Volume 18, Issue 4 pp. 521-525
Article
Full Access

Quantitative prediction of substituted products based on quantum-chemical parameters and neural network method

Wang Xue-Ye

Wang Xue-Ye

College of Chemistry and Chemical Engineering, Xiangtan University, Xiangtan, Hunan 411105, China

Search for more papers by this author
Song Huang

Song Huang

College of Chemistry and Chemical Engineering, Xiangtan University, Xiangtan, Hunan 411105, China

Search for more papers by this author
First published: 27 August 2010

Abstract

The criterion of orientating group of electrophilic aromatic nitration was discussed by means of pattern recognition method with quantum-chemical parameters as features, and the product ratios of the reactions were quantitatively calculated using artificial neural network (ANN) method with the same parameters as inputs, based on the ab initio calculation of quantum chemistry, The quantum-chemical parameters involved orbital energy, orbital electron population, atomic total electron density and atomic net charge. The predicted values are in agreement with experimental results and the predicted error of the ANN with quantum-chemical parameters for the reaction is the smallest among the all methods.

The full text of this article hosted at iucr.org is unavailable due to technical difficulties.