Publications

Assessing Exchange-Correlation Functionals for Heterogeneous Catalysis of Nitrogen Species

Hong-Jik Kim, Neung-Kyung Yu, Nianhan Tian, Andrew J Medford

The Journal of Physical Chemistry C · 2024

Increasing interest in the sustainable synthesis of ammonia, nitrates, and urea has led to an increase in studies of catalytic conversion between nitrogen-containing compounds using heterogeneous catalysts. Density functional theory (DFT) is commonly employed to obtain molecular-scale insight into these reactions, but there have been relatively few assessments of the exchange-correlation functionals that are best suited for heterogeneous catalysis of nitrogen compounds. Here, we assess a range of functionals ranging from the generalized gradient approximation (GGA) to the random phase approximation (RPA) for the formation energies of gas-phase nitrogen species, the lattice constants of representative solids from several common classes of catalysts (metals, oxides, and metal–organic frameworks (MOFs)), and the adsorption energies of a range of nitrogen-containing intermediates on these materials. The results reveal that the choice of exchange-correlation functional and van der Waals correction can have a surprisingly large effect and that increasing the level of theory does not always improve the accuracy for nitrogen-containing compounds. This suggests that the selection of functionals should be carefully evaluated on the basis of the specific reaction and material being studied.

Overcoming the Chemical Complexity Bottleneck in on-the-Fly Machine Learned Molecular Dynamics Simulations

Lucas R Timmerman, Shashikant Kumar, Phanish Suryanarayana, A. Medford

Journal of Chemical Theory and Computation · 2024

We develop a framework for on-the-fly machine learned force field molecular dynamics simulations based on the multipole featurization scheme that overcomes the bottleneck with the number of chemical elements. Considering bulk systems with up to 6 elements, we demonstrate that the number of density functional theory calls remains approximately independent of the number of chemical elements, in contrast to the increase in the smooth overlap of the atomic positions scheme.

Self-consistent convolutional density functional approximations: Application to adsorption at metal surfaces

S. Sahoo, …, A. Medford

ChemPhysChem · 2023

The exchange-correlation (XC) functional in density functional theory is used to approximate multi-electron interactions. A plethora of different functionals are available, but nearly all are based on the hierarchy of inputs commonly referred to as "Jacob's ladder." This paper introduces an approach to construct XC functionals with inputs from convolutions of arbitrary kernels with the electron density, providing a route to move beyond Jacob's ladder. We derive the variational derivative of these functionals, showing consistency with the generalized gradient approximation (GGA), and provide equations for variational derivatives based on multipole features from convolutional kernels. A proof-of-concept functional, PBEq, which generalizes the PBEα framework where α is a spatially-resolved function of the monopole of the electron density, is presented and implemented. It allows a single functional to use different GGAs at different spatial points in a system, while obeying PBE constraints. Analysis of the results underlines the importance of error cancellation and the XC potential in datadriven functional design. After testing on small molecules, bulk metals, and surface catalysts, the results indicate that this approach is a promising route to simultaneously optimize multiple properties of interest.

Formation of Carbon-Induced Nitrogen-Centered Radicals on Titanium Dioxide under Illumination

Po‐Wei Huang, …, Marta C. Hatzell

JACS Au · 2023

Titanium dioxide is the most studied photocatalytic material and has been reported to be active for a wide range of reactions, including the oxidation of hydrocarbons and the reduction of nitrogen. However, the molecular-scale interactions between the titania photocatalyst and dinitrogen are still debated, particularly in the presence of hydrocarbons. Here, we used several spectroscopic and computational techniques to identify interactions among nitrogen, methanol, and titania under illumination. Electron paramagnetic resonance spectroscopy (EPR) allowed us to observe the formation of carbon radicals upon exposure to ultraviolet radiation. These carbon radicals are observed to transform into diazo- and nitrogen-centered radicals (e.g., CHxN2• and CHxNHy•) during photoreaction in nitrogen environment. In situ infrared (IR) spectroscopy under the same conditions revealed C–N stretching on titania. Furthermore, density functional theory (DFT) calculations revealed that nitrogen adsorption and the thermodynamic barrier to photocatalytic nitrogen fixation are significantly more favorable in the presence of hydroxymethyl or surface carbon. These results provide compelling evidence that carbon radicals formed from the photooxidation of hydrocarbons interact with dinitrogen and suggest that the role of carbon-based “hole scavengers” and the inertness of nitrogen atmospheres should be reevaluated in the field of photocatalysis.

Evaluating the Role of Metastable Surfaces in Mechanochemical Reduction of Molybdenum Oxide

Neung-Kyung Yu, Letícia F. Rasteiro, Van Son Nguyen, Kinga Gołąbek, Carsten Sievers, Andrew J. Medford

JACS Au · 2024

Mechanochemistry and mechanocatalysis are gaining increasing attention as environmentally friendly chemical processes because of their solvent-free nature and scalability. Significant effort has been devoted for studying continuum-scale phenomena in mechanochemistry, such as temperature and pressure gradients, but the atomic-scale mechanisms remain relatively unexplored. In this work, we focus on the mechanochemical reduction of MoO3 as a case study. We use experimental techniques to determine the mechanochemical reduction conditions and density functional theory (DFT) simulations to establish an atomistic framework for identifying the metastable surfaces that are most likely to enable this process. Our results show that metastable surfaces can significantly lower or remove thermodynamic barriers for surface reduction and that kinetic energy from milling can facilitate the formation of metastable surfaces that have high surface fracture energies and are not thermally accessible. These findings indicate that metastable surfaces are an important aspect of mechanochemistry along with hot spots and other continuum-scale phenomena.

