NestedAE: interpretable nested autoencoders for multi-scale materials characterization.

We introduce an interpretable machine learning architecture, NestedAE, for multiscale materials using nested supervised autoencoders. We benchmarked the performance of NestedAE on two databases: (1) a synthetic dataset created from nested analytical functions whose dimensionality is therefore known a priori, and (2) a multiscale MHP dataset that is a combination of an open source dataset containing atomic and ionic properties, and a second dataset containing device characterization using current density-voltage (J-V) analysis. The NestedAE architecture was found to have higher noise robustness and lower reconstruction losses when compared to a vanilla autoencoder (AE). Its application on the MHP dataset revealed links between crystal scale properties and device performance in agreement with earlier experimental observations.

PAL 2.0: a physics-driven bayesian optimization framework for material discovery.

The lack of efficient discovery tools for advanced functional materials remains a major bottleneck to enabling advances in the next-generation energy, health, and sustainability technologies. One main factor contributing to this inefficiency is the large combinatorial space of materials (with respect to material compositions and processing conditions) that is typically redolent of such materials-centric applications. Searches of this large combinatorial space are often influenced by expert knowledge and clustered close to material configurations that are known to perform well, thus ignoring potentially high-performing candidates in unanticipated regions of the composition-space or processing protocol. Moreover, experimental characterization or first principles quantum mechanical calculations of all possible material candidates can be prohibitively expensive, making exhaustive approaches to determine the best candidates infeasible. As a result, there remains a need for the development of computational algorithms that can efficiently search a large parameter space for a given material application. Here, we introduce PAL 2.0, a method that combines a physics-based surrogate model with Bayesian optimization. The key contributing factor of our proposed framework is the ability to create a physics-based hypothesis using XGBoost and Neural Networks. This hypothesis provides a physics-based "prior" (or initial beliefs) to a Gaussian process model, which is then used to perform a search of the material design space. In this paper, we demonstrate the usefulness of our approach on three material test cases: (1) discovery of metal halide perovskites with desired photovoltaic properties, (2) design of metal halide perovskite-solvent pairs that produce the best solution-processed films and (3) design of organic thermoelectric semiconductors. Our results indicate that the novel PAL 2.0 approach outperforms other state-of-the-art methods in its efficiency to search the material design space for the optimal candidate. We also demonstrate the physics-based surrogate models constructed in PAL 2.0 have lower prediction errors for material compositions not seen by the model. To the best of our knowledge, there is no competing algorithm capable of this useful combination for materials discovery, especially those for which data are scarce.

Efficient quasi-classical trajectory calculations by means of neural operator architectures.

An accurate description of non-equilibrium chemistry relies on rovibrational state-to-state (StS) kinetics data, which can be obtained through the quasi-classical trajectory (QCT) method for high-energy collisions. However, these calculations still represent one of the major computational bottlenecks in predictive simulations of non-equilibrium reacting gases. This work addresses this limitation by proposing SurQCT, a novel machine learning-based surrogate for efficiently and accurately predicting StS chemical reaction rate coefficients. The QCT emulator is constructed using three independent components: two deep operator networks (DeepONets) for inelastic and exchange processes and a feed-forward neural network (FNN) for the dissociation reactions. SurQCT is tested on the O2 + O system, showing a computational speed-up of 85%. Furthermore, we carry out a StS master equation analysis of an isochoric, isothermal heat bath simulation at various temperatures to study how the predicted rate coefficients impact the accuracy of multiple quantities of interest (QoIs) at the kinetics level (e.g., global quasi-steady state (QSS) dissociation rate coefficients and energy relaxation times). For all these QoIs, the master equation analysis relying on SurQCT data shows an accuracy within 15% across the entire temperature regime.

Comprehensive Study of HCN: Potential Energy Surfaces, State-to-State Kinetics, and Master Equation Analysis.

Understanding the kinetics of the HCN system is critical to several disciplines in science and engineering, including interstellar chemistry, atmospheric reentry, and combustion, to name a few. This paper constructs a rovibrational state-specific kinetic mechanism for the HCN system, leveraging electronic structure calculations, classical scattering dynamics, and state-to-state kinetics. To this aim, three accurate potential energy surfaces (PESs), 1A', 3A', and 3A″, are constructed using multireference configuration interaction (MRCI) calculations for a comprehensive arrangement of the nuclei. Quasi-classical scattering calculations provide elementary reaction rate constants resulting from the interaction between the CN, CH, and NH molecules with H, N, and C atoms, respectively. The rovibrational collisional model developed comprises 50 million bound-bound and free-bound collisional processes. This model is used to study the dynamics of energy transfer and dissociation in an isochoric and isothermal chemical reactor via the solution of the master equation for a wide temperature range from 1000 to 10,000 K. This study unravels the dynamics of dissociation of the molecules in the HCN system, which the PESs primarily control via the formation of short-lived intermediates that shortcut the dissociation pathway. The exchange processes in CH and NH enhance the dissociation by over 80%. The importance of exchange processes is also highlighted in comparing the quasi-steady state and thermal dissociation rates with state-of-the-art rate models and experimental fits.