Indian Institute of Science (IISc) researchers built a data-driven model that maps 9,389 elementary reactions in the conversion of CO2 into fuels on a copper catalyst
The study was reported in Nature Communications; IISc released it on 6 October 2026
A 152-reaction network wrongly predicted formic acid as the main product; the 9,389-reaction network predicted about 40-fold higher CO2 conversion, with methanol and carbon monoxide as the main products, matching experiments
It found that hydrogen can transfer as an intact H2 molecule to some intermediates, not only as single atoms
A catalytic reaction runs through many small elementary steps. Microkinetic modelling uses each step's energy barrier to predict rates and products. If the network leaves out key steps, it predicts the wrong product, as the 152-reaction model did here.
Simple Analogy: A route map missing most roads sends you to the wrong town.
Turning captured CO2 into methanol is a utilisation route; better models help design catalysts for it
Methanol (CH3OH) is a fuel and chemical feedstock; NITI Aayog's Methanol Economy programme promotes it
The authors say it can be adapted to CO2 reduction on other catalysts, nitrogen reduction and water splitting
GS Paper III > Science and Technology; Environment > Carbon utilisation
General Science > Chemistry (formulae, catalysts)
Catalytic reaction of carbon dioxide with hydrogen to give products such as methanol and carbon monoxide
A single molecular step; a full catalytic reaction is a network of many
Energy needed for a step to proceed; here predicted by machine learning