Macro Approach to Molecular Modelling of Linear Polymers Applied to Estimation of Tensile Modulus for New Materials Development

CRAVERO, FIORELLA; MARTINEZ, MARÍA JIMENA; PONZONI, IGNACIO; DIAZ, MÓNICA F.

Aveiro

Conferencia; VIII International Symposium on Materials (Materiais 2017); 2017

Portuguese Materials Society (SPM)

The typical design of new materials has been empirical (formulation, assembling, synthesis, processing, and testing). At present, there has been much progress in the knowledge of relationships between the molecular structure of a material and its properties that lead to the ability to predict material properties previous synthesis. In this way, it could be obtained important reductions in time and cost. However, it is not easy to achieve these predictions for polymers since the involved variables are very complex because of high molecular weight (MW) and polydispersity (PDI). To the best of our knowledge, this is one of the first attempts to investigate the possibility of predicting tensile properties for polymers by using Quantitative Structure-Property Relationship (QSPR) techniques and considering real average MW of polymers (see dataset range in Table), instead of synthetic molecular models (monomer). At present work, main problems for modeling real polymeric materials are described and innovative solutions are proposed. Representing (2D) high MW polymeric molecules cannot be done with available software. We recently developed an informatics tool based on SMILES code that reaches 1x107 g/mol and more. On the other hand, QSPR technique requires calculating molecular descriptors which is also not possible with available software. In this stage, we have worked with different packages of the R programming language, and many of classic descriptors have been calculated. Once these problems were solved, we applied methodology used for micro approach, consisting of 3 steps: A-Feature Selection (DELPHOS), B-Computational Model (VIDEAN, WEKA) including statistical metrics and C- Physic-chemical interpretation. The obtained results are promising and the models look more robust than micro approach ones.