic melt. When run side by side only garnet was produced in addition to granitic melt. Experiments of relatively short duration, however, also contained Al-rich orthopyroxene. We suggest that the large increase in melt fraction in the tonalite is mainly a result of increased activity of Al2O3 in the melt, which lowers the temperature of the biotite dehydration-melting reaction. In the pelite, the increase in the abundance of melt is caused by transport of plagioclase component in the melt from the tonalite-layer to the pelite-layer. This has the effect of changing the bulk composition of this layer in the direction of “minimum-temperature” granitic liquids. Our results show that rocks which are poor melt-producers on their own can become very fertile if they occur in contact with rocks that contain components that destabilize the hydrous phase(s) and facilitate dehydration-melting. Because of this effect, the continental crust may have an even greater potential for granitoid melt production than previously thought. Our results also suggest that many anatectic granites most likely contain contributions from two or more different source rocks, which will be reflected in their isotopic and geochemical compositions.

Notes: Abstract The development of orthopyroxene-Fe/Mg ferrite symplectites associated with olivine is discussed with respect to the chemical reactions by which they form. Previously proposed reactions are presented graphically and the differences between them are reviewed. With the exception of exsolution, these are all discontinuous reactions in the sense that olivine is replaced by the two-phase symplectite assemblage. Olivine-hosted symplectites developed in the margins of lherzolite xenoliths from Kauai, Hawaii, demonstrate a reaction mechanism which has not been previously documented from natural samples. Original Fo90 olivine in these samples oxidized to a new assemblage consisting of orthopyroxene (En92–95)-Fe/Mg ferrite (Mf35–50) symplectites developed within more magnesian olivine (Fo92–96) hosts. Thus, by this mechanism, olivine of a different composition persists as part of a final three-phase assemblage. As oxidation advanced, the compositions of all three product phases became continuously more magnesian and the stoichiometric coefficients of the orthopyroxene and Fe/Mg ferrite continuously increased, whereas those of the product olivine decreased in the mass-balance equations. These characteristics demonstrate that the reaction was controlled by oxygen diffusion into the xenoliths from the highly oxidized alkali picrite melt in which they were entrained. Thermodynamic calculations suggest that a gradient in oxygen fugacity of 100.9 bars existed across the xenolith rims and resulted in compositional gradients of 4 mol% fayalite and ferrosilite and 15 mol% magnetite.

Notes: Abstract The compositions of primary magmas depend to some degree on the dynamic processes occurring in the partially molten region of the mantle. The compositional dependence is estimated for three models which assume either accumulation from a migmatized source region or accumulation entirely by the interstitial flow of magma. Accumulation from a migmatised region results in magmas with higher concentrations of incompatible elements than does batch melting, whereas accumulation by interstitial flow results in magmas with lower concentrations of these elements. The concentrations of refractory elements are almost independent of both the accumulation process and the degree of partial melting and are therefore usefull for the identification of primary magmas.

Notes: Abstract Experiments designed to simulate the interaction of juxtaposed rhyolitic and basic magmas were conducted at 10 kbar with H2O, using reaction-couples consisting of Westerly granite (WG) against basalt (DW-1) and WG against a synthetic mafic glass (SMG, enriched in MgO and Na2O relative to DW-1). Each couple was run with ∼5 and ∼10 wt% H2O corresponding respectively to H2O-undersaturated and H2O-oversaturated conditions. Experiments were run for 42–44 h at 920° C, above the liquidus of WG and within the melting intervals of DW-1 and SMG. WG was run above the basic material in all but one experiment. The composition of the granitic melt was altered only through material exchange with the adjacent basic melts, whereas that of the basic melts also changed (relative to the bulk basic composition) due to partial crystallization. Some crystallization also occurred within the zone of interaction. For control, the basic compositions were also run alone under the same conditions as the reaction-couple experiments. The crystalline phase assemblages in the basic ends of the coupled experiments differed from those produced from the basic materials alone, demonstrating interaction with the granite melt. Moreover, compositional gradients within the basic ends of coupled experiments are indicated by changes in phase assemblage and compositions with distance from the interface with WG. Microprobe analyses of glass collected along the length of the capsules confirm published observations that alkali diffusion is very fast: K2O and Na2O homogenized throughout the capsules in less than the two-day run times. This, coupled with the fact that introduction of K2O into SMG stabilized biotite, produced the result that after interaction the bulk basic material (melt+crystals) contained more K2O than the coexisting felsic melt. Only very gentle gradients for CaO, FeO, and MgO are preserved in our experiments, in contrast with published anhydrous results, suggesting that the difference in activity coefficients for these components between basic and felsic melts is reduced by the introduction of H2O. Gradients for SiO2 and Al2O3 are of comparable length to those of the divalent cations, confirming earlier results that the diffusivities of the network-formers limit the rate of diffusion of Ca, Fe, and Mg.

