As-rich Melts Research Articles

Distribution of noble metals in magmatic sulfide occurrences in the Montagnais Sill Complex, Labrador Trough, Canada

ABSTRACT We have characterized the distribution of noble metals among six styles of magmatic sulfide mineralization in the Montagnais Sill Complex of the Labrador Trough in northern Québec using optical and electron microscopy combined with laser ablation-inductively coupled plasma-mass spectrometry trace element analysis of sulfides. The principal sulfide minerals include pyrrhotite, chalcopyrite, and pentlandite with accessory sphalerite and sulfarsenides. In addition, cubanite, troilite, and mackinawite are present in ultramafic-hosted assemblages. The precious metal mineral assemblages are dominated by tellurides, Ag-rich gold, and sperrylite which generally occur at the margins of sulfides. Few iridium-group platinum group element- and Rh-bearing grains were identified and mass-balance calculations show that these elements are generally hosted in pyrrhotite and pentlandite. Virtually all Pt and Au are hosted in precious metal grains, whereas Pd is distributed between precious metal grains and pentlandite. Where present, sulfarsenides are a key host of iridium-group platinum group element, Rh, Pd, Te, and Au. The presence of troilite, cubanite, and mackinawite and the absence of pentlandite exsolution lamellae in the ultramafic-hosted sulfides indicates an initial sulfide melt with a high metal/S ratio. Sulfarsenides present among globular sulfide assemblages derive from an immiscible As-rich melt that exsolved from the sulfide melt in response to the assimilation of the As-bearing floor rocks. In this study, the composition of sulfides is consistent with those derived from Ni-Cu-dominated deposits and not platinum group element-dominated deposits.

Wetting facilitates late-stage segregation of precious metal–enriched sulfosalt melt in magmatic sulfide systems

Research Article| October 01, 2010 Wetting facilitates late-stage segregation of precious metal–enriched sulfosalt melt in magmatic sulfide systems Andrew G. Tomkins Andrew G. Tomkins School of Geosciences, Monash University, PO Box 28E, Victoria 3800, Australia Search for other works by this author on: GSW Google Scholar Geology (2010) 38 (10): 951–954. https://doi.org/10.1130/G31263.1 Article history received: 29 Mar 2010 rev-recd: 02 Jun 2010 accepted: 04 Jun 2010 first online: 09 Mar 2017 Cite View This Citation Add to Citation Manager Share Icon Share Facebook Twitter LinkedIn MailTo Tools Icon Tools Get Permissions Search Site Citation Andrew G. Tomkins; Wetting facilitates late-stage segregation of precious metal–enriched sulfosalt melt in magmatic sulfide systems. Geology 2010;; 38 (10): 951–954. doi: https://doi.org/10.1130/G31263.1 Download citation file: Ris (Zotero) Refmanager EasyBib Bookends Mendeley Papers EndNote RefWorks BibTex toolbar search Search Dropdown Menu toolbar search search input Search input auto suggest filter your search All ContentBy SocietyGeology Search Advanced Search Abstract The wetting behavior of As-rich sulfosalt melts against monosulfide solid solution (MSS) is investigated using experiments to elucidate late-stage fractionation processes in magmatic sulfide systems, which may control the distribution of platinum group elements. A range of As-rich melt compositions is found to wet MSS, including those that contain significant proportions (1%–45%) of precious metals (Pt, Pd, and Au). However, extremely Au rich or Pt rich sulfosalt melts (>∼40% Au; >∼50% Pt + 5% Au) do not wet MSS. These results imply that if magma contamination and/or fractionation processes were able to produce a late-stage As-rich melt (sulfosalt melts crystallize cooler than MSS) that exceeds ∼0.2% of the rock volume, then an interconnected melt drainage network would be able to form along MSS crystal triple junctions. The dense sulfosalt melt could thereby drain downward, progressively sequestering incompatible Bi, Sb, Te, Pt, Pd, and Au to form sulfosalt melt accumulations that continue to fractionate to form the platinum group minerals. This late-stage fractionation model is consistent with the observed mineral distribution in numerous magmatic sulfide deposits. You do not have access to this content, please speak to your institutional administrator if you feel you should have access.

