^#Konnikov E.G. Influence of water-carbonate fluid on the melting temperature of pyrrhotite.

key words [influence pyrrhotite melt carbon-dioxide]

^# This work was supported by the RFBR (Project N. 97-05-64714)

28

Institute of Experimental Mineralogy, Russian Academy of Sciences, Chernogolovka, Moscow Region

A series (16) experiments on studying the influence of CO₂ on the melting temperature of iron monosulfide was performed. To compare these experiments with those previously performed in the Po--H₂O system, they are carried out at similar thermodynamic parameters, namely P = 1000 atm, T = 1000-1150^oC, and fO₂ was monitored by the NNO buffer. The Fe_0.94S powder of artificially synthesized pyrrhotite was used as the blend. The content of fluid varied from 5 to 30 wt.% of the weight of pyrrhotite, and it was obtained by the thermal decomposition of oxalic acid H₂C₂O₄^.2H₂O resulting in the formation of water-carbonate mixture with the ratio of components 1 : 1.

The experiments showed that when iron monosulfide is melted with the CO₂ + H₂O mixture, the temperature of its solidus decreases by 120-130^oC as compared to the "dry" solidus of pyrrholite with the same composition at the concentration of a water-carbonate mixture of 15-20 wt.% (see figure).

key words [aurostibite sperrylite hydrothermal synthesis ]

Under natural conditions, sperrylite and aurostibite are encountered as individual minerals on contacts with melts of noble metals in ultrabasic rocks, being the result of the high-temperature evolution of the magmatic process. Therefore, the synthesis of these compounds from elements requires high temperatures and, in addition, the melting process should be performed in closed furnaces with an inert atmosphere preventing evaporation and oxidation of the components, especially in the case of arsenic. In the phase diagrams of the Au-Sb and Pt-As systems, the melting temperature for aurostibite is 460^oC, and that for sperrylite is 1510^oC.

On the other hand, the authors of [1] emphasize the role of hydrothermal solutions, especially enriched in volatiles, in the formation of several intermetallides of noble metals at specific stages of the ore process. Third, in the practice of electrochemical studies during reduction of arsenic compounds on the platinum or gold electrodes at the corresponding reduction potentials, surface compounds of arsenic are formed, which is detected by the potential shift of their subsequent anode oxidation in curves I- in the method of inversion voltammetry [2].

Therefore, to develop model conditions of hydrothermal synthesis, we created the reduction medium in situ by the introduction of metallic powdered aluminum in aqueous solutions of dilute hydrochloric acid (0.01-0.1 N), which were placed in autoclaves of the VT-8 titanium alloy. The following reaction with water occurs at 300^oC and 500 atm:

which makes it possible to calculate the stoichiometric hydrogen release, its fugacity, and fugacity with respect to oxygen in order to compare which natural buffers are modeled in such a way.

In experiments, platinum or gold plates were hanged up in the upper part of the autoclave, and crystals of arsenic or antimony in the form of metal were placed in a special container. Prior to experiments, dissolved oxygen was removed from solutions of hydrochloric acid (0.01 and 0.1 N) by bubbling of an inert gas.

This statement of experiments allowed us to observe surface compounds of platinum with arsenic and those of gold with antimony on the platinum and gold (in particular series of experiments) surface. These compounds are stable in air and rather dense, which allowed their analysis by X-ray phase and X-ray spectral methods with a defocused probe. They were identified as sperrylite (PtAs₂) and aurostibite (AuSb₂) from reflections of the X-ray phase analysis.

29

key words[scheelite hubnerite fluid experiment]

The scheelite-hubnerite equilibrium in chloride solutions has been experimentally studied at temperatures 300,400, 500 and 600^oC, pressure 1 kbar, and oxygen fugacity controlled by solid Ni-NiO and Fe-FeO buffers.

