Search for: All records

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. 

We shed light on the mechanism and rate-determining steps of the electrochemical carboxylation of acetophenone as a function of CO 2 concentration by using a robust finite element analysis model that incorporates each reaction step. Specifically, we show that the first electrochemical reduction of acetophenone is followed by the homogeneous chemical addition of CO 2 . The electrochemical reduction of the acetophenone-CO 2 adduct is more facile than that of acetophenone, resulting in an Electrochemical–Chemical–Electrochemical (ECE) reaction pathway that appears as a single voltammetric wave. These modeling results provide new fundamental insights into the complex microenvironment in CO 2 -rich media that produces an optimum electrochemical carboxylation rate as a function of CO 2 pressure. 

The efficient production of green hydrogen via electrochemical water splitting is important for improving the sustainability and enabling the electrification of the chemical industry. One of the major goals of water electrolysis is to utilize non-precious metal catalysts, which can be accomplished with alkaline electrolyzer technologies. However, there is a continuing need for designing catalysts that can operate in alkaline environments with high efficiencies under high current densities. Here we describe a simple, aqueous-based synthesis method to incorporate sulfur into NiFe-based electrocatalysts for the oxygen evolution reaction (OER). Sulfur incorporation was able to reduce the overpotential for the OER from ca. 350 mV on a NiFe catalyst to ca. 290 mV on the NiFeS catalyst at 100 mA cm −2 on a flat supporting electrode. Electrochemical impedance spectroscopy data showed a small decrease in the charge transfer resistance of the NiFeS catalysts, showing that sulfur incorporation may improve the electronic conductivity. Surface-interrogation scanning electrochemical microscopy (SI-SECM) studies combined with Tafel slope analysis suggested that the NiFeS catalyst was able to have vacant surface sites available under OER conditions and was able to maintain a Tafel slope of 39 mV dec −1 . This is in contrast to the NiFe catalyst, for which the SI-SECM studies showed a saturated surface under OER conditions with the Tafel slope transitioning from 39 mV dec −1 to 118 mV dec −1 . The low Tafel slope enabled the NiFeS catalyst to maintain low overpotentials under high current densities, which we attribute to the ability of the NiFeS catalyst to maintain vacant sites during the OER.