Materials Acceleration Platforms (MAPs) Accelerating Materials Research and Development to Meet Urgent Societal Challenges.

Climate Change and Materials Criticality challenges are driving urgent responses from global governments. These global responses drive policy to achieve sustainable, resilient, clean solutions with Advanced Materials (AdMats) for industrial supply chains and economic prosperity. The research landscape comprising industry, academe, and government identified a critical path to accelerate the Green Transition far beyond slow conventional research through Digital Technologies that harness Artificial Intelligence, Smart Automation and High Performance Computing through Materials Acceleration Platforms, MAPs. In this perspective, following the short paper, a broad overview about the challenges addressed, existing projects and building blocks of MAPs will be provided while concluding with a review of the remaining gaps and measures to overcome them.

Deep learning-enhanced characterization of bubble dynamics in proton exchange membrane water electrolyzers.

The ever-increasing utility of imaging technology in proton exchange membrane water electrolyzer research raises the demand for rapid and precise image analysis. In particular, for optical video recordings, the challenge primarily lies in the large number of frames that impede the delineation of bubble dynamics with standard methods. In order to address this problem, the present study supports the automation of data analysis to facilitate swift, comprehensive, and measurable insights from captured imagery. We present a deep learning-based framework to perform high-throughput analyses of bubble dynamics using optical images of proton exchange membrane water electrolyzers. Leveraging a relatively small annotated imaging dataset of just 35 images, various configurations of the U-Net architecture were trained to perform bubble segmentation tasks. The best model achieved a precision of 95%, a recall of 78%, and an F1-score of 86% on the validation set. Subsequent to segmentation, the methodology enabled the rapid extraction of parameters such as time-resolved bubble area, size distributions, bubble position probability density, and individual bubble shape analytics. The findings underscore the potential of deep learning to enhance the analysis of polymer electrolyte membrane water electrolyzer imaging, offering a path toward more efficient and informative evaluations in electrochemical research.

UTILE-Gen: Automated Image Analysis in Nanoscience Using Synthetic Dataset Generator and Deep Learning.

This work presents the development and implementation of a deep learning-based workflow for autonomous image analysis in nanoscience. A versatile, agnostic, and configurable tool was developed to generate instance-segmented imaging datasets of nanoparticles. The synthetic generator tool employs domain randomization to expand the image/mask pairs dataset for training supervised deep learning models. The approach eliminates tedious manual annotation and allows training of high-performance models for microscopy image analysis based on convolutional neural networks. We demonstrate how the expanded training set can significantly improve the performance of the classification and instance segmentation models for a variety of nanoparticle shapes, ranging from spherical-, cubic-, to rod-shaped nanoparticles. Finally, the trained models were deployed in a cloud-based analytics platform for the autonomous particle analysis of microscopy images.

Deep learning for the automation of particle analysis in catalyst layers for polymer electrolyte fuel cells.

The rapidly growing use of imaging infrastructure in the energy materials domain drives significant data accumulation in terms of their amount and complexity. The applications of routine techniques for image processing in materials research are often ad hoc, indiscriminate, and empirical, which renders the crucial task of obtaining reliable metrics for quantifications obscure. Moreover, these techniques are expensive, slow, and often involve several preprocessing steps. This paper presents a novel deep learning-based approach for the high-throughput analysis of the particle size distributions from transmission electron microscopy (TEM) images of carbon-supported catalysts for polymer electrolyte fuel cells. A dataset of 40 high-resolution TEM images at different magnification levels, from 10 to 100 nm scales, was annotated manually. This dataset was used to train the U-Net model, with the StarDist formulation for the loss function, for the nanoparticle segmentation task. StarDist reached a precision of 86%, recall of 85%, and an F1-score of 85% by training on datasets as small as thirty images. The segmentation maps outperform models reported in the literature for a similar problem, and the results on particle size analyses agree well with manual particle size measurements, albeit at a significantly lower cost.

Convolutional neural networks for high throughput screening of catalyst layer inks for polymer electrolyte fuel cells.

