James G. Brophy

Composition gaps, critical crystallinity, and fractional crystallization in orogenic (calc-alkaline) magmatic systems

The recognition of a three-way correlation between magmatic SiO2 content, critical crystallinity, and the size (magnitude) of crystal fractionation-generated composition gaps in calc-alkaline magmatic systems suggests an important control of magmatic critical crystallinity on the formation of such composition gaps. To explain this correlation, it is proposed that fractionation-generated composition gaps are caused by: (1) simultaneous interior (i.e. non-substrate) crystallization and vigorous chamberwide convection which leads to progessive crystal suspension; (2) cessation of convection when the percentage of suspended crystals reaches the critical crystallinity of the magma, and; (3) eventual buoyancy-driven crystal-liquid segregation producing a discrete body of fractionated magma which is separated from the initial magma by a composition gap. This mechanism implies that many, if not most magma bodies are characterized by interior crystallization and vigorous convection, conditions which are not universally agreed upon at present. Given that such conditions characterize natural magma bodies, fractional crystallization through crystal settling in low-velocity boundary layers should be an important mechanism of fractional crystallization. In a crystallizing and convecting body of magma, composition gap formation should represent one endmember of a complete spectrum of possible evolutionary paths governed by the relative rates of crystal settling and crystal retention. As a given volcanic plumbing system matures with time, average settling/retention ratios within individual magma bodies should increase due to higher average wall-rock temperatures. It follows that, within a given volcanic center, early-stage volcanism should be more likely to display fractionation-generated composition gaps than later-stage volcanism. Such a temporal evolution has been documented at at least two Aleutian calc-alkaline volcanic centers.

Copper solubility in a basaltic melt and sulfide liquid/silicate melt partition coefficients of Cu and Fe

The solubility of copper in a sulfur-saturated basaltic melt has been determined at 1245°C as a function of fO2 and fS2. Copper solubilities at log fO2 values between −8 and −11 fall into two distinct populations as a function of fS2. At log fS2 values < −1.65, sulfide liquid that coexists with the basaltic glass quenches to sulfur-poor bornite solid solution. At log fS2 values in excess of −1.65, the sulfide liquid quenches to a complex intergrowth of sulfur-rich bornite and intermediate solid solution. Copper solubilities in the low-fS2 population range from 594 to 1550 ppm, whereas those in the high-fS2 population range from 80 to 768 ppm. Sulfide liquid/silicate liquid partition coefficients (D) for Cu and Fe range from 480 to 1303 and 0.7 to 13.6, respectively. Metal-sulfur complexing in the silicate liquid is shown to be insignificant relative to metal-oxide complexing for Fe but permissible for Cu at high fS2 values. On log DFe (sulfide-silicate) and log DCu (sulfide-silicate) vs. (log fS2 − log fO2) diagrams, both fS2 populations show distinct but parallel trends. The observation of two D values for any fS2/fO2 ratio indicates nonideal mixing of species involved in the exchange reaction. The two distinct trends observed for both Cu and Fe are thought to be due to variations in activity coefficient ratios (e.g., γFeO/γFeS and γCuO0.5/γCuS0.5). Results of the experiments suggest that accurate assessments of fS2/fO2 ratios are required for the successful numerical modeling of processes such as the partial melting of sulfide-bearing mantle and the crystallization of sulfide-bearing magmas, as well as the interpretation of sulfide mineralogical zoning. In addition, the experiments provide evidence for oxide or oxy-sulfide complexing for Cu in silicate magmas and suggest that the introduction of externally derived sulfur to mafic magma may be an important process for the formation of Cu-rich disseminated magmatic sulfide ore deposits.