ARSENOPYRITE MELTING DURING METAMORPHISM OF SULFIDE ORE DEPOSITS

Arsenopyrite is present as a minor phase in many different types of ore deposits. Here we investigate a number of ore deposits metamorphosed to mid-amphibolite facies and above to show that in some environments, arsenopyrite is likely to melt during metamorphism, but in others it will persist until it is converted to lollingite + pyrrhotite. The fate of arsenopyrite is governed by the sulfur fugacity imposed by the surrounding mineralogy during metamorphism. At the Hemlo gold deposit, Canada, which contains a range of disseminated sulfi des, the breakdown of barite promoted conditions of high f(S2) and melting of arsenopyrite during prograde metamorphism. On the other hand, at several massive sulfi de deposits including Osborne Lake, Montauban and Geco in Canada, high f(S2) conditions were instead generated through pyrite breakdown on the pyrite–pyrrhotite buffer, also causing arsenopyrite to melt in favorable parts of the deposits. In contrast, metamorphic processes that inhibit high f(S2) through consumption of sulfur promote the solid-state conversion of arsenopyrite to lollingite and pyrrhotite rather than melting. In most mineral deposits, the strongest infl uence on sulfur fugacity is the pyrite-to-pyrrhotite reaction, which buffers f(S2) to increasingly elevated values as temperature increases. Once pyrite is consumed, however, f(S2) no longer is maintained at elevated values. If rocks hosting arsenopyrite are able to conserve pyrite to middle-amphibolite-facies conditions (beyond 491°C at 1 bar, or ~560°C at 5 kbar), arsenopyrite melting will occur. If not, arsenopyrite melting is unlikely, though still possible. Of the mechanisms that promote pyrite decomposition at metamorphic conditions below arsenopyrite melting, sequestration of sulfur by iron silicates or oxides (or both) to form pyrrhotite may be the most effective in many types of deposit. At the Calumet deposit and in some parts of the Geco deposit, this process was found to be effective in converting pyrite to pyrrhotite in magnetite-rich rocks. Pyrite consumption and low-f(S2) conditions are also promoted to a small extent by incorporation of sulfur in hydrothermal fl uids, such as an introduced fl uid or those generated by dehydration reactions, this effect becoming more signifi cant as temperature rises. Deposits where arsenopyrite is likely to melt during metamorphism include pyrite-rich massive sulfi de deposits as well as disseminated deposits lacking abundant iron silicates and oxides. The As-rich melts that result are highly effective in incorporating and mobilizing other metals, particularly gold and silver, as demonstrated at the Challenger deposit (Australia) and at Hemlo, Ontario.

Influence of Boron in Semi-insulating GaAs Crystals on Their Electrical Activation by Si-Ion Implantation

The influence of boron on the activation of silicon ions implanted in semi-insulating GaAs crystals grown by liquid encapsulated Czochralski (LEC) method from As-rich melt was investigated using crystals with the same carbon concentration and the same deep donor (EL2) concentration. The resistivity of the crystal was not affected by the boron concentration. The sheet resistivity of the implanted layer, however, became higher with increase of the boron concentration. An analysis of the electrical properties of the implanted layer suggested that the occupation rate of the implanted silicon at As site (SiAs), which decides the acceptor level, increased with the boron concentration. The dependence of this occupation rate was influenced by the annealing temperature for activation in the region of low boron concentration under 15×1016 cm-3.