Notes: Abstract A thermodynamic model is developed that describes the compositional variation of coexisting olivine, orthopyroxene, and ferrite (Fe3O4-MgFe2O4) as a function of $$f_{O_2 }$$ and T. The ferrite phase has a cation distribution which varies from nearly inverse to nearly random with increasing T and is described with a model in which the number of sites per formula unit on which mixing occurs varies from 1.67 to 2.0. Given this model and the equilibrium phase composition data for coexisting olivine and ferrite at 1,300° C of Jamieson and Roeder (1984), the ferrite solution is described to an excellent approximation by a symmetric regular solution model with W ft =+14.0 ±0.3 kJ/mole. Orthopyroxene and olivine non-ideality are also considered. The T-dependence of the equilibrium constant for the oxidation reaction 6Fs+2Mt=6Fa+O2 and the two Fe/Mg exchange reactions between olivine-ferrite and olivine-orthopyroxene, are used to calculate the compositional variation of coexisting phases as a function of $$f_{O_2 }$$ and T. The results are summarized on an isobaric (1 bar) $$f_{O_2 }$$ −1/Tplot with the compositional variation of olivine, ferrite, and orthopyroxene shown by sets of isopleths. The ferrite isopleths intersect those of olivine and orthopyroxene at sufficiently high angles for this assemblage to serve as a sensitive geothermometer and oxygen-barometer. The model is applied to orthopyroxene-ferrite symplectite in coronas around olivine in a metamorphosed gabbro, to olivine-hosted orthopyroxene-ferrite symplectite in unmetamorphosed gabbros and norites and to olivine-hosted orthopyroxene-ferrite symplectites developed within the rims of lherzolite xenoliths.

Notes: Abstract We present data on the phase relationships of mixtures between natural tonalite and peridotite compositions with excess H2O at 30 kbar, and on the composition of the piercing point where the peridotite-tonalite mixing line intersects the L(Ga,Opx) reaction boundary. These data, in conjunction with earlier analogous data along peridotite-granite and basalt-granite mixing lines, permit construction of a pseudoternary liquidus projection that is relevant to interaction of peridotite with slab-derived magmas. Knowledge of the liquidus phase and temperature for a range of compositions within this projection enables us to map primary crystallization fields for quartz, garnet, orthopyroxene, clinopyroxene, and olivine, and to estimate the distribution of isotherms across the projection. Using this projection, we explore the consequences of peridotite assimilation by mafic to intermediate (basalt to dacite) hydrous slab-derived melts. Progressive assimilation under isothermal conditions results in garnet precipitation as the melt composition traverses the garnet liquidus surface and then garnet+orthopyroxene crystallization once the melt reaches the L(Ga,Opx) field boundary. The melt is constrained to remain on this field boundary and further assimilation of peridotite simply results in continued precipitation of garnet+orthopyroxene until the melt is consumed. The product is a hybrid solid assemblage consisting of Ga+ Opx. It is noteworthy that this process drives the melt composition in a direction nearly perpendicular to the mixing line between peridotite and the initial melt. If assimilation occurs with increasing temperature (as might occur if a slab-derived magma rises into the hotter mantle wedge), intermediate magmas (e.g. andesites) will again precipitate garnet until they reach the L(Ga,Opx) reaction boundary at which point Ga re-dissolves and orthopyroxene precipitates as the melt composition moves up-temperature along this boundary. The product of this process is a hybrid solid assemblage with garnet subordinate to orthopyroxene. For more mafic initial compositions (e.g. basalts) originally plotting in the Cpx field, it appears possible to avoid field boundaries involving garnet and shift in composition more directly toward peridotite, if assimilation is accompanied by a sharp increase in temperature. Considering published REE evidence (arguing against garnet playing a significant role in the genesis of many subduction-related magmas) in light of our results, it appears unlikely that peridotite assimilation by intermediate magmas under conditions of constant or increasing temperature is an important process in subduction zones. However, if assimilation is accompanied by an increase in temperature, our data do permit the derivation of high-Mg basalts from less refractory precursors (e.g. high-Al basalts) by peridotite assimilation.

