Komatiitic Liquids Research Articles

Physical and chemical constraints on platinum-group element behavior during crystallization of a basaltic komatiite liquid; example of the Proterozoic Delta Sill, New Quebec, Canada

The Delta sill, located in the southern part of the 2.03 to 1.92 Ga Cape Smith fold belt, northern Quebec, is composed of a peridotite, pyroxenite, gabbronorite, and anorthosite suite derived from a basaltic komatiite parental magma (MgO = 13.58-15.66%). Despite its rather limited thickness (215-275 m), it contains three platinum-group elements (PGE) horizons, two slightly enriched horizons, one in the peridotite and one in the lower gabbronorite, and a strongly PGE-enriched horizon (forming a reef) in the upper gabbronorite. The high PGE content in the reef implies a system with an R factor (volume of immiscible sulfide liquid/ volume of silicated magma) of 100 to 1.000. However, the small volume of magma involved suggests that several other processes controlled the PGE behavior. Some experimental works have shown that the CaO and FeO contents in parental magmas influence the sulfide capacity and consequently the sulfur fugacity. The last parameter, as well as the oxygen fugacity, have been previously recognized as having a fundamental influence on the PGE enrichment processes. Because of the relatively high CaO contents (7.21-11.26%) of parental magma and the existence of PGE enrichment, the Delta sill provides an opportunity to test the effects of varying sulfur and oxygen fugacities on the processes of PGE enrichment. Thus, the PGE concentration profiles observed in the two sections transecting the sill are in good agreement with the experimentally determined behavior of these elements in silicate melts as a function of oxygen and sulfur fugacities. First, the relatively high Ca content of the initial magma induced a high sulfide capacity in the liquid. Second, decreasing sulfide capacity (and correlatively, an increase in sulfur fugacity) produced by the precipitation of, calcic phases--notably clinopyroxene and plagioclase--account for the PGE anomalies as well as the late-stage crystallization of sulfides in the reef. In the lowermost PGE anomaly, Ir and Ru enrichment within the peridotitic layer can be clearly related to early ferrochromite crystallization due to increasing oxygen fugacity, whereas subsequent Pt enrichment can be related to the formation of an isoferroplatinum phase (Pt 3 Fe) associated with ferrochromite. In the second PGE anomaly, located in the lower gabbronorite, synchronous Ir and Ru, Rh, Pt, and Pd, and S enrichments are due to the precipitation of small amounts of sulfides under low oxygen fugacity. Both the Ir and Ru and the Rh, Pt, and Pd, left over from the first chromite-forming stage, are complexed by sulfur when the sulfide capacity increases, consequent to CaO removal from the melt by plagioclase fractionation. Finally, the third anomaly in the PGE reef, appears to record the crystallization of a differentiated Fe- and Ti-rich residual liquid containing fluid phases and sulfur. It reflects the migration of an S-bearing liquid following the important lowering of sulfide capacity in the liquid due to the increase of SiO 2 and TiO 2 in the liquid and to the precipitation of Ca-bearing minerals.

Geochemistry of volcanic rocks associated with Cu-Zn and Ni-Cu deposits in the Abitibi Subprovince

