James H. Crocket

The chemical composition of igneous zircon suites: implications for geochemical tracer studies

Abstract Minor and trace element data obtained for ultrapure zircon fractions isolated from a variety of igneous rocks indicate that zircon grains crystallizing in different magmatic environments have unique geochemical signatures. Zircon megacrysts found in kimberlites (mantle zircons?) are characterized by extremely low abundances of U ( 60). In addition, zircon crystallizing in mafic magmas have elevated Sc levels (86 to 230 ppm), strongly fractionated REE patterns, and generally high Th/U ratios (> 1) compared to other igneous zircon suites. The presence of diagnostic geochemical signatures in igneous zircon suites indicates that zircon can be used as a geochemical tracer and may be useful for identifying the geochemical nature of source regions in provenance studies of single detrital zircon grains from sedimentary rocks or for deciphering basement rock compositions, if not exposed, from a geochemical study of xenocrystic zircon grains.

Partitioning of platinum-group elements (Os, Ir, Ru, Pt, Pd) and gold between sulfide liquid and basalt melt

New measurements on partitioning of platinum-group elements (PGE) Os, Ir, Ru, Pt, Pd, and Au between (Fe,Ni)-sulfide liquid and basaltic melt using low pressure, sealed silica glass tube, and controlled atmosphere methods at 1200–1250°C show an extremely wide variation, which spans much of the range reported in the literature. In spite of their geochemical coherence in natural high-temperature assemblages, PGE-Au apparently behave as seven different metals in individual laboratory experiments with silicate melts. Sulfide/silicate partition coefficients, D, generally increase with run time, in a manner consistent with either serial change in chemical state of the metals in the silicate melt or passivation of alloy phases. Ruthenium is the only metal presently investigated that exhibits reversible behaviour. Summary partition coefficients for the lowest PGE contents investigated under reducing conditions and low Ni contents remain unchanged, except for the addition of a value of (4.4 ± 2.4) × 103 for Ru. Absence of interference from the nugget effect and homogeneity and equilibration of silicate melt in the earlier experiments is demonstrated by new measurements on Ir in different grain-size fractions of glass. An additional set of partition coefficients for 37 mol% NiS and 100–1000 ppm PGE in the sulfide is: D(Os) = (30 ± 6) × 103, D(Ir) = (26 ± 11) × 103, D(Ru) = (6.4 ± 2.1) × 103, D(Pt) = (10 ± 4) × 103, and D(Pd) = (17 ± 7) × 103.

Partitioning of palladium, iridium, and platinum between sulfide liquid and basalt melt: Effects of melt composition, concentration, and oxygen fugacity

Abstract The partitioning of Pd, Ir, and Pt between immiscible (Fe, Ni)-monosulfide liquid and basalt melt has been investigated at 1300°C and at low pressure over the concentration range 40 to 20,000 ppm platinum-group element(s) (PGE) in the sulfide liquid and at oxygen fugacities from the C-CO-CO2 to the wustite-magnetite buffers. The experiments used sealed silica-glass tubes with internal oxygen buffers: PGE in glass were determined by radiochemical neutron activation analysis (RNAA). Partition coefficients (D) vary markedly with compositions of the sulfide liquid and silicate melt, increasing with decrease in oxygen fugacity, S, Fe, and possibly Ni, and with increase in total concentration of PGE. For 5 ppb PGE in the silicate melt and the iron-silica phase-fayalite (IQF) buffer, D(Pd) and D(Pt) are about 2 × 103, and D(Ir) is about 3 × 103; whereas, at the maximum concentration of PGE investigated, D(Pd) and D(Pt) are about 2 × 104, and D(Ir) is about 3 × 104. A single experiment confirms the marked fractionation of Pt from Pd predicted for partitioning with alloy in S-bearing and S-saturated silicate melts. The experimental D(PGE) values for low concentration of PGE are similar to D(PGE) calculated for many sulfide ore deposits, but are several orders of magnitude lower than calculated values for concordant sulfide PGE deposits in layered complexes.

Initial 87Sr/86Sr ratios of plutonic and volcanic rocks of the Central Andes between latitudes 26° and 29° south

Initial87Sr/86Sr ratios have been determined for 34 plutonic and volcanic rocks covering the entire age span of magmatic events associated with the Andean orogeny between latitudes 26° and 29° south. The igneous rocks, the majority dated by K/Ar mineral techniques, range in age from Lower Jurassic (190 m.y.) to Quaternary (0.89 m.y.). In addition, initial ratios were determined for three granitoid plutons and one metasediment from the pre-Mesozoic basement which underlies the entire Andean orogen in this transect at shallow depth. The compositions vary from basalt to rhyolite, and from quartz diorite to granodiorite or trondjemite, for the extrusives and intrusives, respectively. Mid-Cretaceous to Quaternary rocks exhibit a systematic west to east increase in mean strontium isotope ratio from 0.7022 to 0.7077, whereas the initial ratios of Jurassic plutons vary from 0.7043 to 0.7059, and do not correlate with age. The existence of unusually low initial ratios (e.g. 0.7022, 0.7023) for several Mesozoic plutonic rocks strongly implies a sub-crustal source for at least some of the Andean magmas. The time-dependent post-Jurassic increase in initial ratio is considered to reflect a systematic change in the composition of partial melts generated in response to the progressive subduction of a lithospheric slab. It is suggested that a systematic change in the locus of melting takes place from along or close to the upper surface of the subduction slab into hanging-wall mantle peridotite as subduction continues.

