Chalcogen Elements Research Articles

X‐ray photoelectron spectroscopy analysis of Ge–Sb–Se pulsed laser deposited thin films

AbstractPulsed laser deposition was used to prepare amorphous thin films from (GeSe2)100−x(Sb2Se3)x system (x = 0, 5, 10, 20, 30, 40, 50, and 60). From a wide variety of chalcogenide glass‐forming systems, Ge–Sb–Se one, especially in thin films form, already proved to offer a great potential for photonic devices such as chemical sensors. This system has a large glass‐forming region which gives the possibility to adjust the chemical composition of the glasses according to required physical characteristics. The chemical composition of fabricated thin films was analyzed via X‐ray photoelectron spectroscopy (XPS) and compared to energy dispersive spectroscopy (EDS) data. The results of both techniques agree well: a small deficiency in chalcogen element and an excess of antimony was found. The structure of as‐deposited thin films has been investigated by XPS. The presence of the two main structural units, [GeSe4] and [SbSe3] proposed by Raman scattering spectroscopy data analysis, was confirmed by XPS. Moreover, XPS core level spectra analysis revealed the presence of M–M bonds (M = Ge, Sb) in (Ge,Sb)–Ge–(Se)3 and (Ge,Sb)–Sb–(Se)2 entities that could correspond to Ge‐based tetrahedra and Sb‐based pyramids where one of its Se atoms at corners is substituted by Ge or Sb ones. The content of depicted M–M bonds tends to increase with introduction of antimony in the amorphous network of as‐deposited thin films from x = 0 to x = 40 and then it decreases. XPS analysis of as‐deposited thin films shows also the presence of the (Ge,Sb)–Se–(Ge,Sb) and Se–Se–(Ge,Sb) entities.

Self-doped iron-based superconductors with intergrowth structures

The key structural unit of iron-based superconductors (FeSCs) is the Fe2X2 (X refers to a pnictogen or a chalcogen element) layer which stacks alternately along the crystallographic c axis with other spacer layers. This structural feature makes it possible to find FeSCs via rational material design. In this paper, we first review the crystal structure of FeSCs along with the relevant progress. Then we summarize several rules for designing the intergrowth structures. The rules include the following points. 1) Lattice match between the intergrowth layers should be good enough. Quantitatively, the lattice mismatch, defined as =2(aA-aB)/(aA + aB), where aA and aB are respectively the lattice parameters of the two constituent compounds, should be no larger than~2%. 2) The charge transfer between the intergrowth layers is mostly essential, which acts as the glue that combines the constituent layers together. Such a charge transfer also induces the extra charge carriers in the superconducting key layer to give rise to superconductivity without extrinsic doping (so-called self doping). 3) For the structure with similar yet crystallographically distinct sites, one needs to avoid forming solid solutions. 4) Each intergrowth layer is preferably thermodynamically stable. 5) The designed structure can be preliminary evaluated with the hard and soft acids and bases conception and ab initio calculations. Following these empirical rules, we introduce and analyze five examples, namely, (Li0.8Fe0.2OH)FeSe, Ba2Ti2Fe4As4O, 42214-type Ln4Fe2As2Te1-xO4 (Ln=Pr, Sm, Gd), 1144-type AkAeFe4As4 (Ak=K, Rb, Cs; Ae=Ca, Sr, Eu), and 12442-type AkCa2Fe4As4F2 and AkLn2Fe4As4O2 (Ak=K, Rb, Cs; Ln=Nd-Ho). For the last 12442-type compounds, we also discuss the unusual relation between superconducting transition temperature and crystallographic parameters. We conclude that the structural-design approach may serve as an effective route, not only for discovering new FeSCs but also for exploring other relevant functional materials with similar crystal structures.

