Tungsten Transport Research Articles

The corrosion of tungsten in metal halide lamps: A review

Tungsten damage in metal halide lamps is caused by chemical attack (corrosion) and physical effects (erosion). This work reviews current understanding and trends in tungsten corrosion research. Despite considerable efforts to elucidate reaction mechanisms in metal halide environments, no quantitative model of tungsten corrosion and transport in discharge lamps has yet been produced. Further progress demands the development of kinetic models and determination of related data to be coupled with the hydrodynamics, thermodynamics and diffusion of these complex systems.

Experimental determination of the solubility of WO 3(s) and the thermodynamic properties of H 2WO 4(aq) in the range 300–600°C at 1 kbar: Calculation of scheelite solubility

Solubility measurements were carried out to assess the importance of chloride to the hydrothermal transport of tungsten. The solubility of WO 3(s) was measured in 10 −3 to 0.49 molal HCl solutions at 1 kbar and temperatures from 300 to 600°C. These experiments were carried out using conventional cold-seal pressure vessels, with WO 3(s) powder contained within a tightly crimped inner Pd capsule surrounded by the fluid and then an outer Au capsule. This arrangement permitted contact between WO 3 and solution at high temperature and pressure, but effectively eliminated back reaction on quench and entrainment of fine particulate WO 3 upon fluid sampling. There is no significant dependence of measured W concentration on initial HCl concentrations between 5 x 10 −3 and 0.49 molal at all temperatures investigated. However, W concentrations were often slightly higher in 10 −3 molal HCl. These results indicate that significant chloride complex formation does not take place with W under supercritical conditions and the species responsible for W transport under the conditions of these experiments is H 2WO 4(aq). The increase in solubility of WO 3 at the lower HCl concentrations can be explained by ionization of H 2WO 4(aq) to HWO 4 − at higher pH values. The logarithms of the equilibrium constants for the reaction: WO 3( s) + H 2 O(1) = H 2 WO 4(aq) are calculated to be −4.0 ± 0.2 (300° C); −3.4 ± 0.4 (400° C); −3.1 ± 0.1 (500° C); −2.7 ± 0.3 (600° C). From these data the values ΔH r ≈ 41 kJ/mol and ΔC p, r ≈ 0 may be derived. Thermodynamic data for H 2WO 4(aq) were also calculated. Using the free energy data for aqueous tungstic acid determined in the solubility study, combined with that in the literature for scheelite, it is possible to show that the solubility of the mineral scheelite as a simple oxyacid species can attain values between 10 and 100 ppm under geologically reasonable conditions, and hence complexation by additional ligands is not required for W transport by hydrothermal solutions. However, W mineral solubility may be further increased by ion pairing of alkali metal cations with tungstate or chloride complexation of Fe, Mn (wolframite), or Ca (scheelite).

A fluid-inclusion study and genetic model of wolframite-bearing quartz veins, Garganta de los Montes, Spanish Central System

AbstractWolframite-bearing quartz veins from Garganta de los Montes, Madrid province, are hosted by banded gneisses that have undergone intense migmatization processes. The ore deposit is closely related to the La Cabrera granitic batholith. The veins strike 075° and dip 75°S. The mineral association includes wolframite, quartz and minor amounts of scheelite and sulphides (sphalerite, chalcopyrite, pyrrhotite, stannite and marcasite). The fluid phases associated with quartz from the vein margin (early barren quartz) and from the vein centre (late wolframite-bearing quartz) have been studied using microthermometry, scanning electron microscopy and crushing test analyses. Four hydrothermal stages have been distinguished.The earliest fluids, only recognized in the barren quartz, contain brine, daughter phase (halite) and trapped minerals. The second hydrothermal stage is characterized by complex carbonic-aqueous inclusions of low salinity (3 to 7 wt.% eq. NaC1) and low density (0.4 to 0.7 g.cm−3). They mainly homogenize into liquid between 300 and 420°C. The third stage is represented by low to moderate salinity inclusions (<9 wt. % eq. NaCl) of moderate density (0.8 to 0.96 g.cm−3), homogenizing between 120° and 330°C. The latest fluids correspond to aqueous solutions of higher salinities (H2O-NaCl, with Ca2+ and Mg2+) and densities (>1 g.cm−3), with TH ranging between 50 and 130°C. The role of the complex-carbonic aqueous fluids in the transport and precipitation of tungsten is highlighted.

