Skip to main content
U.S. flag

An official website of the United States government

Official websites use .gov
A .gov website belongs to an official government organization in the United States.

Secure .gov websites use HTTPS
A lock ( ) or https:// means you’ve safely connected to the .gov website. Share sensitive information only on official, secure websites.

Quantitative Structure-Property Relationship Predictions of Critical Properties and Acentric Factors for Pure Compounds

Published

Author(s)

Wendy H. Carande, Andrei F. Kazakov, Chris D. Muzny, Michael D. Frenkel

Abstract

Knowledge of vapor-liquid equilibrium properties (such as critical constants and phase boundary pressure) is essential to understanding thermodynamic behavior of substances and is often required in practical process design applications. Where experimental data are unavailable, a quantitative structure-property relationship (QSPR) regression method can be used to relate molecular properties (descriptors) to the experimental properties of interest. The relationship is trained and tested using existing experimental data and is dynamic; as new data becomes available, the relationship can be updated to reflect changes. In this work, we use support vector regression (SVR) to predict critical properties and acentric factors for over 900 pure compounds. Using three-dimensional molecular structure information, we calculate over 500 descriptors for each compound. A matrix of descriptor values defines the input vectors for SVR, while experimental values of critical temperature, the ratio of critical temperature to critical pressure, and saturation reduced pressure form the targets. We determine optimal SVR parameters by minimizing the sum of absolute errors between the SVR outputs and the target values. Then, we use a genetic algorithm to find the Pareto Front values that optimize the output fit while reducing the number of input vectors (descriptors). We use a single Pareto front value to make a final evaluation in SVR. To define uncertainties of predicted values, we use error propagation calculations based on a Monte Carlo method that employs latin-hypercube sampling.
Citation
Journal of Chemical and Engineering Data
Volume
60

Keywords

Critical properties, Empirical modeling, Quantitative Structure-Property Relationships, Support Vector Machines, Genetic algorithms

Citation

Carande, W. , Kazakov, A. , Muzny, C. and Frenkel, M. (2015), Quantitative Structure-Property Relationship Predictions of Critical Properties and Acentric Factors for Pure Compounds, Journal of Chemical and Engineering Data, [online], https://doi.org/10.1021/je501093v (Accessed April 19, 2024)
Created April 14, 2015, Updated November 10, 2018