Synthesis and characterization of diruthenaborane analogues of pentaborane(11) and hexaborane(10)

In an attempt to synthesize expanded-cage metallaheteroboranes containing heavier chalcogen atoms, the reaction of diruthenaborane analogue of pentaborane(9), nido-[1,2-(Cp*RuH)2B3H7] (1) with phenyl-chalcogenoborates Li[BH3(EPh)] (E = S, Se or Te) was carried out. Thermolysis of nido-1 with Li[BH3(SPh)] led to the formation of the dimetalla-pentaborane(11) analogue arachno-[(Cp*Ru)2B3H8(SPh)] (2). In parallel to the formation of 2, the reaction also yielded three B-H functionalized compounds, namely [(Cp*Ru)2B4H7(Ph)] (3), [(Cp*Ru)2B4H7(Cl)] (4) and [(Cp*Ru)2B4H6(SPh)(Cl)] (5). On the other hand, reaction of 1 with Li[BH3(SePh)] led to the formation of the diruthenium analogue of hexaborane(10) nido-[(Cp*Ru)2B4H9(SePh)] (6), whereas Li[BH3(TePh)] yielded the capped nido-pentagonal-pyramidal [(Cp*Ru)2B4H6Te] (7). Compound 7 is a rare ruthenaborane cluster containing a heavier chalcogen element (Te). All the compounds were characterized by mass spectrometry and 1H, 1H{11B}, 11B{1H} and 13C{1H}NMR spectroscopy. The solid state X-ray structures of all the compounds were unequivocally established by crystallographic analysis. Additionally, the electronic properties of compound 2 were analyzed.

Chalcogenide glass: Fabrication techniques, properties and applications

ABSTRACTChalcogenides are compound formed predominately from one or more of the chalcogen elements; Sulfur, Selenium, and Tellurium. Although they were studied some fifty years ago, interest in chalcogenide glasses has emerged significantly as glasses, crystals, and alloys find new life in the wide range of optoelectronic devices. Various non-oxide glasses have been prepared and characterized in last few decades, thus widening the groups of chalcogen materials used in various optoelectronic glasses. The present review describes the composition of chalcogenide glasses, their synthesis and fabrication, their physical properties and potential applications in optoelectronics. Subsequently, current research and future prospects of chalcogenide glasses have been sketched. A comprehensive review on chalcogenide glasses shall further help the researchers to expedite their research work in this area.

Current State: The Development of Thin Film Solar Cells Based on Kesteritee Compound

Thin film solar cell experience fast development, especially for thin film solar cell CdTe and Cu(In,Ga)Se2 (CIGS). However, the usage of rare element in the nature such as In, Te, and Ga and toxic such as Cd give limitation in the future development and production growth in big scale. Development of other alternative compound with maintain the profit of electronic and optic character which get from CIGS chalcopyrite compound will be explain. Compound of Cu2ZnSnSe4 (CZTSe) is downward compound from CIGS with substitute the In and Ga element with Zn and Sn. The compound kesterite structure can be modified with variation of chalcogen element to get wanted character in solar cell application. Efficiency record of photovoltaic devices conversion used this compound or downward reach 9.7%.©2016 JNSMR UIN Walisongo. All rights reserved.

Negative capacitance in ZnO1-xChx (Ch = S, Se, Te): Role of localized charge recombination

We demonstrate negative capacitance (NC) dispersion in ZnO by doping lesser electronegative chalcogen (Ch = S, Se, Te) elements at the oxygen (O)-site. Approximately 4.00 ± 0.15 atomic percentage of ChO× is obtained from Rietveld refinement. Using ab initio and dielectric spectroscopy, the tailoring of charge localization around dopants and consequent charge recombination are observed to have a systematic dependence on the stabilization of NC. With the increase of dopant electronegativity difference from SO× to TeO×, the low frequency (<100 Hz) dispersion of NC gradually extends to a larger frequency under lower external bias. Universal Debye response is found to govern the NC dispersion with calculated relaxation time indicating to trap mediated charge recombination.