Capacitance-Voltage-Method Carrier Concentration Depth Profiles in Liquid-Encapsulated-Czochralski-Technique Grown GaAs Wafers from Near-Stoichiometric and Extremely-Composition Melts

ABSTRACTIt was observed that carrier concentration depth profiles under 200μm-diameter gold Schottky electrodes of deeply Si-implanted electrically active layers on undoped semi-insulating GaAs wafers depend on melt composition of liquid-glass encapsulation Czochralski technique (LEC) growth, Fourier-transformed infrared (FT-IR) measurement carbon content of LEC growth, and consecutive cooling thermal cycle of LEC growth. In the range of this experiment, for head position of boules and 2.5 × 10exp.12/cm2 Si dosage, peak carrier concentrations of As-rich melt LEC growth (initial: [As]/([As]+[Ga])≃0.526) were comparable to or slightly higher than those of near-stoichiometric melt LEC growth. For tail position and 3 × 10exp.12/cm2 Si dosage, a peak carrier concentration of a father As-rich melt LEC-growth was clearly higher than that of Ga-rich melt LEC-growth. LEC-grown melt composition effect depended on consecutive cooling thermal cycle. For peak carrier concentration dispersion of head position, there was anticorrelation between carbon content of FT-IR measurement and peak carrier concentration dispersion.

Silicon incorporation in LEC growth of single crystal gallium arsenide

Recent reports suggest that melt stoichiometry has a strong effect on silicon dopant incorporation in the LEC growth of GaAs single crystals. Our studies show that melt stoichiometry has no effect on silicon incorporation. Using a low pressure LEC process in a high volume manufacturing environment we routinely grow 2 inch or 3 inch diameter, up to 4 kg, single crystals of Si-doped GaAs with dislocation densities as low as 100 cm -2. Crystals with Si concentrations ranging from 10 16 to 10 19 cm -3, grown from melts having Ga/As ratios ranging from 0.75 to 1.4 and contained in quartz crucibles, all have a compensation ratio of about 0.5. Contamination of the melt by Si from the quartz crucible as well as removal of Si from the melt by the B 2O 3 encapsulant must be considered when interpreting Si segregation data. In the course of the work with As-rich melts, the excess As in the melt gave rise to constitutional supercooling resulting in interface breakdown similar to that observed for excess dopants. This is the first time that this phenomena has been reported for melt growth of GaAs single crystals.

GaAs phase effect on interfacial drain current of a 400-μm-wide gate Si-implanted MESFET

Using substrates from different composition liquid-encapsulated Czochralski-grown crystal boules, the 1.5 μm×400-μm wide-gate metal-semiconductor field-effect-transistors (MESFETs) drain current (Id)-drain voltage (Vd) characteristic at near pinch-off bias was observed as a low-frequency oscillation (LFO) Id-Vd curve. This process indicates the interfacial behavior of the MESFET’s Id between GaAs crystal bulk and the Si-implanted electrically active channel layer. As a result, both the LFO behavior and the Id-Vd curve at the interfacial layer were dependent on the substrate GaAs composition but were not dependent on its impurity boron concentration. Large Id-LFO was observed for the substrates from near- or stoichiometric GaAs melts and compared with the Id-LFO for As-rich melts. With the As-rich melt samples, small Id-LFO was observed for both low carbon concentration (nondetectable) and high carbon concentration (≂2×1016 cm−3) in the substrates. From the result and discussion of the LFO model, Id-LFO is not caused by the deep center but by the interaction among both the electron carrier (in this case), and the deep center and electrical potential profile. This is formed by imposed electric bias force and nonbias potential profile of both the electrically active impurity doping profile of process and the mixed band structure of substrate bulk, and occurs because the Id-LFO is small on an As-rich substrate which can have many native deep-center EL2s.

Quasi-quantitative estimation of solidus curve on As-rich side of Ga-As system

A quantitative investigation of the solidus curve on the As-rich side of the Ga-As system was done by estimating the amount of excess As in crystals grown from As-rich melts. Excess As atoms in a GaAs crystal are usually subjected to two forms: precipitation and dissolution. To obtain the solidus curve, it was assumed that the EL2 concentration gives the composition of the crystal when almost all the dissolved excess As form EL2. The maximum EL2 concentration was produced in crystals by a new thermal treatment that dissolves As precipitates and is assumed to change almost all the dissolved As into EL2. The maximum EL2 concentration showed a strong linear correlation with the melt composition. A solidus curve has been drawn that has a parabolic shape similar to the liquids curve. The congruent point is shifted from stoichiometry toward the As-rich side by an amount equivalent to an EL2 concentration of 10 16 cm -3.