Implications of composition for experimental partitioning of platinum-group elements and gold between sulfide liquid and basalt melt: The significance of nickel content

Abstract Experimental partitioning of platinum-group elements (PGE) Os, Ir, Ru, Pt, and Pd, and Au between (Fe, Ni)-sulfide liquid and basaltic melt have been carried out at high Ni content with 67 ± 1 mol% NiS and 50 to 400 ppm PGE in the sulfide fraction. The study was designed to evaluate the significance of major element composition, in particular high Ni content, on partitioning of these noble metals. Charges were held at 1250°C for 48 h. Mean partition coefficients D (wt fraction PGE in sulfide/ wt fraction PGE in silicate) obtained were: D( Os ) = (10±2.6 x 10 3 ); D( Ir ) = (51±20 x 10 3 ); D( Ru ) = (7.0±5.4 x 10 3 );D( Pt ) = (16.5±6.3 x 10 3 ); D( Pd ) = (28±12.5 x 10 3 ); D( Au ) = (1.21±0.95 x 10 3 ) and pertain to 7 samples for Os, Ir, and Ru, 6 samples for Pt and Pd, and 5 for Au. The average values for Ir, Ru, Pt, and Pd are comparable with D values reported by Fleet et al. (1996) for a series of 5 samples equilibrated under virtually identical experimental conditions but on charges with lower Ni content. A significant correlation of individual D values with PGE in quenched sulfide was observed for Ru, Pt, and Pd, suggesting that variable concentrations of these metals in initial charges contribute to the inter-experiment variation in D. A marked increase in D(Au) to 13.6 × 1013 was found when the Au content of initial charges was increased from a ppb to a ppm range, a trend comparable to that of Ru, Pt, and Pd. D(PGE) values for ⋍50 ppm PGE in the sulfide liquid are D(Ru) = 1.45 × 103, D(Pt) = 10.5 × 103, and D(Pd) = 17.8 × 103. The significance of NiS for PGE-Au partitioning was investigated by comparing D values for controlled atmosphere experiments ranging from ∼4 to 67 mol% NiS in quenched sulfide. Sulfide liquids with less than ∼38 mol% NiS are associated with more variable D values for all PGE-Au than those of higher NiS content, which yield more consistent D values. Ruthenium is an exception and behaves in a reverse manner. There is no strong correlation of the magnitude of D with composition of the sulfide liquid but rather a correlation of the variance in D. It is suggested that the Ni/Fe proportions of the silicate melt strongly affect the behaviour of PGE-Au in silicate melts, probably through a control on speciation and solubility of noble metals.

Some aspects of the geochemistry of strontium and calcium in the Hudson Bay and the Great Lakes

Abstract The ratio Sr 87 Sr 86 in samples of water and shells of the peleoypod Mytihis edulis, Linne, from the Hudson Bay was found to be 0.7093 ± 0.0003 which is identical to the value of this ratio in the northern Atlantic Ocean. The concentrations of strontium and calcium in four water samples from the eastern Hudson Bay vary linearly with salinity and are appreciably lower than in the open ocean. The ratio Sr × 103/Ca in the Hudson Bay was found to be equal to 18.7 ± 0.3 and does not differ significantly from values obtained for surface water from the Atlantic Ocean. Concentrations of strontium and calcium in Lake Superior are uniform throughout and average Sr = 21.8 ± 0.4 ppb, Ca = 14.3 ± 0.5 ppm. The ratio Sr × 10 3 Ca is 1.53 ± 0.04. Concentrations of the same elements in Lake Huron exhibit regional variations and are significantly higher than in Lake Superior. The Sr × 10 3 Ca ratio is approximately 3.5. The distribution coefficient for Sr+2 for Mytilus edulis in the Hudson Bay ranges from 0.14 to 0.20 and is significantly lower than experimentally determined values in carbonates composed of a mixture of aragonite and calcite. The average distribution coefficient for Sr+2 in the aragonitic shells of Lamposilis is 0.256 ± 0.027 (σ). The concentration of strontium in shells of Lampsilis fluctuates widely, but appears to be controlled, to a first approximation, by the distribution coefficient of Sr+2 in calcium carbonate and water. Attention is drawn to the discrepancy between the Sr × 10 3 Ca ratios of surface run-off water and the modern ocean. The apparent enrichment of strontium over calcium in the oceans is an important aspect of the marine geochemistry of strontium and calcium.

Determination of some precious metals by neutron activation analysis

A neutron activation procedure for the determination of Ru, Pd, Os, Ir, Pt and Au in a single irradiation in silicate rocks, meteorites and sulfide ores has been developed. An alkali fusion was used to dissolve and mix 100 to 200 mg powder samples with appropriate carriers. The individual metals were separated and brought to a state of high radiochemical purity by distillation, ion exchange and solvent extraction techniques. Precious metal activities were counted by both γ and β-methods. The procedure was evaluated by replicate analyses of the granite and diabase rock standards, G-1 and W-1 and a Cu−Ni sulfide matte which had previously been analysed by emission spectrographic and spectrophotometric methods. The results were compared with previously published data. A major discrepancy was found only for Ir in W-1.