Model-Based Design of Experiments for Temporal Analysis of Products (TAP): A Simulated Case Study in Oxidative Propane Dehydrogenation

Adam C. Yonge, G. S. Gusmão, Rebecca Fushimi, Andrew J. Medford

Industrial & Engineering Chemistry Research · 2024

Temporal analysis of products (TAP) reactors enable experiments that probe numerous kinetic processes within a single set of experimental data through variations in pulse intensity, delay, or temperature. Selecting additional TAP experiments often involves an arbitrary selection of reaction conditions or the use of chemical intuition. To make experiment selection in TAP more robust, we explore the efficacy of model-based design of experiments (MBDoE) for precision in TAP reactor kinetic modeling. We successfully applied this approach to a case study of synthetic oxidative propane dehydrogenation (OPDH) that involves pulses of propane and oxygen. We found that experiments identified as optimal through the MBDoE for precision generally reduce parameter uncertainties to a higher degree than alternative experiments. The performance of MBDoE for model divergence was also explored for OPDH, with the relevant active sites (catalyst structure) being unknown. An experiment that maximized the divergence between the three proposed mechanisms was identified and provided evidence that improved the mechanism discrimination. However, reoptimization of kinetic parameters eliminated the ability to discriminate between models. The findings yield insight into the prospects and limitations of MBDoE for TAP and transient kinetic experiments.

Overcoming the Chemical Complexity Bottleneck in on-the-Fly Machine Learned Molecular Dynamics Simulations

Lucas R Timmerman, Shashikant Kumar, Phanish Suryanarayana, A. Medford

Journal of Chemical Theory and Computation · 2024

We develop a framework for on-the-fly machine learned force field molecular dynamics simulations based on the multipole featurization scheme that overcomes the bottleneck with the number of chemical elements. Considering bulk systems with up to 6 elements, we demonstrate that the number of density functional theory calls remains approximately independent of the number of chemical elements, in contrast to the increase in the smooth overlap of the atomic positions scheme.

Assessing Exchange-Correlation Functionals for Heterogeneous Catalysis of Nitrogen Species

Hong-Jik Kim, Neung-Kyung Yu, Nianhan Tian, Andrew J Medford

The Journal of Physical Chemistry C · 2024

Increasing interest in the sustainable synthesis of ammonia, nitrates, and urea has led to an increase in studies of catalytic conversion between nitrogen-containing compounds using heterogeneous catalysts. Density functional theory (DFT) is commonly employed to obtain molecular-scale insight into these reactions, but there have been relatively few assessments of the exchange-correlation functionals that are best suited for heterogeneous catalysis of nitrogen compounds. Here, we assess a range of functionals ranging from the generalized gradient approximation (GGA) to the random phase approximation (RPA) for the formation energies of gas-phase nitrogen species, the lattice constants of representative solids from several common classes of catalysts (metals, oxides, and metal–organic frameworks (MOFs)), and the adsorption energies of a range of nitrogen-containing intermediates on these materials. The results reveal that the choice of exchange-correlation functional and van der Waals correction can have a surprisingly large effect and that increasing the level of theory does not always improve the accuracy for nitrogen-containing compounds. This suggests that the selection of functionals should be carefully evaluated on the basis of the specific reaction and material being studied.

Self-consistent convolutional density functional approximations: Application to adsorption at metal surfaces

S. Sahoo, …, A. Medford

ChemPhysChem · 2023

The exchange-correlation (XC) functional in density functional theory is used to approximate multi-electron interactions. A plethora of different functionals are available, but nearly all are based on the hierarchy of inputs commonly referred to as "Jacob's ladder." This paper introduces an approach to construct XC functionals with inputs from convolutions of arbitrary kernels with the electron density, providing a route to move beyond Jacob's ladder. We derive the variational derivative of these functionals, showing consistency with the generalized gradient approximation (GGA), and provide equations for variational derivatives based on multipole features from convolutional kernels. A proof-of-concept functional, PBEq, which generalizes the PBEα framework where α is a spatially-resolved function of the monopole of the electron density, is presented and implemented. It allows a single functional to use different GGAs at different spatial points in a system, while obeying PBE constraints. Analysis of the results underlines the importance of error cancellation and the XC potential in datadriven functional design. After testing on small molecules, bulk metals, and surface catalysts, the results indicate that this approach is a promising route to simultaneously optimize multiple properties of interest.

Robust and scalable uncertainty estimation with conformal prediction for machine-learned interatomic potentials

Yuge Hu, Joseph Musielewicz, Zachary W. Ulissi, A. Medford

Machine Learning: Science and Technology · 2022

Uncertainty quantification (UQ) is important to machine learning (ML) force fields to assess the level of confidence during prediction, as ML models are not inherently physical and can therefore yield catastrophically incorrect predictions. Established a-posteriori UQ methods, including ensemble methods, the dropout method, the delta method, and various heuristic distance metrics, have limitations such as being computationally challenging for large models due to model re-training. In addition, the uncertainty estimates are often not rigorously calibrated. In this work, we propose combining the distribution-free UQ method, known as conformal prediction (CP), with the distances in the neural network’s latent space to estimate the uncertainty of energies predicted by neural network force fields. We evaluate this method (CP+latent) along with other UQ methods on two essential aspects, calibration, and sharpness, and find this method to be both calibrated and sharp under the assumption of independent and identically-distributed (i.i.d.) data. We show that the method is relatively insensitive to hyperparameters selected, and test the limitations of the method when the i.i.d. assumption is violated. Finally, we demonstrate that this method can be readily applied to trained neural network force fields with traditional and graph neural network architectures to obtain estimates of uncertainty with low computational costs on a training dataset of 1 million images to showcase its scalability and portability. Incorporating the CP method with latent distances offers a calibrated, sharp and efficient strategy to estimate the uncertainty of neural network force fields. In addition, the CP approach can also function as a promising strategy for calibrating uncertainty estimated by other approaches.

Gaussian approximation of dispersion potentials for efficient featurization and machine-learning predictions of metal-organic frameworks

Sihoon Choi, D. Sholl, A. Medford

Journal of Chemical Physics · 2022

Energy-related descriptors in machine learning are a promising strategy to predict adsorption properties of metal-organic frameworks (MOFs) in the low-pressure regime. Interactions between hosts and guests in these systems are typically expressed as a sum of dispersion and electrostatic potentials. The energy landscape of dispersion potentials plays a crucial role in defining Henry's constants for simple probe molecules in MOFs. To incorporate more information about this energy landscape, we introduce the Gaussian-approximated Lennard-Jones (GALJ) potential, which fits pairwise Lennard-Jones potentials with multiple Gaussians by varying their heights and widths. The GALJ approach is capable of replicating information that can be obtained from the original LJ potentials and enables efficient development of Gaussian integral (GI) descriptors that account for spatial correlations in the dispersion energy environment. GI descriptors would be computationally inconvenient to compute using the usual direct evaluation of the dispersion potential energy surface. We show that these new GI descriptors lead to improvement in ML predictions of Henry's constants for a diverse set of adsorbates in MOFs compared to previous approaches to this task.