The experimental technique is analogous to that described earlier for the scheelite-ferberite equilibrium [1]. The initial concentration of CaCl₂ and MnCl₂ was varied broadly from 10^-4 to 1 mole/kg H₂O in order to provide approaching the equilibrium from 'above' and 'below' in the range of different concentrations. The duration of the runs was from 16 to 8 days. After the runs were completed, the solutions were quenched to room temperature for 5-7 min under a pressure. The solution being examined was separated from the mineral charge with a filter after the run. Calcium in the solution was determined on an atomic-absorption spectrophotometer AAS (Carl Zeiss Iena), and manganese on a spectrometer Spekol-11 using potassium metaperoiodate (KJO₄) [2]. The examined reaction can be written as

Scheelite hubnerite

In experiment solid CaWO₄ and MnWO₄, synthesised by us under hydrothermal conditions, in amounts of 15 mg each, and aqueous chloride solutions (CaCl₂, and MnCl₂) of a certain concentration in amounts of 250-450 mg, depending on the temperature, were placed together in a hermetically sealed gold capsule at a preset oxygen fugacity controlled with an inner buffer. The amount of the buffer was 500 mg, and its volume exceeded that of the mineral charge by many times. The equilibrium position was registered only in the case where both scheelite and hubnerite were present in the run products. In individual runs only one mineral was a charge constituent. As a rule, this is characteristic of solutions having a high initial concentration of one of the components (0.1-1 m) at a low concentration of the other which is attributed to a deficit in the buffer capacity of the solid charge required to maintain the equilibrium mCaCl₂:mMnCl₂ in the solution.

The results of the runs throughout the temperature and solution concentration ranges studied are illustrated in the diagram for two oxygen buffers Ni-NiO (shaded figures) and Fe-FeO (unshaded figures). The uncertainty region due to spread in experimental points is maximally 0.2-0.5 as great as the logarithmic value of the concentrations. As seen from the diagram there is a positive dependence of the scheelite-for hubnerite substitution reaction on the temperature. The effect of f_O2 variation (within Ni-NiO and Fe-FeO buffers) on the equilibrium position is in sufficient.

Under the assumption that the activities of the solid phases are unity and the preferential occurrence of the dissolved forms of calcium and manganese at high temperatures are neutral chloride complexes CaCl₂^o and MnCl₂^o, one can write the following approximated relationship for the thermodynamic and concentrational equilibrium constants of the reaction (1):

(2)

where mCaCl₂(aq) and mMnCl₂(aq) correspond to mCa and mMn, measured after the run in quenched solution.

key words[gold solubility fluid sulphide]

In this work, we present the data on the experimental determination of solubility of gold at T = 300^oC, P = 300 atm, P_sat=85 atm, and pH_start.sol. = 7.4, which (Ph) was achieved by the addition of a solution of NaOH with a low concentration. The concentration of H₂S varies in a wide range from 5^.10^-2 to 0.7 m/l. The pH value of the solution with the content of 0.05 m H₂S measured after quenching was 4.15, and for a solution with 0.7 m H₂S, it was 4.0, and that calculated by the "Gibbs" program (Shvarov, 1982) for the system Au-S-Na-H-O at experimental parameters does not virtually change with the concentration H₂S from 0.05 to 0.7 m and equals 5.5.

The experiments were carried out in titanium 20 cm³ autoclaves. A gold (99.99) plate with a size of 10^.5^.0.1 mm was hanged on an obturator in the upper part of the autoclave. Thioacetamide (CH₃CSNH₂) was used as the H₂S source. In aqueous solutions, it is slowly (or rapidly in the presence of acids and alkalis) hydrolyzed to yield H₂S.

^# This work was supported by the RFBR (Project N. 97-05-64646)

32

key words [biotite fluid stability]

Isopleths of annite has been calculated for the reactions

Ann = Kfs + Mag + H₂ (R1)

2Ann + 3Qtz = 2Kfs + 3 Fay + 2H₂O (R2)

xAnn + 1.5S₂ + xKfs + 3Po_ss (R3)

Ann + 0.75O₂ = Kfs + 1.5 Hem _(Ilm) + H₂O (R4)

in equilibrium with C-O-H-S-F fluid (fig.1,2).

