Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “descriptors”

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 433 records · Page 24
All sourcesNASA NTRSDOE OSTI
DOE OSTI2023

Limits of Detection for EXAFS Characterization of Heterogeneous Single-Atom Catalysts

Single-atom catalysts (SACs), consisting of individual metal atoms dispersed on a support, attract attention due to their unique reactivity, efficient use of precious metals, and precise chemical tunability. Characterization of the metal species is crucial to substantiate structure–function relationships. Authors often use—and referees often require—X-ray absorption spectroscopy (XAS) data to prove the absence of clustered metal (or metal oxide) structures after pre-treatment and under in situ or operando conditions. However, there has been no critical assessment of the limitations of XAS in substantiating such conclusive statements, which is particularly important given the potential outsized influence of minority catalyst structures in dictating catalytic activity. Here, in this article, we quantitatively assess the detection limits of XAS to identify metal (or metal oxide) clusters in samples containing predominantly single atoms by modeling the extended X-ray absorption fine structure (EXAFS) of mixtures of structures. We identified that a significant fraction of clusters can coexist with SAC active sites (e.g., ~10% metallic Pt or ~40% oxidized Pt clusters in Pt/CeO 2 SACs), while eluding detection via EXAFS with any statistical significance. To generalize these conclusions, a descriptor-based screening of bulk metal oxides using a continuous Cauchy wavelet transform was proposed that suggests certain materials for which differentiating atomically dispersed metal species and metal oxide clusters would be infeasible by EXAFS (e.g., ReO x ). Based on this analysis, we suggest best practices for the study of SACs using EXAFS and provide recommendations to ensure that conclusions do not outpace the evidence used to support them. In this rapidly expanding research area, rigorous characterization will lead to greater understanding of the behavior of SACs and ultimately improved catalytic materials.

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

Revised Nitrogen Reduction Scaling Relations from Potential-Dependent Modeling of Chemical and Electrochemical Steps

The electrochemical nitrogen reduction reaction (NRR) is a promising route to enable carbon-free ammonia production. However, this reaction is limited by the poor activity and selectivity of current catalysts. The rational design of superior NRR electrocatalysts requires a detailed mechanistic understanding of current material limitations to inform how these might be overcome. The current understanding of how scaling limits NRR on metal catalysts is predicated on a simplified reaction pathway that considers only proton-coupled electron transfer (PCET) steps. Here, we apply grand-canonical density functional theory to investigate a more comprehensive NRR mechanism that includes both electrochemical and chemical steps on 30 metal surfaces in solvent under an applied potential. We applied ..phi..max, a grand-canonical adaptation of the Gmax thermodynamic descriptor, to evaluate trends in catalyst activity. This approach produces a ..phi..max "volcano" diagram for NRR activity scaling on metals that qualitatively differs from the scaling relations identified when only PCET steps are considered. NH3* desorption was found to limit the NRR activity for materials at the top of the volcano and truncate the volcano's peak at increasingly reducing potentials. These revised scaling relations may inform the rational design of superior NRR electrocatalysts. This approach is transferable to study other materials and reaction chemistries where both electrochemical and chemical steps are modeled under an applied potential.

electrocatalysis↗
DOE OSTI2023

Classification of Adsorbed Hydrocarbons Based on Bonding Configurations of the Adsorbates and Surface Site Stabilities

The design of heterogeneous catalysts can be accelerated by identifying relevant descriptors that accurately and effectively link the binding and activation energies to reactivity. Herein, we investigated scaling relations between binding energies of various hydrocarbon-based adsorbates on three different Pt surfaces and metal binding energies estimated via the recently developed α-scheme model. In this study, we find that the scaling slopes are similar for certain groups of adsorbates, which then can be classified based on their spatial and electronic structure enabling fast description of binding strengths for each member of the class. Hence, our findings show that the binding energies of simple hydrocarbons CH x , x = {0,1,2,3,4}, and CHCH 2 can be used to identify the binding energies of more complex hydrocarbon-based adsorbates. We introduce this classification to establish a generalizable scheme in which complex hydrogenation/dehydrogenation processes of higher hydrocarbons can be predicted via the binding energies of simpler hydrocarbon-based species and ultimately through surface site stabilities.