Quantifying the impact of temporal analysis of products reactor initial state uncertainties on kinetic parameters

A. Yonge, …, A. Medford

AIChE Journal · 2021

The temporal analysis of products (TAP) reactor provides a route to extract intrinsic kinetics from transient measurements. Current TAP uncertainty quantification only considers the experimental noise present in the outlet flow signal. Additional sources of uncertainty such as initial surface coverages, catalyst zone location, inert void fraction, gas pulse intensity and pulse delay, are not included. For this reason, a framework for quantifying initial state uncertainties present in TAP experiments is presented and applied to a carbon monoxide oxidation case study. Two methods for quantifying these sources of uncertainty are introduced. The first utilizes initial state sensitivities to approximate the parameter variances and provide insights into the structural certainty of the model. The second generates parameter confidence distributions through an ensemble-based sampling algorithm. The initial state covariance matrix can ultimately be merged with the experimental noise covariance matrix, providing a unified description of the parameter uncertainties for a TAP experiment.

Application of Density Functional Tight Binding and Machine Learning to Evaluate the Stability of Biomass Intermediates on the Rh(111) Surface

Chaoyi Chang, A. Medford

The Journal of Physical Chemistry C · 2021

Ab-Initio Investigation of Finite Size Effects in Rutile Titania Nanoparticles with Semilocal and Nonlocal Density Functionals

S. Sahoo, Xin Jing, Phanish Suryanarayana, A. Medford

The Journal of Physical Chemistry C · 2021

In this work, we employ hybrid and generalized gradient approximation (GGA) level density functional theory (DFT) calculations to investigate the convergence of surface properties and band structure of rutile titania (TiO$_2$) nanoparticles with particle size. The surface energies and band structures are calculated for cuboidal particles with minimum dimension ranging from 3.7 \r{A} (24 atoms) to 10.3 \r{A} (384 atoms) using a highly-parallel real-space DFT code to enable hybrid level DFT calculations of larger nanoparticles than are typically practical. We deconvolute the geometric and electronic finite size effects in surface energy, and evaluate the influence of defects on band structure and density of states (DOS). The electronic finite size effects in surface energy vanish when the minimum length scale of the nanoparticles becomes greater than 10 \r{A}. We show that this length scale is consistent with a computationally efficient numerical analysis of the characteristic length scale of electronic interactions. The surface energy of nanoparticles having minimum dimension beyond this characteristic length can be approximated using slab calculations that account for the geometric defects. In contrast, the finite size effects on the band structure is highly dependent on the shape and size of these particles. The DOS for cuboidal particles and more realistic particles constructed using the Wulff algorithm reveal that defect states within the bandgap play a key role in determining the band structure of nanoparticles and the bandgap does not converge to the bulk limit for the particle sizes investigated.

TAPsolver: A Python package for the simulation and analysis of TAP reactor experiments

A. Yonge, …, A. Medford

arXiv · 2020

An open-source, Python-based Temporal Analysis of Products (TAP) reactor simulation and processing program is introduced. TAPsolver utilizes algorithmic differentiation for the calculation of highly accurate derivatives, which are used to perform sensitivity analyses and PDE-constrained optimization. The tool supports constraints to ensure thermodynamic consistency, which can lead to more accurate parameters and assist in mechanism discrimination. The mathematical and structural details of TAPsolver are outlined, as well as validation of the forward and inverse problems against well-studied prototype problems. Benchmarks of the code are presented, and a case study for extracting thermodynamically-consistent kinetic parameters from experimental TAP measurements of CO oxidation on supported platinum particles is presented. TAPsolver will act as a foundation for future development and dissemination of TAP data processing techniques.

Classification of biomass reactions and predictions of reaction energies through machine learning

Chaoyi Chang, A. Medford

Journal of Chemical Physics · 2020

Elementary steps and intermediate species of linearly structured biomass compounds are studied. Specifically, possible intermediates and elementary reactions of 15 key biomass compounds and 33 small molecules are obtained from a recursive bond-breaking algorithm. These are used as inputs to the unsupervised Mol2Vec algorithm to generate vector representations of all intermediates and elementary reactions. The vector descriptors are used to identify sub-classes of elementary steps, and linear discriminant analysis is used to accurately identify the reaction type and reduce the dimension of the vectors. The resulting descriptors are applied to predict gas-phase reaction energies using linear regression with accuracies that exceed the commonly employed group additivity approach. They are also applied to quantitatively assess model compound similarity, and the results are consistent with chemical intuition. This workflow for creating vector representations of complex molecular systems requires no input from electronic structure calculations, and it is expected to be applicable to other similar systems where vector representations are needed.

Kinetics-Informed Neural Networks

G. S. Gusmão, Adhika Retnanto, Shashwati C. da Cunha, A. Medford

Catalysis Today · 2020

Chemical kinetics consists of the phenomenological framework for the disentanglement of reaction mechanisms, optimization of reaction performance and the rational design of chemical processes. Here, we utilize feed-forward artificial neural networks as basis functions for the construction of surrogate models to solve ordinary differential equations (ODEs) that describe microkinetic models (MKMs). We present an algebraic framework for the mathematical description and classification of reaction networks, types of elementary reaction, and chemical species. Under this framework, we demonstrate that the simultaneous training of neural nets and kinetic model parameters in a regularized multiobjective optimization setting leads to the solution of the inverse problem through the estimation of kinetic parameters from synthetic experimental data. We probe the limits at which kinetic parameters can be retrieved as a function of knowledge about the chemical system states over time, and assess the robustness of the methodology with respect to statistical noise. This surrogate approach to inverse kinetic ODEs can assist in the elucidation of reaction mechanisms based on transient data.