A new algorithm has been proposed for the calculation of the compositions of the C-O-H-S-F (Cl) gas mixture saturated and undersaturated in graphite. As contrasted from ref. [1] where the system of equations models the two-phase graphite-fluid equilibrium, in this work it is made up so as to calculate homogeneous fluid reactions. This solution is then modified for graphite-fluid, fluid-arbitrary mineral association equilibria. The proposed algorithm can be used to calculate oxygen fugacities in the reactions rock-COHSF fluid that involves 19 gas particles. The thermodynamic data for the fluid components were borrowed from ref. [2]. The fugacity coefficients for H₂O and H₂ were calculated from [1,3] those for O₂, CO₂, CO, CH₄, COS , SO₂, S₂, H₂S from [4]. The fugacity calculations for the remaining components were based on the ideal gas model. The thermodynamic constants of annite were calculated using the experimental data of ref. [5]. The calculations were performed by the model of biotite solid solution according to [6] supplemented by the dependence of (Fe²/Fe³+Fe²) in biotite on f_H2 in the fluid [7]. The isopleths of annite in equilibria with graphite, feldspar, fayalite, quartz and/or pyrrhotine in the T-f_O2 space have a pronounced thermal stability maximum (fig.2) complying with the equilibrium with the graphite-saturated COHSF fluid having a maximal molar fraction of water [8].

key words [dissolved nonelectrolite thermodynamic properties calculation]

An approach has been proposed for the description of thermodynamic properties of aqueous nonelectrolytes (Ar, H₂S, CO₂). The approach is based on the 'precise' multiparametric Hill equation [1] for solvent and the formalism closely resembling the virial equation of state. So, the

^# This work was supported by the RFBR (Project N. 98-05-64052)

^## This work was supported by the RFBR (Project N. 98-05-64234)

34

standard chemical potential of a dissolved particle at the given temperature T and pressure P is written in the form

,

where is the chemical potential of the corresponding pure gas under standard pressure, = 1.9872 cal/(mol^.K) is the universal gas constant, f₁^o and ₁^o are the fugacity and density of a pure solvent (H₂O), and B = B₁₂-B₁ is the difference between the virial coefficients characterizing the interactions nonelectrolyte H₂O and H₂O-H₂O. By differentiating this equation with respect to pressure and temperature, we obtain the expression for limiting molal thermodynamic characteristics of the dissolved component:

volume

enthropy and specific heat

Here the index 'i' is pertinent to the solvent at the pressure and temperature, and the index 'g' is pertinent to the dissolved component in the state of ideal gas at the same temperature and standard pressure.

To describe the temperature dependence of B three-parameter equation of the form is proposed.

The empiric parameters ₁, ₂, ₃ can be reconstructed from the experimental data for the Henry gas constant in the subcritical temperature region. These parameters are given in the table.

Gas	1	2	3.102
Ar	2.27132	0.86	-1.50352
H2S	1.5000	0.55	-0.4
CO2	1.76612	0.75	-1.23586

So the knowledge of the thermodynamic characteristics of pure gas together with the three parameters of the model ₁, ₂, ₃, enables one to describe the complete set of thermodynamic properties of a dissolved component (chemical potential, enthropy, molar volume, and specific heat) in a broad range of temperatures (0-500^oC) and pressures (1-2000 bar) involving the near-critical region. As an illustration the figures give the calculated (lines) and experimental (dots) values of various thermodynamic properties of dissolved hydrogen sulphide.

key words [pitch blende sulfur oxidation calculation]

V. I. Vernadskii Institute of Geochemistry and Analytical Chemistry, Russian Academy of Sciences

In the present time, it is established that the formation of pitchblende ores in almost all deposits in the world occurred at the temperature close to 200^oC. This indicates to the internal (for uranium-bearing solutions) temperature reason for ore deposition. Mineralogical studies show that sulfides, sulfo salts are deposited along with pitchblende, and carbonates (ankerite, calcite) are often present.

The experimental works of S. D. Malinin and N. I. Khitarov (1965) and then studies of R. P. Rafal'skii (1985) showed that the lower limit of reduction of sulfate sulfur to sulfide sulfur is 200^oC in near-neutral solutions and 250^oC in acidic solutions. According to the experimental data (Giggenbach, 1977), at temperatures lower than 200^oC, oxidation of sulfide sulfur is strongly suppressed and occurs not to SO₄^2-, but to the intermediate forms of oxidation of sulfur between S^2- and S⁰.

To estimate the effect of the kinetic factor on the redox interaction of different forms of sulfur, the thermodynamic modeling of the process was performed at 200^oC under hydrothermal solutions in the system corresponding to the natural hydrothermal uranium-molybdenum system.

