Bulletin of the American Physical Society
APS March Meeting 2023
Volume 68, Number 3
Las Vegas, Nevada (March 5-10)
Virtual (March 20-22); Time Zone: Pacific Time
Session S03: Machine Learning, Autonomous Experiments, and Big Data in Polymer Physics IFocus
|
Hide Abstracts |
Sponsoring Units: DPOLY Chair: Peter Beaucage, National Institute of Standards and Technology Room: Room 126 |
Thursday, March 9, 2023 8:00AM - 8:36AM |
S03.00001: Cold, warm, warmer, hot! Impact of distance metrics on autonomous experimentation. Invited Speaker: Lilo Pozzo Artificial intelligence (AI), when interfaced with laboratory automation, can accelerate materials optimization and scientific discovery. For example, it may be used to efficiently map a phase-diagram with intelligent sampling along phase boundaries, or in ‘retrosynthesis’ problems where a material with a target structure is desired but its synthetic route is unknown. These AI-driven laboratories are especially promising in polymer physics, where design parameters (e.g. chemical composition, MW, topology, processing) are vast and where properties and function are intimately tied to design features. However, for AI to operate efficiently in these spaces, they must be ‘encoded’ with domain expertise specific to the problems being tackled. In this talk, we focus on the problem of defining appropriate ‘distance’ metrics to describe differences between functions sampled within a design space. Such functions may be spectroscopic (e.g. UV-Vis absorption, fluorescence, impedance) or scattering profiles (SAXS, SANS) of materials, among others. Traditional ‘distance’ metrics, such as Euclidean and parametric definitions, often fail when important features of the measured functions are subtle and/or when sampling takes place far from the target. We have thus developed a new shape-based similarity metric using Riemannian geometry (Phase-Amplitude Distance) that has been successfully implemented in both retrosynthesis and phase mapping problems. This talk will first discuss the definition of the Phase-Amplitude Distance metric. We then demonstrate its implementation in an autonomous batch retrosynthesis problem using spectroscopic signatures in a model system of metal nanostructures. Finally, we implement the new distance metric in phase-mapping problems involving block-copolymers, polymer blends, and inorganic materials to showcase the broad applicability of the method. Mathematically, these phase maps need to be continuous over the design space and correlations are usually defined by shape-based similarity between profiles. We pose both constraints as a geometric feature of the phase map where continuity is obtained by diffusing the shape-based similarity of SAS profiles via a local geometry defined by the linear operators on the design space. |
Thursday, March 9, 2023 8:36AM - 8:48AM |
S03.00002: Interpreting Neutron Reflectivity from Thin Films of Block Copolymers using Neural Networks Miguel Fuentes-Cabrera, Dustin Eby, Mathieu Doucet, Rajeev Kumar Thin films of ionic polymers are of great technological interests due to their relevance to solid polymer electrolytes and energy storage. There is active research on controlling morphology in these films to modify ion transport, a process affected by adsorption of ionic groups on conductive surfaces. Visualization of these adsorbed layers (~5-10 nm) is not a trivial task and requires characterization tools capable of capturing vertical and lateral structure in thin films. Grazing incidence neutron scattering is one such tool, which can provide information about vertical (specular reflection) and lateral structures (Off-specular scattering (OSS)) of adsorbed layers of ionic polymers. |
Thursday, March 9, 2023 8:48AM - 9:00AM |
S03.00003: The Autonomous Formulation Laboratory: Macromolecular Formulation Discovery with Multimodal Measurements Peter Beaucage, Tyler B Martin Liquid formulations are ubiquitous in products ranging from deicing liquids and fuels/lubricants to biologic drugs, shampoo, and food/beverage ingredients. All these products require precise tuning of 10s-100s of components to produce a desired product: viscosity modifiers, surfactants, dyes, fragrances, flammability inhibitors, etc. While scattering methods (SAXS, SANS, WAXS) are workhorse techniques for characterizing model formulations, they have not been widely used to characterize real products, largely because the large number of components often precludes rational mapping between component fractions, structure, and product stability. Multimodal characterization and machine learning (ML) tools promise to greatly reduce the expense of exploring the stability boundaries of a particular, desirable phase in highly multicomponent products. This talk will describe the development of the Autonomous Formulation Laboratory, a highly adaptable platform capable of autonomously synthesizing and characterizing liquid mixtures with varying composition and chemistry using x-ray and neutron scattering in concert with other measurements. Highlights will include our development of model-free analysis approaches for x-ray scattering data, studies of surfactant micelle extension, and closed-loop AI-guided exploration of composite nanoparticle synthesis for coatings development. I will further discuss our ongoing efforts to incorporate a multimodal suite of secondary measurements such as optical imaging, UV-vis-NIR and capillary rheometry to provide greater-than-sum-of-parts materials characterization. |
Thursday, March 9, 2023 9:00AM - 9:12AM |
