Abstract
The individual correlations between nematic transition temperatures and the molecular descriptors derived from density functional theory (DFT) method are investigated. Significant quantitative structure-property relationships (QSPRs) for T
N
of a set of 42 theromtropic liquid crystal were established through principle component analysis (PCA) and multiple linear regression (MLR). A promising six-descriptor linear model with good statistical fitting and predicting capability were developed with the help of DFT. The model is based on E
HOMO
, E
LUMO
, μ, S, W and X. The reliability of this model is clear from its correlation coefficient (R
2
= 0.95) and cross-validation coefficient (R
2
CV
= 0.93).