Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “generalizability”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 235 records · Page 13
All sourcesNASA NTRSDOE OSTI
DOE OSTI2020

Active Learning Accelerated Discovery of Stable Iridium Oxide Polymorphs for the Oxygen Evolution Reaction

The discovery of high-performing and stable materials for sustainable energy applications is a pressing goal in catalysis and materials science. Understanding the relationship between a material’s structure and functionality is an important step in the process, such that viable polymorphs for a given chemical composition need to be identified. Machine-learning-based surrogate models have the potential to accelerate the search for polymorphs that target specific applications. Herein, we report a readily generalizable active-learning (AL) accelerated algorithm for identification of electrochemically stable iridium oxide polymorphs of IrO 2 and IrO 3 . The search is coupled to a subsequent analysis of the electrochemical stability of the discovered structures for the acidic oxygen evolution reaction (OER). Structural candidates are generated by identifying all 956 structurally unique AB2 and AB3 prototypes in existing materials databases (more than 38000). Next, using an active learning approach, we find 196 IrO 2 polymorphs within the thermodynamic amorphous synthesizability limit and reaffirm the global stability of the rutile structure. We find 75 synthesizable IrO 3 polymorphs and report a previously unknown FeF 3 -type structure as the most stable, termed α-IrO 3 . To test the algorithms performance, we compare to a random search of the candidate space and report at least a 2-fold increase in the rate of discovery. Additionally, the AL approach can acquire the most stable polymorphs of IrO 2 and IrO 3 with fewer than 30 density functional theory optimizations. Analysis of the structural properties of the discovered polymorphs reveals that octahedral local coordination environments are preferred for nearly all low-energy structures. Subsequent Pourbaix Ir–H 2 O analysis shows that α-IrO3 is the globally stable solid phase under acidic OER conditions and supersedes the stability of rutile IrO 2 . Calculation of theoretical OER surface activities reveal ideal weaker binding of the OER intermediates on α-IrO 3 than on any other considered iridium oxide. We emphasize that the proposed AL algorithm can be easily generalized to search for any binary metal oxide structure with a defined stoichiometry.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2023

Self-Patterning Tetrathiafulvalene Crystalline Films

Tetrathiafulvalene (TTF) crystals grown from the melt are organized as spherulites in which helicoidal fibrils growing radially from the nucleation center twist in concert with one another. Alternating bright and dark concentric bands are apparent when films are viewed between crossed polarizers, indicating an alternating pattern of crystallographic faces exposed at the film surface. Band-dependent reorganization of the TTF crystals was observed during exposure to methanol vapor. Crystalline growth appears on bright bands at the expense of the dark bands. After a 24 h period of exposure to methanol vapor, the original spherulites were completely restructured, and the films comprise isolated, concentric circles of crystallites whose orientations are determined by the initial TTF crystal fibril orientation. While the surface of these outgrowths appears faceted and smooth, cross-sectional SEM images revealed a semiporous inner structure, suggesting solvent-vapor-induced recrystallization. Collectively, these results show that crystal twisting can be used to rhythmically redistribute material. Crystal twisting is a common and often controllable phenomenon independent of molecular or crystal structure and therefore offers a generalizable path to spontaneous pattern formation in a wide range of materials.

36 MATERIALS SCIENCE↗
DOE OSTI2024

Expanding the Design Space of Polymer–Metal Organic Framework (MOF) Gels by Understanding Polymer–MOF Interactions

The fabrication of polymer-MOF composite gels holds great potential to provide emergent properties for drug delivery, environmental remediation, and catalysis. To leverage the full potential of these composites, we investigated how the presence and chemistry of polymers impact MOF formation within the composites and, in turn, how MOFs impact polymer gelation. We show that polymers with a high density of strongly metal-binding carboxylic acids inhibit MOF formation; however, reducing the density of carboxylic acids or substituting them with weaker metal-binding hydroxyl groups permits both MOF formation and gelation within composites. Preparing composites with poly(ethylene glycol) (PEG), which does not bind MOF zirconium (Zr)-oxo clusters, and observing gelation suggests that MOFs can entrap polymer chains to create cross-links in addition to cross-linking them through polymer-Zr-oxo interactions. Both simulations and experiments show composite hydrogels formed with poly(vinyl alcohol) (PVA) to be more stable than those made with PEG, which can reptate through MOF pores upon heating. We demonstrate the generalizability of this composite formation process across different Zr-based MOFs (UiO-66, NU-901, UiO-67, and MOF-525) and by spin-coating gels into conformable films. PVA-UiO-66 composite hydrogels demonstrated high sorption and sustained release of methylene blue relative to the polymer alone (3× loading, 28× slower release), and PVA-MOF-525 composite hydrogels capably sorb the therapeutic peptide Angiotensin 1–7. By understanding the influence of polymer-MOF interactions on the structure and properties of composite gels, this work informs and expands the design space of this emerging class of materials.