Volcanogenic massive sulfide (VHMS) deposits in the Abitibi subprovince are preferentially associated with volcanic successions containing >150 m thicknesses of felsic volcanic rocks ( approximately 50% by area of volcanic terranes) and are found within volcanic sequences of at least three distinct affinities. Group I, which is host to greater than half of the volcanogenic massive sulfide deposits by tonnage and which comprises only approximately 10 percent by area of volcanic terranes, is composed of bimodal, tholeiitic basalt-basaltic andesite, and high silica rhyolite. The basaltic andesites and high silica rhyolites are characterized by high high field strength element and heavy rare earth element (REE) contents, low light to heavy REE ratios (most with La N /Yb N = 0.8-3), and strong negative Eu anomalies. The Kamiskotia, Matagami, and Chibougamau (Lower cycle) volcanogenic massive sulfide areas, all of which are also underlain by large, synvolcanic gabbroic complexes, are associated with group I volcanic sequences. The Kidd Creek, Potter, Normetal, and Horne deposits are also included in this category. Group II, which is host to one-third of the volcanogenic massive sulfide deposits by tonnage and which is also approximately 10 percent by area of volcanic terranes, is composed of bimodal, transitional tholeiitic to calc-alkalic andesitc and rhyolite, characterized by intermediate high field strength element contents and slightly higher REE ratios (La N /Yb N = 1-4: Noranda camp (excluding the Horne deposit) and Val d'Or camp). Group III is host to only one known deposit, the Selbaie mine, which is unusual in that much of its mineralization cuts stratigraphy. The Selbaie mine sequence contains calc-alkalic andesite-rhyolite with relatively low high field strength element and REE contents, and higher REE ratios (La N /Yb N = 3-9). The vast majority of volcanogenic massive sulfide-bearing mafic and felsic volcanic rocks in the Abitibi subprovince have La N /Yb N < 5.Barren volcanic sequences are group IV--calc-alkalic basaltic andesite to rhyodacite, with low high field strength element and relatively high REE ratios (La N /Yb N = 8-20), represented by the upper Skead Group, the Quebec Hunter Mine Group, and the Upper cycle Chibougamau rocks; and group V, mafic to felsic alkalic volcanic rocks, with high REE ratios (La N /Yb N = 12-62), represented by the Timiskaming, Opemisca, and Ridout series rocks.Group I is most similar to thickened oceanic rift suites (e.g., Galapagos spreading center, Iceland East rift), group II is similar to suites in rifled island arcs (e.g., Hokuroku district, Japan), and group III is comparable to continental arc suites (e.g., Southern Volcanic zone, central Chile). Groups IV and V are comparable to arc-related suites derived from metasomatized mantle, with variable amounts of crustal contamination (e.g., Setouchi area, Japan; Roman province, Italy).Known Ni-Cu deposits in volcanic sequences are hosted exclusively in komatiitic flows and hypabyssal sills, represented by chill compositions with high MgO contents (20-35 wt %, anhydrous), very low incompatible element contents, and depleted signatures (La N /Sm N = 0.5-0.8; Zr/Y < 2.5; Alexo, Marbridge, and Shaw Dome deposits), in comparison to most barren komatiites in the Abitibi subprovince. Their depleted trace element signatures indicate minimal crustal assimilation, which would appear contrary to the widely held view that crustal contamination (e.g., addition of silica, sulfur, oxygen) of a mantle-derived melt is necessary for sulfur saturation and magmatic sulfide segregation. This problem may be reconciled by considering the dynamics of komatiite flows (Lesher and Arndt, 1990) where contaminated* Present address: 23 Euclid Avenue, Ottawa, Ontario, Canada KiS 2W2. komatiite liquid segregated and deposited sulfide and was subsequently swept downstream and replenished by uncontaminated magma over the deposit as the flow ebbed and cooled. When considered together with geologic features such as evidence for thermal erosion in the footwall (Langmuir deposits) and the presence of sulfidic footwall rocks (Shaw Dome deposits), the distinctive, depleted geochemical signature of mineralized komatiites is useful in exploration for Ni-Cu deposits.

Constraining the potential temperature of the Archaean mantle: A review of the evidence from komatiites

The maximum potential temperature of the Archaean mantle is poorly known, and is best constrained by the MgO contents of komatiitic liquids, which are directly related to eruptive temperatures. However, most Archaean komatiites are significantly altered and it is difficult to assess the composition of the erupted liquid. Relatively fresh lavas from the SASKMAR suite, Belingwe Greenstone Belt, Zimbabwe (2.7 Ga) include chills of 25.6 wt.% MgO, and olivines ranging to Fo 93.6, implying eruption at around 1520°C. A chill sample from Alexo Township, Ontario (also 2.7 Ga) is 28 wt.% MgO, and associated olivines range to Fo 94.1, implying eruption at 1560°C. However, inferences of erupted liquids containing 32–33 wt.% MgO, from lavas in the Barberton Greenstone Belt, South Africa (3.45 Ga) and from the Perseverance Complex, Western Australia (2.7 Ga) may be challenged on the grounds that they contain excess (cumulate) olivine, or were enriched in Mg during alteration or metamorphism. Re-interpretation of olivine compositions from these rocks shows that they most likely contained a maximum of 29 wt.% MgO corresponding to an eruption temperature of 1580°C. This composition is the highest liquid MgO content of an erupted lava that can be supported with any confidence. The hottest modern magma, on Gorgona Island (0.155 Ga) contained 18–20% MgO and erupted at circa 1400°C. If 1580°C is taken as the temperature of the most magnesian known eruption, then the source mantle from which the liquids rose would have been at up to 2200°C at pressures of 18 GPa corresponding to a mantle potential temperature of 1900°C. These temperatures are in excess of the mantle temperatures predicted by secular cooling models, and thus komatiites can only be formed in hot rising convective jets in the mantle. This result requires that Archaean mantle jets may have been 300°C hotter than the Archaean ambient mantle temperature. This temperature difference is similar to the 200–300°C temperature difference between present day jets and ambient mantle temperatures. An important subsidiary result of this study is the confirmation that spinifex rocks may be cumulates and do not necessarily represent liquid compositions.