The performance of polymer electrolyte fuel cells decisively depends on the structure and processes in membrane electrode assemblies and their components, particularly the catalyst layers. The structural building blocks of catalyst layers are formed during the processing and application of catalyst inks. Accelerating the structural characterization at the ink stage is thus crucial to expedite further advances in catalyst layer design and fabrication. In this context, deep learning algorithms based on deep convolutional neural networks (ConvNets) can automate the processing of the complex and multi-scale structural features of ink imaging data. This article presents the first application of ConvNets for the high throughput screening of transmission electron microscopy images at the ink stage. Results indicate the importance of model pre-training and data augmentation that works on multiple scales in training robust and accurate classification pipelines.

Three-Dimensional Cathodes for Electrochemical Reduction of CO2: From Macro- to Nano-Engineering.

Rising anthropogenic CO2 emissions and their climate warming effects have triggered a global response in research and development to reduce the emissions of this harmful greenhouse gas. The use of CO2 as a feedstock for the production of value-added fuels and chemicals is a promising pathway for development of renewable energy storage and reduction of carbon emissions. Electrochemical CO2 conversion offers a promising route for value-added products. Considerable challenges still remain, limiting this technology for industrial deployment. This work reviews the latest developments in experimental and modeling studies of three-dimensional cathodes towards high-performance electrochemical reduction of CO2. The fabrication-microstructure-performance relationships of electrodes are examined from the macro- to nanoscale. Furthermore, future challenges, perspectives and recommendations for high-performance cathodes are also presented.

Virtual Materials Intelligence for Design and Discovery of Advanced Electrocatalysts.

Similar to advancements gained from big data in genomics, security, internet of things, and e-commerce, the materials workflow could be made more efficient and prolific through advances in streamlining data sources, autonomous materials synthesis, rapid characterization, big data analytics, and self-learning algorithms. In electrochemical materials science, data sets are large, unstructured/heterogeneous, and difficult to process and analyze from a single data channel or platform. Computer-aided materials design together with advances in data mining, machine learning, and predictive analytics are expected to provide inexpensive and accelerated pathways towards tailor-made functionally optimized energy materials. Fundamental research in the field of electrochemical energy materials focuses primarily on complex interfacial phenomena and kinetic electrocatalytic processes. This perspective article critically assesses AI-driven modeling and computational approaches that are currently applied to those objects. An application-driven materials intelligence platform is introduced, and its functionalities are scrutinized considering the development of electrocatalyst materials for CO2 conversion as a use case.

Inhomogeneous transport in model hydrated polymer electrolyte supported ultrathin films.

The structure of polymer electrolyte membranes, e.g., Nafion, inside fuel cell catalyst layers has significant impact on the electrochemical activity and transport phenomena that determine cell performance. In those regions, Nafion can be found as an ultrathin film, coating the catalyst and the catalyst support surfaces. The impact of the hydrophilic/hydrophobic character of these surfaces on the structural formation of the films and, in turn, on transport properties has not been sufficiently explored yet. Here, we report classical molecular dynamics simulations of hydrated Nafion thin films in contact with unstructured supports, characterized by their global wetting properties only. We have investigated structure and transport in different regions of the film and found evidence of strongly heterogeneous behavior. We speculate about the implications of our work on experimental and technological activity.

Microstructure-based modeling of aging mechanisms in catalyst layers of polymer electrolyte fuel cells.

This work is comprised of a versatile multiscale modeling of carbon corrosion processes in catalyst layers (CLs) of polymer electrolyte fuel cells (PEFCs). Slow rates of electrocatalytic processes in CLs and materials aging are the main sources of voltage loss in PEFCs under realistic operating conditions. We combined microstructure data obtained from coarse-grained molecular dynamics (CGMD) simulations with a detailed description of the nanoscale elementary kinetic processes and electrochemical double-layer effects at the catalyst/electrolyte and carbon/electrolyte interfaces. We exclusively focused on morphology and microstructure changes in the catalyst layer of PEFCs as a result of carbon corrosion. By employing extensive CGMD simulations, we analyzed the microstructure of CLs as a function of carbon loss and in view of ionomer and water morphology, water and ionomer coverage, and overall changes in carbon surface. These ingredients are integrated into a kinetic model, which allows capture of the impact of the structural changes on the PEFC performance decay. In principle, such multiscale simulation studies allow a relation of the aging of CLs to the selection of carbon particles (sizes and wettability), the catalyst loading, and the level of ionomer structural changes during the CL degradation process.

