QSPR studies of polyvinyls by density functional theory Xinliang Yu, Xueye Wang * , Jinwei Gao, Xiaobing Li, Hanlu Wang College of Chemistry, Xiangtan University, Xiangtan, Hunan 411105, People’s Republic of China Received 25 May 2005; received in revised form 10 July 2005; accepted 12 July 2005 Available online 8 August 2005 Abstract Density functional theory (DFT) calculations are carried out for polyvinyls repeating units at the B3LYP/6-31G(d) level, and the calculated results of E T , E int , C v , S, Q ii , m, a and q K are used to predict V (298 K), P s , F d , R LL , c, H vsum , U R and U H . Multiple linear stepwise regression analysis is used to generate eight more physically meaningful quantitative structure–property relationship (QSPR) models having correlation coefficient R of 0.996 for V (298 K), 0.998 for P s , 0.997 for F d , 0.997 for R LL , 0.997 for c, 0.992 for H vsum , 0.992 for U R and 0.991 for U H , and the conclusions are in consistence with theoretical analysis. Investigated results indicate QSPR models given here are easy to apply and have good predictive capability. q 2005 Elsevier Ltd. All rights reserved. Keywords: Polyvinyls; QSPRs; Density functional theory 1. Introduction Quantitative structure–property relationship (QSPR) models of polymers are the theoretical basis for polymeric molecular designs and material designs, and have the ability to survey a list of possible candidates and exclude ones that do not fall into the desired property range for the application. In particular, no time is wasted on synthesis and testing of new materials that are deemed inappropriate by QSPR models. With recent progress in computational hardware and the development of molecular quantum mechanical calculations, the exponential growth in the number of papers dealing with QSPR studies for polymers clearly demonstrates the rapid progress in this area. Hamerton et al. [1] have used molecular modeling and molecular orbital to find molecular descriptors that could be used to derive an empirical equation to describe the glass transition temperature (T g ) of two related classes of poly(arylene ether)s and succeeded in predicting the thermal characteristics of another polymer of the same type using the equation. Katritzky at el. [2] have developed a QSPR model having a coefficient of determination of 0.946 to predict T g s for a diverse set of 88 polymers with comprehensive descriptor for structural and statistical analysis (CODESSA) program. Cao and Lin [3] have tested the same set of 88 polymers against a correlation involving their own chosen descriptors in an attempt to derive a more physically meaningful QSPR with a coefficient of determi- nation of R 2 Z0.9056. Faulon et al. [4] have constructed QSPRs, which could be used to efficiently predict transport properties in amorphous polymeric material from molecular dynamics simulation since only bulk modulus and/or cohesive energy need to be determined from a simulation. Kholodovych et al. [5] have presented a surrogate model for the prediction of cellular response to the surface of biodegradable polymers. The prediction of their model, when tested against experimental results, has shown a high degree of accuracy that was sufficient for rational design of polymeric materials for biomedical applications. Therefore, their models can provide direct guidance to the synthetic chemist in biomaterial design. Zhang et al. [6] have built a QSPR model, based on the structural analysis of polymers, to predict the refractive index of linear polymers. It must be noted that Morrill et al. [7] apply the AM1 method within AMPAC and CODEESSA to calculate molecular properties to derived descriptors, which have strongly predictive nature and is a testimony to the role that semi empirical methods play in studying such large molecular systems. In addition, many authors have studied QSPRs for polymers Polymer 46 (2005) 9443–9451 www.elsevier.com/locate/polymer 0032-3861/$ - see front matter q 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.polymer.2005.07.039 * Corresponding author. Tel.: C86 732 829 2206; fax: C86 732 829 2477. E-mail address: wxueye@xtu.edu.cn (X. Wang).