A larger reaction map changed what a computer model predicted about turning carbon dioxide into fuel: instead of formic acid, it identified methanol and carbon monoxide as major products, while predicting roughly 40 times more CO₂ conversion than a smaller model. The result comes from a computational study of CO₂ hydrogenation over copper—not a claim of a 40-fold increase in industrial output.
What the IISc study investigated
The study examined how carbon dioxide (CO₂) reacts with hydrogen over a copper catalyst. In this process, molecules interact at the catalyst surface through a chain of elementary reactions. Which products a kinetic model predicts depends in part on whether its reaction network includes the steps that actually matter.
As an Amazon Associate I earn from qualifying purchases.
“We began with a worry familiar to anyone who does mechanistic modeling: How do you know that your reaction network has not omitted the one step that matters?” said first author Anand Mohan Verma, whom the IISc-released account identifies as an assistant professor at MNNIT Allahabad at the time of publication.
The Tool Desk
Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →How the team built a network of 9,389 reactions
The researchers began with 152 curated reactions, using quantum-mechanical simulations to assemble the initial database. They then trained machine-learning models to predict activation-energy barriers for additional reactions and used automated tools to enumerate possible single-step reactions among 105 surface species. The combined process expanded the network to 9,389 elementary reactions.
#1 Best Overall
In broad terms, quantum-mechanical calculations supplied reaction data, machine learning helped estimate barriers for additional steps, and automated enumeration widened the set of candidate reactions. Kinetic modelling then used the network to predict how the reaction could proceed and what products might form. The reported reaction counts describe this particular study; they are not general industry statistics.
Why the expanded network changed the predictions
| Model | Network size | Predicted major products | CO₂ conversion | Molecular-H₂ transfer pathways |
|---|---|---|---|---|
| Initial network | 152 reactions | Formic acid | Lower; the model underestimated conversion, according to the corresponding author | The account does not specify their representation in this network |
| Expanded network | 9,389 reactions among 105 surface species | Methanol and carbon monoxide | Approximately 40-fold higher than the initial-network model prediction | Highlighted as potentially important pathways |
The comparison shows why a model’s reaction list matters: adding many candidate steps did not merely refine a number, but changed the predicted major products. According to corresponding author Ananth Govind Rajan, the smaller network predicted formic acid rather than methanol and underestimated conversion; the expanded network’s predictions aligned with experimental observations described by the researchers.
Rank #2
That alignment should be understood as consistency with reported experimental observations, not as a measured 40-fold increase in yield or output. The IISc account attributes experimental validation to collaborators at Hindustan Petroleum Corporation Limited’s Green Research and Development Center and A*STAR in Singapore, but does not give the underlying measurements. It therefore does not establish a specific measured yield or independent replication.
The overlooked possibility: hydrogen can transfer as H₂
A notable pathway in the expanded map involves hydrogen transferring to an intermediate as an intact H₂ molecule, rather than first splitting into separate hydrogen atoms. Co-author Shivam Chaturvedi, an IISc chemical-engineering PhD student, said: “The idea that hydrogen can transfer as an intact molecule, without first splitting into atoms, runs against what most of us were taught.”
The IISc account says the researchers used explicit quantum-mechanical calculations to check these steps and found intact-molecule transfer can be particularly favourable for oxygen-containing intermediates. Chaturvedi said the observation “held up when we went back and computed those steps explicitly.” The finding points to a possible design implication: catalysts that interact more strongly with H₂ could potentially favour methanol-forming pathways. It is a proposed direction for catalyst design, not evidence that a commercial catalyst with that performance has been developed.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.What the result does—and does not—show
- It shows a modelling risk: a compact reaction network can omit steps that change predicted products and conversion.
- It reports a model comparison: the approximately 40-fold conversion difference is between kinetic-model predictions for the expanded and initial networks.
- It reports consistency with experiments: the expanded-network predictions were described as aligning with observations, but the account does not publish measured values.
- It does not establish deployment: the study account does not demonstrate commercial operation, consumer availability, or industrial-scale performance.
- It suggests possible wider use: the authors say the framework may be applied to CO₂ reduction on other catalysts, nitrogen reduction, and water splitting; those applications are prospective, not demonstrated results in this report.
The paper is Anand M. Verma et al., “Data-driven massive reaction networks reveal mechanistic pathways underlying catalytic CO₂ hydrogenation,” published in Nature Communications on 17 September 2026, DOI 10.1038/s41467-026-77080-4. The IISc-released account and a corroborating summary appeared on 6 October 2026: EurekAlert! and Phys.org. The released account supplies the reported counts and interpretation, but not the detailed supplementary methods or numerical experimental data.
Quick Recap
Best Value
Rank #4
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
