N Sukumar - MQSPR modeling in materials informatics: a way to shorten design cycles?

Version 2

      Publication Details (including relevant citation   information):

     

     

    J. Mater. Sci.,47(21), 7703-7715 (2012). DOI: 10.1007/s10853-012-6639-0 (2012).

      Abstract:

      We demonstrate applications of quantitative structure–property   relationship (QSPR) modeling to supplement
      first-principles computations in materials design. We have here   focused on the design of polymers with
      specific electronic properties. We first show that common   materials properties such as the glass transition temperature   (Tg) can be effectively modeled by QSPR to generate highly   predictive models that relate polymer repeat unit
      structure to Tg. Next, QSPR modeling is shown to supplement and   guide first-principles density functional theory
      (DFT) computations in the design of polymers with specific   dielectric properties, thereby leveraging the power of   firstprinciples computations by providing high-throughput   capability. Our approach consists of multiple rounds of
      validated MQSPR modeling and DFT computations to optimize the   polymer skeleton as well as functional group
      substitutions thereof. Rigorous model validation protocols insure   that the statistical models are able to make valid
      predictions on molecules outside the training set. Future work   with inverse QSPRs has the potential to further
      reduce the time to optimize materials properties.

      Address (URL): http://www.doi.org/10.1007/s10853-012-6639-0