36 MATERIALS SCIENCE↗
DOE OSTI2024

Accelerating the Discovery of New, Single Phase High Entropy Ceramics via Active Learning

High-entropy ceramics have garnered interest due to their remarkable hardness, compressive strength, thermal stability, and fracture toughness; yet the discovery of new high-entropy ceramics (out of a tremendous number of possible elemental permutations) still largely requires costly, inefficient, trial-and-error experimental and computational approaches. The entropy forming ability (EFA) factor was recently proposed as a computational descriptor that positively correlates with the likelihood that a 5-metal high-entropy carbide (HECs) will form the desired single phase, homogeneous solid solution; however, discovery of new compositions is computationally expensive. If you consider 8 candidate metals, the HEC EFA approach uses 49 optimizations for each of the 56 unique 5-metal carbides, requiring a total of 2744 costly density functional theory calculations. Here, we describe an orders-of-magnitude more efficient active learning (AL) approach for identifying novel HECs. To begin, we compared numerous methods for generating composition-based feature vectors (e.g., magpie and mat2vec), deployed an ensemble of machine learning (ML) models to generate an average and distribution of predictions, and then utilized the distribution as an uncertainty. Here we then deployed an AL approach to extract new training data points where the ensemble of ML models predicted a high EFA value or was uncertain of the prediction. Our approach has the combined benefit of decreasing the amount of training data required to reach acceptable prediction qualities and biases the predictions toward identifying HECs with the desired high EFA values, which are tentatively correlated with the formation of single phase HECs. Using this approach, we increased the number of 5-metal carbides screened from 56 to 15,504, revealing 4 compositions with record-high EFA values that were previously unreported in the literature. Our AL framework is also generalizable and could be modified to rationally predict optimized candidate materials/combinations with a wide range of desired properties (e.g., mechanical stability, thermal conductivity).

36 MATERIALS SCIENCE↗
DOE OSTI2024

Effect of H + Exchange and Surface Impurities on Bulk and Interfacial Electrochemistry of Garnet Solid Electrolytes

Contact loss and current constriction pose significant challenges at the Li metal interface of solid-state batteries. For garnet-structured Li 7 La 3 Zr 2 O 12 (LLZO), these effects are amplified by Li + /H + exchange and surface contamination reactions, which lead to conductivity losses and poor Li wetting. In this study, we utilize a variety of surface treatment processes across 37 cells to selectively induce proton exchange and contamination reactions in LLZO. The resulting bulk and surface chemistry is systematically characterized and correlated to changes in electrochemical properties. Additionally, we combine impedance analysis and finite element method modeling to deconvolute sources of impedance contributions at the Li metal interface. Specifically, we show that constriction impedance at the Li metal interface arises not solely from voids, but also from ionically-resistive surface contaminants. Further, these findings emphasize the connection between ionic conductivity and constriction, demonstrating that micron-scale ionically-resistive components increase constriction even with identical contact geometries. Finally, we leverage our comprehensive dataset to highlight unstable overpotential growth as a failure mechanism, additionally showing that the phase of a cell’s impedance is a sensitive indicator for the onset of interfacial instability. Overall, this study clarifies the impacts of proton exchange and surface contamination on electrochemical properties at the Li|solid electrolyte interface and elucidates insights that are generalizable to other solid-state battery systems.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2025