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

Calorimetric Energies and Chemical Potentials of Metal Atoms in Catalytic Nanoparticles on Oxide and Carbon Supports: Improved Size Dependencies and Adhesion Energies

The chemical potential of metal atoms, μM, in supported metal nanoparticles is an important descriptor related to both the catalytic activity and the stability of the nanoparticles. Here, we derive an expression relating μM to the radius of the particle’s contact area with the support and the adhesion energy at the metal/support interface (E adh ) that assumes the particles have the shape of spherical caps but of arbitrary contact angle with the support (θ c ) and includes an empirical correction for the increase in metal surface energy and adhesion energy with decreasing radius of curvature. We then show that, at any assumed contact angle, we can simultaneously fit previously reported measurements of both calorimetric μ M (from heats of metal vapor adsorption during nanoparticle growth by vapor deposition) versus metal coverage data and the He + low-energy ion scattering (LEIS) intensities for the metal and/or support versus metal coverage (using our recently developed spherical cap model for quantitative LEIS intensities), to determine the particle size versus coverage and E adh . Only one choice of contact angle gives a pair of values for contact angle and E adh , which is consistent with the Young–Dupré equation for the equilibrium shape of a spherical particle. At this equilibrium shape, we then applied this spherical cap model (SCM) to reanalyze microcalorimetric metal chemical potentials and LEIS signals versus coverage data for nine metal/support combinations that were previously analyzed by assuming that the particles had the shape of hemispherical caps, i.e., with a contact angle of 90°. We show that this revised approach gives close agreement with the calorimetric and LEIS data; the best-fit contact angles vary from 64 to 84°, correcting the earlier assumption of 90°. Furthermore, these results provide significant accuracy improvements in particle size versus coverage, metal chemical potential versus size and coverage, metal/support adhesion energies and contact angles for Cu, Ag and Au on CeO 2 (111), Ni on MgO(100), Ag on Fe 3 O 4 (111) and TiO 2 (100), and Ag, Ni and Pd on Ni-supported graphene. This revised approach is much more broadly applicable than the earlier hemispherical cap model (HCM).

99 GENERAL AND MISCELLANEOUS↗
DOE OSTI2023

Correlating Experimentally Determined Hydrogen Binding Energy with Hydrogen Evolution Activity over Metal Monolayers on Molybdenum Nitride

It is well established that hydrogen binding energy (HBE) is a key descriptor for hydrogen evolution reaction (HER) activity, and such a relationship is a useful tool for searching efficient and cost-effective HER catalysts. However, in almost all cases, the HBE values are obtained from density functional theory (DFT) calculations. Here in this study, temperature programmed desorption (TPD) was used to experimentally determine the HBE values of metal monolayers supported on molybdenum nitride (Mo 2 N), and electrochemical measurements were performed on the same surfaces. Combined DFT and kinetic Monte Carlo (kMC) simulations were used to validate the trend observed with TPD and the electrochemical HER activity. Depositing one monolayer of Pt on Mo 2 N led to similar HBE values seen for bulk Pt, and electrochemical measurements showed that monolayer Pt on Mo 2 N had similar HER activity to that of bulk Pt. Similar studies were also performed for monolayer Pd on Mo 2 N to confirm the correlation. This work also demonstrates that Mo 2 N is a promising support to reduce precious metal loading in HER catalysts.

08 HYDROGEN↗
DOE OSTI2024

Data Analytics for Catalysis Predictions: Are We Ready Yet?

Catalysis informatics has received tremendous attention in recent years as a tool to design catalysts and discover unique descriptors that capture the relationships between chemical properties and catalytic performance. One of the stop-gaps in understanding catalytic effects, which is often ignored and limits the deployment of data science tools, relates to the lack of uniform data. The catalytic cleavage of C–X (X= H, C, N, and O) bonds is relevant to many fundamental catalytic processes. In this Perspective, we performed data analytics on four groups of C–X cleavage reactions that are common in production, upcycling, or reactive separation: the C–C cleavage in cyclopropyl alcohol, the C–H cleavage in hydroacylation reactions, the C–O cleavage in β-O-4 linkages, and the C–N cleavage in amides, using experimental data collected from the literature to understand their underlying correlations. Experimental variables of high impact are identified for each reaction by dimensionality reduction methods. We highlight the urgent need for experimental data sets that include full details on the reaction conditions, such as reagent concentration, reaction temperature, or time in machine-readable forms. We discuss the potential improvement of the data of these reactions and promising approaches such as autonomous experiments to fill the gaps in unbiased experimental data. Finally, we also address the early stage consideration of separation aspects in the experimental design of efficient catalytic systems for these fundamental examples of chemical reactivity.