The main task of the modeling performed was the consideration of possibilities of changing valence states of sulfur to form all known forms of occurrence of this element in solutions and minerals and elucidation of conditions of existence of sulfur in the sulfide and disulfide forms only. The amount of sulfur, which was in solution in the sulfate form, was replaced by artificial anions with the physicochemical properties similar to those of the forms of sulfuric acid in solution. This was performed to retain complex formation and acidity in the solutions. These anions had no effect on the redox conditions in the system. The choice of the state of the system at which this replacement was performed is caused by the maximum amount of the sulfate ion within the stability field of molybdenite.

The results of the theoretical experiments show that the following occurs when the mechanism of oxidation of sulfide sulfur is changed (sulfide/sulfate by sulfide/disulfide):

- a sharp increase in the stability field of pyrite and molybdenite;

- a shift of the redox parameters of the system to a more reduced region.

The estimation of possibilities of formation of anhydrite in equilibrium with calcite in a similar system showed:

- when S^2- is oxidized to S⁶⁺, anhydrite stable in association with pyrite and magnetite should be formed in equilibrium with calcite.

The presented results of mineralogical experimental and thermodynamic studies indicate that in natural hydrothermal (including uranium-bearing) solutions at temperatures close to 200^oC and lower, the formation of the sulfate ion is almost ceased. Sulfide sulfur is oxidized only to S- due to the kinetic peculiarities of the oxidation reactions.

The suggested mechanism of redox reactions of sulfur compounds explains the natural phenomena at the certain stage of the hydrothermal process (200^oC and lower) by:

- formation of pseudomorphoses of pitchblende by sulfides;

- deposition of pitchblende in association with disulfides in any rocks in the absence of strong reducing agents;

- a higher content of uranium in ores in regions of development of strong reducing agents, because the buffer action of sulfate sulfur is eliminated;

- formation of ferrous molybdenite;

- deposition of sulfo salts of Ag, As, Sb.

key words [pore solution freezing] Moscow State University

Based on a complex analysis of defrosting thermograms, calorimetric studies, and determination of thermophysical parameters of rocks containing Na and K chloride solutions, we discovered a thermal effect that we assigned to the process of cryohydrate formation in disperse rocks [1,2]. Later the existence of this effect was confirmed by NMR studies of the system.

In this work we have carried out a quantitative estimation of cryohydrate formation processes. We used a differential scanning calorimeter (DSC) 'Mettler TA-2000B' in the temperature range from +50 to 100^oC, the heating and cooling rates varied from 0.1 to 2-5L/min, the samples (both soils and solutions) were compressed into aluminium crucible, the charges were varied from 30 mg (in small crucibles) to 140mg (in large crucibles). The rocks of disturbed construction, i.e. sands, kaolin, Na-montmorillonite, hydromicaceous clay, containing NaCl and KCl solutions, and, also sea salt and NaCl solutions of various concentrations themselves were studied.

Two endothermal effects were documented on differential thermal curves upon heating NaCl, KCl, and sea salt solutions from 100 to +20^oC. The first effect was observed at temperatures close to 0^oC, the second one was observed near the eutectic temperature = -21^oC for NaCl and 11^oC for KCl) (fig.1a).

As the temperature is decreased from 100 to +20^oC the solutions get supercooled as a result of which the temperature of the first effect shifts by 10-20^oC (freezing point), and that of the second effect by 20-25^oC (cryohydrate formation temperature) (fig.1b).

The effect of cryohydrate formation was also observed in wet salinated specimens of sand and kaolin, but it is absent in Na-montmorillonite and hydromicaceous clay (fig.2), which may be related to a strong energy effect of the developed specific surface on a pore solution (S=600m²/g).

The values of freezing heats of salt solutions of various concentrations measured by the DSC vary from 320-330 J/g for small concentrations (5 g/l and 32.3g/l) to 220-250 J/l for strong concentrations (273 g/l) for all the investigated types of salinations.

^# This work was supported by the RFBR (Project N. 98-05-64864)

36

key words [freezing temeprature moisture in rock] Moscow State University

No unique approach exists presently for the description of various categories of moisture in rocks. There are, however, partly universal models to be proposed.

