Geochimtca ez Cosmochchrmico Acta Vol. S3, pp. 2581-2590 Copyright Q 1989 Pcr@mon Press plc. Pnntcd in U.S.A. zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 0016.7037/89/$3.00 + .oO The thermodynamics of the aqueous carbon dioxide fluid within thin pores A. B. BELONOSHKO Institute of Experimental Mineralogy, U.S.S.R. Academy of Sciences, 142432 Moscow District, Chemogolovka, U.S.S.R. zyxwvutsrqponm (ReceivedAugust 17, 1988; accepted in revised form June 15, 1989) Abstract-The density functionaI method was used to study the properties of a binary mixture water- carbon dioxide confined within an infinitely long cylindrical pore of 8 to 25 A in diameter, over a wide range of temperatures, pressures and compositions. Interaction with pore walls was calculated through the use of van der Waals and Kazimir-Polder attractive forces; the inte~oi~u~ interaction was calculated by means of the spheric-symmetrical approximation. The caiculations show that the pore fluid differs markedly in composition from the bulk fluid with which it is in equilibrium at the same temperatures and pressures. The difference diminishes regularly with increasing pore size to become negligible at a pore diameter of 25 A. The results of the work presented here provide a physical explanation of the JOHANNES-SCHREYER (1981) experiments. Reasons are offered for the predominancy carbon dioxide composition of fluid inclusions formed from the fluid of aqueous bulk composition. The relationship between the size of fluid-filled pores and the geometry of the MgO-SiOz-HzO-CO2 diagram in the T-&o, coordinates is discussed. The low activity of water during metamorphism has been explained. 1. INTRODUCTION THE PROBLEM OF peculiar fluid properties at the fluid-solid interface has long attracted a great deal of attention from many scientists in various fields, including Earth Science. This interest is well warranted because many phenomena taking place near the interface (dissolution, crystallization, adsorption, condensation and many others) can only be ex- plained from detailed analysis of the physico-chemical pro- cesses going on in the near-surface layers of the fluid. It is little wonder, therefore, that these topics attract much research effort in the Earth sciences and that the number of publica- tions has considerably increased. As far back as 1929, VERNADSKY (1929), in his work on different forms of water found in nature, recognized a special type, “hair water”, i.e., water to be found in the microcapillary form. Since that time, the unusual properties of a fluid con- fined in a thin mineral pore have been studied in many geol- ogy-related works. Those are experimental studies conducted by Blokh on the thermodynamics of microporous systems water-rocks (BLOKH, 1980, 198 I). Other investigations ad- dressed themselves to the issue of mineral strength (ETH- ERIDGE, 1983, 1984) and mechanically induced hydrolysis (ANDERSON and GREW, 1977). The properties of supercritical water in quartz pores were studied by the molecular dynamics method and through sets of integral equations (BELONOSHKO and SHMULOVICH, 1986, 1987; BELONOSHKO, 1988). The computer simulation method was also used to study ion solvation in thin slit-like pores (SPOHR and HEINZINGER, 1985). The kinetics of dis- solution and formation of etch pits on the quartz surface was investigated by means of the Monte Carlo method (LASAGA and B~~,1986). It has been found in al1 these works that the properties of the surface and the very occurrence of the surface have a marked effect on the fluid properties near the surface. At the same time, porosimetry measurements ( VIKENTJEV and TOPOR, 1987) and direct rock permeability (VITOVTOVA and SHMONOV, 1982) indicate that small pores, 25 A (1 ii = lo-” m) and less in diameter, account for most of the pore space. Most studies regard the Buid as purely aqueous, whereas the actual fluid in the Earth’s crust consists mainly of two components-water and carbon dioxide (FURRY, 1976, 1979). The bulk properties of such fluids are relatively well understood (KERRICK and JACOBS,198 11, whereas confined water-carbon dioxide fluids have been given insufficient at- tention. We can mention an experimental study on the ad- sorption of the water-carbon dioxide fluid on cordierite (JO- HANNES and SCHREYER, 198 1). SHMONOV et al. (1984) an- alyzed the partitioning isotherms for the water-carbon dioxide fluid by extrapolating the experimental adsorption measure- ments on NaXzeolite (ZHUKOV, 197 1; DUBININet al., 1966) for pores of arbitrary size. Hence, it is very important for high-temperature and high-pressure mineral equilibrium studies that the actual composition of confined water-carbon dioxide fluid in equilibrium with the bulk fluid at the same T and P be determined. In the present work, an attempt has been made to apply the results of mean-field density func- tional calculations obtained for binary confined solutions (TAN et al., 1987; BELONOSHKO,1989) to calculating the composition of the water-carbon dioxide fluid which is in equilib~um with the bulk fluid and confined in cylindrical pores of 8, 10 and 25 A in diameter. The calculations were carried out for the whole compositional range at 400°C < T < 1000°C and 30 MPa < P < 1000 MPa. The paper is arranged as follows: Sec. 2 discusses briefly the density functional method as used in the present study. In Sec. 3 we present the calculation of fluid component in- teraction with the mineral surface and intermolecular inter- action. Sec. 4 analyzes the calculated composition of the wa- ter-carbon dioxide fluid confined in a pore. In Sec. 5 the T-X,,, phase diagram for the system MgO-SiOz-HzO-CO2 is used to assess the effect of compositional changes in the con- fined fluid. We conclude in Sec. 6 with a discussion of possible lines of research on the surface fluid in geological systems. 2581