Fractionation of REEs by olivine and the origin of Kambalda komatiites, Western Australia

Analysis of trace elements in spinifex and cumulate samples from a single komatiite lava flow from Kambalda (Western Australia) reveals significant variations in the relative proportions of REE and other elements normally considered insensitive to olivine fractionation. Variations of LREE result from a combination of mobility during alteration and variable degrees of wall-rock assimilation. Relative concentrations of HREE also vary: in spinifex-textured lavas, the heavier elements are depleted [(Gd/ Yb) n ∼ 1.08]; in olivine cumulates these elements are enriched [(Gd/Yb) n ∼ 0.92]. Although the influence of alteration during metamorphism cannot be ruled out completely, we prefer to attribute these effects to olivine crystallization and accumulation, a process that can effectively fractionate REE only if: 1. (a) heavier REE are much more compatible in olivine than lighter REE; and 2. (b) olivine partition coefficients for komatiitic liquids are ∼ten times higher than published values for basaltic liquids. This interpretation is supported by analyses of REE and other trace elements (by IDMS, INAA, and ion microprobe) in olivine separated from the Kambalda komatiite. These analyses reveal extreme fractionation of HREE [(Gd/Yb) n ∼ 0.28] and unusually high concentrations of REE ( Sm = 0.013 ppm, Yb = 0.06 ppm) in the olivine. Partition coefficients calculated by relating the olivine to an appropriate spinifex lava using Mg-Fe partitioning are: La to Sm ≤ 0.02, Gd = 0.02, and Yb = 0.09. A possible role of chromite is eliminated by analyses of a chromite separate, which yields partition coefficients for all REE < 0.01. These partitioning data provide a basis for discussing the origin of komatiites. Munro-type komatiites (chondritic Al 2O 3/TiO 2 and HREE) result from relatively high degrees of mantle melting, 30% or more. The melt segregates at pressures around 8 GPa, the pressure at which melt becomes less dense than the ol ± opx matrix. Barberton-type komatiites (low Al 2O 3/TiO 2 and depleted HREE) form by lower degrees of melting at depths greater than 15 GPa, and their segregation is driven by the density difference between melt and a majorite-rich matrix.

Orthopyroxene stability along the peridotite solidus and the origin of cratonic lithosphere beneath southern Africa

The stability of orthopyroxene within the melting interval of natural fertile peridotite between 4.5 and 6.5 GPa was investigated. Orthopyroxenes rich in CaO and Al 2O 3 were found stable at and above the fertile peridotite solidus in this pressure range. These results contradict a previous melting study of fertile peridotite, which found orthopyroxene absent in the melting interval above 4.5 GPa, but are consistent with phase equilibrium investigations which show orthopyroxene stable with olivine between 4.5 and 6.0 GPa along the liquidi of komatiites in both simple and complex systems, and near the solidus of depleted abyssal and cratonic peridotites. The new melting relationships for fertile peridotite are used in a simplified model for the origin of low-temperature, coarse-grained, garnet peridotite xenoliths from the Kaapvaal craton, southern Africa. The model depicts cratonic peridotites as former harzburgitic residues of Archean komatiitic liquid extraction which cooled to become lherzolites at ambient mantle temperatures within the mantle root of the craton. The large and varied abundance of orthopyroxene, and published petrographical and bulk chemical observations for cratonic peridotites from the Bultfontein and Matsoku pipes, southern Africa, are generally explainable by this ‘exsolution’ origin. Cratonic peridotites from other localities in the Kaapvaal craton, however, may contain excess modal pyroxene and/or garnet of uncertain origin, which cannot easily be accounted for by the ‘exsolution’ hypothesis.