Prospects and Challenges for Solar Fertilizers

Benjamin M. Comer, …, A. Medford

Joule · 2019

Summary Using solar energy to convert triple bonded molecular dinitrogen from the air into fixed nitrogen products that act as nutrients for plants presents an opportunity to develop “solar fertilizers.” The approach has much in common with solar fuels and chemicals but also has some unique advantages and challenges. The possibility of producing nitrogen fertilizers at a country’s regional level may be able to effectively compete with existing technology by reducing the per unit cost of N nutrient production, removing or reducing transportation costs within or across international borders, and integrating with existing infrastructure. Furthermore, solar fertilizer technologies can reduce the energy and carbon footprint currently associated with the Haber-Bosch process for ammonia synthesis. Deploying solar fertilizer technology in developing countries can also improve access to fertilizers for farmers in remote regions and help achieve the United Nations Sustainable Development Goal 2 of ending hunger and ensuring access to safe, nutritious, and sufficient food for all people, in particular the poor. However, there are also substantial challenges that must be overcome in identifying active catalytic materials and effectively integrating solar fertilizer processes with agricultural infrastructure. This paper outlines the agronomic considerations that drive the development of decentralized solar fertilizer production and explores their implications on fertilizer prices with an emphasis on the developing world. The work also provides an overview of the technical strategies that may enable photo(electro)chemical fertilizer production processes, including the use of fertigation or production of enhanced biochar, and identifies some target metrics and testing considerations to promote efficient development of new fertilizer materials and processes.

ElectroLens: Understanding Atomistic Simulations through Spatially-Resolved Visualization of High-Dimensional Features

Xiangyun Lei, Fred Hohman, Duen Horng Chau, A. Medford

IEEE VIS · 2019

In recent years, machine learning (ML) has gained significant popularity in the field of chemical informatics and electronic structure theory. These techniques often require researchers to engineer abstract "features" that encode chemical concepts into a mathematical form compatible with the input to machine-learning models. However, there is no existing tool to connect these abstract features back to the actual chemical system, making it difficult to diagnose failures and to build intuition about the meaning of the features. We present ElectroLens, a new visualization tool for high-dimensional spatially-resolved features to tackle this problem. The tool visualizes high-dimensional data sets for atomistic and electron environment features by a series of linked 3D views and 2D plots. The tool is able to connect different derived features and their corresponding regions in 3D via interactive selection. It is built to be scalable, and integrate with existing infrastructure.

Design and analysis of machine learning exchange-correlation functionals via rotationally invariant convolutional descriptors

Xiangyun Lei, A. Medford

Physical Review Materials · 2019

In this work we explore the potential of a new data-driven approach to the design of exchange-correlation (XC) functionals. The approach, inspired by convolutional filters in computer vision and surrogate functions from optimization, utilizes convolutions of the electron density to form a feature space to represent local electronic environments and neural networks to map the features to the exchange-correlation energy density. These features are orbital free, and provide a systematic route to including information at various length scales. This work shows that convolutional descriptors are theoretically capable of an exact representation of the electron density, and proposes Maxwell-Cartesian spherical harmonic kernels as a class of rotationally invariant descriptors for the construction of machine-learned functionals. The approach is demonstrated using data from the B3LYP functional on a number of small-molecules containing C, H, O, and N along with a neural network regression model. The machine-learned functionals are compared to standard physical approximations and the accuracy is assessed for the absolute energy of each molecular system as well as formation energies. The results indicate that it is possible to reproduce B3LYP formation energies to within chemical accuracy using orbital-free descriptors with a spatial extent of 0.2 A. The findings provide empirical insight into the spatial range of electron exchange, and suggest that the combination of convolutional descriptors and machine-learning regression models is a promising new framework for XC functional design, although challenges remain in obtaining training data and generating models consistent with pseudopotentials.

Scalable approach to high coverages on oxides via iterative training of a machine‐learning algorithm

Fuzhu Liu, Shengchun Yang, A. Medford

ChemCatChem · 2019

Understanding the interaction of multiple types of adsorbate molecules on solid surfaces is crucial to establishing the stability of catalysts under various chemical environments. Computational studies on the high coverage and mixed coverages of reaction intermediates are still challenging, especially for transition‐metal compounds. In this work, we present a framework to predict differential adsorption energies and identify low‐energy structures under high‐ and mixed‐adsorbate coverages on oxide materials. The approach uses Gaussian process machine‐learning models with quantified uncertainty in conjunction with an iterative training algorithm to actively identify the training set. The framework is demonstrated for the mixed adsorption of CHx, NHx and OHx species on the oxygen vacancy and pristine rutile TiO2(110) surface sites. The results indicate that the proposed algorithm is highly efficient at identifying the most valuable training data, and is able to predict differential adsorption energies with a mean absolute error of ∼0.3 eV based on <25 % of the total DFT data. The algorithm is also used to identify 76 % of the low‐energy structures based on <30 % of the total DFT data, enabling construction of surface phase diagrams that account for high and mixed coverage as a function of the chemical potential of C, H, O, and N. Furthermore, the computational scaling indicates the algorithm scales nearly linearly (N1.12) as the number of adsorbates increases. This framework can be directly extended to metals, metal oxides, and other materials, providing a practical route toward the investigation of the behavior of catalysts under high‐coverage conditions.

Computational Study of Transition-Metal Substitutions in Rutile TiO2 (110) for Photoelectrocatalytic Ammonia Synthesis

Benjamin M. Comer, Max H. Lenk, Aradhya Rajanala, Emma L. Flynn, A. Medford

Catalysis Letters · 2019

Synthesis of ammonia through photo- and electrocatalysis is a rapidly growing field. Titania-based catalysts are widely reported for photocatalytic ammonia synthesis and have also been suggested as electrocatalysts. The addition of transition-metal dopants is one strategy for improving the performance of titania- based catalysts. In this work, we screen d-block transition- metal dopants for surface site stability and evaluate trends in their performance as the active site for the reduction of nitrogen to ammonia on TiO2. We find a linear relationship between the d-band center and metal substitution energy of the dopant site, while the binding energies of N2, N2H, and NH2 all are strongly correlated with the cohesive energies of the dopant metals. The activity of the metal-doped systems shows a volcano type relationship with the NH2 and N2H energies as descriptors. Some metals such as Co, Mo, and V are predicted to slightly improve photo- and electrocatalytic performance, but most metals inhibit the ammonia synthesis reaction. The results provide insight into the role of transition-metal dopants for promoting ammonia synthesis, and the trends are based on unexpected electronic structure factors that may have broader implications for single-atom catalysis and doped oxides.