PEMFC catalyst layers: the role of micropores and mesopores on water sorption and fuel cell activity.

The effects of carbon microstructure and ionomer loading on water vapor sorption and retention in catalyst layers (CLs) of PEM fuel cells are investigated using dynamic vapor sorption. Catalyst layers based on Ketjen Black and Vulcan XC-72 carbon blacks, which possess distinctly different surface areas, pore volumes, and microporosities, are studied. It is found that pores <20 nm diameter facilitate water uptake by capillary condensation in the intermediate range of relative humidities. A broad pore size distribution (PSD) is found to enhance water retention in Ketjen Black-based CLs whereas the narrower mesoporous PSD of Vulcan CLs is shown to have an enhanced water repelling action. Water vapor sorption and retention properties of CLs are correlated to electrochemical properties and fuel cell performance. Water sorption enhances electrochemical properties such as the electrochemically active surface area (ESA), double layer capacitance and proton conductivity, particularly when the ionomer content is very low. The hydrophilic properties of a CL on the anode and the cathode are adjusted by choosing the PSD of carbon and the ionomer content. It is shown that a reduction of ionomer content on either cathode or anode of an MEA does not necessarily have a significant detrimental effect on the MEA performance compared to the standard 30 wt % ionomer MEA. Under operation in air and high relative humidity, a cathode with a narrow pore size distribution and low ionomer content is shown to be beneficial due to its low water retention properties. In dry operating conditions, adequate ionomer content on the cathode is crucial, whereas it can be reduced on the anode without a significant impact on fuel cell performance.

On the micro-, meso-, and macroporous structures of polymer electrolyte membrane fuel cell catalyst layers.

In this work, N(2) adsorption was employed to investigate the effects of carbon support, platinum, and ionomer loading on the microstructure of polymer electrolyte membrane fuel cell catalyst layers (CLs). Brunauer-Emmett-Teller and t-plot analyses of adsorption isotherms and pore-size distributions were used to study the microstructure of carbon supports, platinum/carbon catalyst powders, and three-component platinum/carbon/ionomer CLs. Two types of carbon supports were chosen for the investigation: Ketjen Black and Vulcan XC-72. CLs with a range of Nafion ionomer loadings were studied in order to evaluate the effect of an ionomer on the CL microstructure. Regions of adsorption were differentiated into micropores associated with the carbon primary particles (<2 nm), mesopores ascribed to the void space inside agglomerates (2-20 nm), and meso- to macroporous space inside aggregates of agglomerates (>50 nm). Ketjen Black was found to possess a significant fraction of micropores, 25% of the total pore volume, in contrast to Vulcan XC-72, for which the corresponding fraction of micropores was 15% of the total pore volume. The microstructure of the carbon support was found to be a significant factor in the formation of the microstructure in the three-component CLs, serving as a rigid porous framework for distribution of platinum and the ionomer. It was found that platinum particle deposition on Ketjen Black occurs in, or at the mouth of, the support's micropores, thus affecting its effective microporosity, whereas platinum deposition on Vulcan XC-72 did not significantly affect the support's microstructure. The codeposition of ionomer in the CL strongly influenced its porosity, covering pores < 20 nm, which are ascribed to the pores within the primary carbon particles (pore sizes < 2 nm) and to the pores within agglomerates of the particles (pore sizes of 2-20 nm).

Molecular dynamics simulations of adsorption and diffusion of gases in silicon-carbide nanotubes.

