Poster: Polymer Informatics Tools for Sustainable Polymers and 3D printing
Bacova, Petra; Christofi, Eleftherios; Harmandaris, Vagelis; Molina, Sergio I.
- Publisher
- Zenodo
- Language
- en
Abstract
Poster presented at 2nd Spanish Soft Matter 1 1/2 Day (https://www.benasque.org/2024ssm/) which was held in Benasque, Spain from November 3 to November 6, 2024. Abstract: Biodegradable polymers are intended to replace the synthetically produced polymers, however, their wider usage is limited by their moderate thermal and mechanical properties. In order to elucidate the structure-properties-performance relationship of these materials, we design a bottom-up approach, combining multiple computational tools, with the ultimate goal to facilitate their characterization and property optimization during the additive manufacturing. We use the poly(lactic acid) (PLA) as the model polymer, which due to its chirality and wide popularity in 3D printing represents an ideal candidate to test the developed methodology. In the first stage of the bottom-up approach, we produced an extensive data set consisting of atomistic trajectories of different stereoisomers of PLA in melt. We used those data to build a chemistry-specific coarse-grained (CG) model to extend the time and length scales to those relevant in the experimental studies. In addition, to close the loop, we implement a machine-learning based methodology to reinsert the atomistic details into CG models of different stereochemistry [1]. Since the computational techniques are considered to be a more sustainable alternative to the experimental characterization, we aim to extend the simulation practices commonly used for synthetic polymers to more complex bio-based polymers. By combining different computational techniques, we provide a consistent set of open-access tools [2,3] with the ultimate goal to facilitate the usage of multiscale computational analysis in the fast-growing field of biodegradable materials and additive manufacturing. [1]A Physics-informed Deep Learning Approach for Re-introducing Atomic Detail in Coarse-Grained Configurations of Multiple Poly(lactic Acid) Stereoisomers, E. Christofi, P. Bačová, V. Harmandaris, Journal of Chemical Information and Modeling, 2024 64 (6), 1853-1867 [2]https://github.com/SimEA-ERA/PLABackMap-CG[3]https://github.com/pbacova/PLA_analysis_tools.git
Full text
Polymer Informatics Tools for Sustainable Polymers and 3D printing P. Bačová1*, E. Christofi2, V. Harmandaris2,3,4,S. I. Molina1 1Departamento de Ciencia de los Materiales e Ingeniería Metalúrgica y Química Inorgánica, Facultad de Ciencias, IMEYMAT, Campus Universitario Río San Pedro s/n., Universidad de Cádiz, 11510 Puerto Real, Cádiz, Spain 2Computation-Based Science and Technology Research Center, The Cyprus Institute, Nicosia, Cyprus 3Department of Mathematics and Applied Mathematics, University of Crete, Heraklion, Crete, Greece 4 Institute of Applied and Computational Mathematics, Foundation for Research and Technology - Hellas, GR-71110 Heraklion, Crete, Greece *e-mail: petra.baco[email protected] •Extend the methodologies commonly used for synthetic polymers to more sustainable materials MOTIVATION METHODOLOGY PRE-PROCESSING: ATOMISTIC SIMULATIONS This research was funded by the European Union’s Horizon 2020 research and innovation program under Grant no. 810660 and under the Marie Skłodowska-Curie Horizon Europe action (grant agreement no. 101105208). Computing time awarded on the Cyclone supercomputer of The Cyprus Institute under project ID pro22a107s1 and computing time granted from the National Infrastructures for Research and Technology S.A. (GRNET S.A.) in the National HPC facility ARIS under the project name SPASA are also acknowledged. MODEL EVALUATION TRAINING: INTRODUCING PRIOR KNOWLEDGE ℒ =1 𝑛|𝐛𝐢−𝐛 𝐢| ℒ = ℒ +ℒ •Versatile and open-access [1,2] approach based only on relative positions, avoids tedious bookkeeping of molecular details during the reconstruction •Approach capable of predicting atomistic structures, resembling those in equilibrium, of an arbitrary composition and a chain length [3] •Easily extended to a broad class of polymeric materials given the CG description CONCLUSIONS ACKNOWLEDGEMENTS REFERENCES [1] P. Bačová, PLA Analysis Tools , 2023. https://github.com/pbacova/PLA_analysis_tools.git [2] E. Christofi, PLA Backmapping , 2023 . https://github.com/SimEA-ERA/PLA-BackMap-CG [3] E. Christofi et al., Journal of Chemical Information and Modeling 2024, 64 (6), 1853-1867 DOI: 10.1021/acs.jcim.3c01870 Label # of mers Microstructure PLLA(%) PDLA(%) PLLA100 100 100 0 PDLA100 100 0 100 Copo100 100 45 55 PLLA30 30 100 0 PDLA30 30 0 100 Copo30 30 84 16 ℒ =1 𝑛| 𝐛𝐢− 𝐛 𝐢| Target Initial prediction runEQ output PLLA100 PDLA100 Copo100 PLLA30 PDLA30 Copo30 Copolymer’s initial prediction from a system with different microstructure Copolymer’s initial prediction from a model trained solely on homopolymers Target copolymer Copo100 •Bond vectors term: •Bond lengths term: TRANSFERABILITY ACROSS MOLECULAR WEIGHT CHEMICAL TRANSFERABILITY •Develop computational tools to facilitate the characterization of these materials prior processing and thus to reduce the carbon footprint of the process •Connect theoretical and experimental data to elucidate the structure-properties-performance relationship