Spectroscopic Investigation into P(NDI2OD-T2) Charge Localization

P(NDI2OD-T2), commonly referred to as N2200, stands out as a promising electron-transporting (n-type) polymer for low-cost, flexible (photo)electrochemical applications due to its reversible two-electron reduction and high electron mobility. UV–vis spectroelectrochemistry in the tetrabutyl ammonium hexafluorophosphate/acetonitrile electrolyte shows two sets of chemically reversible redox signals in the cyclic voltammetry corresponding to the reduction of the neutral polymer film to polaronic and bipolaronic species. These electrochemical signatures suggest a distinct electronic reorganization upon reduction (polaron/bipolaron formation), highlighting the need for molecular-level insights into how charges are accommodated within the polymer backbone. While it has been previously hypothesized that charge predominantly localizes on the naphthalene diimide (NDI) unit during reductive charging, specific changes in atomic environments that confirm this localization have not been characterized in n-type polymers. Herein, we use near-edge X-ray absorption fine structure (NEXAFS) spectroscopy to probe electronic transitions in an electrochemically charged polymer to deduce charge localization. The O K-edge (1s) spectra exhibit two distinct π* peaks; the intensity of the lower-energy π* a peak that corresponds to an excitation to a largely localized carbonyl state decreases with reductive potentials relative to the higher-energy π* b peak. We corroborate this with Raman spectroscopy at different potentials, which shows a decrease in intensity on the C–C/C═C and C═O stretching bands of NDI as well as a red shift of the carbonyl band due to the formation of a polaron on the NDI. Additionally, new Raman active NDI signals associated with elongated C═O and C═C bonds are observed at lower energy during the formation of charged states. Together with theoretical calculations, these findings show that the injected charge spatially localizes on the NDI units and is dominantly distributed on the carbonyl groups. In conclusion, the combination of NEXAFS, optical and vibrational spectroscopies, and theoretical calculations is generalizable to other pi-conjugated polymers and can identify charge localization for the further development of organic semiconductors.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2022

Modified Energy Span Analysis Reveals Heterogeneous Catalytic Kinetics

Mechanistic modeling is a cornerstone of catalyst development generally conducted with microkinetic models or density functional theory-based energy profiles. We extend the energy span model of homogeneous catalysis to heterogeneous systems by introducing the modified energy span analysis (MESA) model by implementing collision theory and gas-phase concentration effects. We determine analytically turnover frequencies, coverages, rate-determining steps, apparent activation energies, and reaction orders in agreement with microkinetic and kinetic Monte Carlo simulations. The model applies to discrete catalyst state systems, including single-atom catalysts, homogeneous systems, and microporous materials. A generalizable rate expression is derived for reactor modeling or mechanistic insights from solely experimentally measured reaction orders. Here, we illustrate MESA on three published mechanisms and reveal unexpected phenomena absent in traditional energy span profiles.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2020

General Protocol for the Accurate Prediction of Molecular 13 C/ 1 H NMR Chemical Shifts via Machine Learning Augmented DFT

An accurate prediction of NMR chemical shifts at affordable computational cost is very important for different types of structural assignments in experimental studies. Density functional theory (DFT) and gauge-including atomic orbital (GIAO) are two of the most popular computational methods for NMR calculation, yet they often fail to resolve ambiguities in structural assignments. In this work, we present a new method that uses machine learning (ML) techniques (DFT + ML) that significantly increases the accuracy of 13 C/ 1 H NMR chemical shift prediction for a variety of organic molecules. The input of the generalizable DFT + ML model contains two critical parts: one is a vector providing insights into chemical environments, which can be evaluated without knowing the exact geometry of the molecule; the other one is the DFT-calculated isotropic shielding constant. The DFT + ML model was trained with a data set containing 476 13 C and 270 1 H experimental chemical shifts. For the DFT methods used here, the root mean square deviations (RMSDs) for the errors between predicted and experimental 13 C/ 1 H chemical shifts can be as small as 2.10/0.18 ppm, which is much lower than those from simple DFT (5.54/0.25 ppm), or DFT + linear regression (LR) (4.77/0.23 ppm) approaches. It also has a smaller maximum absolute error than two previously proposed NMR-predicting ML models. The robustness of the DFT + ML model is tested on two classes of organic molecules (TIC10 and hyacinthacines), where the correct isomers were unambiguously assigned to the experimental ones. Overall, the DFT + ML model shows promise for structural assignments in a variety of systems, including stereoisomers, that are often challenging to determine experimentally.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2020

