Machine learning (ML)-based protocol for selecting highly relevant papers, extracting important experimental data, and analyzing research trends & directions focusing on the field of CO 2 reduction reactions (CO 2 RRs).

The materials science community has been increasingly interested in harnessing the power of deep learning to solve various domain challenges. However, despite their effectiveness in building highly predictive models, e.g., predicting material properties from microstructure imaging, due to their opaque nature fundamental challenges exist in extracting meaningful domain knowledge from the deep neural networks. In this work, we propose a technique for interpreting the behavior of deep learning models by injecting domain-specific attributes as tunable "knobs" in the material optimization analysis pipeline. By incorporating the material concepts in a generative modeling framework, we are able to explain what structure-to-property linkages these black-box models have learned, which provides scientists with a tool to leverage the full potential of deep learning for domain discoveries.

The exponential growth and success of machine learning (ML) has resulted in its application in all scientific domains including material science. Advancement in experimental techniques has led to an increase in the volume of material science data encouraging material scientists to investigate data-driven solutions to scientific problems. While the resources available to get started with ML are ever increasing, there is little literature on traversing through the space of decisions that need to be made to implement a robust and trustworthy ML solution. A lack of such resources leads to researchers wading through articles and papers trying to determine the best approach for their problem and sometimes also falling prey to pitfalls in a real-world scenario. This paper aims to act as a guide for researchers who want to strategically approach a ML solution to their problem through the use of domain knowledge and systematic evaluation of the major aspects of a ML pipeline. We focus on four aspects of the ML pipeline: (1) problem formulation, (2) data curation, (3) feature representation and model selection, and (4) model generalizability and real-world performance. In each case, we discuss the space of decisions, provide examples from scientific literature, and illustrate how different choices can affect the outcome through a case study of predicting compressive strength of uniaxially pressed molecular solid, 2,4,6-triamino-1,3,5-trinitrobenzene (TATB) samples. Using a similar approach of critical thinking along with rigorous evaluation and diagnostics, researchers can be assured of the reliability of predictions from their ML models.

Machine learning models are increasingly used in materials studies because of their exceptional accuracy. However, the most accurate machine learning models are usually difficult to explain. Remedies to this problem lie in explainable artificial intelligence (XAI), an emerging research field that addresses the explainability of complicated machine learning models like deep neural networks (DNNs). This article attempts to provide an entry point to XAI for materials scientists. Concepts are defined to clarify what explain means in the context of materials science. Example works are reviewed to show how XAI helps materials science research. Challenges and opportunities are also discussed.

Programmable positioning of 2 μm polystyrene (PS) beads with single particle precision and location selective, "on-demand", particle deposition was demonstrated by utilizing patterned electrodes and electrophoretic deposition (EPD). An electrode with differently sized hole patterns, from 0.5 to 5 μm, was used to illustrate the discriminatory particle deposition events based on the voltage and particle-to-hole size ratio. With decreasing patterned hole size, a larger electric field was required for a particle deposition event to occur in that hole. For the 5 μm hole, particle deposition began to occur at 10 V/cm where as an electric field of 15 V/cm was required for particles to begin depositing in the 2 μm holes. The likelihood of particle depositions continued to increase for smaller sized holes as the electric field increased. Eventually, a monolayer of particles began to form at approximately 20 V/cm. In essence, a voltage threshold was found for each hole pattern of different sizes, allowing fine adjustments in pattern hole size and voltage to control when a particle deposition event took place, even with the patterns on the same electrode. This phenomenon opens a route toward controlled, multimaterial deposition and assembly onto substrates without repatterning of the electrode or complicated surface modification of the particles. An analytical approach using the theories for electrophoresis and dielectrophoresis found the former to be the dominating force for depositing a particle into a patterned hole. Ebeam lithography was used to pattern spherical holes in precise configurations onto electrode surfaces, where each hole accompanied a polystyrene (PS) particle placement and attachment during EPD. The versatility of e-beam lithography was utilized to create arbitrary pattern configurations to fabricate particle assemblies of limitless configurations, enabling fabrication of unique materials assemblies and interfaces.