The equation of state of a molten komatiite: 1 Shock wave compression to 36 GPa

The equation of state (EOS) of an initially molten (1550°C) komatiite (27 wt % MgO) was determined in the 5–36 GPa pressure range via shock wave compression. Shock wave velocityUsand particle velocityUp(kilometers/second) follow the linear relationshipUs= 3.13(±0.03) + 1.47(±0.03)Up. Based on a calculated density at 1550°C, 0 bar of 2.745±0.005 g/cm3, thisUs‐Uprelationship gives the isentropic bulk modulusKS= 27.0 ± 0.6 GPa, and its first and second isentropic pressure derivatives,K′S= 4.9 ±0.1 andK″S= −0.109 ± 0.003 GPa−1. The calculated liquidus compression curve agrees within error with the static compression results of Agee and Walker (1988) to 6 GPa but is less dense than their extrapolated values at higher pressures. We determine that olivine (Fo94) will be neutrally buoyant in komatiitic melt of the composition that we studied near 8.2 GPa. Clinopyroxene would also be neutrally buoyant near this pressure. Liquidus garnet‐majorite may be less dense than this komatiitic liquid in the 20–24 GPa interval; however, pyropic‐garnet and perovskite phases are denser than this komatiitic liquid in their respective liquidus pressure intervals to 36 GPa. Liquidus perovskite may be neutrally buoyant near 70 GPa. At 40 GPa, the density of shock‐compressed molten komatiite would be approximately equal to the calculated density of an equivalent mixture of dense solid oxide components. This observation supports the model of Rigden et al. (1989) for compressibilities of liquid oxide components. Using their theoretical EOS for liquid forsterite and fayalite, we calculate the densities of a spectrum of melts from basaltic through peridotitic that are related to the experimentally studied komatiitic liquid by addition or subtraction of olivine. At low pressure, olivine fractionation lowers the density of basic magmas, but above 13–14 GPa this trend is reversed. All of these basic to ultrabasic liquids are predicted to have similar densities at 13–14 GPa, and this density is approximately equal to the density of the bulk (preliminary reference Earth model) mantle in this pressure range. This suggests that melts derived from a peridotitic mantle may be inhibited from ascending from depths greater than 400 km.

The equation of state of a molten komatiite: 2. Application to komatiite petrogenesis and the Hadean Mantle

New experimental data for the equation of state of a komatiitic liquid were used to model adiabatic melting in a peridotitic mantle. If komatiites are formed by &gt;30% partial melting of a peridotitic mantle, then komatiites generated by adiabatic melting come from source regions that began their unmelted ascent in the lower transition zone (≈500–670 km) or the lower mantle (&gt;670 km). The great depth of incipient melting implied by this model suggests that komatiitic liquids may form in a pressure regime where they are denser than their coexisting crystals, possibly the bulk mantle. Although komatiitic magmas are thought to separate from residual crystals in the mantle at a temperature ≈200°C greater than modem mid‐ocean ridge basalts (MORBs), their ultimate sources are predicted to be diapirs that, if adiabatically decompressed from initially solid mantle, were more than 700°C hotter than the sources of MORBs and were derived from great depth. We also studied the evolution of an initially molten mantle, i.e., a magma ocean. Our model considers the thermal structure of the magma ocean, density constraints on crystal segregation, and approximate phase relationships for a nominally chondritic mantle. Crystallization will begin at the core‐mantle boundary. Perovskite buoyancy at &gt;70 GPa may lead to a compositionally stratified lower mantle with iron‐enriched magnesiowüstite content increasing with depth. Large convective velocities in the magma ocean would prohibit crystal‐liquid fractionation by settling or flotation until quiescent boundary layers form. Such boundary layers could form when the crystal content of the magma reaches a critical value (near 44 vol %) and, in the late stages of crystallization, around unmelted blocks of foundered protocrust. Matrix compaction and diapirism could also lead to fractionation effects. The upper mantle could be depleted or enriched in perovskite components relative to the bulk mantle. Olivine neutral buoyancy may lead to the formation of a dunite septum in the upper mantle, partitioning the ocean into upper and lower reservoirs, but this septum must be permeable.

Melting in an Archaean mantle plume: heads it's basalts, tails it's komatiites

THE lower part of most Archaean greenstone sequences is dominated by interlayered basaltic and komatiitic (ultrabasic) flows. These two magma types are petrologically and geochemically distinct, yet they display a close spatial and temporal association. The origin of the anomalously high-temperature komatiitic liquids has been much debated because of the implications for the thermal structure and composition of the Archaean mantle. Here we argue that both the basalts and komatiites are produced by a starting thermal plume rising in a warmer Archaean mantle. Fluid-dynamics studies show that a starting plume consists of a hot axial jet, capped by a large head into which cooler surrounding mantle is entrained. Our calculations for such a flow indicate that komatiites could form by melting in the high-temperature axis of the plume and basalts by melting in the cooler head. The sudden onset and limited duration of the basalt/komatiite sequences, seen in the greenstone belts of Western Australia and elsewhere, are explained by this model.

Mass balance and phase density constraints on early differentiation of chondritic mantle

A surprisingly simple and precise major element mass balance is consistent with derivation of average upper mantle peridotite from a partially molten chondritic Earth by subtraction of perovskite and addition of olivine. Majorite involvement is precluded unless some as yet unidentified components play a role. Perovskite subtraction during a primordial melting event is expected to occur by crystal fractionation at depth, while olivine addition is accomplished by a combination of buoyancy mechanisms: crystal flotation from a deep layer of melt buried by its own compressibility to the base of the solidifying upper mantle and subsequent solid state convection of this buoyant magnesian olivine upward. These processes are consistent with known density relations of crystals and liquid at very high pressure. Mass balance predicts that the residual magma body at depth after supplying olivine by flotation upward can be komatiitic. Distribution of originally C1 chrondritic bulk Earth material a few 100 m.y. after primordial differentiation is solid peridotite upper mantle, perovskite lower mantle, and a komatiitic liquid sandwich horizon.