Thermochemical Data Fusion Using Graph Representation Learning

Large databases are required for “Big Data” applications in catalysis and materials science. Thermochemical databases can be created by combining data from various sources and by correcting low-fidelity datasets to higher accuracy with minimal computation. To achieve this “data fusion”, thermochemical quantities of interest, calculated at various levels of density functional theory (DFT), need to be mapped to the same, high levels of theory. In this work, a graph theoretical, statistical framework is proposed for such tasks. Subgraph frequencies are shown to provide a natural representation for learning these fusion maps. The maps are linear and are learnt with automated descriptor selection. Using a dataset of as few as ~1% from the QM9 database of 133,885 molecules, these models can predict multiple thermochemical quantities at a higher level of theory with an accuracy of 1 kcal/mol. Here, the method is explainable, generalizable, and provides a diagnostic tool for outlier identification

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2024

Ice Phase Classification Made Easy with Score-Based Denoising

Accurate identification of ice phases is essential for understanding various physicochemical phenomena. However, such classification for structures simulated with molecular dynamics is complicated by the complex symmetries of ice polymorphs and thermal fluctuations. For this purpose, both traditional order parameters and data-driven machine learning approaches have been employed, but they often rely on expert intuition, specific geometric information, or large training data sets. In this work, we present an unsupervised phase classification framework that combines a score-based denoiser model with a subsequent model-free classification method to accurately identify ice phases. Further, the denoiser model is trained on perturbed synthetic data of ideal reference structures, eliminating the need for large data sets and labeling efforts. The classification step utilizes the smooth overlap of atomic position (SOAP) descriptors as the atomic fingerprint, ensuring Euclidean symmetries and transferability to various structural systems. Our approach achieves a remarkable 100% accuracy in distinguishing ice phases of test trajectories using only seven ideal reference structures of ice phases as model inputs. This demonstrates the generalizability of the score-based denoiser model in facilitating phase identification for complex molecular systems. The proposed classification strategy can be broadly applied to investigate structural evolution and phase identification for a wide range of materials, offering new insights into the fundamental understanding of water and other complex systems.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2026

Commutative Algebra Modeling in Materials Science – A Case Study on Metal–Organic Frameworks (MOFs)

Metal-organic frameworks (MOFs) are a class of important crystalline and highly porous materials whose hierarchical geometry and chemistry hinder interpretable predictions in materials properties. Commutative algebra is a branch of abstract algebra that has been rarely applied in data and material sciences. We introduce the first ever commutative algebra modeling and prediction in materials science. Specifically, category-specific commutative algebra (CSCA) is proposed as a new framework for MOF representation and learning. It integrates element-based categorization with multiscale algebraic invariants to encode both local coordination motifs and global network organization of MOFs. These algebraically consistent, chemically aware representations enable compact, interpretable, and data efficient modeling of MOF properties such as Henry’s constants and uptake capacities for common gases. Compared to traditional geometric and graph-based approaches, CSCA achieves comparable or superior predictive accuracy while substantially improving interpretability and stability across data sets. By aligning commutative algebra with the chemical hierarchy, the CSCA establishes a rigorous and generalizable paradigm for understanding structure and property relationships in porous materials and provides a nonlinear algebra-based framework for data-driven material discovery.

Khaemba, Caleb S.↗
DOE OSTI2023

Probing Accuracy-Speedup Tradeoff in Machine Learning Surrogates for Molecular Dynamics Simulations

The performance promise of machine learning surrogates of molecular dynamics simulations of soft materials is significant but generally comes at the cost of acquiring large training datasets to learn the complex relationships between input soft material attributes and output properties. Under the constraint of limited high-performance computing resources, optimizing the size of the training datasets becomes paramount. Using an artificial neural network based surrogate for molecular dynamics simulations of confined electrolytes, we explore the tradeoff between surrogate accuracy and computational gains. Accuracy is assessed by computing the root-mean-square errors between the surrogate predictions and the ground truth results obtained via molecular dynamics simulations. The computational performance is judged by evaluating the speedup which incorporates the training dataset creation time. Improvement in accuracy occurs with a loss of speedup, which scales as the inverse of the training dataset size. Furthermore, the link between surrogate generalizability and the accuracy-speedup tradeoff is assessed by examining the errors incurred in surrogate predictions on unseen, interpolated input variables and developing a net speedup metric to capture the associated gains.