Selectivity loss during tungsten chemical vapor deposition: The role of tungsten pentafluoride

Our previous studies found evidence for a selectivity loss mechanism that is initiated by desorption of tungsten subfluorides formed by the reaction of WF6 with metallic tungsten surfaces. When a hot (≂600 °C) tungsten surface is exposed to WF6, a nonvolatile subfluoride, apparently WF4 , will accumulate on a nearby SiO2 surface held at ≂30 °C. Disproportionation of this tungsten subfluoride upon heating above 350 °C produces a reactive state of tungsten that induces rapid selectivity loss when exposed to chemical vapor deposition conditions. The net effect of this tungsten subfluoride desorption–disproportionation mechanism is the transport of tungsten from a tungsten surface to the silicon dioxide surface. Isolation of a higher tungsten subfluoride, apparently WF5, has been accomplished by cooling the SiO2 surface to ≂−45 °C under otherwise similar experimental conditions. Preliminary Auger spectroscopic results yield a stochiometry of F/W=4.9±0.5 for this subfluoride, and temperature programmed desorption results are consistent with the behavior of tungsten pentafluoride. Upon heating WF5 multilayers, most of the compound desorbs, but some reaction and disproportionation occurs leaving a reduced state of tungsten on the SiO2 surface. It seems likely that tungsten pentafluoride is the key volatile species in a tungsten transport mechanism that can initiate selectivity loss.

A Mechanism for Selectivity Loss during Tungsten CVD

We have investigated possible mechanisms for the loss of selectivity (i.e., deposition on silicon dioxide) during tungsten CVD by reduction of tungsten hexafluoride and found strong evidence that selectivity loss is initiated by desorption of tungsten subfluorides formed by the reaction of with metallic tungsten surfaces. Adsorption and disproportionation of the tungsten subfluorides on the silicon dioxide surface produces a reactive state of tungsten that can lead directly to selectivity loss. The key feature of the experimental setup is the ability to independently heat a tungsten foil and a nearby oxide‐covered silicon sample in the presence of tungsten hexafluoride. With the tungsten foil at 600°C and the sample at ∼30°C under a ambient, a tungsten subfluoride was found to deposit on the surface. Auger electron spectroscopy was used to measure a F/W ratio of . Heating this tungsten subfluoride overlayer resulted in disproportionation to yield gas‐phase and metallic tungsten which remained on the surface. With the tungsten foil at 600°C and the sample at 300°C in the presence of , metallic tungsten deposited directly on the without stopping at the subfluoride adsorption step. The net effect of this tungsten subfluoride desorption‐disproportionation mechanism is the transport of tungsten from tungsten surfaces to silicon dioxide surfaces as well as other regions in the deposition chamber. Extrapolated rates for this process are high enough to explain the magnitude of the selectivity loss seen at normal CVD temperatures.

Fluid inclusion and stable isotope studies at the Chicote tungsten deposit, Bolivia

The Chicote tungsten deposit is situated along a major north-northwest to south-southeast structural trend in lower Paleozoic sediments of the eastern cordillera of Bolivia. Kami and Amutara are ore deposits similar to Chicote; they occur along the same structural trend over a distance of 35 km. No igneous rocks are exposed in the vicinity of these mines. The tungsten-bearing veins at Chicote are restricted to a zone of pervasive hydrothermal alteration extending over an area of 3,250 by 1,000 m with a vertical range of over 1,600 m. Whole-rock K-Ar dating of the sericite-rich wall rock indicates a maximum age of 26.1 m.y. for the mineralization. The mineral paragenesis follows an oxide-sulfide-carbonate sequence. Peripheral to the tungsten mineralization is a zone of tin-bearing veins.Fluid inclusion studies demonstrate that the ore fluids are complex brines with salinities of 5 to 48 equiv wt percent NaCl and trapping temperatures of 195 degrees to 530 degrees C. The salinities and temperatures of the fluids display a progressive decrease with time. Stable isotope analyses for delta 18 O and delta D on vein quartz and siderite suggest that the mineralizing fluids were magmatic in origin.Despite strong similarities in mineralogy between Chicote and other tungsten deposits, such as Panasqueira in Portugal and Pasto Bueno and San Cristobal in Peru, there are important differences. The ore fluids were far more saline than the typical worldwide trend and were dominantly magmatic. The distal setting of the Chicote mineralization from the parent felsic intrusion is explained by migration of ore fluids along the major Amutara-Chicote-Kami structural feature. The hypersaline Chicote ore fluids provided a suitable medium for the transport of tungsten and precipitation was probably caused by a decrease in temperature and pressure.