The Perseverance Ultramafic Complex, Western Australia: The Product of a Komatiite Lava River

The Perseverance ultramafic complex is a body of olivine-rich komatiitic rocks spatially associated with the Agnew nickel deposit, in the Agnew-Wiluna greenstone belt of the Archaean Yilgarn Block in Western Australia. The complex consists of a central lenticular body, up to 700 m thick, of olivine adcumulates, flanked by laterally extensive sheet-like bodies of olivine orthocumulates and spinifextextured komatiite flows. Rocks progressively further away from the central lens have chemical compositions reflecting higher original proportions of komatiite liquid to cumulus olivine. Parent liquids had MgO contents between 25 and 32% MgO, approximately chondritic Al/Ti ratios and HREE patterns, and moderate depletion in LREE. Olivines within the adcumulate lens show a progressive increase in forsterite content from Fo93 at the bottom to Fo94..5 at the top. Calculated original olivine compositions in ihe flanking rocks are similar to those at the base of the central lens. Original olivine nickel contents show a symmetrical variation from maximum values of 3500 ppm at the top of the central lens, through minimum values of 1000 ppm at the base and margins of the central lens to intermediate values in the distal rocks. The complex as a whole shows evidence for nickel depletion relative to other komatiite suites. These observations are explained in terms of prolonged eruption and flow of komatiitic lava down a major flow channel or lava river. Adcumulates crystallized on the floor and sides of the central channel, which was formed at an early stage by thermal erosion of floor rocks. Episodic overflow of the central channel produced distal 'flood plain' rocks consisting of olivine orthocumulates and layered flows. Lavas became more magnesian and nickel-rich with time, giving rise to the observed spatial variation in primary olivine composition. Nickel depletion of the earliest lavas is attributed to pre-eruption segregation of large volumes of immiscible Fe-Ni-sulfide, which were concentrated to form the underlying Agnew nickel deposit.

Densities of Na2O-K2O-CaO-MgO-FeO-Fe2O3-Al2O3-TiO2-SiO2 liquids: New measurements and derived partial molar properties

The densities of 27 liquids in the system Na2O-K2O-MgO-MgO-FeO-Fe2O3-Al2O3-TiO2-SiO2 have been measured using the double-bob Archimedean technique. These results indicate that multicomponent silicate liquid volumes have a linear dependence on composition with the exception of the TiO2 component. The equation: V(T) = ∑Xi(T)Vi(T) + XNa2OXTiO2VNa2O-TiO2 was used to derive values of the oxide partial molar volumes (V̄i) by the method of least squares. Regressions were made separately at 1573, 1673, 1773 and 1873 K (as liquid ferric-ferrous ratios change with temperature) with relative standard errors for each fit of 0.38%, 0.32%, 0.30% and 0.32%, respectively. Derived dV̄i/dT values by separate least squares regression for each oxide component reproduce the measured dV/dT of the experimental liquids by 20.21% on average. The effects of iron redox state on the density of a variety of natural liquids are demonstrated and at most amount to a variation of 1%.These new data on silicate liquid volumes were used to re-derive oxide partial molar compressibilities, dV̄i,T/dP and \\́gbiT, at 1 bar from ultrasonic velocity and calorimetric data from the literature. The fits for dV̄i,T/dP and \\́gbi,T at 1673 K have relative standard errors of 3.9% and 1.8%, respectively, which represent substantial improvements over previous fits. Two applications of these volume data are given: firstly, the fusion curve of diopside is calculated up to 95 kbr using an equation of state for liquid volume expressed as: V(T, P) = V(T) exp ∝ -β(T)(1 − bP+cP2)dP (where β = K−1 and therefore dK/dP = K′ = −β−2dβ/dP); secondly, using a range of K′ appropriate to a komatiite liquid, the pressure-temperature conditions where the density of the liquid equals that of its olivine phenocrysts are calculated.

