Affordable Access

Access to the full text

Modeling of the solubility of H2S in [bmim][PF6] by molecular dynamics simulation, GA-ANFIS and empirical approaches

  • Dashti, Amir1
  • Zargari, Farshid2
  • Harami, Hossein Riasat3
  • Mohammadi, Amir H.4, 5
  • Nikfarjam, Zahra6
  • 1 Islamic Azad University, Young Researchers and Elites Club Science and Research Branch, Tehran, Iran , Tehran (Iran)
  • 2 University of Sistan and Baluchestan (USB), Department of Chemistry, Faculty of Science, Zahedan, Iran , Zahedan (Iran)
  • 3 University of Kashan, Department of Chemical Engineering, Kashan, Iran , Kashan (Iran)
  • 4 Institut de Recherche en Genie Chimique et Petrolier (IRGCP), Paris Cedex, France , Paris Cedex (France)
  • 5 University of KwaZulu-Natal, Discipline of Chemical Engineering, School of Engineering, Howard College Campus, King George V Avenue, Durban, 4041, South Africa , Durban (South Africa)
  • 6 Chemistry and Chemical Engineering Research Center of Iran, Department of Molecular and Supramolecular Modelling, Tehran, Iran , Tehran (Iran)
Published Article
Korean Journal of Chemical Engineering
Publication Date
Oct 03, 2019
DOI: 10.1007/s11814-019-0330-y
Springer Nature


Predicting the solubility of acid gases in ionic liquids (ILs), has lately appeared as advantageous for natural gas purifying, which is equipped by powerful models considering technical and economic aspects. Important issue in the assessment of ILs for potential utilization in gas sweetening process is estimating the H2S solubility at various temperatures and pressures Experimental measurements are costly and take considerable time and effort. As a result, proposing methods for predicting the behavior of this system over a wide range of conditions is vital. In this regard, molecular dynamics simulation (MD) technique as well as artificial intelligence knowledge of hybrid genetic algorithm-adaptive neuro fuzzy inference system (GA-ANFIS) and an empirical polynomial regression (PR) model were employed to estimate the solubility of H2S in [bmim][PF6] IL. Pressure and temperature are considered as the independent input variables and H2S solubility as the dependent output variable. The results of this study reveal that the simple fourth-order PR model and GA-ANFIS have the highest accuracy. As a result of the simplicity and accuracy of PR model, it can be used without any prior knowledge about MD and artificial intelligence (AI). According to the accuracy and precision of model proved by the obtained result, the solubility of H2S in ILs has been estimated. The results show that the PR method is more trustworthy than other models.

Report this publication


Seen <100 times