Machine learning enables polymer cloud-point engineering via inverse design

Inverse design is an outstanding challenge in disordered systems with multiple length scales such as polymers, particularly when designing polymers with desired phase behavior. Here we demonstrate high-accuracy tuning of poly(2-oxazoline) cloud point via machine learning. With a design space of four repeating units and a range of molecular masses, we achieve an accuracy of 4 °C root mean squared error (RMSE) in a temperature range of 24–90 °C, employing gradient boosting with decision trees. The RMSE is >3x better than linear and polynomial regression. We perform inverse design via particle-swarm optimization, predicting and synthesizing 17 polymers with constrained design at 4 target cloud points from 37 to 80 °C. Our approach challenges the status quo in polymer design with a machine learning algorithm, that is capable of fast and systematic discovery of new polymers.

[1]  C. D. Gelatt,et al.  Optimization by Simulated Annealing , 1983, Science.

[2]  Andreas Holzinger,et al.  Data Mining with Decision Trees: Theory and Applications , 2015, Online Inf. Rev..

[3]  Françoise M. Winnik,et al.  Microcalorimetric Study of the Temperature-Induced Phase Separation in Aqueous Solutions of Poly(2-isopropyl-2-oxazolines) , 2004 .

[4]  Anand Chandrasekaran,et al.  Polymer Genome: A Data-Powered Polymer Informatics Platform for Property Predictions , 2018, The Journal of Physical Chemistry C.

[5]  Richard Hoogenboom,et al.  Tuning the LCST of poly(2‐cyclopropyl‐2‐oxazoline) via gradient copolymerization with 2‐ethyl‐2‐oxazoline , 2014 .

[6]  Ulrich S. Schubert,et al.  Poly(2-cyclopropyl-2-oxazoline): From Rate Acceleration by Cyclopropyl to Thermoresponsive Properties , 2011 .

[7]  Guigang Zhang,et al.  Deep Learning , 2016, Int. J. Semantic Comput..

[8]  Kurt Kremer,et al.  Molecular dynamics simulation of a polymer chain in solution , 1993 .

[9]  M. C. Stuart,et al.  Emerging applications of stimuli-responsive polymer materials. , 2010, Nature materials.

[10]  Shyue Ping Ong,et al.  Deep neural networks for accurate predictions of crystal stability , 2017, Nature Communications.

[11]  G. C. Tiao,et al.  Bayesian inference in statistical analysis , 1973 .

[12]  Corinna Cortes,et al.  Support-Vector Networks , 1995, Machine Learning.

[13]  Juan Carlos Rueda,et al.  Synthesis and characterization of block copolymers from 2-oxazolines , 2015 .

[14]  Arun Mannodi-Kanakkithodi,et al.  Scoping the polymer genome: A roadmap for rational polymer dielectrics design and beyond , 2017, Materials Today.

[15]  Ulrich S. Schubert,et al.  Thermoresponsive polymers with lower critical solution temperature: from fundamental aspects and measuring techniques to recommended turbidimetry conditions , 2017 .

[16]  Eric R. Ziegel,et al.  The Elements of Statistical Learning , 2003, Technometrics.

[17]  Michael E. Paulaitis,et al.  Monomer Hydrophobicity as a Mechanism for the LCST Behavior of Poly(ethylene oxide) in Water , 2006 .

[18]  Kazunori Kataoka,et al.  Precise control of lower critical solution temperature of thermosensitive poly(2-isopropyl-2-oxazoline) via gradient copolymerization with 2-ethyl-2-oxazoline as a hydrophilic comonomer , 2006 .

[19]  Petra Ostermann.  Self Organized Nanostructures Of Amphiphilic Block Copolymers I , 2016 .

[20]  Santiago J. Garcia,et al.  Effect of polymer architecture on the intrinsic self-healing character of polymers , 2014 .

[21]  Richard Hoogenboom,et al.  Tuning the LCST of poly(2-oxazoline)s by varying composition and molecular weight: alternatives to poly(N-isopropylacrylamide)? , 2008, Chemical communications.