Noble metal abundances in komatiite suites from Alexo, Ontario and Gorgona Island, Colombia

The distribution of the chalcophile and siderophile metals Cu, Ni, Au, Pd, Ir, Os and Ru in an Archaean komatiite flow from Alexo, Ontario and in a Phanerozoic komatiitic suite of Gorgona Island, Colombia, provides new information about the geochemical behaviour of these elements. Copper, Au and Pd behave as incompatible elements during the crystallization of these ultramafic magmas. In contrast, Ni, Ir, Os and Ru concentrations systematically decrease with decreasing MgO contents, a pattern characteristic of compatible elements. These trends are most probably controlled by olivine crystallization, which implies that Ir, Os and Ru are compatible in olivine. Calculated partition coefficients for Ir, Os and Ru between olivine and the melt are about 1.8. Compared to primitive mantle, parental komatiitic liquids are enriched in (incompatible) Cu, Au and Pd and depleted in (compatible) Ir, Os and Ru. Within both Archaean and Phanerozoic komatiites, noble metal ratios such as Au Pd , Ir Os , Os Ru and Ru Ir and ratios of lithophile and siderophile elements such as Ti/Pd, Ti/Au are constant and similar to primitive mantle values. This implies that Au and Pd are moderately incompatible elements and that there has been no significant fractionation of siderophile and lithophile elements since the Archaean. Platinum-group element abundances of normal MORB are highly variable and always much lower than in komatiites, because MORB magma is saturated with sulfur and a variable but minor amount of sulfide segregated during mantle melting or during the ascent of magma to the surface. Sulfide deposits associated with komatiites display similar chalcophile element patterns to those of komatiites. Noble metal ratios such as Pd Ir , Au Ir , Pd Os and Pd Ru can be used to determine the composition of the host komatiite at the time of sulfide segregation.

The role of fluids in the metamorphism of komatiites, Agnew nickel deposit, western Australia

The Agnew nickel sulfide deposit is spatially associated with a lenticular body of ultramafic rocks which shows a concentric zonation in metamorphic mineralogy. Olivine + tremolite + chlorite + cummingtonite ±enstatite assemblages occur at the margin of the ultramafic lens, giving way to olivine + anthophyllite, olivine + talc and olivine + antigorite assemblages successively inwards. These rocks are interpreted as having crystallized from komatiitic lavas, and exhibit a spectrum of compositions from those of original flow tops to pure olivine adcumulates. The relative modal abundances of metamorphic olivine, tremolite and chlorite reflect original proportions of cumulus olivine and komatiite liquid in the protolith. Peak metamorphic conditions are estimated at 550° C, based on garnet-biotite thermometry, at a maximum pressure of 3 kb. This temperature falls within the narrow range over which metamorphic olivine may co-exist with enstatite, anthophyllite, talc or antigorite depending upon the fugacity of water in the metamorphic fluid. The observed mineralogical zonation is therefore attributed to infiltration by CO2-rich fluids, generated by decarbonation of talc-carbonate rocks formed during pre-metamorphic marginal alteration of the ultramafic lens. Metamorphic fluids were essentially binary mixtures of water and CO2, with minor H2S having a maximum partial pressure less than 1 percent of total pressure. Enstatite-bearing assemblages formed in the presence of CO2-rich fluids at fluid: rock volume ratios close to one, while anthophyllite, talc and antigorite bearing assemblages formed in the presence of progressively more water-rich fluids at progressively lower fluid-rock ratios.

Early Proterozoic Metavolcanic Suites of the Northernmost Part of the Baltic Shield

Summary Metavolcanic rocks of Proterozoic age occur in three major supracrustal belts in the northernmost part of the Baltic Shield, in Norway and the USSR. In the west, the Kautokeino Belt , and its continuation into the tectonic windows of Finnmark, contains thick sequences of tholeiitic metabasalt the age of which is poorly constrained at present. In the Repparfjord Window, these basalts are strongly enriched in Fe, Ti, LIL and LREE and are chemically similar to modern continental flood basalts. Stratigraphic equivalents in the Alta area and Kautokeino Belt are less enriched in incompatible elements, and more like MORB in composition, although a within-plate setting is indicated by the field evidence. These differences are believed to reflect heterogeneity of the mantle source region, with a locally ‘enriched’ source in the Finnmark windows. The Karasjok Belt of central Finnmark forms the northward continuation of the Kittilä Belt of northern Finland. Amphibolitic rocks derived by metamorphism of tholeiitic basalts dominate the tectonostratigraphy of the greenstone belt. These are associated with metamorphosed komatiitic rocks which have recently yielded a Sm-Nd age of 2.1 Ga. The Karasjok Belt is interpreted as an early Proterozoic supracrustal sequence thrust westwards as an allochthonous sheet during uplift of the Lapland Granulite Complex in the Svecokarelian Orogeny at about 1.9 Ga. The Pechenga Complex of the USSR continues westward into Finnmark, and overlies the Archaean gneisses of the Kola Complex. The Kola Superdeep Borehole penetrated a metamorphosed supracrustal sequence 7 km thick, dominated by basalts and andesites of calcalkaline composition. The MORB-normalized geochemical patterns of these lavas indicate that they were erupted in a mature continental volcanic arc. The geochemical data support geotectonic models invoking rifting of pre-existing continental crust in the early Proterozoic. Komatiitic liquids were locally able to reach the earth’s surface, but probably only in the rift zones suffering greatest extension. Calcalkaline arc magmatic suites along the western edge of the Kola Craton reflect subduction of‘oceanic’ lithosphere, although no trace of this is preserved along the suture now marked by the Karasjok Belt.