S03.00004: Combining Flory-Huggins Theory and Machine Learning for Improved Polymer Solution Phase Behavior Predictions Jeffrey G Ethier, Debra J Audus, Devin C Ryan, Richard A Vaia Polymer solution macrophase separation predicted by Flory-Huggins theory often shows disagreement with experiment. Improvements rely on empirical parameters fit to modified Flory-Huggins equations that require extensive experimental data, complex optimization, and are ultimately chemistry-dependent (i.e., useful for a small range of polymer solutions). Recently, machine learning (ML) models trained on experimental data have shown predictability of the phase boundary to within 1-3oC across a broad range of polymer-solvent systems and extrapolation to new systems with as little as 20 additional data points. However, the physics underlying these ML predictions are obscured, which hinders human interpretability and trust. In this work, we build a framework combining Flory-Huggins theory with ML models to (i) improve predictability and generalization with less experimental data, and (ii) provide interpretability. To achieve this, we estimate theoretical, molecular weight dependent parameters using neural networks. Using polystyrene-cyclohexane as a focus, we compare the ability of these hybrid theory-ML models relative to current theory and traditional ML to predict macrophase separation, and interpolate to new molecular weights as a function of training data set size. Our framework showcases the benefits of physics-interpretable ML models for polymers. |
Thursday, March 9, 2023 9:12AM - 9:24AM |
S03.00005: Application of Deep Learning to Polymer Solutions Ryan Sayko, Michael S Jacobs, Marissa Dominijanni, Andrey V Dobrynin Characterizations of the molecular interactions in polymer solutions is a fundamental problem of polymer physics. We developed a framework that utilizes a scaling relationship between solution correlation length ξ=lgν/B and number of monomers g per correlation volume for chains with monomer projection length l and a deep learning approach for evaluation of the B-parameters. The values of Bg and Bth corresponding to exponents ν = 0.588 and 0.5 uniquely describe a solvent quality for the polymer backbone. Applying a convolutional neural network (CNN), we obtained the set {Bg, Bth,} from solution specific viscosity, ηsp, as a function of concentration, c. The CNN was trained by generating a large number of sparse images representing the normalized specific viscosity ηsp/Nw(cl3)1/(3ν-1) in solutions of chains with the weight-average degree of polymerization, Nw. This approach is capable of predicting the B-parameters with a mean absolute percentage error less than 6%. The calculated B-parameters were used to obtain the packing number, Pe and to predict the onset of entanglements in solutions of synthetic polymers and polysaccharides in water, organic solvents, and ionic liquids. |
Thursday, March 9, 2023 9:24AM - 9:36AM |
S03.00006: Sequence, phase behavior and dynamics in protein condensates: an eternal triangle revealed by machine learning Michael A Webb While biomolecular condensates have been identified in a wide range of intracellular compartments, debate continues over the utility of such “droplets” and the connection between sequence characteristics, phase behavior, and condensate dynamics. Prior experimental observations and physical arguments suggest that the phase-separating propensity of a protein is strongly correlated to the dynamics in a protein-rich phase. Inspired by this work, we endeavor to understand to what extent we can exploit sequence patterns to break this correlation and what are the characteristics of any sequences that “break the mold.” To navigate the complexity in both sequence design and understanding the characteristics of designed condensates, we heavily rely on and demonstrate the utility of techniques from machine learning. In aggregate, we hope to augment our understanding of existing biological systems and importantly highlight opportunities for functional soft materials design, including phase-separating protein condensates with “tunable” dynamical properties. |
Thursday, March 9, 2023 9:36AM - 9:48AM |
S03.00007: Quantitative high-throughput measurement of bulk mechanical properties using commonly available equipment Muzhou Wang, Justin Griffith, Yusu Chen, Qingsong Liu, Qifeng Wang, Jeffrey J Richards, Danielle Tullman-Ercek, Kenneth R Shull In the era of Big Data and Machine Learning, it remains a large challenge to measure mechanical properties at high throughput, as these assays often require custom or expensive equipment or can only measure a few samples at a time. In this work, we explore mechanical properties that can be measured using a novel high-throughput colorimetric method using a common laboratory centrifuge, multiwell plates, and microparticles. The use of centrifugation is key, as it enables the application of a homogeneous mechanical force across many samples in a multiwell plate, that can be arranged in a completely arbitrary fashion. To measuring bulk mechanical properties of soft materials, we embed microparticles inside samples that are loaded within multiwell plates, and then we determine the centrifugal speed that enables the particles to break out of the materials. These results were then correlated to mechanical properties such as modulus and yield strength, and we establish quantitative agreement with more standard one-at-a-time test methods through experimentation and analytical solutions of the underlying mechanics. We then demonstrate the throughput of our method, which is limited only by the number of wells in the plates. |
Thursday, March 9, 2023 9:48AM - 10:00AM |