Silicon carbide nanotubes (SiCNTs) are new materials with excellent properties, such as high thermal stability and mechanical strength, which are much improved over those of their carboneous counterparts, namely, carbon nanotubes (CNTs). Gas separation processes at high temperatures and pressures may be improved by developing mixed-matrix membranes that contain SiCNTs. Such nanotubes are also of interest in other important processes, such as hydrogen production and its storage, as well as separation by supercritical adsorption. The structural parameters of the nanotubes, i.e., their diameter, curvature, and chirality, as well as the interaction strength between the gases and the nanotubes' walls, play a fundamental role in efficient use of the SiCNTs in such processes. We employ molecular dynamics simulations in order to examine the adsorption and diffusion of N(2), H(2), CO(2), CH(4), and n-C(4)H(10) in the SiCNTs, as a function of the pressure and the type of the nanotubes, namely, the zigzag, armchair, and chiral tubes. The simulations indicate the strong effect of the nanotubes' chirality and curvature on the pressure dependence of the adsorption isotherms and the self-diffusivities. Detailed comparison is made between the results and those for the CNTs. In particular, we find that the adsorption capacity of the SiCNTs for hydrogen is higher than the CNTs' under the conditions that we have studied.

Nanophase segregation and water dynamics in hydrated Nafion: molecular modeling and experimental validation.

Reported results of coarse-grained molecular dynamics simulations rationalize the effect of water on the phase-segregated morphology of Nafion ionomers. We analyzed density maps and radial distribution functions and correlated them with domain structures, distributions of protogenic side chains, and water transport properties. The mesoscopic structures exhibit spongelike morphologies. Hydrophilic domains of water, protons, and anionic side chains form a random three-dimensional network, which is embedded in a matrix of hydrophobic backbone aggregates. Sizes of hydrophilic domains increase from 1 to 3 nm upon water uptake. At low water content, hydrophilic domains are roughly spherical and poorly connected. At higher water content, they convert into elongated cylindrical shapes with high connectivity. Further structural analysis provides a reasonable estimate of the percolation threshold. Radial distribution functions from coarse-grained and atomistic molecular dynamics models exhibit a good agreement. Water cluster size distributions from coarse-grained molecular dynamics and dissipative particle dynamics are consistent with small angle x-ray scattering data. Moreover, we calculated the water diffusivity by molecular dynamics methods and corroborated the results by comparison with pulsed field gradient NMR.

Molecular simulations of solute transport in xylose isomerase crystals.

Cross-linked enzyme crystals (CLECs) enclose an extensive regular matrix of chiral solvent-filled nanopores, via which ions and solutes travel in and out. Several cross-linked enzyme crystals have recently been used for chiral separation and as biocatalysts. We studied the dynamics of solute transport in orthorhombic d-xylose isomerase (XI) crystals by means of Brownian dynamics (BD) and molecular dynamics (MD) simulations, which show how the protein residues influence the dynamics of solute molecules in confined regions inside the lattice. In the BD simulations, coarse-grained beads represent solutes of different sizes. The diffusion of S-phenylglycine molecules inside XI crystals is investigated by long-time MD simulations. The computed diffusion coefficients within a crystal are found to be orders of magnitude lower than in bulk water. The simulation results are compared to the recent experimental studies of diffusion and reaction inside XI crystals. The insights obtained from simulations allow us to understand the nature of solute-protein interactions and transport phenomena in CLECs, which is useful for the design of novel nanoporous biocatalysts and bioseparations based on CLECs.

Reply to Comment on 'Diffusion of water and sodium counter-ions in nanopores of ß-lactoglobulin crystal: a molecular simulation study'.

The analysis in Hu and Jiang's Comment to our paper cannot reveal long-time diffusion, and incorrectly led the authors to conclude that the diffusion in beta-lactoglobuline is anomalous. In this context, the limitations of applying a mean-square displacement analysis to short, heterogeneous pore channels are discussed. A more appropriate approach based on first-passage time analysis is illustrated by a detailed analysis of water motion in a natural membrane protein channel. The partitioning and the motion of water molecules between core and surface hydration layers is discussed. Finally, the calculation of the water density profile is commented upon.

Water diffusion through a membrane protein channel: a first passage time approach.