Comparison of Archean dunites and komatiites associated with nickel mineralisation in Western Australia: Implications for dunite genesis

The two important classes of Archean Fe-Ni-Cu sulphide deposits in Western Australia are those hosted by (1) coarse-grained adcumulate dunites (e.g. Agnew, Mount Keith) and (2) komatiitic lavas (e.g. Kambalda). Close similarities in compositions of relict olivine, pyroxene, and chromite from dunitic and komatiitic host rocks and compositional and textural similarities between chilled margins of some dunites and some komatiites indicate origins from similar komatiitic liquids which contained 20%–32% MgO. As field relations suggest that the dunites are probably lateral stratigraphic equivalents of spinifextextured komatiites, they are interpreted as the slowcooled end member of a continuum from spinifex-textured komatiites (former liquids) through cumulate komatiites (40%–90% olivine) to adcumulate komatiitic dunite (>90% olivine). Previous classification of these Ni deposits as “intrusive”- and “volcanic”-related deposits now appears unwarranted, and the terms komatiitic dunite-hosted and komatiite-hosted deposits are preferred.

The continental and oceanic crust and lithosphere in the Archaean: isostatic, thermal, and tectonic models

Simple isostatic models of the development of selected Archaean basins and greenstone belts can be used to constrain models of the Archaean lithosphere. Geophysical, geochemical, and thermal data imply that in southern Africa the late Archaean crust was probably comparable in thickness to the modern continental crust, and the whole lithosphere was perhaps over 80–100 km thick. Two Archaean belts studied may have developed by stretching of the order of 1.3–1.6:1. Isostatic and thermal data and petrological assumptions allow the construction of simple models of the Archaean oceanic lithosphere. Oceans were probably deep: mid-ocean ridges were probably subaqueous and fed by komatiitic liquids. Calc-alkaline volcanism in the Archaean may have been produced as a consequence of the subduction of hydrated komatiitic oceanic crust into a hot, low-viscosity asthensophere.

The komatiite nickel sulfide association at Scotia; a petrochemical investigation of the ore environment

The Scotia nickel deposit is typical of the volcanic peridotite-associated type of Archean sulfide deposit. It occurs at the basal contact of a massive dunitic unit which is itself the basal unit of a thick extrusive komatiitic sequence. The ore zone is marked by a thickening of the overall ultramafic sequence and is located in a depression at the footwall contact. Sulfide textures in the massive ore indicate that it has been strongly deformed, which is in marked contrast to the overlying matrix ore and komatiitic sequence. Textural and mineralogical preservation of olivine in the ultramafic sequence shows that the sequence is made up of a series of thin flow units with gravity or flow-differentiated olivine-rich zones. Between these flows are thicker flows, characterized by massive olivine harrisites, commonly with ocelli structures. Throughout the sequence there are spinifex-textured olivine zones and well-preserved ultramafic volcanic breccias.Major element data indicate that the primary komatiitic magma had 30 wt percent MgO (anhydrous) and this gave rise to all the flow types. Microprobe data on relic brown olivines show that the spinilex-textured olivines are strongly zoned (Fo (sub 75-93) ) while the harrisitic olivines are remarkably constant at about Fo 92 . Olivine chemistry within the ore zone reflects the position of the olivine with respect to the massive sulfides and also subsolidus reequilibration. A particular feature of the ore zone is the occurrence of colorless olivine which appears to be degraded after brown igneous olivine. The colorless olivine has abnormally low Ni concentrations associated with Fo- and Mn-rich cores. Rims to these grains are Mn depleted, Ni enriched, and more fayalitic than the cores.Various possible models are discussed for the origin of the orebody and it is proposed that the sulfide was erupted as a dispersed liquid phase within an intratelluric mush of olivine and komatiitic liquid. The sulfides are thought to fill topographic lows or rifts on the sea floor in which the supernatant komatiite liquid is decanted off as the trough fills with later erupted olivine grains. Within the ore-bearing ultramafic unit there is little evidence for in situ olivine crystallization.

Palaeogene komatiites from Gorgona island, East Pacific - A primary magma for ocean floor basalts?