The Role of Adventitious Carbon in Photo-catalytic Nitrogen Fixation by Titania

Benjamin M. Comer, …, A. Medford

Journal of the American Chemical Society · 2018

Photo-catalytic fixation of nitrogen by titania catalysts at ambient conditions has been reported for decades, yet the active site capable of adsorbing an inert N2 molecule at ambient pressure and the mechanism of dissociating the strong dinitrogen triple bond at room temperature remain unknown. In this work in situ near-ambient-pressure X-ray photo-electron spectroscopy and density functional theory calculations are used to probe the active state of the rutile (110) surface. The experimental results indicate that photon-driven interaction of N2 and TiO2 is observed only if adventitious surface carbon is present, and computational results show a remarkably strong interaction between N2 and carbon substitution (C*) sites that act as surface-bound carbon radicals. A carbon-assisted nitrogen reduction mechanism is proposed and shown to be thermodynamically feasible. The findings provide a molecular-scale explanation for the long-standing mystery of photo-catalytic nitrogen fixation on titania. The results suggest that controlling and characterizing carbon-based active sites may provide a route to engineering more efficient photo(electro)-catalysts and improving experimental reproducibility.

Analysis of Photocatalytic Nitrogen Fixation on Rutile TiO₂(110)

Benjamin M. Comer, A. Medford

ACS Sustainable Chemistry & Engineering · 2017

Photocatalytic nitrogen fixation provides a promising route to produce reactive nitrogen compounds at benign conditions. Titania has been reported as an active photocatalyst for reduction of dinitrogen to ammonia; however there is little fundamental understanding of how this process occurs. In this work the rutile (110) model surface is hypothesized to be the active site, and a computational model based on the Bayesian error estimation functional (BEEF-vdW) and computational hydrogen electrode is applied in order to analyze the expected dinitrogen coverage at the surface as well as the overpotentials for electrochemical reduction and oxidation. This is the first application of computational techniques to photocatalytic nitrogen fixation, and the results indicate that the thermodynamic limiting potential for nitrogen reduction on rutile (110) is considerably higher than the conduction band edge of rutile TiO$_2$, even at oxygen vacancies and iron substitutions. This work provides strong evidence against the most commonly reported experimental hypotheses, and indicates that rutile (110) is unlikely to be the relevant surface for nitrogen reduction. However, the limiting potential for nitrogen oxidation on rutile (110) is significantly lower, indicating that oxidative pathways may be relevant on rutile (110). These findings suggest that photocatalytic dinitrogen fixation may occur via a complex balance of oxidative and reductive processes.

Boron-tuned tetrahedral Co (II) sites in zeolite beta enhance propane dehydrogenation

Kim Y, Locht H, Medford AJ, Jones CW

Applied Catalysis B: Environment and Energy · 2026

Optimization of random phase approximation calculations for improved energies of molecules, solids, and surfaces

2026

Modeling Batch Crystallization under Uncertainty Using Physics-informed Machine Learning

2026

Prospects for using artificial intelligence to understand intrinsic kinetics of heterogeneous catalytic reactions

Medford AJ, Whittaker TN, Kreitz B, Flaherty DW, Kitchin JR

Current Opinion in Chemical Engineering · 2026

Universal electronic manifolds for extrapolative alloy discovery

2026

Roadmap for transforming heterogeneous catalysis with artificial intelligence

Xin H, …, Wang X

Nature Catalysis · 2026

SPARC-X-API: Versatile Python Interface for Real-space Density Functional Theory Calculations

Tian T, Timmerman LR, Kumar S, Comer B, Medford AJ, Suryanarayana P

Journal of Open Source Software · 2025

D–MOPH–25: diverse MOF–molecule pairs for Henry’s constants prediction

Choi S, Sholl DS, Medford AJ

Machine Learning: Science and Technology · 2025

Rational design of selective catalysts for ethylene hydroformylation via microkinetic modeling and universal neural network potentials (vol. 450, 116253, 2025)

2025

Ab initio study of strain-driven vacancy clustering in aluminum

2025

Real-space Hubbard-corrected density functional theory

Bhowmik S, Medford AJ, Suryanarayana P

The Journal of Chemical Physics · 2025

Spectral scheme for atomic structure calculations in density functional theory

Bhowmik S, Pask JE, Medford AJ, Suryanarayana P

Computer Physics Communications · 2025

Comparing classical and machine learning force fields for modeling deformation of metal–organic frameworks relevant for direct air capture

Brabson LM, Medford AJ, Sholl DS

The Journal of Physical Chemistry C · 2025

Benchmarking photocatalysts for dinitrogen photoreduction reaction

Po‐Wei Huang, Danae A. Chipoco Haro, Hakhyeon Song, Andrew J. Medford, Marta C. Hatzell

Chem Catalysis · 2024

Micro-kinetic modeling of temporal analysis of products data using kinetics-informed neural networks

Nai D, Gusmão GS, Kilwein ZA, Boukouvala F, Medford AJ

Digital Discovery · 2024

Innovative approach to sustainable fertilizer production: leveraging electrically assisted conversion of sewage sludge for nutrient recovery

2024

Evaluating the Role of Metastable Surfaces in Mechanochemical Reduction of Molybdenum Oxide

Neung-Kyung Yu, Letícia F. Rasteiro, Van Son Nguyen, Kinga Gołąbek, Carsten Sievers, Andrew J. Medford

JACS Au · 2024

Electrocatalysts for inorganic and organic waste nitrogen conversion

Chipoco Haro DA, …, Hatzell MC

ACS Catalysis · 2024

Maximum-likelihood estimators in physics-informed neural networks for high-dimensional inverse problems

Gusmão GS, Medford AJ

Computers & Chemical Engineering · 2024

SPARC v2.0.0: Spin-orbit coupling, dispersion interactions, and advanced exchange–correlation functionals

Zhang B, …, Suryanarayana P

Software Impacts · 2024

Impact of local microenvironments on the selectivity of electrocatalytic nitrate reduction in a BPM‐MEA system