The principal assumptions used in deriving the formula (1) are as follows: ice that forms in pores liberates and occurs under normal pressure, the freezing rates are such that large ice crystals tend to form in pores, and in salt-rich rocks ice crystals force salt ions out into water films; the values of differential wetting heats, adsorption and dissolution are temperature independent; the energy influence of a specific active surface of rocks and of dissolved ions proceeds in an additive manner; the specific heat of the pore moisture for saltless and salt-containing rocks are taken both for supercooled water and analogous salt solutions in a volume, respectively.

Given that the water state in the liquid phase is chosen as standard, the principal formula for the freezing temperature estimation will be written in the form

(1)

where a is the activity; =T_o-T_f; T_f is the freezing temperature; T_o=273.15K; T_a is the rock temperature; L is the phase transition heat; R is the universal gas constant; [H_i^o-H_i] is the enthalpy difference, taken depending on the equilibrium type, as the differential wetting heat (adsorption, dissolution), the parameters , , for the adsorption equilibrium are found from specific heat of ice and supercooled water presented as series with respect to temperature, as proposed by Low P.E., Anderson D.M., Hoekstra P., and in the case of chemical equilibrium with respect to specific heat of ice and solutions with account taken of the dependence of specific heat on the concentration and type of salt.

Calculation from the adsorption equilibrium data. The phase equilibrium parameters were found by taking the values of differential heat adsorption and chemical potential from sorption isotherms for kaolinite and Na-montmorrilonite clay at various moistures. Θ was then calculated from (1) and the unfrozen water content (W) vs the temperature curves were plotted (fig.1). The curves were compared with the experimental data for those rocks. A satisfactory agreement was obtained.

key words [mineral solubility chloride fluid]

Vernadsky Institute of Geochemistry and Analytical Chemistry, RAS 117975 Moscow Kosygin Str., 19

An extremal scantiness and scatter in the available data on mineral solubility in fluids have impelled the authors to conducting systematic studies of mineral solubility in water-chloride solutions under the parameters of the fluid state. Over several years the authors were concerned with experimental studies of the system CaF_2(k) NaCl-CaCl₂-H₂O and CaWO_4(k) NaCl- CaCl₂- H₂O at temperatures 400-800^oC, pressures 0.5-2 kbar, with the concentrations of individual chlorides and their mixtures from0.05 to 34.7 mole/kg H₂O in the regions of homogeneous and heterogeneous fluid states [1-5].

The solubility studies were carried out by a capsule technique using a weight loss method, the runs were performed in a two-chamber high pressure cell provided with an external heater and a rapid isobaric quenching system.

The results of the studies are reported here.

The following regularities have been established for the homogeneous state of the chloride fluids in question.

1. Solubility of the both minerals grows drastically with the chloride concentration, the dependence being stronger in CaCl₂ solution despite the effect of the common calcium ion. Combined with the total high solubility, this fact attests the complex formation in the systems.

2. The dependence of solubility on the chloride concentration in a logarithmic scale is linear in the first approximation. For the solubility of CaF₂ in NaCl solutions at 800^oC and 2 kbar the linearity is found in the range of NaCl concentrations from 0.2 to 0.35 mole/kg H₂O. The slope of the isotherms remains herewith constant within 500-800^oC for NaCl and CaCl₂ solutions alike.

3. The diagrams of the both systems are close in appearance despite the significantly different chemical nature of the minerals. The isotherms slope stability within 500-800^oC gives evidence for the invariability of the composition of dissolved particles in the temperatural range.

4. The solubility data have been used to identify particles of the dissolved mineral substance by a method of fitting on the base of a number of assumptions. The forms (Na₂WO₄NaCl)^o and (CaWO₄CaCl₂)^o have been proposed for the system with scheelite, and (NaFNaCl)^o (CaF₂CaCl₂)^o for the system with fluorite. The approach employed does not enable one to solve uniquely the problem of estimation of particular form of particles, but it enables one to estimate the type of particles, in this case neutral in the charge, multiple-mixed in the composition.

The following regularities have been established for the mineral solubility in heterogeneous fluids at 600-800^oC and 0.5-2 kbar.

1. At the fluid decomposition in the system with NaCl, as a result of decompression, the mineral solubility changes in a jump-like manner (by approximately one order of magnitude), that is stipulated by the specific feature of mineral solubility in fluids, namely, by the drastic dependence of the solubility on the concentration of NaCl in a fluid. At 800^oC the effect of the solubility jump is more pronounced than at 600^oC.