Gorgona island is the only place in the world where young ultramafic (pyroxenitic) komatiitic lava flows are known to occur. The island, which appears to be a small southern remnant of the Colombian Coastal Cordillera, is built up of serpentinised peridotites, gabbros, basalts and associated Palaeogene oceanic sediments. Spinifex texured rocks occur within the basaltic and doleritic complex, closely associated with pillow basalts of tholeiitic composition. During rapid cooling the order of crystallisation was: olivine (Fo90-91), high Al-calcic pyroxene and Cr-spinel as the major quenched phases; then olivine (Fo86-82). low Al-calcic pyroxene, plagioclase (An75-81) and Ti-magnetite as quenched products in the interstitial groundmass. Bulk chemistry: MgO ∼ 16wt.% (Mg/Mg+Fe2+ = .74), Cr ∼ 1250ppm, Ni ∼ 700ppm, Sc ∼ 33ppm, Ba ∼ 5 ppm, Zr ∼ 30 ppm, Y ∼ 15 ppm; flat REE pattern (5 to 8 x chondrites) with slight depletion of LREE. The associated basalts and dolerites yield similar pattern, although slightly enriched in LREE. Model calculations incorporating major and trace elements show that fractionation of 20-25% olivine + 0.5% Cr-spinel from the komatiitic liquid could produce olivine tholeiites similar to those of Palaeogene Nazca plate. The unique structural setting of the Gorgona rocks probably results from a rapid uplift of a central part of the Nazca ridge including immature magma chambers.

The geology and petrogenesis of a greenstone belt and related nickel sulfide mineralization at Yakabindie, Western Australia

Two major rock sequences, the Lower and Upper Greenstones, separated by an unconformity and associated conglomerate, are exposed in a narrow (5–30 km) elongate (130 km) greenstone belt in the Lawlers—Mt Keith area of the Eastern Goldfields of Western Australia. The Upper Greenstones, which contain zones of significant nickel sulfide mineralization, appear to have been emplaced in a graben which cuts across folded strata of the Lower Greenstones. Exposed portions of the Lower Greenstones consist of over 90% mafic and less than 10% felsic rocks, whereas the Upper Greenstones consist of approximately one third mafic and two thirds felsic rocks. The mafic rocks of the latter sequence all appear to belong to the komatiitic rock series. ∗∗ ∗∗ Using the revised definition of Arndt et al., 1977. They are divided into two groups of flows and thin sills ranging in composition from peridotite to basalt and a number of strata-bound podiform dunitic bodies which are regarded as feeder chambers for some of the overlying flows. Mafic rocks of the Lower Greenstones include Fe-rich tholeiitic basalts, gabbros, komatiitic ∗∗ ultramafic rocks and some basalts which appear chemically to be transitional between the two rock series. At Yakabindie in the centre of the Lawlers—Mt Keith belt, Western Australia, the nickel sulfide mineralization occurs mainly within the dunite pods as large (in excess of 70·10 6 tonnes) low grade (approx. 0.6 wt.% Ni) zones composed largely of disseminated pentlandite. Textures in much of the ore suggest that magmatic sulfides were trapped interstitial to olivine grains, although textures and chemical variations in some of the ore indicate that metals and sulfides have been redistributed during metamorphism or alteration of the host rocks. Chemical variations within the komatiites are consistent with their being the result of low pressure fractional crystallization of a liquid containing in excess of 33 wt. % MgO. In order to produce such a magnesian liquid, it is postulated that the source region of the mantle must have been melted twice, once to give rise to a basaltic liquid and again to give rise to the magnesian, komatiitic liquid. This would require diapiric ascent of mantle material from a depth of perhaps 200–250 km. Consideration of the melting curves of sulfides and mantle periodotite with pressure suggests that sulfides might have become concentrated at these depths, thus accounting for their almost universal association, in Western Australia, with very magnesian komatiites.

Genesis of Archean komatiites from Munro Township, Ontario: Trace-element evidence

Trace elements, particularly rare earths, determined in Archean komatiitic and tholeiitic rocks of Munro Township, Ontario, appear mainly to reflect the processes that produced the lavas, rather than the metamorphic processes. Pyroxenitic to basaltic komatiites may have originated by fractional crystallization of peridotitic komatiite liquid or by partial melting of plagioclase peridotite in the mantle. Peridotitic komatiite magmas may be partial melts of a residual mantle assemblage that was depleted in largeion lithophile (LIL) elements, possibly through earlier extraction of liquids. Picritic tholeiites may have originated from melting of garnet peridotite. A multistage melting model may relate both tholeiites and komatiites to a common source, perhaps in a rising mantle diapir. In this model tholeiites would be the first-formed liquids extracted whereas residual mantle might be further melted to produce komatiitic liquids.