訂正

訂正

Fluid inclusion study of Xihuashan tungsten deposit in the southern Jiangxi province, China

The Xihuashan tungsten deposit is closely related to a small highly evolved granitic intrusion. The fluid phases associated with the wolframite-bearing quartz veins have been investigated using microthermometry and the Raman microprobe; they are highly variable in density and composition. The earlier fluids are low-density and low-salinity CO2-bearing aqueous solutions circulating at temperatures up to 420 °C, and low-salinity (2–3 equiv. wt% NaCl) aqueous solutions without traces of CO2 circulating at high temperatures 280°–400 °C) involved in a specific hydrothermal fracturing event; limited unmixing occurs at 380 °C and 200–100 bar in response to a sudden pressure drop. The second types of fluids related to deposition of idiomorphic drusy quartz are typical CO2-bearing aqueous solutions with low salinity (2.5 equiv. wt% NaCl) homogenizing at low to moderate temperatures (180°–340 °C). The late fluids characterize the sulfide deposition stage; they are aqueous fluids with variable salinities homogenizing in the liquid phase between 100° and 275 °C. The Xihuashan hydrothermal evolution resulted from a discontinuous sequence of specific events occurring between 420° and 150 °C and during a continuous hydrothermal evolution of the system during cooling. The role played by the CO2-rich fluids in the transport and deposition of tungsten in the hydrothermal environment is discussed.

Importance of tungstate species for the transport of tungsten in hydrothermal solution: Case studies on major Japanese tungsten deposits.

Thermodynamic calculations on the stability relations among various tungsten species are performed in order to elucidate the possible form of tungsten during the transportation in hydrothermal solution, using the knowledge of the fluoride and chloride concentrations. Based on the fluorine contents in micas and topaz and the fluid inclusion data, it is calculated that HF/H2O and HC1/H2O fugacity ratios in the natural hydrothermal solutions responsible for major Japanese tungsten deposits at the Kaneuchi, Ohtani, Takatori, Fujigatani, Kiwada, and Kuga range from 10-3.15 to 10-5.91 and are less than 10-3.33, respectively. Pressure and temperature conditions for the formation of the studied deposits are summarized and calculated to be from 1 to 2 kbars, and from 300° to 600°C, respectively. Oxygen fugacities are assumed to be those defined by the C-CO-CO2 buffer as the most reducing condition encountered. Based on the pressure, temperature, f(O2), f(HF)/f(H2O), and f(HCl)/f(H2O) conditions evaluated for the studied deposits, thermodynamic calculations are performed to compare the stabilities of halogen-bearing species with those of tungstates. Calculated results show that H2WO4 may be a predominant species in the natural hydrothermal solutions responsible for the formation of major Japanese tungsten deposits.

ChemInform Abstract: Thermodynamic Calculations on the Transport of Tungsten in Systems Containing Fluorine Compounds

ChemInform Abstract: Thermodynamic Calculations on the Transport of Tungsten in Systems Containing Fluorine Compounds

ChemInform Abstract: Experimental Investigations on the Tungsten Transport in Systems Containing Fluorine Compounds.

AbstractFrom investigations of the chemical processes in tungsten incandescent lamps containing gaseous fluorides, e.g. C2Br2F4, as the transportation agent by IR and UV spectroscopic and MS analysis of the gas phase as well as by analysis of the deposits at the wall and changes at the filament it is concluded, that fluorine containing gases are converted irreversibly into SiF4 which also maintains the chemical W transport.

Physical and chemical controls (fO2,T,pH) of the opposite behaviour of U and Sn-W as examplified by hydrothermal deposits in France and Great-Britain, and solubility data

The aim of this paper is to determine the physical and chemical parameters which control the opposite behaviour of uranium and tin-tungsten between 300° and 500 °C. In uranium deposits, fO2 and fS2 of mineralizing fluids are higher than values fixed by the pyrite-hematite-magnetite triple point, as shown by uraninite-hematite and/or pyrite mineral association. The stability of quartz-K feldspar-muscovite paragenesis in the wall-rocks of hydrothermal U deposits indicates weakly acid pH. By contrast, in the Sn-W occurrences from the French Southern Massif Central, the fO2 of mineralizing fluids is between Ni-NiO and Q-F-M buffers as shown by CO2-CH4-H2O-NaCl bearing fluid inclusions. The pH of these fluids is weakly acid to weakly basic as shown by the stability of muscovite in presence or absence of quartz and/or feldspar. Sn-W mineralizing fluids from Cornwall are by contrast purely aqueous and acid, as indicated by the mineral assemblage muscovite-quartz which is typical of greisens. Experimental data on UO2, SnO2, FeWO4, CaWO4 solubility and metal species in fluids show that fO2 > H-M are required for uranium transport whereas fO2 ≤ Ni-NiO favours Sn transport. The fluid oxidation state has no direct influence on the transport and deposition of tungsten. The fO2 control on the hydrothermal transport properties of these three metals is related on the one hand to the fluid and rock composition, and on the other hand to the minimal 320 °C temperature required for homogeneous equilibria in the C-O-H system to control the oxidation state at low values. At high temperatures, Sn, Fe and Ca chloride complexes are more stable than carbonate and phosphate uranium complexes ; this is attributable to their structure and to the dielectric content of the fluid. The presence of dissolved gases at high concentration, which are produced by devolatilization reactions at high temperatures, is emphasized since they lower the dielectric constant of the fluid which enhances the stability of chloride complexes. All these results show that temperature and fO2 account for the opposite behaviour of uranium and tin-tungsten in hydrothermal systems between 300° to 500 °C.

