Premium
An open and extensible sigma‐profile database for COSMO‐based models
Author(s) -
Ferrarini F.,
Flôres G. B.,
Muniz A. R.,
de Soares R. P.
Publication year - 2018
Publication title -
aiche journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.958
H-Index - 167
eISSN - 1547-5905
pISSN - 0001-1541
DOI - 10.1002/aic.16194
Subject(s) - sigma , set (abstract data type) , computer science , dilution , software , binary number , experimental data , basis (linear algebra) , extensibility , computational science , spec# , cosmo rs , range (aeronautics) , theoretical computer science , database , data mining , chemistry , mathematics , physics , thermodynamics , engineering , quantum mechanics , programming language , statistics , aerospace engineering , biochemistry , geometry , arithmetic , ionic liquid , catalysis
COSMO‐based activity coefficient models have become an interesting alternative to the prediction of the behavior of substances in mixture. These models depend on the pure substance information known as sigma‐profile. The present work aims to create and freely distribute a sigma‐profile database for a wide range of molecules using the GAMESS software. Different quantum chemistry theories for the calculation of the electronic structure and basis sets were tested, in order to define an accurate and efficient approach for computing the profiles. To check the accuracy of the profiles obtained, infinite dilution activity coefficients calculated by the COSMO‐SAC model were compared with experimental data for binary mixtures. Among the studied alternatives, the Hartree‐Fock method with the TZVP basis set led to results with the best agreement with the experimental data, within a reasonable computational cost. Using this approach, a database was created based solely free computational tools. © 2018 American Institute of Chemical Engineers AIChE J , 64: 3443–3455, 2018