2. As the concentration of NaCl is changed, in the heterogeneous region under the isobaric conditions the decompression of the phases and the solubility of fluorite are quantitatively described by the equation of the rule of the gravity center which is manifested in the proportional dependence of solubility on the fraction of the liquid phase in the heterogeneous system and gives evidence for the hydrolytic stability of NaCl solutions. This makes it possible to quantitatively estimate the mineral solubility in different parts of the system both in the heterogeneous and homogeneous regions (near the interphase boundary) based on minimal experimental data.

3. Studies of fluorite solubility in CaCl₂ and CaCl₂+NaCl solutions at 600^oC and 0.5-2 kbar establish no relationship between the supposed phase transition and fluorite solubility which is attributed to deep hydrolysis in these systems.

4. The major criterion of manifestation of the effect of solubility increase under the phase decomposition is the character of the solubility dependence on the salt compo-

^# This work was supported by the RFBR (Project N. 97-05-64036)

38

nent concentration in the fluid. The realization of the effect necessitates that the solubility should increase in proportion with the degree of the salt component concentration value higher than unity, the effect being manifested the stronger the higher is the degree.

The obtained results are beyond the scope of particular systems and are suggestive of a general conception of regularities of mineral solubility in supercritical solutions. Based on the general regularities of mineral solubility in fluids, we can propose the following factors for mineral deposition from solutions:

1. Simple cooling of a mineral saturated solution.

2. Isothermal dilution of a saturated solution with water.

3. Combined case dilution of a solution with subsequent cooling.

4. Change in the solution composition: extraction of Ca²⁺ from a solution (replacement by Na).

Decompression leading to the fluid heterogenization cannot be a deposition factor, as it was commonly believed. On the contrary this process increases the solvent action of a fluid and contributes to stabilization of the dissolved mineral substance.

1. Malinin S.D., Kurovskaya N.A. (1992) // Geochim., N.7., pp.993-1006.
2. Malinin S.D., Kurovskaya N.A. (1992) // Geochim., N.11., pp.1473-1482.
3. Malinin S.D., Kurovskaya N.A. (1996) // Geochim., N.12., pp.1183-1187.
4. Malinin S.D., Kurovskaya N.A. (1996) // Geochim., N.1., pp.51-58.
5. Malinin S.D., Kurovskaya N.A. (1993) // Geochim., N.3., pp.1-5.

Institute of Experimental Mineralogy, Russian Academy of Sciences, Chernogolovka, 142432 Moscow Region

key words [thermodynamic model solid solution]

The development of thermodynamic models from data on phase equilibria between solid solutions is based on equality of chemical potentials of the components for co-existing phase compositions. Experimental data compare points of compositions for which this equality is fulfilled, and the values of chemical potentials themselves are unknown, only their equality in these points is determined, i.e., the model is based on relative data.

In the general case, this is a quite correct mathematical model that has the thermodynamic form, but values of thermodynamic parameters obtained by the model are expressed in the inherent calculation scale. Of course, this model can be used for studying equilibria of the phase studied (for example, to determine the temperature or pressure of the mineral formation process from the co-existing compositions), but its use for obtaining quantitative values of thermodynamic properties (for example, for calculation of reactions with other phases) is not substantiated.

To develop a model with the quantitative thermodynamic content, one should synthesize a model based on the agreement of data on phase equilibria with those of thermochemical and crystallochemical studies. Efficient agreement requires, first, that the model form would provide a sufficient flexibility and, second, it would assume the direct accordance with the very wide range of studies. In particular, for agreement with data of low-temperature calorimetry, it is desirable that the model form assume a correct extrapolation to the low-temperature region.

