Generaked site-bond percolation theory is uttied to provide an analog for a two-component liquid-vapor system in the phase transltlon region. The percolation model exlubits a behaviour typical of a non-ideal real hqrud-vapor system with features such as separate curves for the hquld and vapor phases, and mimmum and mavrmum azeotrope formation. The analogy between phase transltion processes and percolation phenomena has been deternuned m recent years. Kasteleyn and Fortum [l] have shown that the bond percolation model is a special case of the Ashkin-Teller [2] and Potts [3] model for S-state atoms; thus provxhng a theoretical foundation in utlkzmg the percolatlon model as an analog [4] for phase transition phenomena. Recently Stauffer [s] has emphasized the analogy between percolation, magnetization, hquldvapor transItIon and polymer gelatlon.
In this letter I would like to explore the analogy between the generalized site-bond problem [6] and a two-component hquid-vapor system. In this system the fimte clusters and the mfiite cluster provide the vapor and the condensed phase analogs [S], respectxvely. The percolatron threshold is reached when finite clusters (LLvapor") coalesce to form an infmte cluster ("hquld").
Let us now consider a percolation model for a bmary rmxture of A and B atoms, randomly &stnbuted over a lattice. Each lattice site JS occupied with an A atom or a B atom with a probabtity CA or C,, respectively such that CA + CB = 1. In a real hquid-vapor system CA and C, would denote the mole fraction of the A and B components.
Kasteleyn and Fortuin's [1] prescription for the bond probabihties can be extended to the generalized site-bond problem. Thus, one could wnte the following bond probabihties for the A and B atoms:
* Supported by NSF Grant DMR7507832
AOl.