Wolframite deposition in a hydrothermal vein system; the Grey River tungsten prospect, Newfoundland, Canada

Mineralization in the Grey River tungsten prospect, Newfoundland, is contained within a sheared Devonian K-feldspar megacrystic granite and metamorphic rocks which are typical of the gneissic terrain of the Gander zone. The mineralization is genetically related to a suite of post-tectonic, highly differentiated, alkali-rich leucogranite dikes.The tungsten-bearing hydrothermal veins range from tensional veinlets to lodes containing several injections of quartz. Major composite lodes have greisen alteration envelopes which are characterized by enrichment in volatiles, K 2 O, Al 2 O 3 , Rb, Li, F, heavy rare earth elements, and W. The heavy rare earth elements enrichment appears to be related to deposition of carbonate and fluorite in the greisen. The mineralization is divided into four stages: stage I, characterized by quartz + feldspar + molybdenite veins; stage II, composite lode formation, consisting of five vein types (in paragenetic order), quartz + bismuthinite (IIa), quartz + Fe-rich wolframite (IIb), greisen (IIc), quartz + sulfide (IId), and quartz + Fe-poor wolframite veins (IIe); stage III, characterized by silver-bearing quartz + galena + sphalerite veins; and stage IV, composed of fluorite + calcite + barite veins. A spatial mineral zonation from south to north matches the temporal sequence outlined above, with the exception of the stage IV veins which crosscut stage II veins.Fluid inclusion data indicate a complex evolutionary history for the hydrothermal fluid. The simultaneous trapping of CO 2 -rich and H 2 O-rich fluid inclusions, as well as solid inclusions of calcite, in quartz of the quartz+feldspar+ molybdenite and quartz + bismuthinite vein types, is evidence for the existence of a heterogeneous fluid during stage I and early stage II mineralization. Fluid phase equilibria indicate that the inclusions were trapped at temperatures of 330 degrees to 360 degrees C and fluid pressures of less than 500 bars.In composite lodes, quartz + sulfide and quartz + Fe-poor wolframite veins were deposited in open spaces created by normal faulting. These movements prompted rapid decrease in fluid pressure and temperature and caused retrograde boiling of the hydrothermal fluid (at 200 bars and 300 degrees to 330 degrees C) during deposition of quartz + sulfide veins. Deposition of wolframite in quartz + Fe-poor wolframite veins occurred at 270 degrees to 300 degrees C from an aqueous fluid of low salinity (<0.5 wt % NaCl) and low CO 2 content (<1.0 wt % CO 2 ) and after separation of a CO 2 vapor phase by retrograde boiling. Isotopic and fluid inclusion data suggest that deposition of Fe-rich wolframite may have occurred at a slightly higher temperature (330 degrees C) and from a fluid of higher CO 2 content than the bulk of the wolframite which has an Fe-poor composition.Progressive CO 2 loss from the hydrothermal fluid by immiscibility and retrograde boiling had a marked effect on the solution pH which shifted from 5 weakly acid) to 6 (weakly alkaline at 300 degrees to 350 degrees C) during tungsten deposition. Thermodynamic data on the solubility of wolframite suggest that a pH increase would provide an effective mechanism for tungsten deposition and explain the variation in wolframite composition, from Fe rich to Fe poor, with time.The association of hematite, Fe-rich wolframite, and muscovite in the greisen suggests high f (sub O 2 ) (10 (super -20) -10 (super -30) ) conditions and is supported by Eu behavior during alteration. Sulfides formed after tungsten deposition at temperatures less than 300 degrees C and under f (sub S 2 ) conditions of less than 10 (super -11) .CO 2 -rich fluids are commonly associated with tungsten deposits from a variety of environments and the evidence from the Grey River tungsten prospect shows that CO 2 plays an important role in the transport and deposition of tungsten in the hydrothermal environment.