Despite the growing interest in generative adversarial networks (GANs), training GANs remains a challenging problem, both from a theoretical and a practical standpoint. To address this challenge, in this paper, we propose a novel way to exploit the unique geometry of the real data, especially the manifold information. More specifically, we design a method to regularize GAN training by adding an additional regularization term referred to as manifold regularizer. The manifold regularizer forces the generator to respect the unique geometry of the real data manifold and generate high quality data. Furthermore, we theoretically prove that the addition of this regularization term in any class of GANs including DCGAN and Wasserstein GAN leads to improved performance in terms of generalization, existence of equilibrium, and stability. Preliminary experiments show that the proposed manifold regularization helps in avoiding mode collapse and leads to stable training.

To expedite new molecular compound development, a long-sought goal within the chemistry community has been to predict molecules' bulk properties of interest a priori to synthesis from a chemical structure alone. In this work, we demonstrate that machine learning methods can indeed be used to directly learn the relationship between chemical structures and bulk crystalline properties of molecules, even in the absence of any crystal structure information or quantum mechanical calculations. We focus specifically on a class of organic compounds categorized as energetic materials called high explosives (HE) and predicting their crystalline density. An ongoing challenge within the chemistry machine learning community is deciding how best to featurize molecules as inputs into machine learning models-whether expert handcrafted features or learned molecular representations via graph-based neural network models-yield better results and why. We evaluate both types of representations in combination with a number of machine learning models to predict the crystalline densities of HE-like molecules curated from the Cambridge Structural Database, and we report the performance and pros and cons of our methods. Our message passing neural network (MPNN) based models with learned molecular representations generally perform best, outperforming current state-of-the-art methods at predicting crystalline density and performing well even when testing on a data set not representative of the training data. However, these models are traditionally considered black boxes and less easily interpretable. To address this common challenge, we also provide a comparison analysis between our MPNN-based model and models with fixed feature representations that provides insights as to what features are learned by the MPNN to accurately predict density.

We recently developed a deep learning method that can determine the critical peak stress of a material by looking at scanning electron microscope (SEM) images of the material's crystals. However, it has been somewhat unclear what kind of image features the network is keying off of when it makes its prediction. It is common in computer vision to employ an explainable AI saliency map to tell one what parts of an image are important to the network's decision. One can usually deduce the important features by looking at these salient locations. However, SEM images of crystals are more abstract to the human observer than natural image photographs. As a result, it is not easy to tell what features are important at the locations which are most salient. To solve this, we developed a method that helps us map features from important locations in SEM images to non-abstract textures that are easier to interpret.

A molecular crystal structure prediction workflow, based on the random structure generator, Genarris, and the genetic algorithm (GA), GAtor, is successfully applied to two energetic materials and a chiral arene.

Machine-learning (ML) techniques hold the potential of enabling efficient quantitative micrograph analysis, but the robustness of ML models with respect to real-world micrograph quality variations has not been carefully evaluated. We collected thousands of scanning electron microscopy (SEM) micrographs for molecular solid materials, in which image pixel intensities vary due to both the microstructure content and microscope instrument conditions. We then built ML models to predict the ultimate compressive strength (UCS) of consolidated molecular solids, by encoding micrographs with different image feature descriptors and training a random forest regressor, and by training an end-to-end deep-learning (DL) model. Results show that instrument-induced pixel intensity signals can affect ML model predictions in a consistently negative way. As a remedy, we explored intensity normalization techniques. It is seen that intensity normalization helps to improve micrograph data quality and ML model robustness, but microscope-induced intensity variations can be difficult to eliminate.

Gaussian Processes (GPs) are known to provide accurate predictions and uncertainty estimates even with small amounts of labeled data by capturing similarity between data points through their kernel function. However traditional GP kernels are not very effective at capturing similarity between high dimensional data points. Neural networks can be used to learn good representations that encode intricate structures in high dimensional data, and can be used as inputs to the GP kernel. However the huge data requirement of neural networks makes this approach ineffective in small data settings. To solves the conflicting problems of representation learning and data efficiency, we propose to learn deep kernels on probabilistic embeddings by using a probabilistic neural network. Our approach maps high-dimensional data to a probability distribution in a low dimensional subspace and then computes a kernel between these distributions to capture similarity. To enable end-to-end learning, we derive a functional gradient descent procedure for training the model. Experiments on a variety of datasets show that our approach outperforms the state-of-the-art in GP kernel learning in both supervised and semi-supervised settings. We also extend our approach to other small-data paradigms such as few-shot classification where it outperforms previous approaches on mini-Imagenet and CUB datasets.

