Quantitative Structure Property Relations (QSPRs) for Predicting Standard Absolute Entropy, S°298, of Inorganic Compounds

摘要

For predicting the standard entropy of inorganic compound, two novel connectivity indexes mQ, mG and their converse indexes mQ', mG' based on adjacency matrix of molecular graphs are proposed as follows: mQ=∑(qi.qj.qk...)0.5, mG = ∑ (gi.gj.gk...)0.5, mG=∑(g1.gj.gk...)0.5, mQ' =∑(qj.qj.qk...)-0.5 The qi and qj of adjacency matrix are defined as qi=(1+Zi) /(1+ni), gi=(1+Zi) /(1+ri+ri-1), where Zi, ni, ri are the charge number, the outer electronic shell primary quantum number, and the radii of ionic / respectively. The excellent QSPR models for the standard entropies can be constructed from 0Q, 0Q', 0G, 1G, and 1G', by using multivariate linear regression (MLR) method and artificial neural network (NN) method. The correlation coefficient r, the standard error, and the average error of the MLR model and NN model are 0.990, 0.992, 8.88 J.K-1.mol-1, 7.83 J.K-1.mol-1, 7.10 % and 6.36%, respectively, for 590 inorganic compounds. The cross-validation by using the leave-one-out method demonstrates that the MLR model is highly reliable from the point of view of statistics. The results show that the current method is more effective than literatme metliods for estimating the standard entropy of inorganic compound. Both MLR and NN methods can provide acceptable models for the prediction of the standard entropies. The NN model for the standard entropy appears to be more reliable than the MLR model.