Huang P, …, Hatzell MC

Advanced Energy Materials · 2024

2023 roadmap on ammonia as a carbon-free fuel

David WIF, …, Valera-Medina A

Journal of Physics: Energy · 2024

Model-Based Design of Experiments for Temporal Analysis of Products (TAP): A Simulated Case Study in Oxidative Propane Dehydrogenation

Adam C. Yonge, G. S. Gusmão, Rebecca Fushimi, Andrew J. Medford

Industrial & Engineering Chemistry Research · 2024

Overcoming the Chemical Complexity Bottleneck in on-the-Fly Machine Learned Molecular Dynamics Simulations

Lucas R Timmerman, Shashikant Kumar, Phanish Suryanarayana, A. Medford

Journal of Chemical Theory and Computation · 2024

Assessing Exchange-Correlation Functionals for Heterogeneous Catalysis of Nitrogen Species

Hong-Jik Kim, Neung-Kyung Yu, Nianhan Tian, Andrew J Medford

The Journal of Physical Chemistry C · 2024

Version 2.0.0: SPARC: Simulation Package for Ab-initio Real-space Calculations

2023

AmpTorch: A Python package for scalable fingerprint-based neural network training on multi-element systems with integrated uncertainty quantification

Shuaibi M, …, Ulissi Z

Journal of Open Source Software · 2023

Phase stability of large-size nanoparticle alloy catalysts at ab initio quality using a nearsighted force-training approach

Zeng C, Sahoo SJ, Medford AJ, Peterson AA

The Journal of Physical Chemistry C · 2023

Atomistic learning in the electronically grand-canonical ensemble

Chen X, El Khatib M, Lindgren P, Willard A, Medford AJ, Peterson AA

npj Computational Materials · 2023

Automated generation of microkinetics for heterogeneously catalyzed reactions considering correlated uncertainties

Kreitz B, Lott P, Studt F, Medford AJ, Deutschmann O, Goldsmith CF

Angewandte Chemie International Edition · 2023

The Open DAC 2023 Dataset and Challenges for Sorbent Discovery in Direct Air Capture

Anuroop Sriram, …, D. Sholl

ACS Central Science · 2023

Surface Interactions of Erythrose on Assorted Metal Oxides: A Solid-State NMR Study

Sean Najmi, C. Liotta, A. Medford, Carsten Sievers

The Journal of Physical Chemistry C · 2023

Role of Catalyst Domain Size in the Hydrogenation of CO2 to Aromatics over ZnZrOx/ZSM-5 Catalysts

Iman Nezam, …, Christopher W. Jones

The Journal of Physical Chemistry C · 2023

Formation of Carbon-Induced Nitrogen-Centered Radicals on Titanium Dioxide under Illumination

Po‐Wei Huang, …, Marta C. Hatzell

JACS Au · 2023

Kohn-Sham accuracy from orbital-free density functional theory via Δ-machine learning

Shashikant Kumar, Xin Jing, J. Pask, A. Medford, Phanish Suryanarayana

Journal of Chemical Physics · 2023

Self-consistent convolutional density functional approximations: Application to adsorption at metal surfaces

S. Sahoo, …, A. Medford

ChemPhysChem · 2023

Online graduate certificate in data science for the chemical industry

Medford A

Chemical Engineering Education · 2022

Training stiff dynamic process models via neural differential equations

Computer Aided Chemical Engineering · 2022

A universal framework for featurization of atomistic systems

Lei X, Medford AJ

The Journal of Physical Chemistry Letters · 2022

Supported Molybdenum Oxides for the Aldol Condensation Reaction of Acetaldehyde

M. Rasmussen, Sean Najmi, Giada Innocenti, A. Medford, Carsten Sievers, J. Will Medlin

Journal of Catalysis · 2022

Robust and scalable uncertainty estimation with conformal prediction for machine-learned interatomic potentials

Yuge Hu, Joseph Musielewicz, Zachary W. Ulissi, A. Medford

Machine Learning: Science and Technology · 2022

Perspectives on the Competition between the Electrochemical Water and N2 Oxidation on a TiO2(110) Electrode

Ebrahim Tayyebi, …, Egill Skúlason

The Journal of Physical Chemistry Letters · 2022

Internal Calibration of Transient Kinetic Data via Machine Learning

M. Kunz, …, R. Fushimi

Catalysis Today · 2022

A Career in Catalysis: Jens Kehlet Nørskov

A. Medford, P. G. Moses, K. Jacobsen, A. Peterson

ACS Catalysis · 2022

Gaussian approximation of dispersion potentials for efficient featurization and machine-learning predictions of metal-organic frameworks

Sihoon Choi, D. Sholl, A. Medford

Journal of Chemical Physics · 2022

Soft and transferable pseudopotentials from multi-objective optimization

M. F. Shojaei, J. Pask, A. Medford, Phanish Suryanarayana

Computer Physics Communications · 2022

Quantifying the impact of temporal analysis of products reactor initial state uncertainties on kinetic parameters

A. Yonge, …, A. Medford

AIChE Journal · 2021

Methods for nitrogen activation by reduction and oxidation

Haldrian Iriawan, …, Y. Shao-horn

Nature Reviews Methods Primers · 2021

Efficient Models for Predicting Temperature-Dependent Henry’s Constants and Adsorption Selectivities for Diverse Collections of Molecules in Metal–Organic Frameworks

Xiaohan Yu, Sihoon Choi, D. Tang, A. Medford, D. Sholl

The Journal of Physical Chemistry C · 2021

Direct aromatization of CO2 via combined CO2 hydrogenation and zeolite-based acid catalysis

Iman Nezam, …, Christopher W. Jones

Journal of CO₂ Utilization · 2021

Application of Density Functional Tight Binding and Machine Learning to Evaluate the Stability of Biomass Intermediates on the Rh(111) Surface

Chaoyi Chang, A. Medford

The Journal of Physical Chemistry C · 2021

Ab-Initio Investigation of Finite Size Effects in Rutile Titania Nanoparticles with Semilocal and Nonlocal Density Functionals

S. Sahoo, Xin Jing, Phanish Suryanarayana, A. Medford

The Journal of Physical Chemistry C · 2021

TAPsolver: A Python package for the simulation and analysis of TAP reactor experiments

