An official website of the United States government
The conversion and utilization of carbon dioxide (CO2) have dual significance for reducing carbon emissions and solving energy demand. Catalytic reduction of CO2 is a promising way to convert and utilize CO2. However, high-performance catalysts with excellent catalytic activity, selectivity and stability are currently lacking. High-throughput methods offer an effective way to screen high-performance CO2 reduction catalysts. Here, recent advances in high-throughput screening of electrocatalysts for CO2 reduction are reviewed. First, the mechanism of CO2 reduction reaction by electrocatalysis and potential catalyst candidates are introduced. Second, high-throughput computational methods developed to accelerate catalyst screening are presented, such as density functional theory and machine learning. Then, high-throughput experimental methods are outlined, including experimental design, high-throughput synthesis, in situ characterization and high-throughput testing. Finally, future directions of high-throughput screening of CO2 reduction electrocatalysts are outlooked. This review will be a valuable reference for future research on high-throughput screening of CO2 electrocatalysts.
more »
« less
Accelerating Catalysis Understanding via Large Language Model Data Extraction and Shallow Machine Learning Techniques
Catalysis is inherently complex. The lack of precise knowledge available to experimental researchers about the microenvironment, catalytic sites, mechanisms, and changes that occur under reaction conditions has hindered the effectiveness of deep-learning artificial intelligence algorithms to predict catalyst behavior under reaction conditions. Given the type and quality of data available in the scientific literature, there are still open questions on how machine learning can be used by experimentalists working in the field of catalysis to accelerate catalyst design. Here, we present a framework that leverages large language models to extract textual data from known and trusted sources to automatically generate large, but relatively low-fidelity, experimental catalysis data sets across many research groups. We also show that instead of using deep-learning models, which require highquality data, shallow learning models with posthoc interpretability can extract valuable information about experimental catalytic systems from these low-fidelity data sets. The innovation of this work lies not in the model development but in the prompt engineering, data encoding, and question architectures employed to extract meaningful information. We applied this framework to two different model reactions: the electrocatalytic reduction of carbon dioxide and the electrocatalytic oxygen reduction reaction. We showcase that this framework has the ability to uncover known and established facts within the catalysis community, such as the catalytic properties of Cu, as well as novel insights, including the critical role of voltages above a certain threshold in producing multicarbon products from CO2. We anticipate that this proposed framework will serve as an entryway for experimental catalytic researchers to utilize machine learning to rapidly process literature, generate novel hypotheses, and design experiments to accelerate catalyst development.
more »
« less
- Award ID(s):
- 1922649
- PAR ID:
- 10653632
- Publisher / Repository:
- American Chemical Society
- Date Published:
- Journal Name:
- JACS Au
- Volume:
- 5
- Issue:
- 11
- ISSN:
- 2691-3704
- Page Range / eLocation ID:
- 5578 to 5589
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Polymer-encapsulated cobalt phthalocyanine (CoPc) is a model system for studying how polymer-catalyst interactions in the electrocatalytic systems influence performance for the CO2 reduction reaction. In particular, understanding how bulk electrolyte and proton concentration influences polymer protonation, and in turn how the extent of polymer protonation influences catalytic activity and selectivity, is crucial to understanding polymer-catalyst composite materials. We report a study of the dependence of bulk pH and electrolyte concentration on the fractional protonation of poly-4-vinylpyridine and related polymers with both electrochemical and spectroscopic evidence. In addition, we show that the fractional protonation of the polymer is directly related to both the activity of the catalyst and the reaction selectivity for the CO2 reduction reaction over the competitive hydrogen evolution reaction. Of particular note is that the fractional protonation of the film is related to electrolyte concentration, which suggests that the transport of counterions plays an important role in regulating proton transport within the polymer film. These insights suggest that electrolyte concentration and pH play an important in the electrocatalytic performance for polymer-catalyst composite systems, and these influences should be considered in both experimental preparation and analysis.more » « less
-
Obtaining useful insights from machine learning models trained on experimental datasets collected across different groups to improve the sustainability of chemical processes can be challenging due to the small size and heterogeneity of the dataset. Here we show that shallow learning models such as decision trees and random forest algorithms can be an effective tool for guiding experimental research in the sustainable chemistry field. This study trained four different machine learning algorithms (linear regression, decision tree, random forest, and multilayer perceptron) using different sized datasets containing up to 520 unique reaction conditions for the nitrogen reduction reaction (NRR) on heterogeneous electrocatalysts. Using the catalyst properties and experimental conditions as the features, we determined the ability of each model to regress the ammonia production rate and the faradaic efficiency. We observed that the shallow learning decision tree and random forest models had equal or better predictive power compared to the deep learning multilayer perceptron models and the simple linear regression models. Moreover, decision tree and random forest models enable the extraction of feature importance, which is a powerful tool in guiding experimental research. Analysis of the models showed the complex interaction between the applied potential and catalysts on the effective rate for the NRR. We also suggest some underexplored catalysts–electrolyte combinations to experimental researchers looking to improve both the rate and efficiency of the NRR reaction.more » « less
-
Abstract Electrocatalysis plays a key role in the development of renewable energy technologies. Along with the design of electrocatalysts, the microenvironment around catalytic sites has received increasing attention because it affects the distribution and mass transport of reaction species and impacts the reaction kinetics. In this Concept article, we highlight some mechanistic insights into the effect of microenvironment on gas‐involving electrocatalytic reactions, including CO2 reduction, 2‐electron oxygen reduction, and hydrazine oxidation, demonstrating their sensitivity to the wetting properties of microenvironment. For reactions with a gaseous reactant, a moderately hydrophobic microenvironment can greatly enhance the mass transport of gaseous species to accelerate the reaction kinetics while improving the stability of gas‐diffusion electrodes. In contrast, for reactions with a liquid reactant and gaseous product, a hydrophilic microenvironment improves the exposure of catalytic sites to the reactant, while a hydrophobic microenvironment benefits the reaction on the other end by accelerating the diffusion and detachment of generated gas bubbles, which would otherwise block the catalytic sites from the reactant. These understandings and insights can provide important guidelines on the control and optimization of catalyst microenvironment for the development of efficient electrolyzers.more » « less
-
Electrocatalytic CO 2 reduction is an attractive strategy to mitigate the continuous rise in atmospheric CO 2 concentrations and generate value-added chemical products. A possible strategy to increase the activity of molecular systems for these reactions is the co-catalytic use of redox mediators (RMs), which direct reducing equivalents from the electrode surface to the active site. Recently, we demonstrated that a sulfone-based RM could trigger co-electrocatalytic CO 2 reduction via an inner-sphere mechanism under aprotic conditions. Here, we provide support for inner-sphere cooperativity under protic conditions by synthetically modulating the mediator to increase activity at lower overpotentials (inverse potential scaling). Furthermore, we show that both the intrinsic and co-catalytic performance of the Cr-centered catalyst can be enhanced by ligand design. By tuning both the Cr-centered catalyst and RM appropriately, an optimized co-electrocatalytic system with quantitative selectivity for CO at an overpotential ( η ) of 280 mV and turnover frequency (TOF) of 194 s −1 is obtained, representing a three-fold increase in co-catalytic activity at 130 mV lower overpotential than our original report. Importantly, this work lays the foundation of a powerful tool for developing co-catalytic systems for homogeneous electrochemical reactions.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript
- Journal Article:
- https://doi.org/10.1021/jacsau.5c01087