Origin of highly siderophile and chalcogen element fractionations in the components of unequilibrated H and LL chondrites

Osmium isotopic compositions, abundances of highly siderophile elements (HSE: platinum group elements, Re and Au), the chalcogen elements S, Se and Te and major and minor elements were analysed in physically separated size fractions and components of the ordinary chondrites WSG 95300 (H3.3, meteorite find) and Parnallee (LL3.6, meteorite fall). Fine grained magnetic fractions are 268-65 times enriched in HSE compared to the non-magnetic fractions. A significant deviation of some fractions of WSG 95300 from the 4.568 Ga 187Re-187Os isochron was caused by redistribution of Re due to weathering of metal. HSE abundance patterns show that at least four different types of HSE carriers are present in WSG 95300 and Parnallee. The HSE carriers display (i) CI chondritic HSE ratios, (ii) variable Re/Os ratios, (iii) lower than CI chondritic Pd/Ir and Au/Ir and (iv) higher Pt/Ir and Pt/Ru than in CI chondrites. These differences between components clearly indicate the loss of refractory HSE carrier phases before accretion of the components. Tellurium abundances correlate with Pd and are decoupled from S, suggesting that most Te partitioned into metal during the last high-temperature event. Tellurium is depleted in all fractions compared to CI chondrite normalized Se abundances. The depletion of Te is likely associated with the high temperature history of the metal precursors of H and LL chondrites and occurred independent of the metal loss event that depleted LL chondrites in siderophile elements. Most non-magnetic and slightly magnetic fractions have S/Se close to CI chondrites. In contrast, the decoupling of Te and Se from S in magnetic fractions suggests the influence of volatility and metal-silicate partitioning on the abundances of the chalcogen elements. The influence of terrestrial weathering on chalcogen element systematics of these meteorites appears to be negligible.

Preferential S/Se occupation in an anisotropic ReS2(1-x)Se2x monolayer alloy.

Band structure engineering of two-dimensional (2D) metal dichalcogenides (TMDs) is crucial for their light-matter interaction and optoelectronic applications. Alloying of different metal or chalcogen elements with different stoichiometries in TMDs provides a versatile and efficient approach for modulating the electronic structure and properties of 2D materials. In 2D alloys, quantification of spatial distribution and local coordination of atoms facilitates the establishment of the structure-property relationship at the atomic scale. Here, we have imaged and analyzed the atomic configuration of sulfur and selenium atoms in anisotropic ReS1.4Se0.6 by scanning transmission electron microscopy (STEM). In Z-contrast images, we have realized the identification and quantification of Re, Se and S at different coordination sites. Different from the random distribution of metal and chalcogen elements in MoS2(1-x)Se2x and Mo1-xWxS2, we find that Se atoms preferentially locate inside of Re4 diamonds in ReS2(1-x)Se2x. Further density function theory (DFT) calculations reveal electronic structure modulation for Se occupation at different sites.

Chalcogenide microstructured optical fibres for mid-IR applications

Abstract Compared to oxide-based glasses, vitreous materials composed of chalcogen elements (S, Se, Te) show large transparency windows in the infrared. Indeed, chalcogenide glasses can be transparent from the visible up to 12 – 18 μm , depending on their compositions. In addition, chalcogenide glasses contain large polarisable atoms and external lone electron pairs that induce exceptional non-linear properties. Consequently, the non-linear properties can be 100 or 1000 times as high as the non-linearity of silica. An original way to obtain single-mode fibres is to design microstructured optical fibres (MOFs). These fibres present unique optical properties thanks to the high degree of freedom in the design of their geometrical structure. Various chalcogenide MOFs operating in the IR range have been elaborated in order to associate the high non-linear properties of these glasses with the original MOF properties. Indeed, chalcogenide MOFs might lead to new devices with unique optical properties in the mid-infrared domain, like multimode or endlessly single-mode transmission of light, small or large mode area fibres, highly birefringent fibres and non-linear properties for wavelength conversion or generation of supercontinuum sources.