A. Yonge, …, A. Medford

arXiv · 2020

SPARC: Simulation Package for Ab-initio Real-space Calculations

Qimen Xu, …, Phanish Suryanarayana

SoftwareX · 2020

Pretreatment Effects on the Surface Chemistry of Small Oxygenates on Molybdenum Trioxide

Sean Najmi, …, Carsten Sievers

ACS Catalysis · 2020

Mechanocatalytic Ammonia Synthesis over TiN in Transient Microenvironments

Andrew W. Tricker, …, Carsten Sievers

ACS Energy Letters · 2020

Data Driven Reaction Mechanism Estimation via Transient Kinetics and Machine Learning

M. Kunz, …, R. Fushimi

Chemical Engineering Journal · 2020

Continuous Liquid-Phase Upgrading of Dihydroxyacetone to Lactic Acid over Metal Phosphate Catalysts

Giada Innocenti, E. Papadopoulos, G. Fornasari, F. Cavani, A. Medford, Carsten Sievers

ACS Catalysis · 2020

Classification of biomass reactions and predictions of reaction energies through machine learning

Chaoyi Chang, A. Medford

Journal of Chemical Physics · 2020

Heterogeneity in susceptibility dictates the order of epidemic models

Christopher Rose, A. Medford, C. Goldsmith, T. Vegge, J. Weitz, Andrew A. Peterson

Journal of Theoretical Biology · 2020

Kinetics-Informed Neural Networks

G. S. Gusmão, Adhika Retnanto, Shashwati C. da Cunha, A. Medford

Catalysis Today · 2020

Does Blanket Fertilizer Recommendation Still Work? A Case Study of Maize Production in Northern Ghana

2019

Prospects and Challenges for Solar Fertilizers

Benjamin M. Comer, …, A. Medford

Joule · 2019

ElectroLens: Understanding Atomistic Simulations through Spatially-Resolved Visualization of High-Dimensional Features

Xiangyun Lei, Fred Hohman, Duen Horng Chau, A. Medford

IEEE VIS · 2019

Design and analysis of machine learning exchange-correlation functionals via rotationally invariant convolutional descriptors

Xiangyun Lei, A. Medford

Physical Review Materials · 2019

Scalable approach to high coverages on oxides via iterative training of a machine‐learning algorithm

Fuzhu Liu, Shengchun Yang, A. Medford

ChemCatChem · 2019

Computational Study of Transition-Metal Substitutions in Rutile TiO2 (110) for Photoelectrocatalytic Ammonia Synthesis

Benjamin M. Comer, Max H. Lenk, Aradhya Rajanala, Emma L. Flynn, A. Medford

Catalysis Letters · 2019

The Role of Adventitious Carbon in Photo-catalytic Nitrogen Fixation by Titania

Benjamin M. Comer, …, A. Medford

Journal of the American Chemical Society · 2018

Selectivity of Synthesis Gas Conversion to C2+ Oxygenates on fcc(111) Transition-Metal Surfaces

Julia Schumann, …, J. Nørskov

ACS Catalysis · 2018

Extracting Knowledge from Data through Catalysis Informatics

A. Medford, M. Kunz, Sarah M. Ewing, Tammie L. Borders, R. Fushimi

ACS Catalysis · 2018

A Highly Active Molybdenum Phosphide Catalyst for Methanol Synthesis from CO and CO2

MeIis S. Duyar, …, T. Jaramillo

Angewandte Chemie International Edition · 2018

Thermodynamic Limitations of the Catalyst Design Space for Methanol Production from Methane

Jennifer N. Jocz, A. Medford, Carsten Sievers

ChemCatChem · 2018

Database of Computation-Ready 2D Zeolitic Slabs

O. Knio, A. Medford, S. Nair, D. Sholl

Chemistry of Materials · 2018

To address surface reaction network complexity using scaling relations machine learning and DFT calculations

Ulissi ZW, Medford AJ, Bligaard T, Nørskov JK

Nature Communications · 2017

Scaling-Relation-Based Analysis of Bifunctional Catalysis: The Case for Homogeneous Bimetallic Alloys

M. Andersen, A. Medford, J. Nørskov, K. Reuter

ACS Catalysis · 2017

Photon-Driven Nitrogen Fixation: Current Progress, Thermodynamic Considerations, and Future Outlook

A. Medford, Marta C. Hatzell

ACS Catalysis · 2017

Extracting knowledge from molecular mechanics simulations of grain boundaries using machine learning

Joshua A. Gomberg, A. Medford, S. Kalidindi

Acta Materialia · 2017

Analysis of Photocatalytic Nitrogen Fixation on Rutile TiO₂(110)

Benjamin M. Comer, A. Medford

ACS Sustainable Chemistry & Engineering · 2017

Vision for Data and Informatics in the Future Materials Innovation Ecosystem

S. Kalidindi, A. Medford, D. McDowell

JOM · 2016

Intrinsic Selectivity and Structure Sensitivity of Rhodium Catalysts for C(2+) Oxygenate Production

Nuoya Yang, …, J. Nørskov

Journal of the American Chemical Society · 2016

Framework for Scalable Adsorbate–Adsorbate Interaction Models

M. Hoffmann, A. Medford, T. Bligaard

The Journal of Physical Chemistry C · 2016

Analyzing the Case for Bifunctional Catalysis

M. Andersen, A. Medford, J. Nørskov, K. Reuter

Angewandte Chemie International Edition · 2016

Computational Insight into catalytic hydrogenation of nitrogen and carbon monoxide

2015

From the Sabatier principle to a predictive theory of transition-metal heterogeneous catalysis

A. Medford, …, J. Nørskov

Journal of Catalysis · 2015

Degree of rate control approach to computational catalyst screening

Christopher A. Wolcott, A. Medford, F. Studt, C. Campbell

Journal of Catalysis · 2015

CatMAP: A Software Package for Descriptor-Based Microkinetic Mapping of Catalytic Trends

A. Medford, …, J. Nørskov

Catalysis Letters · 2015

Thermochemistry and micro-kinetic analysis of methanol synthesis on ZnO (0001)

