A side chain pulled-along model of polymers is presented in this paper. It is believed that the stereo effect of side chains of polymers comes from their axis cross-area, which correlates to the number of carbon branches of side chains. The numbers of carbon branches of side chains are used directly as the descriptors of their stereo effect. Based on this assumption, quantitative structure-property relationships ( QSPR) are developed to correlate the glass transition temperature ( T-g) to the chemical structure of 13 polyacrylates and 9 polymethacrylates, respectively. Good three- parameter models, which consist of the number of carbon branches of side chains B, the group charge of side chains C and the topological length of side chains L, are obtained with regression coefficients R-2 = 0.989 ( s = 3.8 K) and R-2 = 0.993 ( s = 4.8 K) respectively. For a common three-descripter model 2 covering both the 13 polyacrylates and the 9 polymethacrylates, the regression coefficient is R-2 = 0.980 ( s = 8.3 K) At can be seen from the statistics analysis results for all models that the descriptors B and C have positive correlation to T-g of the title compounds and the descriptor L has negative correlation. Furthermore, the descriptor B has greater influence on the models than the other two. So it is comparatively reasonable to use the number of carbon branches of side chains as the descriptor of the stereo effect for polyacrylates and polymethacrylates. All the descriptors in this paper are easy to be calculated. The models have good stability and applicability, whose standard errors are all less than or approximate to the experimental error.