Ab-initio structural search in solid oxygen at high pressure: from zero to finite temperature

The crystal structure of solid oxygen in the terapascal (TPa) regime has been investigated with Density Functional Theory and the Random Search algorithm at zero temperature. We also considered the effect of the entropy at finite temperatures using the QuasiHarmonic Approximation, and we found that the regime of stability of solid oxygen differs strongly from the results predicted at zero temperature. Finally, we provide some insights of oxygen as a chalcogen element.

Iron-sulfur gel provides new route to ammonia

To produce the large quantities of fertilizer needed to sustain global food production, industry relies on the Haber-Bosch process to split dinitrogen from the air to synthesize ammonia. But to generate the high temperatures and pressures needed for the reaction, ammonia plants consume fossil fuels and, as a result, contribute significantly to greenhouse gas emissions. Researchers want to develop a green alternative to the Haber-Bosch process. In a new effort, Mercouri G. Kanatzidis and George C. Schatz of Northwestern University and coworkers used iron-sulfur (Fe4S4) clusters, found naturally in the nitrogenase enzymes bacteria use to split dinitrogen, to make a light-driven catalyst that converts N2 to NH3 in water at ambient temperature and pressure (Proc. Natl. Acad. Sci. USA 2016, DOI: 10.1073/pnas.1605512113). After synthesizing the Fe4S4 clusters, the researchers link them with [Sn2S6]4– anions and form the complex into a chalcogel, a foamy gel containing the chalcogen element sulfur. The

Highly siderophile and chalcogen element constraints on the origin of components of the Allende and Murchison meteorites

187Re-187Os systematics, abundances of highly siderophile elements (HSE: Re, PGE, and Au), chalcogen elements (Te, Se, and S), and some major and minor elements were determined in physically separated components of the Allende (CV3) and Murchison (CM2) carbonaceous chondrites. Substantial differences exist in the absolute and relative abundances of elements in the components, but the similarity of calculated and literature bulk rock abundances of HSE and chalcogens indicate that chemical complementarity exists among the components, with CI chondrite-like ratios for many elements. Despite subsequent alteration and oxidation, the overall cosmochemical behavior of most moderately to highly siderophile elements during high-temperature processing has been preserved in components of Allende at the sampling scale of the present study. The 187Re-187Os systematics and element variations of Allende are less disturbed compared with Murchison, which reflects different degrees of oxidation and alteration of these meteorites. The HSE systematics (with the exception of Au) is controlled by two types of materials: Pd-depleted condensates and CI chondrite-like material. Enrichment and heterogeneous distribution of Au among the components is likely the result of hydrothermal alteration. Chalcogen elements are depleted compared with HSE in all components, presumably due to their higher volatility. Small systematic variations of S, Se, and Te in components bear the signature of fractional condensation/partial evaporation and metal–sulfide–silicate partitioning.

Understanding the optical properties of ZnO1−xSx and ZnO1−xSex alloys

ZnO1−xYx with chalcogen element Y exhibits intriguing optoelectronic properties as the alloying strongly impacts the band-gap energy Eg(x). In this work, we analyze and compare the electronic structures and the dielectric responses of Zn(O,S) and Zn(O,Se) alloys by means of the density functional theory and the partially self-consistent GW approach. We model the crystalline stability from the total energies, and the results indicate that Zn(O,S) is more stable as alloy than Zn(O,Se). We demonstrate also that ion relaxation strongly affects total energies, and that the band-gap bowing depends primarily on local relaxation of the bonds. Moreover, we show that the composition dependent band-gap needs to be analyzed by the band anti-crossing model for small alloying concentration, while the alloying band-bowing model is accurate for strong alloying. We find that the Se-based alloys have a stronger change in the band-gap energy (for instance, ΔEg(0.50) = Eg(ZnO) – Eg(x = 0.50) ≈ 2.2 eV) compared with that of the S-based alloy (ΔEg(0.50) = 1.2 eV), mainly due to a stronger relaxation of the Zn–anion bonds that affects the electronic structure near the band edges. The optical properties of the alloys are discussed in terms of the complex dielectric function ε(ω) = ε1(ω) + iε2(ω) and the absorption coefficient α(ω). While the large band-gap bowing directly impacts the low-energy absorption spectra, the high-frequency dielectric constant ε∞ is correlated to the intensity of the dielectric response at energies above 4 eV. Therefore, the dielectric constant is only weakly affected by the non-linear band-gap variation. Despite strong structural relaxation, the high absorption coefficients of the alloys demonstrate that the alloys have well-behaved optoelectronic properties.