In this paper, we leverage predictive uncertainty of deep neural networks to answer challenging questions material scientists usually encounter in machine learning-based material application workflows. First, we show that by leveraging predictive uncertainty, a user can determine the required training data set size to achieve a certain classification accuracy. Next, we propose uncertainty-guided decision referral to detect and refrain from making decisions on confusing samples. Finally, we show that predictive uncertainty can also be used to detect out-of-distribution test samples. We find that this scheme is accurate enough to detect a wide range of real-world shifts in data, e.g., changes in the image acquisition conditions or changes in the synthesis conditions. Using microstructure information from scanning electron microscope (SEM) images as an example use case, we show that leveraging uncertainty-aware deep learning can significantly improve the performance and dependability of classification models.

While Deep Neural Networks (DNNs) achieve state-of-the-art accuracy in various applications, they often fall short in accurately estimating their predictive uncertainty and, in turn, fail to recognize when these predictions may be wrong. Several uncertainty-aware models, such as Bayesian Neural Network (BNNs) and Deep Ensembles have been proposed in the literature for quantifying predictive uncertainty. However, research in this area has been largely confined to the big data regime. In this work, we show that the uncertainty estimation capability of state-of-the-art BNNs and Deep Ensemble models degrades significantly when the amount of training data is small. To address the issue of accurate uncertainty estimation in the small-data regime, we propose a probabilistic generalization of the popular sample-efficient non-parametric kNN approach. Our approach enables deep kNN classifier to accurately quantify underlying uncertainties in its prediction. We demonstrate the usefulness of the proposed approach by achieving superior uncertainty quantification as compared to state-of-the-art on a real-world application of COVID-19 diagnosis from chest X-Rays. Our code is available at this https URL

The scientific community has been increasingly interested in harnessing the power of deep learning to solve various domain challenges. However, despite the effectiveness in building predictive models, fundamental challenges exist in extracting actionable knowledge from the deep neural network due to their opaque nature. In this work, we propose techniques for exploring the behavior of deep learning models by injecting domain-specific actionable concepts as tunable ``knobs'' in the analysis pipeline. By incorporating the domain knowledge with generative modeling, we are not only able to better understand the behavior of these black-box models, but also provide scientists with actionable insights that can potentially lead to fundamental discoveries.

Given the emergence of data science and machine learning throughout all aspects of society, but particularly in the scientific domain, there is increased importance placed on obtaining data. Data in materials science are particularly heterogeneous, based on the significant range in materials classes that are explored and the variety of materials properties that are of interest. This leads to data that range many orders of magnitude, and these data may manifest as numerical text or image-based information, which requires quantitative interpretation. The ability to automatically consume and codify the scientific literature across domains—enabled by techniques adapted from the field of natural language processing—therefore has immense potential to unlock and generate the rich datasets necessary for data science and machine learning. This review focuses on the progress and practices of natural language processing and text mining of materials science literature and highlights opportunities for extracting additional information beyond text contained in figures and tables in articles. We discuss and provide examples for several reasons for the pursuit of natural language processing for materials, including data compilation, hypothesis development, and understanding the trends within and across fields. Current and emerging natural language processing methods along with their applications to materials science are detailed. We, then, discuss natural language processing and data challenges within the materials science domain where future directions may prove valuable.

This paper studies the problem of post-hoc calibration of machine learning classifiers. We introduce the following desiderata for uncertainty calibration: (a) accuracy-preserving, (b) data-efficient, and (c) high expressive power. We show that none of the existing methods satisfy all three requirements, and demonstrate how Mix-n-Match calibration strategies (i.e., ensemble and composition) can help achieve remarkably better data-efficiency and expressive power while provably maintaining the classification accuracy of the original classifier. Mix-n-Match strategies are generic in the sense that they can be used to improve the performance of any off-the-shelf calibrator. We also reveal potential issues in standard evaluation practices. Popular approaches (e.g., histogram-based expected calibration error (ECE)) may provide misleading results especially in small-data regime. Therefore, we propose an alternative data-efficient kernel density-based estimator for a reliable evaluation of the calibration performance and prove its asymptotically unbiasedness and consistency. Our approaches outperform state-of-the-art solutions on both the calibration as well as the evaluation tasks in most of the experimental settings. Our codes are available at this https URL.