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

Machine Learning Accelerated First-Principles Study of the Hydrodeoxygenation of Propanoic Acid

The complex reaction network of catalytic biomass conversions often involves hundreds of surface intermediates and thousands of reaction steps, greatly hindering the rational design of metal catalysts for these conversions. Here, we present a framework of machine learning (ML)-accelerated first-principles studies for the hydrodeoxygenation (HDO) of propanoic acid over transition metal surfaces. The microkinetic model (MKM) is initially parametrized by ML-predicted energies and iteratively improved by identifying the rate-determining species and steps (RDS), computing their energies by density functional theory (DFT), and reparameterizing the MKM until all the RDS are computed by DFT. The Gaussian process (GP) model performs significantly better than the linear ridge regression model for predicting both the adsorption free energies and transition state free energies. Parameterized with energies from the GP model, only 5–20% of the full reaction network has to be computed by DFT for the MKM to possess DFT-level accuracy for the TOF and dominant reaction pathway. While the linear ridge regression model performs worse than the GP model, its performance is greatly improved when only transition states are predicted by the regression model and adsorption energies are computed by DFT. Overall, we find that a high accuracy in adsorption free energies is more important for a reliable MKM than a high accuracy in TS free energies. Lastly, based on the GP model with GOH and GCHCHCO as catalyst descriptors, we build two-dimensional volcano plots in activity and selectivity that can help design promising alloy catalysts for HDO reactions of organic acids.

adsorption↗
DOE OSTI2025

Nitrate Reduction Modeling under Acidic Conditions with Late Transition Metals

The electrochemical reduction of nitrate (NO 3 R) to ammonia is a bold yet conceivable way of producing ammonia using renewable electricity. However, serious challenges remain in finding optimal electrocatalysts for the process. An atomistic understanding of the surface energetics behind the NO 3 R is needed in order to design an efficient catalyst. Herein, we combine energetics from density functional theory and microkinetic modeling to demonstrate how surface descriptors can help simplify the search for efficient NO 3 R electrocatalysts. We illustrate the strong correlations between transition-state energetics and O* binding energies for adsorbed nitrate and nitrite on transition metals. For intermediates from NO* and beyond, we compare the benefits of using either the N* or H* binding energies to predict reduction onset potentials. These insights enable us to develop a simple microkinetic model that elucidates the surface coverages of intermediates and the product selectivity of NO 3 R across a range of potentials and transition metals. As a result, we show that the model adequately corroborates with quasi-steady-state rates observed experimentally.

ammonia↗
DOE OSTI2025

Unraveling Adsorbate-Induced Structural Evolution of Iron Carbide Nanoparticles

Iron carbide (Fe x C y ) nanoparticles (NPs) are promising candidates for replacing platinum group metals in industrial applications, such as high-temperature Fischer–Tropsch synthesis. However, due to their amorphous nature, characterization of the active sites has been challenging experimentally and computationally. Here, using a combined density functional theory (DFT), neural network interatomic potential-assisted global optimization, and ensemble learning study, we evaluate dynamic surface changes associated with syngas (H and CO) interactions. For this purpose, we have developed a general procedure that we use to model an experimentally relevant 270-atom Fe 182 C 88 NP using the neural network-assisted stochastic surface walk global optimization algorithm (SSW-NN). Once generated, the Fe 182 C 88 NP active sites and particle morphology are thoroughly characterized before the effects of syngas adsorbate interactions are explored by using DFT and molecular dynamics simulations. Lastly, we explore correlations between geometric and electronic features of the active sites and the adsorption of H (H ads ), using a regularized random forest machine learning algorithm. In doing so, we identified the Fe–C coordination number and p orbital occupancy as the most important descriptors affecting H ads . Furthermore, using a combined ML and quantum chemistry approach, our work demonstrates a general and efficient procedure for generating and probing complex surface phenomena on binary nanoparticles.