Water diffusion through OmpF, a porin in the outer membrane of Escherichia coli, is studied by molecular dynamics simulation. A first passage time approach allows characterizing the diffusive properties of a well-defined region of this channel. A carbon nanotube, which is considerably more homogeneous, serves as a model to validate the methodology. Here we find, in addition to the expected regular behavior, a gradient of the diffusion coefficient at the channel ends, witness of the transition from confinement in the channel to bulk behavior in the connected reservoirs. Moreover, we observe the effect of a kinetic boundary layer, which is the counterpart of the initial ballistic regime in a mean square displacement analysis. The overall diffusive behavior of water in OmpF shows remarkable similarity with that in a homogeneous channel. However, a small fraction of the water molecules appears to be trapped by the protein wall for considerable lengths of time. The distribution of trapping times exhibits a broad power law distribution psi(tau) approximately tau (-2.4), up to tau=10 ns, a bound set by the length of the simulation run. We discuss the effect of this distribution on the dynamic properties of water in OmpF in terms of incomplete sampling of phase space.

Solute transport in orthorhombic lysozyme crystals: a molecular simulation study.

Long-time equilibrium molecular dynamics simulations were performed to study the passage of a substrate, L: -arabinose, through nanopores of orthorhombic hen egg white lysozyme crystals. Cross-linked protein crystals (CLPC), as novel biological nanoporous media, consist of an extensive regular matrix of chiral solvent-filled nanopores via which ions and solutes, e.g. sugars and amino acids, travel in and out. We studied the diffusive motion of arabinose inside protein channels. The computed diffusion coefficients within the crystal were orders of magnitudes lower relative to the diffusion coefficient of the solute in water. This study is valuable for understanding the nature of solute-protein interactions and transport phenomena in CLPCs and provides an understanding of biocatalytic and bioseparation processes using CLPC.

Translocation and interactions of L-arabinose in OmpF porin: A molecular dynamics study.

The passage of a natural substrate, L-arabinose (L-ARA) through Escherichia coli porin embedded in an artificial bilayer, is studied by equilibrium molecular dynamics simulations. We investigate the early stage of translocation process of L-ARA from intra-cellular to extra-cellular side (Int-to-Ext) across the bilayer. The average trajectory path over all L-ARA molecules along with quantum-mechanical configuration-optimizations at PM3 level predict the existence of at least three trapping zones. The common feature within all these zones is that L-ARA remains perpendicular to the channel axis. It is remarkable how the orientation and translational-rotational motion of L-ARA molecule play a role in its transport through OmpF channel. These simulations are important for better understanding of permeation process in OmpF channel. They also provide an insight into the chiral recognition of translocation process in protein nanochannels from substrate and protein prospects and help interpret experiments on permeation process of small dipolar molecules across biological membranes.

Diffusion of water and sodium counter-ions in nanopores of a ß-lactoglobulin crystal: a molecular dynamics study.

The dynamics of water and sodium counter-ions (Na(+)) in a C222(1) orthorhombic ß-lactoglobulin crystal is investigated by means of 5 ns molecular dynamics simulations. The effect of the fluctuation of the protein atoms on the motion of water and sodium ions is studied by comparing simulations in a rigid and in a flexible lattice. The electrostatic interactions of sodium ions with the positively charged LYS residues inside the crystal channels significantly influence the ionic motion. According to our results, water molecules close to the protein surface undergo an anomalous diffusive motion. On the other hand, the motion of water molecules further away from the protein surface is normal diffusive. Protein fluctuations affect the diffusion constant of water, which increases from 0.646 ± 0.108 to 0.887 ± 0.41 nm(2) ns(-1), when protein fluctuations are taken into account. The pore size (0.63-1.05 nm) and the water diffusivities are in good agreement with previous experimental results. The dynamics of sodium ions is disordered. LYS residues inside the pore are the main obstacles to the motion of sodium ions. However, the simulation time is still too short for providing a precise description of anomalous diffusion of sodium ions. The results are not only of interest for studying ion and water transport through biological nanopores, but may also elucidate water-protein and ion-protein interactions in protein crystals.