Chemical equilibrium calculations of tungsten–halogen systems

The performance of the compound bromotrifluoromethane (CBrF3) as a tungsten–halogen bulb fill gas was evaluated theoretically by chemical equilibrium calculations. It was found to be an ideal fill gas in terms of maintaining clear bulb walls and redepositing evaporated tungsten on hot spots in the filament. The calculations were done at different CBrF3 and water concentrations. The compounds most responsible for tungsten transport were identified; fluorine compounds of tungsten dominate over bromine compounds. Finally, CBrF3 is compared to other possible fill gases such as CBr2F2, CBr3F, CBr4, and also HBr, since HBr is presently used commercially.

Effect of a magnetic filter on the formation and deposition of high-Z impurities in a multicusp ion source

The presence of a magnetic filter in a multicusp ion source reduces the transport of tungsten to the extractor region by forming a uniform plasma potential region in the source chamber. Because of the reduced numbers of primary electrons in the extraction chamber, the plasma grid floats near the plasma potential and this reduces the amount of impurity ions formed from sputtered grid material.

Fluid inclusion evidence for the transport of tungsten by carbonate complexes in hydrothermal solutions

Fluid inclusion evidence from the Grey River Tungsten Prospect, Newfoundland, and other tungsten deposits indicates that CO2 is an important component of the hydrothermal fluid. Carbon dioxide is enriched in fluids evolved from granitic melts under high fluid pressure, while lower pressure fluids are chloride-rich. The association of tungsten deposits with these carbon dioxide rich hydrothermal fluids suggests that carbonate/bicarbonate complexes may be important in tungsten transport at very high fluid pressures.

Life and luminous flux of halogen incandescent lamps related to filament temperature, pressure and CH2Br2 content

An experimental study has been made of 20 V, 100 W halogen incandescent lamps using different filling pressures for the halogen additive (0. 25-16 torr CH2Br2) and the inert gas Kr (4-16 atm.), and burning at two voltages. The maximum luminous flux and life were found for lamps with halogen filling pressures of 1-4 torr CH2Br2. Lamps with low halogen filling pressures blacken. At high pressures, attack on the cold coil ends determines life. The blackening or non-blackening in general accords with the transport criterion derived from the solubility of tungsten in the gaseous phase. Halogen lamps with properly operating transport cycles fail due to a local temperature increase leading to burning-through of the filament. The experimentally found relation between life, filament temperature and filling pressure agrees approximately with the theory of the radial transport of tungsten in these halogen lamps. A correlation was found between lamp life and the temperature increase of the hot spot.

The reaction of tungsten with CO 2 and D 2O at high temperatures and low pressures

The reactions of tungsten with CO 2 and D 2O at high temperatures and low pressures ( T >2000 K, 10 −5 < P < 10 −2 Torr) has been studied using mass spectrometry. Under these conditions both reactants produce the same volatile oxide products which result in significant transport of tungsten compared to sublimation. The quasi-equilibrium model of gas surface reactions provides a good qualitative description of the rates of product formation. The fit is quantitatively satisfactory if the JANAF thermochemical values for the products are adjusted within their reported error. A comparison is made between these reactions and the reaction of oxygen with tungsten under similar conditions.

Numerical calculation of radial transport of mass and heat as applied to incandescent lamps

Numerical calculations of the radial transport of tungsten and heat are performed with a view to the practical application to incandescent lamps of the refined theory of the Langmuir model as given by Almer and de Ridder. From the results of these numerical calculations approximate formulae are derived for the transport of tungsten and heat per unit length of the filament, the formulae being power functions of the filament temperature and the filling-gas pressure. The values of the exponents in the power functions agree with those published by other author s.

A thermodynamical study of halogen lamps with carbon additive

A thermodynamical calculation scheme is worked out to study radial and axial transport of tungsten in a halogen incandescent lamp. The approximations of a common diffusion coefficient for all gaseous components and local chemical equilibrium are used. Non-blackening is connected to an inequality forbidding deposition of tungsten onto the bulb wall. Chemical attack or growth of a given part of the coil is characterized by the second derivative of the global pressure of tungsten along the coil. The gettering effect of carbon on oxygen is shown to be limited by partial deposition of carbon as well as the formation of CO2 that decomposes at the coil releasing active oxygen.