Adsorption↗
DOE OSTI2025

Facet-Dependent Hydrogen Evolution Reaction on M 2 P (M = Ni, Co, Fe) Single Crystals

Transition-metal phosphides (MPs) are promising earth-abundant catalysts for hydrogen evolution reactions (HERs) due to their remarkable activity and stability. To further improve their properties, facet control is a key strategy. The growth of shape-selected nanoparticles may substantially enhance electrocatalytic activity, but this approach requires fundamental studies of facet-specific catalytic properties. There are only a few reports on the facet effects of MPs, which leads to a limited understanding of the activity of each facet and hampers catalyst design. Here, in this study, we grew large hexagonal-prism-shaped single crystals of three representative M 2 P (M = Ni, Co, and Fe) catalysts using metal flux routes. Two facets of M 2 P single crystals were tested to study facet-dependent HER activities, and it was consistently demonstrated that for all M 2 P crystals, a tip facet [(0001) for Ni 2 P/Fe 2 P and (010) for Co 2 P] had a higher activity than the side facet [(101̅0) for Ni 2 P/Fe 2 P and (100) for Co 2 P]. HER activity between the same facet elucidated the activity ordered between different transition metals as Fe 2 P > Co 2 P > Ni 2 P under low-potential regions. At high applied potentials, this trend is reversed due to the differences in Tafel slopes, with Ni 2 P becoming the most active catalyst, such that the activity of the (0001) facet of Ni 2 P approaches that of Pt. The calculated surface density of states (DOS) of each facet and its local curvature were found to be a useful descriptor for the activity trends among different transition metals of the same facets.

Co2P↗
DOE OSTI2026

Group-Additivity–Embedded Multiscale Modeling for Electric Field-Enhanced Nanocatalysis

Elucidating structure-performance relationships remains a central challenge in field-enhanced catalysis, where nanoparticles exhibit nonuniform surface sites with site-dependent responses to electric fields. Low-coordination sites (edges, corners, and tips) are particularly electric field-sensitive (EF), leading to nonuniform charge distribution, adsorption energies, and catalytic activity. Here, using ammonia decomposition on a ruthenium cluster as a model system, we develop a transferable multiscale framework integrating density functional theory, group additivity (GA), Brønsted-Evans-Polanyi scaling, and microkinetic modeling to predict EF-dependent activity across nonuniform cluster sites. Across sites and fields, the nitrogen adsorption energy (E N ) emerges as the governing descriptor, yielding robust volcano relationships whose optimum shifts systematically with field: negative fields strengthen N binding via electron accumulation, while positive fields weaken N binding via charge depletion, moving the optimal E N toward weaker binding. Microkinetic analysis shows that N≡N bond formation remains the key kinetic bottleneck over most conditions; positive fields lower the effective barrier and, critically, increase the fraction of near-optimal active sites, leading to a net enhancement in overall activity relative to zero-field and negative-field cases. By capturing EF- and site-dependent energetics with high accuracy and low computational cost, this GA-embedded multi-scale simulation workflow provides a physically interpretable route to predict and design field-enhanced nanocatalysis.

ammonia decomposition↗
DOE OSTI2026

Quantifying the Impact of Metal Population Distribution in MFI-Supported Mo Catalysts for Methane Dehydroaromatization

Precise evaluation of intrinsic kinetic behavior in Mo/MFI catalysts for methane dehydroaromatization (MDA) is confounded by variations in Mo dispersion and speciation, which are influenced by metal loading and zeolite acidity. This work advances the utility of H 2 -temperature programmed reduction (H 2 -TPR) for characterizing Mo/MFI catalysts by enabling quantitative comparison of MoO x populations distinguished by reduction behavior and linked to initial catalytic performance. A comprehensive H 2 reduction pathway is established through systematic H 2 - TPR studies varying catalyst composition (1–10 wt % Mo, Si/Al = 15, 40, ∞), supplemented by UV-Raman spectroscopy, X-ray powder diffraction, N 2 physisorption, NH 3 -TPD, and advanced spectroscopic analysis (in situ XAS with principal component analysis/multivariate curve resolution—alternating least squares). Two low-temperature H 2 -TPR regions capture distinct Mo populations undergoing initial Mo(VI)→Mo(IV) reduction: a lower-temperature population (Mo-RI) associated primarily with highly dispersed, anchored MoO x species expected to predominantly reside within MFI channels, and a higher-temperature population (Mo-RII) corresponding to a broader set of MoO x species that becomes increasingly bulk-like/extrazeolitic at higher Mo loading. Quantification of these populations provides practical, kinetically relevant descriptors for comparing initial MDA rates across catalysts with varying Mo loading, Si/Al ratio, and MoO x heterogeneity. Application of lower-temperature Mo-RI estimates to kinetic measurements reveals a minimum threshold of ∼0.12 × 10 –3 mol Mo-RI/g cat , above which initial forward benzene rates normalized to this population converge despite differences in metal loading and zeolite Brønsted acidity. This threshold coincides with a transition toward a common C 2 -mediated benzene-forming regime, as indicated by approach-to-equilibrium analysis of methane-to-ethane, ethane-to-ethylene, and ethylene-to-benzene reaction steps. Above this threshold, normalized initial benzene rates are nearly invariant with increasing Mo-RII/Mo-RI population ratio, indicating that excess Mo-RII populations, including bulk-like/extrazeolitic MoO x domains present at higher loading, do not measurably suppress benzene formation associated with anchored and mostly channel-confined Mo population under the initial-rate conditions examined.