Anions↗
DOE OSTI2020

Photosynthesis of a Photocatalyst: Single Atom Platinum Captured and Stabilized by an Iron(III) Engineered Defect

Single atom (SA), noble metal catalysts are of interest due to high projected catalytic activity while minimizing cost. Common issues facing many synthesis methodologies include complicated processes, low yields of SA product, and production of mixtures of SA and nanoparticles (NPs). Herein we report a simple, room-temperature synthesis of single Pt-atom decorated, anatase Fe-doped TiO 2 particles that leverages the Fe dopant as an engineered defect site to photodeposit and stabilize atomically dispersed Pt. Both particle morphology and Fe dopant location are based on thermodynamic principles (Gibbs-Wulff construction). CO-DRIFTS (diffuse reflectance infrared Fourier transform spectroscopy) reveals absence of bridge-bonded CO signal, confirming atomically dispersed Pt. XAS (x-ray absorption spectroscopy) of both Pt and Fe indicates Fe-O-Pt bonding that persists through catalytic cycling. Mass balance indicates Pt loading on single particles is 2.5 wt% Pt; single Pt-atom decorated nanoparticle yield is 17%. Pt containing particles show more than an order-of-magnitude increased photo-oxidation efficiency relative to particles containing only Fe. High single-atom-Pt yield, ease of synthesis, and high catalytic activity demonstrate the utility and promise of this method. In conclusion, the principles of this photodeposition synthesis allow for its generalizability towards other SA metals of catalytic interest.

36 MATERIALS SCIENCE↗
DOE OSTI2021

Localization of Nonblinking Point Sources Using Higher-Order-Mode Detection and Optical Heterodyning: Developing a Strategy for Extending the Scope of Molecular, Super-resolution Imaging

While the stochastic, “blinking” nature of fluorescent systems has enabled the super-resolution of their localization by the fitting of their point-spread functions (PSFs), this strategy cannot be exploited for similar resolution of “nonblinking” systems, such as those that might be encountered in a coherent Raman experiment. An alternative method for subdiffraction-limited imaging lies in the exploitation of optical heterodyning. For example, if a Gaussian PSF (a TEM00 mode) of a point emitter is displaced with respect to the origin of the optical system, photons in the higher-order TEM modes carry information about that displacement. Information concerning the displacement can be extracted from photons in these higher-order modes. These photons can be collected by optical heterodyning, which exploits the large gain in a detector’s response to an optical signal from an emitter coupled to a local oscillator, which is prepared in the TEM of interest, e.g., TEM10. We have generalized and developed the heterodyning technique to localize point emitters via the detection of higher-order spatial modes. We have developed a theoretical approach to find a practical estimation limit of the localization parameters using a realistic model that accounts for shot noise, background noise, and Gaussian noise. To demonstrate the applicability of the method, we designed experiments in which a laser is a surrogate for one and two point emitters. Using the Fisher information and its accompanying Cramér-Rao lower bound, we demonstrate super-resolution localization in these cases: we show that objects can be localized to roughly 2–3 orders of magnitude of their point-spread function’s size for a given optical system. Finally and most importantly, it is suggested that the results will ultimately be generalizable to multiple emitters and, most importantly, to “nonblinking” molecular systems, which will be essential for broadening the scope of super-resolution measurements beyond the limits of fluorescence-based techniques.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2023

Thermodynamic Basis for the Stabilization of Helical Peptoids by Chiral Sidechains

Peptoids are a class of highly customizable biomimetic foldamers that retain properties from both proteins and polymers. It has been shown that peptoids can adopt peptide-like secondary structures through the careful selection of sidechain chemistries, but the underlying conformational landscapes that drive these assemblies at the molecular level remain poorly understood. Given the high flexibility of the peptoid backbone, it is essential that methods applied to study peptoid secondary structure formation possess the requisite sensitivity to discriminate between structurally similar yet energetically distinct microstates. In this work, a generalizable simulation scheme is used to robustly sample the complex folding landscape of various 12mer polypeptoids, resulting in a predictive model that links sidechain chemistry with preferential assembly into one of 12 accessible backbone motifs. Using a variant of the metadynamics sampling method, four peptoid dodecamers are simulated in water: sarcosine, N-(1-phenylmethyl)glycine (Npm), (S)-N-(1-phenylethyl)glycine (Nspe), and (R)-N-(1-phenylethyl)glycine (Nrpe)–to determine the underlying entropic and energetic impacts of hydrophobic and chiral peptoid sidechains on secondary structure formation. Our results indicate that the driving forces to assemble Nrpe and Nspe sequences into polyproline type-I helices in water are found to be enthalpically driven, with small benefits from an entropic gain for isomerization and steric strain due to the presence of the chiral center. The minor entropic gains from bulky chiral sidechains in Nrpe- and Nspe-containing peptoids can be explained through increased configurational entropy in the cis state. However, overall assembly into a helix is found to be overall entropically unfavorable. Furthermore, these results highlight the importance of considering the many various competing interactions in the rational design of peptoid secondary structure building blocks.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2021

