Computer analysis of phase diagrams and thermodynamic properties of cryolite based systems: I. The AIF 3 -LiF-NaF system
- PDF / 1,219,485 Bytes
- 9 Pages / 583.28 x 777.28 pts Page_size
- 100 Downloads / 221 Views
FOR about a century, the Hall-Heroult electrolytic process has been used for the production of aluminum. Nominally, the electrolyte consists of an appropriate mixture of A1F3, A1203, LiF, NaF, and CaF/, with a composition that varies slightly from one industrial plant to the other. In addition, consideration must be given to other constituents which result either from impurities present in the raw materials used (e.g., MgFz) or from chemical reactions between the basic constituents of the electrolyte, the liquid aluminum, and the graphite container. Therefore, the complexity of a study of phase equilibria in the electrolyte is evident. There have been numerous investigations of phase diagrams of limited regions of the above mentioned electrolyte. A recent review by Grjotheim, Krohn, Malinovsky, Matiasovsky and Thonstad ~gives an almost complete survey of the experimental work. Large disparities in the data are often encountered owing to the limitations inherent in the techniques used and to the chemical nature of the melts (e.g., a significant tendency to supercool). Selection of the most reliable data among those available is likely to be aided by using a method that is thermodynamically self-consistent. In the present paper, we report on computations and analyses of the phase diagrams for the following systems, A1F3-LiF, A1F3-NaF , LiF-NaF, Li3A1F6NasA1F6, and A1F3-LiF-NaF. From available thermodynamic data of the pure compounds and of the binary solutions (e.g., the known phase diagram data, the calorimetric enthalpies of mixing reported by Holm 2 and by Hong and Kleppa 3,4and the excess free energies of mixing measured by DewingS,6), the phase diagrams of the above systems are calculated. The calculations are thus founded on a broad data base and are useful in assessing the validity of experimental results. Finally, the derived numerical values can be used to define an M. L. SABOUNGI is Chemist at Argonne National Laboratory, Argonne, IL 60439. P. L. LIN is Assistant Research Professor, Universite de Montreal, Montreal, Canada, P. CERISIER is Maitre-Assistant, University of Provence, Marseille, France, and A. D. PELTON is Associate Professor, Universite de Montreal, Montreal, Canada. Manuscript submitted July 13, 1979.
optimum range of temperatures in the operation of production cells, which is an important consideration in energy conservation. In what follows, we first describe briefly the principles and the numerical method used in the phase diagram computations. Secondly, we review the thermodynamic data available for the components and binary solutions involved in this study, and compare the calculated phase equilibria of the binary subsystems with the various experimental results available in the literature. Thirdly, we compute the phase diagram of a portion of the ternary A1F3-LiF-NaF system and compare it with published data. Finally, the relative importance of the selected data on the topology of the phase diagram is discussed. II. PHASE DIAGRAM COMPUTATIONS A. Basic Considerations The phase diagram of
Data Loading...
