Experimental measurement and modelling of vapor-liquid equilibrium for 3,3,3- Trifluoropropene (R1243zf) and trans-1,3,3,3-Tetrafluoropropene (R1234ze(E)) binary system

Abstract

• Experimental measurement of VLE data for R1243zf + R1234ze(E) by a static analytical method. • The R1243zf + R1234ze(E) system is a zeotropic system within 283.15 to 323.14 K. • Experimental data show good agreement with the predicted results of PR-MC-vdW and PR-MC-MHV2 models. The environmental performance becomes the critical factor of the selection of refrigerant. 3,3,3-trifluoropropene (R1243zf) and trans-1,3,3,3-tetrafluoropropene (R1234ze(E)) are environmental friendly candidates to replace R134a. Whereas, in the consideration of environmental factor, safety, and energetic efficiency at the same time, none of the pure refrigerants has been proved to be a suitable solution. Mixed refrigerants provide an approach to this problem because the characteristics of refrigerant blends can be optimized for particular applications in the way pure component refrigerant alone cannot be. The purpose of this work is to explore an R1243zf/R1234ze(E) blend which can be possibly used as a long-term alternative to replace hydrofluorocarbons (HFCs) and hydrochlorofluorocarbons (HCFCs). We report the measurements of saturated vapor pressures of R1234ze(E) and R1243zf ranging from 273.17 to 353.13 K, and isothermal vapor-liquid equilibrium of R1243zf + R1234ze(E) systems from 283.15 to 323.14 K. The measurements are performed by static-analytic type apparatus coupled with two electromagnetic capillary samplers (ROLSI®, patent of Armines). The experimental data are correlated by the Peng-Robinson (PR) equation of state (EoS) associated with Mathias-Copeman (MC) alpha function and classical mixing rules. For the comparison of conventional mixing rules and excess free energy ( G E ) mixing rules models, classic van der Waals one fluid mixing rules and modified Huron-Vidal second-order (MHV2) mixing rules are employed respectively to the correlation of experimental VLE data. The modeling results are in good agreement with the measured data, while the PRMC-MHV2 model exhibits better performance in VLE prediction.