Quasi-Chalcogen Characteristics of Al12Be: A New Member of the Three-Dimensional Periodic Table

The creative concept of superatom brings a new dimension to the conventional periodic table, which has been gradually enriched by both theoretical and experimental research. In this article, we propose a new member, namely, Al12Be, to the superatom family. The amazing similarity between the Al12Be cluster and the chalcogen elements makes the former an excellent superatom counterpart of the latter. In addition, Al12Be exhibits more exothermic first electron affinity (EA) and less endothermic second EA values due to its size advantage over the chalcogen atoms, showing the superatom superiority in this respect. The stable compounds formed between Al12Be and other atoms, such as carbon, beryllium, calcium, and lithium, provide further evidence to support the quasi-chalcogen identity of Al12Be.

Chalcogen–chalcogen secondary bonding interactions in trichalcogenaferrocenophanes

The solid-state structures of all members in the series of trichalcogenaferrocenophanes [Fe(C5H4E)2E′] (E, E′ = S, Se, Te) (1–9) have been explored to understand the trends in secondary bonding interactions (SBIs) between chalcogen elements sulfur, selenium, and tellurium. To complete the series, the crystal structures of the four hitherto unknown complexes [Fe(C5H4S)2Te] (3), [Fe(C5H4Se)2S] (4), [Fe(C5H4Se)2Te] (6), and [Fe(C5H4Te)2S] (7) have been determined in this contribution. The packings of all complexes 1–9 were considered by DFT calculations at the PBE0/pob-TZVP level of theory using periodic boundary conditions. The intermolecular close contacts were considered by QTAIM and NBO analyses. The isomorphous complexes [Fe(C5H4S)2S] (1), [Fe(C5H4S)2Se] (2), and [Fe(C5H4Se)2Se] (5a) form dimers via weak interactions between the central chalcogen atoms of the two trichalcogena chains of adjacent complexes. In the second isomorphous series consisting of [Fe(C5H4Se)2S] (4) and 5b, the complexes are linked together into continuous chains by short contacts via the terminal selenium atoms. The intermolecular chalcogen–chalcogen interactions are significantly stronger in complexes [Fe(C5H4S)2Te] (3), [Fe(C5H4Se)2Te] (6), and [Fe(C5H4Te)2E′] (E′ = S, Se, Te) (7–9), which contain tellurium. The NBO comparison of donor–acceptor interactions in the lattices of [Fe(C5H4S)2S] (1), [Fe(C5H4Se)2Se] (5a and 5b), and [Fe(C5H4Te)2Te] (9) indeed shows that the n(5pTe)2 → σ*(Te–Te) interactions in 9 are the strongest. All other interaction energies are significantly smaller even in the case of tellurium. The computed natural charges of the chalcogen atoms indicate that electrostatic effects strengthen the attractive interactions in the case of all chalcogen atoms.

