Platinum catalysts work as electrical networks, not isolated hotspots

For decades, the prevailing view of how catalysts work went something like this: certain spots on the surface are highly active, others are not, and the overall reaction rate is determined by the sum of these independent hotspots. Optimize the hotspots and you optimize the catalyst.

That picture turns out to be incomplete. In a study published in Nature Catalysis, researchers at the University of Warwick and MIT have produced the first-ever electrochemical images of a thermochemical catalytic reaction at work -- and what they found looks less like a collection of isolated spots and more like an interconnected electrical circuit.

Seeing the surface for the first time

The team used a technique called scanning electrochemical cell microscopy (SECCM), which had never previously been applied to thermochemical reactions. SECCM uses a tiny droplet of electrolyte at the tip of a nanopipette to probe electrochemical activity at individual points on a surface. By scanning across the catalyst, the researchers built detailed activity maps showing where reactions occurred and how different regions behaved.

The catalyst they studied was platinum, widely used in fuel production, clean energy, and industrial chemical processes. When they combined the SECCM activity maps with crystallographic data revealing the orientation and boundaries of individual crystal grains on the surface, a pattern emerged that contradicted the hotspot model.

"Catalysts were long thought to work through individual hotspots where reactions happen fastest," said Dr. Xiangdong Xu, Research Fellow in Chemistry at Warwick and first author. "Our work shows that the surface behaves more like an interconnected electrical network, with different regions sharing electrons and working together to drive the overall reaction."

Specialized grains, cooperative chemistry

The key discovery is that individual crystal grains on the platinum surface specialize in different chemical steps. Some grains favored oxidation reactions; others favored reduction. Rather than each grain independently performing the complete reaction, they divided the labor and communicated through electron flows across grain boundaries.

"Catalyst surfaces are not just a patchwork of individual sites," said Dr. Yogesh Surendranath, Associate Professor at MIT and co-author. "We saw that different regions communicate through electron flow, and that connectivity helps make the overall reaction more efficient."

The team also observed what they call chemical crosstalk -- reactions in one region influenced activity in neighboring areas, sometimes enhancing and sometimes suppressing it. This means that the local environment of any given catalytic site matters not just for that site's performance, but for the behavior of surrounding regions.

Redesigning catalysts from the network up

The implications for catalyst design are significant. If catalytic surfaces function as coordinated networks rather than collections of independent sites, then optimizing individual sites in isolation misses the bigger picture. Instead, researchers could engineer how different regions interact -- tuning grain sizes, orientations, and boundary structures to create surfaces where cooperative electron flows enhance overall performance.

"For the first time, we can see how catalytic activity is organised across a real surface," said Prof. Pat Unwin FRS, from the Department of Chemistry at Warwick. "This opens the door to designing better catalysts by engineering how different regions interact, instead of focusing on single active sites."

What the technique cannot yet show

SECCM provides exceptional spatial resolution, but the current study examines the surface under specific laboratory conditions. Real-world catalytic processes involve high temperatures, pressures, and complex gas mixtures that differ substantially from the controlled environment of the measurement. Whether the cooperative grain behavior observed here persists under industrial operating conditions is an open question.

The study also focused on platinum, one of the most studied and most expensive catalytic materials. Whether the network behavior extends to cheaper alternative catalysts -- a critical question for practical applications -- remains to be tested. The technique itself, however, is transferable and could be applied to other materials and reaction systems.

The findings are most immediately relevant to the design of catalysts for fuel production and clean energy technologies, where even modest improvements in efficiency translate to substantial economic and environmental benefits.