(272e) Materials Informatics for Process Optimization: Case Studies Using P3HT and PP Composites | AIChE

(272e) Materials Informatics for Process Optimization: Case Studies Using P3HT and PP Composites


McBride, M. - Presenter, Georgia Institute of Technology
Persson, N., Georgia Institute of Technology
Reichmanis, E., Georgia Institute of Technology
Grover, M., Georgia Tech
The traditional paradigm in materials process optimization is energy-intensive and time consuming. Generally, single factor, trial-and-error based experimentation is used to screen a wide array of complex processing conditions. This generates a vastly multidimensional design space that is impossible to fully characterize. Furthermore, results are difficult to reproduce because of a lack of standardization due to the exploratory nature of the field. As a result, it is challenging to develop global, robust process-structure-property relationships to generate novel material formulations. Instead, informatics based methodologies can leverage existing knowledge from the literature to guide high-throughput screening of promising candidate experimental conditions. In this work two case studies of applying materials informatics methodologies to polymer systems will be presented.

In the first study, the impact of processing conditions on the organic field effect mobility of the semiconducting polymer, poly(3-hexylthiophene) (P3HT), will be presented. A database of over 200 devices was created to identify significant trends and to identify an optimal operating space. Processing conditions were sorted and filtered to identify a standard device, with reported mobility values that span two orders of magnitude. Furthermore, both integer and categorical processing conditions resulting in mobility values exceeding 0.1 cm2/V-s were identified to reduce the dimensionality of the design space. This enabled the generation of new hypotheses for future high-throughput experimentation.

In the second study, a database with over 140 entries quantifying the development of polypropylene and talc composites for high strength materials was created. However, this database suffered from inconsistent reporting of processing conditions. Physical based models, for example relating melt flow index to molecular weight, were proposed to help fill in missing data. Varying thresholds of Young’s Modulus were applied to identify the span of processing conditions that result in high strength composites. In both studies quantified analysis of all processing conditions has identified key design spaces to enable high through-put experimentation.