Engineering polymeric optical fibers with desired properties

Polymeric materials are finding increasing application in commercial optical communication systems. Taking advantage of techniques from the field of artificial intelligence, the goal of our research is to construct systems that can computationally design polymer formulations, including polymer optical fibers, with specified desirable consumer characteristics. Through the use of an extensive structure - property correlation database, properties of polymers can be predicted by an artificial neural network and the structure of novel polymers with desired properties can be optimized by a genetic algorithm. In this paper we are focusing on one of the parameters, glass transition temperature (T_g) that influences a desired outcome in polymer optical fibers. Performance of such fibers can be optimized by engineering a polymer to exhibit a lower refractive index and T_g. This paper compares and discusses an artificial neural network model and a multiple linear regression model that have been developed to correlate glass transition temperature (T_g) and repeating units of polymers. A set of descriptors, chosen based on previous studies on the relations between T_g and polymer structure, was used to describe the structure of repeating units, individual bond energies and intermolecular forces, especially hydrogen bonding, which is the strongest intermolecular force and exerts the greatest influence on T_g comparing with other intermolecular interactions. Comprehensive neural network models with 28 and 10 descriptors were developed to predict T_g values of 6 randomly selected polymers from a database containing 71 polymers.