S03.00008: Predicting microstructure of a polymer nanocomposite using machine learning Tarak K Patra, Kumar Ayush Polymer nanocomposites (PNCs) offer a broad range of thermophysical properties that are linked to their compositions. However, it is challenging to establish a universal composition-property relation of PNCs due to their enormous composition and chemical space. Here we address this problem and develop a new method to model the composition-microstructure relation of a PNC through a machine learning pipeline, named as nanoNET. The inputs to the model are five composition parameters viz., NP size, polymer chain length, NP-polymer interaction strength, NP-NP interaction, and NP concentration of a PNC. The output is radial distribution function (RDF) of NPs in the polymer matrix. We conduct molecular dynamics simulations of a model PNC within a carefully selected range of compositional parameters, which span a wide region of the phase space, including experimentally known phases, and utilize the data to establish and validate the nanoNET. Within this framework, the grayscale images of NPs RDF in a polymer matrix are encoded to a latent space using a convolutional neural network (CNN) autoencoder. Subsequently, a random forest regressor establishes a correlation between the composition of the PNC and the latent space representation of its NPs' RDF. The nanoNET predicts NPs distribution in many unknown PNCs very accurately. This method is very generic and can accelerate the design, discovery, and fundamental understanding of composition-microstructure relations of PNCs and other molecular systems. |
Thursday, March 9, 2023 10:00AM - 10:12AM |
S03.00009: Fast and Accurate Prediction of Polymer Viscoelasticity via Physics-Based Ensemble Learning Umi Yamamoto, Kenji Yoshimoto We present a machine learning framework to construct a predictive model for dynamic moduli of linear entangled homopolymers. Using well accepted computational data as a training data set, it is shown that a straightforward supervised learning with standard algorithms (support vector machine, kernel ridge regression, etc.) provides reasonable prediction accuracy when an input parameter, the number of entanglements in the present case, is extrapolated. However, the predictive power is further improved to a non-trivial extent by integrating a polymer physic s idea with the learning procedure. Namely, by constructing individual predictive models that specialize in the representation of the distinct relaxation behavior at short, intermediate and long time scale, and merging them into a single model using a frequency dependent ensemble method, we can predict the storage and loss modulus in quantitative agreement with the training data. |
Thursday, March 9, 2023 10:12AM - 10:24AM |
S03.00010: Predicting the Glass Transition of Complex Polymers via Integration of Machine Learning, Theory and Molecular Modeling Wenjie Xia Semiconducting conjugated polymers (CPs) are attractive organic electronic materials for a wide range of applications due to their unique properties such as easy processability, tunable electrical performance, and mechanical flexibility. Despite tremendous efforts, design and prediction of Tg remain notably challenging for CPs due to their complex chain architecture associated with diverse chemical building blocks. In this work, we establish an integrated framework based on machine learning (ML) and molecular simulations to predict Tg for a diverse set of CPs and other polymers with a drastic difference in their chemical structures. Informed from informatics and molecular theory, the developed ML model takes the geometry of diverse chemical building block to define simplified structural features to make Tg prediction, which is further validated by experimental measurement. Moreover, the use of molecular modeling and theory in conjunction with ML uncovers the critical roles of key molecular features in influencing the glass transition temperature as well as dynamics heterogeneity associated with glass formation of complex polymers. The established predictive framework and ML model could be ready to use for design of high-performance CPs and relevant materials via molecular engineering. |
Thursday, March 9, 2023 10:24AM - 10:36AM |
S03.00011: Machine learning-assisted discovery of high-performance polymer membranes for gas separation Jiaxin Xu, Agboola Suleiman, Gang Liu, Meng Jiang, Ruilan Guo, Tengfei Luo High-performance polymer membranes have achieved remarkable success in gas separations. Contrary to the traditional Edisonian trial-and-error approach, the growing machine learning (ML) technique possesses great promise to accelerate the discovery and development of innovative polymer membrane materials yet is obstructed by the insufficiency and label imbalance of training data. We demonstrate the success of a state-of-the-art semi-supervised graph regression framework leveraging unlabeled polymers to improve the performance of ML models even if only very limited and imbalanced permeability data are available based on an open-source polymer gas permeability experimental database for six major industrial gases. Using the trained models, gas permeability prediction is conducted on over 12,000 existing unlabeled homopolymer candidates and the prediction accuracy is tested by experimentally synthesizing two of the most promising polymer membranes, which are found to possess extraordinary H2/CH4, H2/N2, and O2/N2 separation performance, representing an advanced way to explore the unknown chemical space for high-performance gas separation polymer membrane design. |
Follow Us |
Engage
Become an APS Member |
My APS
Renew Membership |
Information for |
About APSThe American Physical Society (APS) is a non-profit membership organization working to advance the knowledge of physics. |
© 2024 American Physical Society
| All rights reserved | Terms of Use
| Contact Us
Headquarters
1 Physics Ellipse, College Park, MD 20740-3844
(301) 209-3200
Editorial Office
100 Motor Pkwy, Suite 110, Hauppauge, NY 11788
(631) 591-4000
Office of Public Affairs
529 14th St NW, Suite 1050, Washington, D.C. 20045-2001
(202) 662-8700