Precious metal abundances in some carbonaceous and enstatite chondrites

Abstract Neutron activation analysis has been used to determine the concentrations of Ru, Pd, Os, Ir, Pt and Au in seven carbonaceous and two enstatite chondrites. The average atomic abundances relative to 10 6 silicon atoms are tabulated below: Chondrite Ru Pd Os Ir Pt Au Group Carbon- aceous 1.84±0.09 6 1.32± 0.21 0.79± 0.12 0.76 ± 0.28 1.4 ± 0.25 0.18 ± 0.040 Enstatite 1.60 ± 0.08 1.23 ± 0.06 0−66 ± 0.01 0.25 ± 0.03 1.1 ± 0.00 0−19 ± 0.010 The data are compared with previously published precious metal values for olivine bronzite and olivine hypersthene chondrites. Abundances in olivine bronzite and carbonaceous chondrites are generally comparable except for Ir which is lower in the former group. Both the enstatite and olivine hypersthene chondrites have lower atomic abundances than the carbonaceous chondrites except for Au in the enstatite chondrites. The average depletion of precious metals in the olivine hypersthene chondrites relative to the carbonaceous chondrites is approximately 40%. As no obvious mechanisms for fractionation of precious metals during condensation of meteoritic matter are apparent, their abundances in carbonaceous chondrites should provide good estimates of the cosmic abundances of these elements.

Distribution of noble metals across the Cretaceous/Tertiary boundary at Gubbio, Italy: Iridium variation as a constraint on the duration and nature of Cretaceous/Tertiary boundary events

Iridium, Pd, Pt, and Au were determined in sections from the Bottaccione Gorge and Contessa Valley, Gubbio, Italy, by radiochemical neutron activation. Shales and limestones were sampled from 2.85 m above to 219 m below the Cretaceous/Tertiary (K/T) boundary. Metal enrichment was evaluated by comparing the boundary shale region with the lower part of the section (background). Iridium is concentrated by 63 times in the boundary shales in comparison with the background, whereas other metals are enriched by no more than 2.2 times. The enrichment of Ir is not confined to the boundary shales but extends approximately 2 m above and below this horizon. Within this Ir-rich region there are four distinct Ir maxima in addition to the major Ir enrichment in the K/T boundary shales. Iridium maxima are stratigraphically coincident with maxima in abundances of shocked minerals characteristic of explosive volcanism. Limestones are much lower in noble metals than shales, and their Ir contents in the K/T boundary region are largely accounted for by their minor clay mineral contents. The time represented by the 4 m of Ir-rich section is at least 3 x 10 5 yr, if published sedimentation rates are used. To sustain an increased Ir flux over this period and to account for the Ir distribution near the K/T boundary, intense volcanic activity is a preferred alternative to impact of extraterrestrial material.

PGE in fresh basalt, hydrothermal alteration products, and volcanic incrustations of Kilauea volcano, Hawaii

Abstract The concentrations of Os, Ir, Pd, and Au in fresh unaltered Kilauean tholeiite were determined by radiochemical neutron activation analysis. For a suite of 18 samples, averages were: Os = 0.38 ± 0.23, Ir = 0.38 ± 0.14, Pd = 2.40 ± 1.04, and Au = 1.78 ± 0.57 (in ppb with a 1σ SD). Correlations of these metals with Co, Cr, Cu, Ni, and MgO in fresh basalts, and petrographic observations, indicate that Os and Ir are carried mainly in chromite, much of which occurs as inclusions in olivine phenocrysts. Palladium correlations suggest its occurrence partly in olivine and partly in the matrix whereas Au seems to be predominantly a matrix constituent. Altered basalts were analyzed for Ir, Pd, and Au in a suite of 19 samples from five different locations. Minor changes only in either concentrations or element ratios were found for Ir and Pd when fresh and altered rock data were compared. However, Au was consistently enriched in altered relative to fresh rocks. These results imply that Pd and Ir, in contrast to Au, will likely retain their eruptive signatures upon burial in a subaerial eruptive setting. High-temperature sulfate-dominated condensates generate incrustations enriched in Ir, Os, Au, and Pd by approximately 50, 20, 10, and 3×, respectively, relative to fresh rocks. In contrast, low-temperature native sulfur deposits are the most depleted material found in the study with Ir, Pd, and Au lower by factors of 10, 4, and 5 compared with fresh rock averages. The strong enrichments of Os and Ir in the high-temperature suite are attributed mainly to enhanced volatility in highly oxygenated magmatic hydrothermal fluids contaminated by meteoric water near the structural top of volcanic conduits. The relatively smaller Pd enrichment, which is dependent on the chloride content of fluids, implies that PGE partitioning into volcanic fume may fractionate these metals (e.g., Pd versus Ir) relative to host basalt in the eruptive process.