IIT Gandhinagar Maps Carbon Structures That Could Help Hydrogen Catalysis

Backlog edition: 17 September 2026. Reviewed and supplemented on 20 September with the institute's 18 September release.

IIT Gandhinagar researchers have identified a possible way to make carbon more useful in hydrogen-producing reactions: control the disorder in its atomic structure. Their computational study examines monolayer amorphous carbon, a sheet whose atoms form an irregular network rather than graphene's repeating hexagons.

The result is a set of design clues for future catalysts. It is not a demonstration of a working electrolyser, a measured hydrogen-production rate or a reduction in the price of hydrogen.

Finding useful binding sites

A catalyst surface must bind hydrogen strongly enough to support a reaction, while still allowing it to leave. The researchers studied hydrogen adsorption free energy, a descriptor for that balance; values near zero are promising.

They generated carbon structures through melt-quench simulations and used density functional theory to examine 30 local environments. Calculated adsorption energies ranged from −0.02 to +1.35 electronvolts. The variation matters: a disordered sheet contains many distinct sites, so an average description can miss its most useful regions.

A fine-tuned MACE machine-learning potential extended the screening. The paper links more favourable adsorption to local geometry, including seven-membered rings, curvature and ripples. The implication is that researchers could seek particular structural features instead of treating all disorder as equivalent.

What the study establishes

The work is by Sreehari M S, Ashutosh Krishna Amaram and Raghavan Ranganathan. It appeared in npj 2D Materials and Applications on 26 August 2026; IITGN highlighted it in a release on 18 September. The preprint was first submitted in May.

The institute reports that the machine-learning screen examined approximately 1,183 sites. That broadens the search, but model predictions still need experimental validation. Neither a favourable adsorption energy nor a large number of simulated sites establishes durability, reaction speed or performance in a complete electrochemical device.

Why it matters

Carbon-based catalysts could reduce reliance on scarce catalytic materials if their practical performance proves competitive. This study addresses one part of that engineering problem. Electricity use, equipment and operating conditions also influence hydrogen economics; the paper does not establish that catalysts alone determine the cost.

The next useful milestone is a physical material with controlled structural features and measured hydrogen-evolution performance. Reproducible tests would need to connect the predicted sites to activity and stability under stated operating conditions. Until then, the contribution is a computational guide for choosing what to make and test.

Image: the research team, from left: Sreehari M S, Ashutosh Krishna Amaram and Raghavan Ranganathan. Credit: Indian Institute of Technology Gandhinagar. Republished with reference to the research paper, as required by the institute's media terms.

Sources

• Research paper, npj 2D Materials and Applications, 26 August 2026
• Author preprint and computational results
• IIT Gandhinagar release, 18 September 2026
• Research team photograph, caption and reuse conditions