Determining the Structure-Property Relationships of Quasi-Two-Dimensional Semiconductor Nanoplatelets.

We report a theoretical study of CdSe nanoplatelets aimed at identifying the main factors determining their photophysical properties. Using atomic configurations optimized with density functional theory calculations, we computed quasiparticle and exciton binding energies of nanoplatelets with two to seven monolayers. We employed many body perturbation theory at the GW level and solved the Bethe-Salpeter equation to obtain absorption spectra and excitonic properties. Our results, which agree well with recent experiments, were then used to design a model that allows us to disentangle the effects of quantum confinement, strain induced by passivating ligands, and dielectric environment on the electronic properties of nanoplatelets. We found that, for the model to accurately reproduce our first principle results, it is critical to account for surface stress and consider a finite potential barrier and energy-dependent effective masses when describing quantum confinement. Our findings call into question previous assumptions on the validity of an infinite barrier to describe carrier confinement in nanoplatelets, suggesting that it may be possible to optimize interfacial charge transfer and extraction by appropriately choosing passivating ligands. The model developed here is generalizable to core-shell platelets and enables the description of system sizes not yet directly treatable by first-principles calculations.

Greenwood, Arin R.↗
DOE OSTI2022

Selective Ion Enrichment and Charge Storage through Transport Hysteresis in Conical Nanopipettes

Greater selectivity and controls in the ion transport dynamics are essential in fields such as charge storage, separation, energy storage and conversion, neuromorphic computing and learning, electrochemistry, to name a few. Mechanistic insights into the intriguing hysteresis effects in the rectified electrokinetic transport through single conical nanopipettes are unveiled by combining time-resolved electroanalytical experiments with numeric simulation. Cations as counterions for surface charges are found to dominate not just the through-nanopore flux but also the hysteresis charges, that is, the net enriched or expelled charges during the transport process. Built on our earlier report on the through-nanopore ion flux dominated by counterions for surface charges, the “trapped ions” or hysteresis charges are analyzed herein. Cation selectivity is almost 100% in the hysteresis charges during the potential scans in low conductivity states driven by the combined applied and intrinsic surface electrical fields. Surprisingly, the cation selectivity in the total hysteresis charges remains high at 70–80% over a wide bulk concentration range in the high conductivity (HC) states, where higher ionic strength due to ion enrichment would decrease electrostatistic effects and thus ion selectivity. The retained high selectivity at HC is explained by the competition effects of electroosmotic flow against the co-ion migration. Furthermore, the respective cation and anion portions in the total hysteresis charges over a wide range of ionic strength and measurement conditions provide generalizable strategies for improvements in both transport throughput and selectivity.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2022

How Well Does Kohn–Sham Regularizer Work for Weakly Correlated Systems?

Kohn–Sham regularizer (KSR) is a differentiable machine learning approach to finding the exchange-correlation functional in Kohn–Sham density functional theory that works for strongly correlated systems. Here we test KSR for a weak correlation. We propose spin-adapted KSR (sKSR) with trainable local, semilocal, and nonlocal approximations found by minimizing density and total energy loss. We assess the atoms-to-molecules generalizability by training on one-dimensional (1D) H, He, Li, Be, and Be 2+ and testing on 1D hydrogen chains, LiH, BeH 2 , and helium hydride complexes. The generalization error from our semilocal approximation is comparable to other differentiable approaches, but our nonlocal functional outperforms any existing machine learning functionals, predicting ground-state energies of test systems with a mean absolute error of 2.7 mH.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