Dislocations and microdefects in LEC-grown silicon-doped gallium arsenide crystals

This contribution refers to systematic research on silicon-doped LEC-grown GaAs crystals, aimed at lowering the dislocation density. By Si-doping in the range (1-5) × 10 18 atoms/cm 3 large dislocation-free regions can be obtained. It is shown that Si-doping entails large boron contamination. Results on the effect of deviations in the melt stoichiometry from which the crystals are grown are reported as regards dislocations and microdefects. From etching and photoetching studies, evidence is given that (i) most of the dislocations are not of a glide type and (ii) only dislocations in crystals grown from As-rich melts show characteristic As-type precipitates, while in crystals grown from stoichiometric and Ga-rich melts the nature of decorating microdefects is not yet clear.

Structural and electrical properties ofn-type bulk gallium arsenide grown from non-stoichiometric melts

The pulling of Si-doped GaAs from melts either Ga- or As-rich resulted in crystals having substantially different properties. The crystals grown from As-rich melts exhibited higher electron mobility with respect to those grown under Ga-rich conditions. These higher mobilities have been ascribed to more pronounced incorporation of silicon in donor sites and to lower density of acceptor-like complexes involving Si. The Ga-rich crystals were characterized by a lower concentration of dislocations. Such dislocation reduction is explained in terms of reduction of point defects (As interstitials) which can generate microloops and subsequently dislocations. An extensive TEM investigation further supports the idea of Asi as the point defect which more largely influences the structural properties of LEC GaAs. It is shown that the concentration and size of microdefects is drastically reduced when pulling from Ga-rich melts.

Influence of melt composition on electron mobility in Si-doped LEC GaAs

A number of GaAs single crystals has been grown by means of the LEC technique starting from silicon doped melts which were Ga-rich, As-rich and stoichiometric, respectively. The electrical properties were assessed in order to establish a correlation between silicon incorporation, electron mobility and melt composition. It was found that the electron mobility is higher in crystals grown from As-rich melts. The chemical incorporation of silicon is independent of the melt composition though the distribution of the dopant atoms in acceptors, donors and, possibly, complexes changes drastically with stoichiometry.

Thermal annealing of semi-insulating GaAs under controlled arsenic pressure

Bulk, undoped, semi-insulating, liquid encapsulated Czochralski GaAs crystals have been annealed under controlled arsenic pressures, and the uniformities of EL2 concentration and resistivity were investigated. Crystals were grown from different melt compositions from Ga rich to As rich. Arsenic pressure was changed from 0 (vacuum) to 3.4×103 Torr. As a result, it was found that both EL2 concentration and resistivity in the crystal grown from As-rich melt have the lowest fluctuations around 530 Torr arsenic pressure.

Growth condition dependence of EL2 concentrations in magnetic field applied liquid-encapsulated Czochralski GaAs crystals

In the magnetic field applied liquid-encapsulated Czochralski (MLEC) GaAs crystals, midgap level EL2 concentrations have been found to be strongly affected by the melt composition and crystal versus crucible rotational conditions. Although the EL2 concentrations only varied from 1 to 4×1016 cm−3 in the conventional LEC GaAs crystals, they ranged from far below 1×1015 cm−3 in Ga-rich melt to 1×1017 cm−3 in As-rich melt for the MLEC GaAs crystals.