A. Medford, …, P. G. Moses

Journal of Catalysis · 2014

Methanol-to-hydrocarbons conversion: The alkene methylation pathway

R. Brogaard, …, U. Olsbye

Journal of Catalysis · 2014

High Pressure CO Hydrogenation Over Bimetallic Pt–Co Catalysts

J. Christensen, A. Medford, F. Studt, A. Jensen

Catalysis Letters · 2014

Exploring the limits: A low-pressure, low-temperature Haber–Bosch process

A. Vojvodić, …, J. Nørskov

Chemical Physics Letters · 2014

Departures from the Adsorption Energy Scaling Relations for Metal Carbide Catalysts

R. Michalsky, Yin-Jia Zhang, A. Medford, A. Peterson

The Journal of Physical Chemistry · 2014

Assessing the reliability of calculated catalytic ammonia synthesis rates

A. Medford, …, J. Nørskov

Science · 2014

Activity and Selectivity Trends in Synthesis Gas Conversion to Higher Alcohols

A. Medford, …, F. Studt

Topics in Catalysis · 2014

On the effect of coverage-dependent adsorbate–adsorbate interactions for CO methanation on transition metal surfaces

Adam C. Lausche, …, F. Studt

Journal of Catalysis · 2013

Finite-Size Effects in O and CO Adsorption for the Late Transition Metals

A. Peterson, …, J. Nørskov

Topics in Catalysis · 2012

Corrigendum to “Elementary steps of syngas reactions on Mo2C(0 0 1): Adsorption thermochemistry and bond dissociation” [J. Catal. 290 (2012) 108–117]

A. Medford, A. Vojvodić, F. Studt, F. Abild-Pedersen, J. Nørskov

Journal of Catalysis · 2012

Elementary Steps of Syngas Reactions on Mo2C(001): Adsorption Thermochemistry and Bond Dissociation

A. Medford, A. Vojvodić, F. Studt, F. Abild-Pedersen, J. Nørskov

Journal of Catalysis · 2012

Electrocatalytic interaction of nano-engineered palladium on carbon nanofibers with hydrogen peroxide and β-NADH

Zhan Lin, Liwen Ji, A. Medford, Q. Shi, W. Krause, Xiangwu Zhang

Journal of Applied Electrochemistry · 2011

An inter-laboratory stability study of roll-to-roll coated flexible polymer solar modules

S. Gevorgyan, …, F. Krebs

Solar Energy Materials and Solar Cells · 2011

Ultra fast and parsimonious materials screening for polymer solar cells using differentially pumped slot-die coating

Jan Alstrup, M. Jørgensen, A. Medford, F. Krebs

ACS Applied Materials & Interfaces · 2010

The effect of post-processing treatments on inflection points in current–voltage curves of roll-to-roll processed polymer photovoltaics

M. Lilliedal, A. Medford, M. Madsen, K. Norrman, F. Krebs

Solar Energy Materials and Solar Cells · 2010

Grid-connected polymer solar panels: initial considerations of cost, lifetime, and practicality

A. Medford, …, F. Krebs

Optics Express · 2010

Formation and electrochemical performance of copper/carbon composite nanofibers

Liwen Ji, …, Xiangwu Zhang

Electrochimica Acta · 2010

Fabrication of carbon nanofiber-driven electrodes from electrospun polyacrylonitrile/polypyrrole bicomponents for high-performance rechargeable lithium-ion batteries

Liwen Ji, …, Xiangwu Zhang

Journal of Power Sources · 2010

Assembly of carbon-SnO2 core-sheath composite nanofibers for superior lithium storage

Liwen Ji, Zhan Lin, Bingkun Guo, A. Medford, Xiangwu Zhang

Chemistry – A European Journal · 2010

Porous carbon nanofibers loaded with manganese oxide particles: Formation mechanism and electrochemical performance as energy-storage materials

Liwen Ji, A. Medford, Xiangwu Zhang

Physical Chemistry Chemical Physics · 2009

Porous carbon nanofibers from electrospun polyacrylonitrile/SiO2 composites as an energy storage material

Liwen Ji, Zhan Lin, A. Medford, Xiangwu Zhang

Carbon · 2009

In-situ encapsulation of nickel particles in electrospun carbon nanofibers and the resultant electrochemical performance

Liwen Ji, Zhan Lin, A. Medford, Xiangwu Zhang

Chemistry – A European Journal · 2009

Fabrication of carbon fibers with nanoporous morphologies from electrospun polyacrylonitrile/poly(L-lactide) blends

Liwen Ji, A. Medford, Xiangwu Zhang

Journal of Polymer Science Part B: Polymer Physics · 2009

Electrospun polyacrylonitrile/zinc chloride composite nanofibers and their response to hydrogen sulfide

Liwen Ji, A. Medford, Xiangwu Zhang

Polymer · 2009

Electrospun polyacrylonitrile fibers with dispersed Si nanoparticles and their electrochemical behaviors after carbonization

Liwen Ji, Kyung-Hye Jung, A. Medford, Xiangwu Zhang

Physical Chemistry Chemical Physics · 2009

Unifying thermochemistry concepts in computational heterogeneous catalysis

Kreitz B, …, Medford AJ

preprint · 2024

Screening and Discovery of Metal Compound Active Sites for Strong and Selective Adsorption of N2 in Air

Tian N, Comer B, Medford A

preprint · 2023

The Open DAC 2025 Dataset for Sorbent Discovery in Direct Air Capture

Anuroop Sriram, …, D. Sholl

2025

Identifying useful sorbent materials for direct air capture (DAC) from humid air remains a challenge. We present the Open DAC 2025 (ODAC25) dataset, a significant expansion and improvement upon ODAC23 (Sriram et al., ACS Central Science, 10 (2024) 923), comprising nearly 60 million DFT single-point calculations for CO$_2$, H$_2$O, N$_2$, and O$_2$ adsorption in 15,000 MOFs. ODAC25 introduces chemical and configurational diversity through functionalized MOFs, high-energy GCMC-derived placements, and synthetically generated frameworks. ODAC25 also significantly improves upon the accuracy of DFT calculations and the treatment of flexible MOFs in ODAC23. Along with the dataset, we release new state-of-the-art machine-learned interatomic potentials trained on ODAC25 and evaluate them on adsorption energy and Henry's law coefficient predictions.