Andesites from northeastern Kanaga Island, Aleutians

Kanaga island is located in the central Aleutian island arc. Northeastern Kanaga is a currently active late Tertiary to Recent calc-alkaline volcanic complex. Basaltic andesite to andesite lavas record three episodes (series) of volcanic activity. Series I and Series II lavas are all andesite while Series III lavas are basaltic andesite to andesite. Four Series II andesites contain abundant quenched magmatic inclusions ranging in composition from high-MgO low-alumina basalt to low-MgO highalumina basalt. The spectrum of lava compositions is due primarily to fractional crystallization of a parental low-MgO high-alumina basalt but with variable degrees of crustal contamination and magma mixing. The earliest Series I lavas represent mixing between high-alumina basalt and silicic andesite with maximum SiO2 contents of 65–67 wt %. Later Series I and all Series II lavas are due to mixing of andesite magmas of similar composition. The maximum SiO2 content of the pre-mixed andesites magmas is estimated at 60–63 wt %. The youngest lavas (Series III) are all non-mixed and have maximum estimated SiO2 contents of 59 wt %. The earliest Series I lavas contain a significant crustal component while all later lavas do not. It is concluded that the maximum SiO2 contents of silicic magmas, the contribution of crustal material to silicic magma generation, and the role of magma mixing all decrease with time. Furthermore, silicic magmas generated by fractional crystallization at this volcanic center have a maximum SiO2 content of 63 wt %. All of these features have also been documented at the central Aleutian Cold Bay Volcanic Center (Brophy 1987). Based on data from these two centers a model of Aleutian calc-alkaline magma chamber development is proposed. The main features are: (1) a single low pressure magma chamber is continuously supplied by primitive low-alumina basalt; (2) non-primary high-alumina basalt is formed along the chamber margins by selective gravitational settling of olivine and clinopyroxene and retention of plagioclase; (3) sidewall crystallization accompanied by crustal melting produces buoyant silicic (>63 wt % SiO2) liquids that pond at the top of the chamber, and; (4) continued sidewall crystallization, now isolated from the chamber wall, produces silicic liquids with ≤63 wt % SiO2 that increase the thickness and lowers the overall SiO2 content of the upper silicic zone. It is suggested that the maximum SiO2 content of 63% imposed on fractionation-generated magmas is due to a rheological barrier that prohibits the extraction of more silicic liquids from a crystal-liquid mush along the chamber wall.

The Cold Bay Volcanic Center, Aleutian Volcanic arc

The Cold Bay Volcanic Center has experienced two major stages of eruptive activity. Early (M-Series) acitivity produced bimodal Hi-Alumina basalt and calc-alkaline andesite lavas while later (FPK-Series) activity produced only calc-alkaline andesite. The spectrum of basalt compositions is believed to be due to high pressure (∼8 kb) fractionation at or near the base of the crust. Abundant mineralogical and geochemical evidence support a lower pressure mixing origin for all andesites. Inspection of the mineralogical data has shown that the earliest (M-Series) andesites were produced by mixing of basalt (<53 wt% SiO2) and silicic andesite (60.5 to 62.5 wt%) while later (FPK-Series) andesites resulted from the mixing of basaltic-andesite (53 to 56 wt%) and less silicic andesite (58.5 to 60.0 wt%). The major element and trace element geochemical data are consistent with a low pressure fractionation origin for the silicic endmember magmas and support the temporal variations in both mafic and silicic endmember compositions. The complete lack of crustal inclusions in all lavas is taken as evidence for a minimal crustal melting and/or assimilation role in the origin of the silicic endmembers. Many of the features of all andesites, including the important long term convergence of endmember magma compositions, are consistent with the process of liquid fractionation, accompanied by large scale magma mixing. A deduced upper limit of 62.5 wt% SiO2 for the silicic endmember magmas suggests that liquid fractionation, in the absence of major crustal melting, cannot produce more silicic magmas. A possible explanation is the presence of a rheological barrier, based on the concept of critical crystallinity (Marsh 1981), which prohibits more silicic liquids from being extracted from a crystal-liquid suspension.

Can high-alumina arc basalt be derived from low-alumina arc basalt? Evidence from Kanaga Island, Aleutian Arc, Alaska

The coexistence of both high-MgO, low-alumina basalt (LAB) and low-MgO, high- alumina basalt (HAB) magmatic inclusions within a single andesitic host lava on Kanaga Island provides the opportunity to rigorously evaluate the presence (or absence) of a fractionation relation between the two basalt-magma types. Traditional mass-balance calculations, Pearce ratio arguments, comparative phenocryst modes and compositions, and incompatible element (Rb, Ba, Zr) abundances all indicate that selective clinopyroxene and orthopyroxene fractionation from the plagioclase-clinopyroxene-orthopyroxene-titanomagnetite-saturated LAB magma yielded the plagioclase-phyric HAB magma. The same data are consistent with a model of low-pressure HAB generation calling upon selective mafic crystal fractionation along the margins of a crystallizing and convecting LAB magma chamber. The ability to demonstrate that nonliquid HAB magma can be produced from LAB through crystal fractionation eliminates many of the arguments used in support of a primary eclogite melting origin for HAB.