Deep Neural Networks (DNNs) are known to make highly overconfident predictions on Outof-Distribution data. Recent research has shown that uncertainty-aware models, such as, Bayesian Neural Network (BNNs) and Deep Ensembles, are less susceptible to this issue. However research in this area has been largely confined to the big data setting. In this work, we show that even state-of-the-art BNNs and Ensemble models tend to make overconfident predictions when the amount of training data is insufficient. This is especially concerning for emerging applications in physical sciences and healthcare where overconfident and inaccurate predictions can lead to disastrous consequences. To address the issue of accurate uncertainty (or confidence) estimation in the small-data regime, we propose a probabilistic generalization of the popular sample-efficient non-parametric kNN approach. We demonstrate the usefulness of the proposed approach on a realworld application of COVID-19 diagnosis from chest X-Rays by (a) highlighting surprising failures of existing techniques, and (b) achieving superior uncertainty quantification as compared to state-of-the-art.

Nanomaterials of varying compositions and morphologies are of interest for many applications from catalysis to optics, but the synthesis of nanomaterials and their scale-up are most often time-consuming and Edisonian processes. Information gleaned from scientific literature can help inform and accelerate nanomaterials development, but again, searching the literature and digesting the information are time-consuming manual processes for researchers. To help address these challenges, we developed scientific article-processing tools that extract and structure information from the text and figures of nanomaterials articles, thereby enabling the creation of a personalized knowledgebase for nanomaterials synthesis that can be mined to help inform further nanomaterials development. Starting with a corpus of ca. 35k nanomaterials-related articles, we developed models to classify articles according to the nanomaterial composition and morphology, extract synthesis protocols from within the articles’ text, and extract, normalize, and categorize chemical terms within synthesis protocols. We demonstrate the efficiency of the proposed pipeline on an expert-labeled set of nanomaterials synthesis articles, achieving 100% accuracy on composition prediction, 95% prediction on morphology prediction, 0.99 AUC on protocol identification, and up to 0.87 F1-score on chemical entity recognition. In addition to processing articles’ text, microscopy images of nanomaterials within articles are also automatically identified and analyzed to determine nanomaterials’ morphologies and size distributions. To enable users to easily explore the database, we developed a complementary browser-based visualization tool that provides flexibility in comparing across subsets of articles of interest. We use these tools and information to identify trends in nanomaterials synthesis, such as the correlation of certain reagents with various nanomaterial morphologies, which is useful in guiding hypotheses and reducing the potential parameter space during experimental design.

Scanning Electron Microscopy (SEM) images provide a variety of structural and morphological information of nanomaterials. In the material informatics domain, automatic recognition and quantitative analysis of SEM images in a high-throughput manner are critical, but challenges still remain due to the complexity and the diversity of image configurations in both shape and size. In this paper, we present a generally applicable approach using computer vision and machine learning techniques to quantitatively extract particle size, size distribution and morphology information in SEM images. The proposed pipeline offers automatic, high-throughput measurements even when overlapping nanoparticles, rod shapes, and core-shell nanostructures are present. We demonstrate effectiveness of the proposed approach by performing experiments on SEM images of nanoscale materials and structures with different shapes and sizes. The proposed approach shows promising results (Spearman coefficients of 0.91 and 0.99 using fully automated and semi-automated processes, respectively) when compared with manually measured sizes. The code is made available as open source software at https://github.com/LLNL/LIST.

Packing motifs-patterns in how molecules orient relative to one another in a crystal structure-are an important concept in many subdisciplines of materials science because of correlations observed between specific packing motifs and properties of interest. That said, packing motif data sets have remained small and noisy due to intensive manual labeling processes and insufficient labeling schemes. The most prominent labeling algorithms calculate relative interplanar angles of nearest neighbor molecules to determine the packing motif of a molecular crystal, but this simple approach can fail when neighbors are naively sampled isotropically around the crystal structure. To remedy this issue, we propose an optimization algorithm, which rotates the molecular crystal structure to find representative molecules that inform the packing motif. We package this algorithm into an automated framework-Autopack-which both optimally rotates the crystal structure and labels the packing motif based on the appropriate neighboring molecules. In this work, we detail the Autopack framework and its performance, which shows improvements compared to previous state-of-the-art labeling methods, providing the first quantitative point of comparison for packing motif labeling algorithms. Furthermore, using Autopack (available at https://ipo.llnl.gov/technologies/software/autopack), we perform the first large-scale study of potential relationships between chemicals' compositions and packing motifs, which shows that these relationships are more complex than previously hypothesized from studies that used only tens of polycyclic aromatic hydrocarbon molecules. Autopack's capabilities help pose next steps for crystal engineering research focusing not only on a molecule's adoption of a specific packing motif but also on new structure-property relationships.