Low dislocation, semi-insulating In-doped GaAs crystals

Elemental indium doping of GaAs melts to a concentration in the 10 20 cm -3 range was found to be highly effective in reducing dislocation densities in large diameter GaAs crystals grown by the high-pressure liquid encapsulated Czochralski technique. Nominally 50 mm diameter 〈 100 〉-grown crystals exhibit dislocation densities of less than 500 cm -2 over 80% of the central crystal diameter compared to densities greater than 10 4 cm -2 in undoped GaAs crystals. In other respects, In-doped GaAs grown from stoichiometric or slightly As-rich melts are indistinguishable from undoped GaAs, showing stable resistivities in the 10 7 to 10 8 ohm cm range, measured mobilities approaching 5000 cm 2/V·s and only slightly modified 29Si implantation characteristics.

Dislocation studies in 3-inch diameter liquid encapsulated Czochralski GaAs

We have characterized the density and distribution of dislocations in 3-inch diameter, undoped GaAs crystals grown by the liquid encapsulated Czochralski technique. The radial distribution across wafers follows a “W”-shaped profile indicating excessive thermal gradient-induced stress as the primary cause of dislocations. The density along the body of each crystal increases continuously from front to tail. In contrast, the longitudinal distribution in the cone region is inverted, first increasing, and then decreasing as the crystal expands from the neck to full diameter. Growth parameters favoring reduced dislocation densities include good diameter control, the use of thick B 2O 3 encapsulating layers, slightly As-rich melts, and low ambient pressures. The dislocation density in the body of the crystal is practically independent of cone angle Θ for 20° < Θ < 70°. However, high densities result for flat-top (0° < Θ < 20°) crystals. Dash-type seed necking works to reduce the dislocation density only when high-density seeds (>5000 cm -2) are used. Dislocation densities below 1X10 4 cm -2 can be achieved routinely in the large annulus region at the front of a crystal through proper control of these parameters. The lowest measured dislocation density was 6000 cm -2. Further, we show that convective heat transfer from the crystal to the high pressure ambient plays a dominant role in controlling the dislocation density.

Evidence of the role of boron in undoped GaAs grown by liquid encapsulated Czochralski

We provide experimental evidence of electrical activity correlated with residual boron impurities and native point defects in undoped liquid-encapsulated Czochralski GaAs crystals. In Ga-rich samples containing ≥1017 cm−3 boron, an 0.073-eV acceptor level is observed in which the concentration increases with Ga and B content. An approximately quadratic increase in the concentration of the 0.073-eV defect acceptor is observed with increasing boron concentration, suggesting that a complex involving boron with an intrinsic defect (VAs, Gai, or GaAs ) is responsible for the observed acceptor behavior. No evidence of electrically active boron or boron complexes was found in semi-insulating GaAs pulled from stoichiometric or As-rich melts.

Effects of stoichiometry on thermal stability of undoped, semi-insulating GaAs

We provide independent evidence of the effects of melt composition on the semi-insulating behavior and thermal stability of undoped GaAs crystals pulled from pyrolytic boron nitride crucibles by the liquid-encapsulated Czochralski technique. Semi-insulating crystals pulled from stoichiometric or As-rich melts show excellent stability to heat treatments, in contrast to semi- insulating Ga-rich material which exhibits p-type conversion after prolonged annealing at 860 °C. Deep level transient spectroscopy and Hall-effect measurements suggest that p-conversion is caused by outdiffusion of native As and/or Ga defects associated with the EL2 deep donors, resulting in uncompensated residual acceptors, predominantly carbon impurities at the surface. The measured mobility of undoped, semi-insulating GaAs is found empirically to be a sensitive parameter for qualifying substrates for direct ion implantation.

Stoichiometry of melt-grown N-type GaAs as revealed by photoluminescence measurements

A comparative study of the 77K PL spectra of n-type GaAs single crystals, grown by the horizontal or Czochralski technique, shows that the former are crystallized from Ga-rich melts (although in equilibrium with 1 atm As-pressure) whereas the latter are pulled from As-rich melts (although the starting loads are mostly As-deficient). Besides the VGa, copper appears as the major compensating acceptor in these crystals. Association of Cu and/or VGa with VAs and/or donor atoms tends to neutralize this compensation. This work shows how the above interaction explains the empirically determined relation ND∝(ND-NA), which is shown to hold for crystals of various sources.