Solubility of copper in a sulfur-free mafic melt

The solubility of Cu in S-free mafic melts has been measured at a series of ƒO2 values and temperatures of 1245 and 1300°C. At constant temperature Cu solubility increases from 0.04 wt% at log ƒO2 = −11.9 to 1.10 wt% at log ƒO2 = −7.4. Copper solubilities were in excess of 8 wt% in two runs controlled at very high ƒO2 conditions of 10−1.4 and 10−22 Partitioning of Cu between metal and glass shows a strong ƒO2 dependence, with DCumetgl ranging from 90 at log ƒO2 = −7.4 to 2190 at log ƒO2 = −11.9. Slopes of Cu solubility and DCumet/gl vs. log ƒO2 suggest that Cu dissolves predominantly in the +1 valence state. Copper solubility decreases with increasing temperature at constant ƒO2, similar to experimental results for Ni, Co, and Mo (Dingwell et al., 1994; Holzheid et al., 1994). The data are consistent with Cu dissolution as an oxide (represented by CuO0.5) and suggest that changes in ƒO2 (Fe2+Fe3+ variations and Cu1+ complexation with Fe3+) may have large effects on the distribution of Cu between silicate and sulfide magmas. Results also suggest that the extraction of oxide-bonded Cu in mafic magmas by externally derived S may be an important mechanism in the generation of Cu-rich sulfide ores.

In situ ion-microprobe determination of trace element partition coefficients for hornblende, plagioclase, orthopyroxene, and apatite in equilibrium with natural rhyolitic glass, Little Glass Mountain Rhyolite, California

Abstract Partially crystalline hornblende gabbro inclusions from the Little Glass Mountain Rhyolite contain euhedral plagioclase, orthopyroxene, hornblende, and apatite crystals in contact with interstitial rhyolitic (71-76% SiO2) glass. Textural and mineral compositional data indicate that the gabbros crystallized sufficiently slowly that surface equilibrium was closely approached at the interface between crystals and the liquid. This rare occurrence represents a natural dynamic crystallization experiment with a “run time” that is not realistically achievable in the laboratory. SIMS analysis of mineral rim-glass pairs have permitted the determination of high-quality, equilibrium trace-element partition coefficients for all four minerals. These data augment the limited partition coefficient database for minerals in high-SiO2 rhyolitic systems. For all minerals, the D values are consistent with those anticipated from crystal-chemical considerations. These data further support a liquid SiO2 control on the REEs (and presumably other elements) partitioning wherein D values systematically increase with increasing liquid SiO2 content.

Oxygen isotope constraints on the petrogenesis of Aleutian arc magmas

The first measurement of 18 O/ 16 O ratios of plagioclase, clinopyroxene, orthopyroxene, and titanomagnetite phenocrysts from modern Aleutian island-arc lavas provides new insight and independent constraints on magma sources and intracrustal processes. Basalts are heterogeneous on the scale of the entire arc (δ 18 Oplag = 5.5‰-6.7‰) and individual volcanic centers (ranges of 0.3‰ to 0.8‰). Combined with Sr isotope and trace element data, δ 18 Oplag values suggest a variable magma source characterized by differences in the mantle wedge or the subducted sediment component along the volcanic front. Seven tholeiitic basalt to rhyodacite (50%-71% SiO 2 ) lavas from the Seguam volcanic center have nearly identical δ 18 Oplag values of 6.0‰ ±0.2‰, reflecting extensive closed-system plagioclase-dominated crystal fractionation. Oxygen isotope thermometry and pyroxene and oxide equilibria indicate that differentiation occurred between 1150 ±100 °C (basalt) and 950 ±100 °C (rhyodacite). In contrast, δ 18 Oplag values of 12 calc-alkalic basaltic andesites and andesites (53%-62% SiO 2 ) from the smaller Kanaga volcanic center span a broader range of 5.9‰-6.6‰, and consist of mostly higher values. Isotopic disequilibrium in the Kanaga system is manifest in two ways: two types of basaltic inclusions with contrasting δ 18 O values (6.0‰ and 6.5‰) occur in one andesite (δ 18 Oplag = 6.6‰), and in two other andesites plagioclase-titanomagnetite and clinopyroxene-titanomagnetite oxygen isotope temperatures are inconsistent. One andesite, however, yields concordant oxygen isotope temperatures of ∼850 ±100 °C, and a two-pyroxene temperature of 920 ±50 °C. Assimilation of isotopically heavy crustal wall rock and mixing off basaltic and andesitic magmas explains the variable and high δ 18 Oplag values and low eruptive temperature of the Kanaga andesites.