Notes: Abstract We report results of anhydrous 1 atm and piston-cylinder experiments on ID16, an Aleutian high-magnesia basalt (HMB), designed to investigate potential petrogenetic links between arc high-alumina basalts (HABs) and less common HMBs. ID16 is multiply saturated with a plagioclase/spinel iherzolite mineral assemblage (olivine, plagioclase, clinopyroxene, orthopyroxene, spinel) immediately beneath the 12 kbar liquidus. Derivative liquids produced at high temperatures in the 10–20 kbar melting interval of ID16 have compositions resembling those published of many moderate-CaO HABs, although lower-temperature liquids are poorer in CaO and richer in alkalies than are typical HABs. Isomolar pseudoternary projections and numerical mass-balance modeling suggest that derivative melts of ID16 enter into a complex reaction relationship with olivine at 10 kbar and 1,200° C–1,150° C. We sought to test such a mechanism to explain the lack of liquidus olivine in anhydrous experiments on mafic high-alumina basalts such as SSS. 1.4 (Johnston 1986). These derivative liquids, however, do not resemble typical arc high-alumina basalts, suggesting that olivine-liquid reaction does not account for Johnston's (1986) observations. Instead, we suggest that olivine can be brought onto the liquidus of such compositions only through the involvement of H2O, which will affect the influence of bulk CaO, MgO, and Al2O3 contents on the identity of HAB liquidus phases (olivine or plagioclase) at pressures less than ∼12 kbar.

Notes: Abstract We report the T-X(H2O) phase relations for the trondhjemitic Nûk gneiss which comprises the principal component of the second phase of Archean (3.0–2.8 by) igneous activity in the Godthåb region of southwestern Greenland. A pressure of 15 kbar was chosen to place constraints on possible protoliths for trondhjemitic melts at lower crustal depths. Under H2O-saturated conditions, a melting interval of ∼135° C separates the solidus at ∼610° C from the liquidus at 745° C. H2O-saturation at 15 kbar occurs at approximately 15.5 wt % H2O. The H2O-undersaturated liquidus extends along a curved path from ∼745° C at 15.5 wt % H2O to ∼1100° C at 2% H2O. Lower H2O contents were not investigated. At low H2O contents (〈6%) sodic plagioclase (Pl, An32) is the liquidus phase followed at lower but still near-liquidus temperatures by quartz (Qz) and then garnet (Ga). At 6% H2O, Ga replaces Pl on the liquidus and is joined at slightly lower temperatures by Pl and hornblende (Hb). The field for liquidus Ga extends to only ∼7.5% H2O where it is replaced by Hb which is the liquidus phase up to 13% H2O. At all higher H2O contents, epidote (Ep) is the first phase to crystallize, followed by biotite (Bi) at slightly lower temperatures. Following the standard inverse approach, the near-liquidus phase assemblages are interpreted as potential residues from which trondhjemitic melts could be extracted. At high melt H2O contents (〉7%), mafic residues consisting of some combination of Hb, Ga, Ep, and Bi are possible and could correspond to amphibolitic source rocks. At lower melt H2O contents (〈 5%), possible residues consist of Na-Pl+Qz±Ga and could correspond to an earlier generation of tonalitic-trondhjemitic rocks. However, such residues would not impart the highly fractionated REE patterns characteristic of Archean trondhjemites. If a first generation of tonalitic-trondhjemitic melts was generated by higher pressure partial fusion of eclogite and emplaced at 55 km depth, it would crystallize to an assemblage consisting almost entirely of Na-Pl+Qz with highly fractionated REE patterns. These rocks in turn could be partially melted to yield a second generation of trondhjemites which would inherit the highly fractionated REE patterns because neigher Pl nor Qz is capable of significantly fractionating HREE from LREE.