Mo/MFI↗
DOE OSTI2026

Ultra-Low Amounts of Transition Metals in Carbon-Based Catalysts Improve Their Alkaline OER Performance: A Systematic Study

The pursuit of green hydrogen production highlights a persistent gap in electrocatalyst research: while academic efforts prioritize cost-efficiency via activity enhancement, industrial viability demands greater emphasis on electrochemical stability. Carbon-based electrocatalysts, particularly those incorporating transition metals, have shown promise in alkaline oxygen evolution (OER) due to their high activity, cost-efficiency, and resource-efficiency. However, these catalysts suffer from insufficient stability under oxidizing conditions compared to pure transition metal catalysts due to erosion of the carbon support resulting from carbon corrosion, among other degradation mechanisms. In this systematic study, the influence of ultra-low amounts (<1 wt %) of iron, cobalt, nickel, and their most common combinations on the stability of a hydrothermally derived, N-doped carbon support and the overall catalyst performance during alkaline OER is systematically explored. By identifying critical stability descriptors and correlating them with synthesis conditions and catalyst properties, primary and secondary corrosion pathways are unraveled. Subsequently, through careful adjustment of carbonization temperature and composition of incorporated transition metals, overall catalyst corrosion can be suppressed immensely. Especially, the inclusion of Ni and Fe is paramount for the formation of stable catalyst materials under laboratory conditions (10 mA/cm 2 in 0.1 M KOH), which is surprisingly unconstrained by the degree of graphitization of the carbon support. Mixing this electrochemically stable material with a graphitic carbon powder results in an excellent stability of over 400 h at 100 mA/cm 2 in 1 M KOH, implying great potential for the future improvement of carbon-based electrodes under oxidizing conditions toward industrial application.

N-doped carbon↗
DOE OSTI2026

Surface Charge Predicts the Presence of Cation Effects in Electrocatalysis

The identity of electrolyte cations has an important influence on the rates of many electrocatalytic reactions, but these effects are not always observed. Recently, we found that the surface charge of the catalyst, quantified by the potential of zero total charge (PZTC), is a useful heuristic for predicting when cation effects will be observed for the oxygen reduction reaction (ORR). Here, we demonstrate that this descriptor allows us to rationalize the observation or absence of cation effects across a range of conditions, reactions (the hydrogen evolution reaction, the ORR, methanol oxidation, ethylene glycol oxidation, glycerol oxidation, and glucose reduction) and metal surfaces (Pt, Pd, Ag, and Au). These results suggest that when the reaction’s operating potential is negative of the metal’s PZTC, electrolyte cations accumulate at the catalyst surface and influence reaction rates. As a result, when reactions occur positive of the PZTC, cation effects are not observed.

Cations↗
DOE OSTI2026

Enabling Stable and Salt-Free Formic Acid Production via Reactive Interface Optimization in Forward-Bias Bipolar-Membrane CO2 Electrolyzers

Electrochemical conversion of CO2 to liquid products is limited by parasitic reactions that consume products and destabilize operation. Here, we show that salt-free formic acid synthesis in a forward-bias bipolar-membrane CO2 electrolyzer is governed by a coupled reaction-transport instability driven by parasitic anode formic acid oxidation. Operando mass spectrometry reveals that this process proceeds through CO-mediated poisoning, identifying catalyst tolerance to CO as a key descriptor of anode selectivity. Among the catalysts examined, PtRu/C preferentially promotes hydrogen oxidation over formic acid oxidation. Coupling this catalyst with transport-interface engineering improves product removal, enhances Faradaic and energy efficiency, and enables stable operation at 200 mA/cm2 for 190 h with a voltage decay of 0.64 mV/h. Techno-economic analysis indicates an 18% reduction in minimum selling price and highlights product concentration as the dominant cost lever. These results establish a general framework for suppressing product oxidation in liquid-product electrolyzers.