Possible atomic structures responsible for the sub-bandgap absorption of chalcogen-hyperdoped silicon

Single-crystal silicon was hyperdoped with sulfur, selenium, and tellurium using ion implantation and nanosecond laser melting. The hyperdoping of such chalcogen elements led to strong and wide sub-bandgap light absorption. Annealing the hyperdoped silicon, even at low temperatures (such as 200–400 °C), led to attenuation of the sub-bandgap absorption. To explain the attenuation process, we modeled it as chemical decomposition reaction from an optically absorbing structure to a non-absorbing structure. Attenuation of the experimental absorption coefficient was fit using the Arrhenius equation. From the fitted data, we extracted the reaction activation energies of S-, Se-, and T-hyperdoped silicon as 0.338 ± 0.029 eV, 0.471 ± 0.040 eV, and 0.357 ± 0.028 eV, respectively. We discuss these activation energies in terms of the bond energies of chalcogen–Si metastable bonds, and suggest that several high-energy interstitial sites, rather than substitutional sites, are candidates for the atomic structures that are responsible for the strong sub-bandgap absorption of chalcogen hyperdoped silicon.

Chalcogenide glass fibers: Optical window tailoring and suitability for bio-chemical sensing

Glassy materials based on chalcogen elements are becoming increasingly prominent in the development of advanced infrared sensors. In particular, infrared fibers constitute a desirable sensing platform due to their high sensitivity and versatile remote collection capabilities for in-situ detection. Tailoring the transparency window of an optical material to the vibrational signature of a target molecule is important for the design of infrared sensor, and particularly for fiber evanescent wave spectroscopy. Here we review the basic principles and recent developments in the fabrication of chalcogenide glass infrared fibers for application as bio-chemical sensors. We emphasize the challenges in designing materials that combine good rheological properties with chemical stability and sufficiently wide optical windows for bio-chemical sensing. The limitation in optical transparency due to higher order overtones of the amorphous network vibrations is established for this family of glasses. It is shown that glasses with wide optical window suffer from higher order overtone absorptions. Compositional engineering with heavy elements such as iodine is shown to widen the optical window but at the cost of lower chemical stability. The optical attenuations of various families of chalcogenide glass fibers are presented and weighed for their applications as chemical sensors. It is then shown that long-wave infrared fibers can be designed to optimize the collection of selective signal from bio-molecules such as cells and tissues. Issues of toxicity and mechanical resistance in the context of bio-sensing are also discussed.

Density Functional Theory (DFT) Calculation of Band Structure of Kesterite

The kesterite, Cu2ZnSnS4 has a big potential as a future solar material in replacing current material. Although the kesterite and copper indium gallium selenide, CIGS has almost same structure but the constituent elements of kesterite are earth-abundance, cheaper and non-toxic. The chalcogen elements existed inside the kesterite compound are selenium and sulphur, Cu2ZnSnSe4 / Cu2ZnSnS4. Therefore, the structural flexibility of kesterite opens up an avenue to develop light-absorber material with suitable properties and applications. The density functional theory (DFT) has been used to calculate the total energy of Kesterite developed from Material Studio - CASTEP. The general gradient approximation (GGA) has been choosing to treat the exchange-correlation. The structure of kesterite has been developed by determining its space group, I4 and Pc and its coordination of each atom. The previous calculated shown that the energy of its band gap is around 1.0-1.5 eV.

Fractionation of highly siderophile and chalcogen elements in components of EH3 chondrites