[22]  J. Friedman.  Greedy function approximation: A gradient boosting machine. , 2001 .

[23]  Melanie Mitchell,et al.  An introduction to genetic algorithms , 1996 .

[24]  Kazunori Kataoka,et al.  Comprehensive and Accurate Control of Thermosensitivity of Poly(2-alkyl-2-oxazoline)s via Well-Defined Gradient or Random Copolymerization , 2007 .

[25]  C. Brabec,et al.  A Bayesian Approach to Predict Solubility Parameters , 2018, Advanced Theory and Simulations.

[26]  P. Chattopadhyay.  Fifty Years of Research , 2015 .

[27]  Chiho Kim,et al.  A polymer dataset for accelerated property prediction and design , 2016, Scientific Data.

[28]  Jürgen Schmidhuber,et al.  Deep learning in neural networks: An overview , 2014, Neural Networks.

[29]  H. Tenhu,et al.  Non-ionic Thermoresponsive Polymers in Water , 2010 .

[30]  T. W. Anderson,et al.  An Introduction to Multivariate Statistical Analysis , 1959 .

[31]  Richard Hoogenboom,et al.  Continuous poly(2-oxazoline) triblock copolymer synthesis in a microfluidic reactor cascade. , 2015, Chemical communications.

[32]  Martin Kröger,et al.  Poly(N-isopropylacrylamide) Phase Diagrams: Fifty Years of Research. , 2015, Angewandte Chemie.

[33]  Gerbrand Ceder,et al.  Rapid Photovoltaic Device Characterization through Bayesian Parameter Estimation , 2017 .

[34]  Rainer Storn,et al.  Differential Evolution – A Simple and Efficient Heuristic for global Optimization over Continuous Spaces , 1997, J. Glob. Optim..

[35]  Ryan P. Adams,et al.  Design of efficient molecular organic light-emitting diodes by a high-throughput virtual screening and experimental approach. , 2016, Nature materials.

[36]  Daniel Huster,et al.  Amyloid Beta Aggregation in the Presence of Temperature-Sensitive Polymers , 2016, Polymers.

[37]  Alán Aspuru-Guzik,et al.  Automatic Chemical Design Using a Data-Driven Continuous Representation of Molecules , 2016, ACS central science.

[38]  Alán Aspuru-Guzik,et al.  Neural Networks for the Prediction of Organic Chemistry Reactions , 2016, ACS central science.

[39]  Arun Mannodi-Kanakkithodi,et al.  Machine Learning Strategy for Accelerated Design of Polymer Dielectrics , 2016, Scientific Reports.

[40]  Sanket A. Deshmukh,et al.  Machine-Learning Enabled New Insights into the Coil-to-Globule Transition of Thermosensitive Polymers Using a Coarse-Grained Model. , 2018, The journal of physical chemistry letters.

[41]  Ulrich S. Schubert,et al.  The influence of polymer architecture on in vitro pDNA transfection. , 2015, Journal of materials chemistry. B.

[42]  Trevor Hastie,et al.  The Elements of Statistical Learning , 2001 .

[43]  Paul Raccuglia,et al.  Machine-learning-assisted materials discovery using failed experiments , 2016, Nature.

[44]  David Paramelle,et al.  Photothermally responsive gold nanoparticle conjugated polymer-grafted porous hollow silica nanocapsules. , 2016, Chemical communications.

[45]  Françoise M. Winnik,et al.  Versatile Synthesis of End-Functionalized Thermosensitive Poly(2-isopropyl-2-oxazolines) , 2004 .

[46]  Martin Fechner,et al.  Best bang for your buck: GPU nodes for GROMACS biomolecular simulations , 2015, J. Comput. Chem..

[47]  Baohui Li,et al.  Phase Diagram of Poly(ethylene oxide) and Poly(propylene oxide) Triblock Copolymers in Aqueous Solutions , 2006 .