Hu, Leiming↗
DOE OSTI2026

Quantitative Analysis and Prediction of Thermal Runaway Metrics of High-Nickel Oxide Cathodes by Machine Learning Models

The pursuit of higher energy density in lithium-ion batteries has made high-nickel (Ni) layered oxides leading cathode candidates for next-generation electric vehicles. However, their poor thermal stability, particularly at Ni contents ≥ 90%, increases the risk of cathode-initiated thermal runaway. Furthermore, we present a data-driven framework combining linear and nonlinear machine learning models to predict key thermal runaway descriptors from a high-throughput differential scanning calorimetry database. With cathode composition and state of charge (SOC) as input features, the ensemble model accurately predicts peak temperature, heat release, and peak heat flow. SHAP analysis identifies Ni content and SOC as the dominant factors controlling thermal runaway temperature, while SOC primarily governs heat release and peak heat flow. Al, Mg, and Mn improve thermal stability by strengthening metal–oxygen bonding and delaying structural transformation, whereas B mainly reduces heat release through surface passivation. Validation with a new cathode composition confirms accurate prediction of SOC-dependent thermal runaway behavior and critical SOC.

25 ENERGY STORAGE↗
DOE OSTI2021

Toward the Development of a High-Voltage Mg Cathode Using a Chromium Sulfide Host

Development of Mg-ion batteries as advanced electrochemical energy storage systems relies on the design and discovery of high-voltage positive electrode (cathode) materials. To date, a variety of sulfide cathodes have been reported (e.g., Mg x Mo 6 S 8 , Mg x Ti 2 S4, etc.), but the voltages of these materials are too low to prepare a high energy density Mg cell. Theoretical computations predicted that Mg x Cr 2 S 4 operating with the high-voltage Cr 3+/4+ redox couple would serve as a suitable cathode candidate, but experimental attempts to extract Mg 2+ from the lattice have been largely unsuccessful. Here, we show that reversible electrochemical activity within a thiospinel framework (AB 2 S 4 ) relies on a redox-active transition metal present in the B site; otherwise, anionic redox activity triggers decomposition of the spinel structure. Since Cr and S states are highly coupled in MgCr 2 S 4 , the Cr 3+/4+ redox couple is inaccessible so that reversible (de)intercalation of Mg 2+ cannot occur and charging leads to dissolution of the active material. These findings point to an insufficiency in the screening criteria previously used to identify MgCr 2 S 4 as a promising Mg cathode. Thus, a computable descriptor based on the electronic structure of the discharged material is proposed to predict the prevalence of cation vs anion redox and improve future surveys of potential candidates. It is unlikely that the high-voltage Cr redox couple will be accessible to oxidation in the presence of sulfur within the restrictions of a spinel framework; however, it is possible that a suitable layered Mg x CrS 2 structure could serve as a reversible high-voltage Mg cathode.

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

Spectroscopic Probe Molecule Selection Using Quantum Theory, First-Principles Calculations, and Machine Learning

Probe molecule vibrational spectra have a long history of being used to characterize materials including metals, oxides, metal-organic frameworks, and even human proteins. Furthermore, recent advances in machine learning have enabled computationally generated spectra to aid in detailed characterization of complex surfaces with probe molecules. Despite widespread use of probe molecules, the science of probe molecule selection is underdeveloped. Here, we develop physical concepts, including orbital interaction energy and the energy overlap integral, to explain and predict the ability of probe molecules to discriminate structural descriptors. We resolve the crystal orbital overlap population (COOP) to specific molecular orbitals and quantify their bonding character, which directly influences vibrational frequencies. Using only a single adsorbate calculation from density function theory (DFT), we compute the interaction energy of individual adsorbate molecular orbitals with adsorption site atomic orbitals across many different sites. Combining the molecular orbital resolved COOP and changes in orbital interaction energy enables probe molecule selection for improved discrimination of various sites. We demonstrate these concepts by comparing the predicted effectiveness of carbon monoxide (CO), nitric oxide (NO), and ethylene (C 2 H 4 ) to probe Pt adsorption sites. Finally, using a previously developed machine learning framework, we show that models trained on hundreds of thousand C 2 H 4 spectra, computed from DFT, which regress surface binding-type and generalized coordination number (GCN), outperform those trained using CO and NO spectra. Lastly, a python package, pDOS_overlap, for implementing the electron density based analysis on any combination of adsorbates and materials, is also made available.

77 NANOSCIENCE AND NANOTECHNOLOGY↗