Abundances of highly siderophile elements (HSE: Re, platinum group elements and Au), chalcogens (Te, Se and S), 187Os/188Os and the major and minor elements Mg, Ca, Mn, Fe, Ni and Co were determined in the components of Sahara 97072 (EH3, find) and Kota Kota (EH3, find) in order to understand the element fractionation processes. In a 187Re–187Os isochron diagram, most magnetic components lie close to the 4.56Ga IIIA iron meteorite isochron, whereas most other components show deviations from the isochron caused by late redistribution of Re, presumably during terrestrial weathering. Metal- and sulfide rich magnetic fractions and metal-sulfide nodules are responsible for the higher 187Os/188Os in bulk rocks of EH chondrites compared to CI chondrites. The HSE and chalcogens are enriched in magnetic fractions relative to slightly magnetic and nonmagnetic fractions and bulk compositions, indicating that Fe–Ni metal is the main host phase of the HSE in enstatite chondrites. HSE abundance patterns indicate mixing of two components, a CI chondrite like end member and an Au-enriched end member. Because of the decoupled variations of Au from those of Pd or the chalcogens, the enrichment of Au in EH metal cannot be due to metal–sulfide–silicate partitioning processes. Metal and sulfide rich nodules may have formed by melting and reaction of pre-existing refractory element rich material with volatile rich gas. A complex condensation and evaporation history is required to account for the depletion of elements having very different volatility than Au in EH chondrites. The depletions of Te relative to HSE, Se and S in bulk EH chondrites are mainly caused by the depletion of Te in metal. S/Se and S/Mn are lower than in CI chondrites in almost all components and predominantly reflect volatility-controlled loss of sulfur. The latter most likely occurred during thermal processing of dust in the solar nebula (e.g., during chondrule formation), followed by the non-systematic loss of S during terrestrial weathering.

Fractionation of highly siderophile and chalcogen elements during magma transport in the mantle: Constraints from pyroxenites of the Balmuccia peridotite massif

Sulfide segregation from sulfur saturated basic magmas affects the compositions of chalcophile elements such as the highly siderophile elements (HSE) and the chalcogens S, Se, Te to variable extent. Whether this process predominantly occurs in the lower crust or in the mantle and how segregation of liquid sulfide and accumulation affects concentrations and ratios of these elements at different mantle depths and in presumed primitive basic magmas remains uncertain. Abundances of the HSE, S, Se and Te and Os isotopes in websterites and spinel clinopyroxenites of the Balmuccia peridotite massif (Ivrea-Verbano Zone, Italian Alps) provide new insight on sulfide segregation and the compositional change of melt and peridotite during magma transport in the mantle.Balmuccia websterites and clinopyroxenites formed from late Paleozoic and Mesozoic melt influx into stretched continental lithospheric mantle of the Ivrea-Verbano Zone, respectively. The HSE and chalcogen element compositions of websterites and clinopyroxenites reflect the segregation and accumulation of sulfide melt from S saturated silicate melts with different abundances and ratios of chalcogens and the HSE. The pyroxenites display large variations in abundances of the platinum group elements (PGE) and Te whereas abundances of less chalcophile elements S, Se and Re are much less variable. The fractionation between the PGE and fractionation of Re/Os, S/Se and Se/Te in the mantle pyroxenites are consistent with sulfide melt–silicate melt partitioning with a sequence of apparent coefficients of DPGE>DAu⩾DTe>DSe⩾DS≈DRe. Concentrations in ocean ridge basalts and in gabbros of the lower oceanic crust are also consistent with such fractionation.Websterites which have formed during refertilization of depleted peridotites display ratios of the HSE and moderately suprachondritic initial 187Os/188Os similar to interstitial sulfides of refertilized peridotites. These compositions are different from ‘residual’ compositions of sulfide inclusions in such peridotites. These observations suggest that preferential dissolution of interstitial sulfides from peridotites into migrating silicate melts, and thus chemical and isotopic disequilibrium, may be common during open-system melt extraction and melt-peridotite reaction in the mantle. Precipitation of sulfides and pyroxenes from such migrating melts may lead to refertilization of peridotites and formation of pyroxenites elsewhere in the mantle. The enrichment of the PGE and Te in early precipitated sulfides in these cumulates explains the depletion of the PGE and Te in many basalts and lower crustal gabbros relative to peridotites. Similar ranges of Re/Os and Pd/Ir and suprachondritic initial 187Os/188Os in pyroxenites and in oceanic crust indicate that most of the compositional variation of these elements in basic magmas may reflect repeated cycles of assimilation of mainly interstitial, peridotite-hosted sulfide liquid and subsequent liquid sulfide segregation during magma transport in the mantle.