In the case of decomposition of a solid solution, the very popular models of solid solutions, which represent thermodynamic functions of mixing in the form of third-power polynomials of the components of the solution composition, relate rigidly data on phase ratios to quantitative values of the excessive energy of mixing of the solid solution. The necessary flexibility of the model can be achieved by the supplement of the polynomials with terms with higher powers. This procedure is described in detail in [1]. The values of excessive enthalpy (H_e), entropy (S_e), and volume (V_e) are usually calculated by these polynomials for the specified composition of the solution, and based on these data, the excessive energy of mixing (G_e) is calculated for the specified temperature (T) and pressure (P):

G_e = H_e TS_e +PV_e

Formally, in this expression the excessive enthalpy has a sense of the excessive enthalpy at T = 0 K, and the excessive entropy has a sense of the configuration entropy, but in fact this model works correctly in some specified temperature range and does not assume extrapolation to zero values of temperature; therefore, the formal sense of these parameters is fictitious.

Somewhat different form of the model can be suggested.

Based on the following:

1) the excessive isobaric heat capacity (C_p) approaching T = 0 K tends to zero;

2) the temperature dependence of the excessive heat capacity can have a maximum and allow inversion; we can present the temperature dependence of the excessive heat capacity in the form of a fractional function with two parameters (A, B):

and using this function and the configuration entropy (S_con) (in the case of successive agreement of experimental data with the model, this value can be expected to be not fictitious.), we can determine the other excessive thermodynamic parameters:

_{T T}

The approaches supposed were successfully approbated for modeling of the three-component solution of feldspars based on the agreement of experimental data of different characters. The B parameter was accepted to be constant, and the A, S_con, and V_e values were determined by the parameters of the fourth-power polynomials on the components of the solution composition by the procedure described in [1].

1. Petukhov P. A. (1997) Detailed Empirical Models of Solid Solutions. // Geokhimiya, N.7, pp. 764-771 (in Russian).

39

key words [antimony complexes solution Raman spectrocopy] Moscow State Unibersity

The problem of transfer of ore components, specifically of gold, has drawn attention to thiocomplexes of antimony. S.A.Wood studied water solutions of thioantimonate by the Raman spectroscopy method (S.A.Wood, 1989) He examined the dissolution of antimonite (Sb₂S₃) at room temperature in a 1m Na₂S water solution, the concentration of Sb being from 0.005 to 0.1 m/l.

We have studied this system at temperatures to 130^oC, Sb concentration from 0.005 to 0.1 m/l and Na₂S concentrations from 0.1 to 0.8 m/l. The reactants were placed into a microautoclave (0.6 ml in volume) fabricated from a melted quartz or sapphire monoblock. The spectrum was recorded with an optical multichannel analyser.

At every concentration of antimony the spectrum demonstrates at least two RS lines the intensive well polarized 367 cm^-1 line and the depolarized 378 cm^-1 line. At Na₂S concentrations in excess of 0.5 m/l the third 350 cm^-1 line is sometimes observed. Besides, pure Na₂S solution is characterized by the weak 316 cm^-1 line.

S.Wood observed four lines at frequencies 380, 369, 350, and 314 cm^-1 which he totally assigned to a vibrational spectrum of a Sb₂S₄^-2 dimer. In doing so he was erroneous in the determination of the degree of depolarization of the 380 cm^-1 line and in the assignment of the 314 cm^-1 frequency vibrations to the thiocomplex.

Based on those results and his own quantum chemical calculations J.Tossell assigned the 380 cm^-1 and 369 cm^-1 frequencies to vibrations of H₂SbS₃^- and HSbS₃^-2 groups, and the 350 cm^-1 and 314 cm^-1 frequencies to vibrations of a H₂Sb₂S₄ dimer (J.A.Tossell, 1994).

At the same time there are Raman spectra of Na₃SbS₄ salt and its water solution (W.Mikenda, A.Preisinger, 1980) containing the polarized 367 cm^-1 line and the depolarized 380cm^-1 line which, most probably, correspond to vibrations of a SbS₄^-3 group or its protonic forms.

According to our quantum chemical calculation (program GAMESS) the SbS₄^-3 complex of symmetry T_d is characterized by the vibrational frequencies: 139 cm^-1 (E), 187cm^-1 (F2), 334 cm^-1 (A), and 360 cm^-1 (F2). So, the frequency of a polarized RS line (type A) ought to be lower than that of a depolarized line (type F2) which, precisely, is observed in experiment. The calculated interatomic distances Sb-S=2.32Å agree with those known for a salt Na₃SbS₄ (Von H.A.Graf, H.Schafer, 1976). The calculated atomic charges (according to Malliken) equal: Sb=(+1.164); S=(-1.041).