Notes: Abstract Anhydrous P-T phase relations, including phase compositions and modes, are reported from 10–31 kbar for a near-primary high-alumina basalt from the South Sandwich Islands in the Scotia Arc. The water content of natural subduction-related basalt is probably 〈0.5 wt.% and thus, these results are relevant to the generation of primary basaltic magmas in subduction zones. At high pressures (〉27 kbar) garnet is the liquidus phase followed by clinopyroxene, then quartz/coesite at lower temperatures. At intermediate pressures (17–27 kbar), clinopyroxene is the liquidus phase followed by either garnet, quartz, plagioclase, then orthopyroxene or plagioclase, quartz, garnet, then orthopyroxene depending on the pressure within this interval. At all lower pressures, plagioclase is the liquidus phase followed at much lower temperatures (∼100° C at 5 kbar) by clinopyroxene. The absence of olivine from the liquidus suggests that the composition studied here could not have been derived from a more mafic parent by olivine fractionation at any pressure investigated, and supports the interpretation that it is primary. If so, these results also preclude an origin for this melt by partial melting of olivine-rich mantle periddotite and suggest instead that it was generated by partial melting of the descending slab (quartz eclogite) leaving clinopyroxene, garnet, or both in the residue. The generally flat REE patterns for low-K series subduction related basalts argue against any significant role for garnet, however, and it is thus concluded that the composition studied here was extracted at 20–27 kbar after sufficiently high degrees of partial melting (∼50%) to totally consume garnet in the eclogite source. Melting experiments on three MORB composition, although not conclusive, are in agreement with this mechanism. Results at 30 kbar support an origin for tonalite/trondhjemite series rocks by lower degrees of melting (15–30%), leaving both garnet and clinopyroxene in the residue.

Notes: Abstract Peraluminous granitoid magmas are a characteristic product of ultrametamorphism leading to anatexis of aluminous metasedimentary rocks in the continental crust. The mechanisms and characteristic length-scales over which these magmas can be mobilized depend strongly on their melt fraction, because of their high viscosities. Thus, it is of fundamental importance to understand the controls exerted by pressure, temperature and bulk composition of the source material on melt productivity. We have studied experimentally the vapour-absent melting behaviour of a natural metapelitic rock and our results differ greatly from those of previous experimental and theoretical investigations of melt productivity from metamorphic rocks. Under H2O-undersaturated conditions, bulk composition of the source material is the overriding factor controlling melt fraction at temperatures on the order of 850–900° C. Granitoid melts formed in this temperature interval by the peritectic dehydration-melting reaction: $$\begin{gathered} Biotite + plagioclase + aluminosilicate + quartz \hfill \\ = melt + garnet \hfill \\ \end{gathered} $$ have a restricted compositional range. As a consequence, melt fractions will be maximized from protoliths whose modes coincide with the stoichiometry of the melting reaction. This “optimum mode” (approximately 38% biotite, 32% quartz, 22% plagioclase and 8% aluminosilicate) reflects the fact that generation of low-temperature granitoid liquids requires both fusible quartzo-feldspathic components and H2O (from hydrous minerals). Metapelitic rocks rich in mica and aluminosilicate and poor in plagioclase contain an excess of refractory material (Al2O3, FeO, MgO) with low solubility in low-temperature silicic melts, and will therefore be poor magma sources. Melt fraction varies inversely with pressure in the range 7–13 kbar, but the effect is not strong: the decrease (at constant temperature) over this pressure range is of at most 15 vol% (absolute). The liquids produced in our experiments are silicarich (68–73 wt% SiO2), strongly peraluminous (2–5 wt% normative corundum) and very felsic (MgO+FeO* +TiO2 less than 3 wt%, even at temperatures above 1000° C). The last observation suggests that peraluminous granitoids with more than 10% mafic minerals (biotite, cordierite, garnet) contain some entrained restite. Furthermore, because liquids are also remarkably constant in composition, we believe that restite separation is more important than fractional crystallization in controlling the variability within and among peraluminous granitoids. We present liquidus phase diagrams that allow us to follow the phase relationships of melting of silica-and alumina-saturated rocks at pressures corresponding to the mid- to deep-continental crust. Garnet, aluminosilicate, quartz and ilmenite are the predominant restitic phases at temperatures of about 900° C, but Ti-rich biotite or calcic plagioclase can also be present, depending on the bulk composition of the protolith. At temperatures above 950–1050° C (depending on the pressure) the restitic assemblage is: hercynitic spinel+ilmenite+quartz±aluminosilicate. Our results therefore support the concept that aluminous granulites (garnet-spinel-plagioclase-aluminosilicate-quartz) can be the refractory residuum of anatectic events.