Solid-Liquid Equilibria Of Diphenyl Carbonate With Low Carbonate Alcohols

Diphenyl carbonate is one of the main monomers of polycarbonate (PC), and we call it DPC for short. At present, the need for PC by the world industry is increasing day by day. The purity of the DPC has an important impact on the quality of the PC, so more study on the purification for DPC is needed. DPC is a substance which is prone to decompose at high temperature, so cooling crystallization is used for the purification.The solid-liquid equilibrium data is necessary for the study of cooling crystallization, so the solubility of DPC in seven solvents were determined by static method in this paper. The solvents all remain with alcohol, including methanol, ethanol, 1-propanol, isopropanol, 1-butanol, isobutanol, sec-Butanol. As the result of experiments, cooling crystallization can be used to refine DPC because the solubility of DPC increases as the temperature ascends. The solubility data of DPC is not only the base for the solvents selection and process conditions study of cooling crystallization but also the base data for the study of dynamics and industrial magnification.The study on solid-liquid equilibrium not only can avoid many unnecessary experiments, but also can provide direction to the industry. The systems in this paper were studied by three solid-liquid equilibrium models, and the results are all very pleasant.At first, the UNIFAC group contribution method was used in the calculation of the systems in this paper. A new group C(O) O2 was defined in this paper, and the group volume and surface-area parameters were calculated. The interaction parameters of C(O)2O with CH3OH, CH2, OH and ACH were regressed with the data of the seven binary systems in this paper. The systems of this paper were predicted with UNIFAC, and the average correlation error was less than 2.1%.The experimental data were also correlated by Wilson equation andλh equation. Both of two equations have two parameters that have to be regressed with the experimental data. The largest average correlation error of the two equation were 0.36% and 0.87%, and they both fit the experimental data well.