Controlling boron concentration at grain boundaries in platinum nanocatalysts prevents structural failure and preserves catalytic performance under prolonged high temperature conditions.
(Nanowerk Spotlight) Most industrial catalysts, the materials that accelerate chemical reactions in everything from car exhaust systems to hydrogen fuel cells, are designed to enable high reaction rates under demanding operating conditions. But over time, they always degrade. Their internal structures change, surfaces smooth out, and the very features that make them active begin to disappear. This quiet failure, happening at the atomic scale, limits the efficiency and lifespan of technologies that depend on them.
A large part of this problem comes down to boundaries. In nanoscale metals like platinum, which are widely used in advanced catalysts, the material is not a single crystal but a patchwork of tiny crystalline grains. These structures, often just a few nanometers across, form internal interfaces known as grain boundaries. These regions are chemically reactive and often enhance performance, especially in nanostructured materials. But they are also fragile. Under heat, the atoms in these boundaries can shift. Grains merge. Boundaries disappear. When that happens, the performance benefits vanish with them.
Researchers have tried to slow this process by adding other elements that settle into the grain boundaries and act as stabilizers. This tactic, called segregation, has been studied for years. But it presents a tradeoff. While some additives strengthen the boundaries, they may also interfere with the catalyst’s chemical behavior. Others improve stability for a while but then cause new problems, such as weakening the structure or creating regions that poison the reaction. The underlying physics is complex, and without a clear strategy to predict which elements work under what conditions, progress has remained slow.
In a new study published in Advanced Science (“Rational Grain Boundary Segregation Enables Long‐Term Thermally Stable, Catalytically Active, and CO‐Tolerant Nanograined Metals”), researchers from the Max Planck Institute for Sustainable Materials in Germany offer a more precise and predictive approach. Instead of choosing stabilizers by trial and error, they combined atomic scale simulations with high resolution experiments to find a way to both protect and preserve the most active parts of a platinum nanocatalyst.
Their work focuses on how a common additive, boron, can either stabilize or destabilize a catalyst depending entirely on where and how much of it is used. The findings point to a method for building catalysts that not only perform well initially but also maintain that performance over time, even under high temperature conditions.
a) Electron microscope image showing a network of platinum nanoparticles joined at grain boundaries. b) Individual platinum nanoparticles before assembly. c) Comparison of catalytic performance in the hydrogen oxidation reaction: isolated nanoparticles show low activity and strong deactivation by carbon monoxide, while grain boundary–rich assemblies exhibit much higher activity and strong resistance to CO poisoning. d) Theoretical calculations show that grain boundaries lower the energy barriers of key reaction steps, explaining the enhanced activity. e–g) Models and simulations reveal that grain boundaries change the electronic structure of platinum, reducing the strength of CO binding and thereby improving CO tolerance. (Image: Reprinted from DOI:10.1002/advs.202512730, CC BY)
The researchers began by assembling platinum nanoparticles about two nanometers wide into porous networks that naturally formed many grain boundaries. Compared to isolated particles, these networks showed significantly higher activity in a key reaction known as hydrogen oxidation. This reaction is central to hydrogen fuel technologies and is especially sensitive to the structure of the catalyst surface. Simulations confirmed that the grain boundaries lowered the energy needed for the reaction to proceed, making them the most active sites in the material.
To test the long-term stability of these structures, the team heated the platinum networks to four hundred degrees Celsius for extended periods. The results showed a steady loss of catalytic performance and a visible coarsening of the grain structure. Boundaries merged. Nanoparticles fused. The total surface area shrank. These changes matched what would be expected from thermal degradation and confirmed the need for a method to keep the grain boundaries intact.
To identify a suitable stabilizing element, the researchers used density functional theory, a computational method for modeling atomic interactions. They screened more than thirty elements to find ones that would naturally migrate to platinum grain boundaries without disrupting the reaction. Boron stood out. It had a strong thermodynamic preference for boundary sites and a neutral effect on the key reaction steps. This meant it could potentially stabilize the structure without harming performance.
To add boron selectively, the researchers adsorbed boron-containing molecules onto the surfaces of the platinum particles before assembly. As the particles connected to form grain boundaries, the boron atoms became trapped at the interfaces. They then washed away any remaining boron from the surface, ensuring that only the boundary regions remained enriched. Using atom probe tomography, they confirmed the location and concentration of boron within the network.
Initial tests showed that boron-stabilized structures retained nearly all of their catalytic activity after twenty hours of heating. But after forty hours, performance dropped sharply. Imaging showed that the grain boundaries had begun to separate. Nanoparticles were no longer connected. Structural degradation had accelerated. This result pointed to a critical threshold. While moderate boron levels preserved the boundaries, higher concentrations caused them to fail.
The researchers traced the problem to clustering. As boron concentration increased beyond the equilibrium limit for the boundary, the atoms began to group together. These clusters introduced strain and repulsive forces at the interface. Simulations showed that when boron atoms were spaced more than about zero point four nanometers apart, they strengthened the boundary. But when they became too close, they pushed the surrounding atoms apart, weakening cohesion. After enough time at high temperature, the excess boron created localized stress that fractured the boundary and allowed the grains to separate.
To test whether keeping boron below the critical concentration would avoid this failure, the team created a new version of the platinum network with a reduced amount of boron at the grain boundaries. After five hundred hours of heating, the structure remained stable. There was no visible coarsening, and catalytic performance declined by less than seven percent. This result showed that stability is not just a matter of choosing the right element but also of controlling how much of it accumulates and where it goes.
Based on these findings, the authors proposed a three-part strategy for designing stable and active grain boundaries. First, use computational screening to identify elements that will segregate to boundaries without interfering with reaction energetics. Second, use thermodynamic models to calculate the concentration limit beyond which the element becomes destabilizing. Third, develop experimental methods that introduce the element in a way that ensures it accumulates only at the boundary and does not exceed the safe threshold.
This approach challenges the assumption that segregation is always beneficial. It shows that the same element can either protect or destroy a boundary depending on its local environment. The work also offers a path toward designing catalysts that retain their activity over time rather than declining with use. While this study focused on platinum and hydrogen oxidation, the principles apply more broadly. Any material that relies on grain boundaries for performance, whether in catalysis, electronics, or energy storage, could benefit from this kind of precise, predictive control.
By combining modeling, advanced imaging, and careful synthesis, the researchers have shown how to move from empirical guesswork toward rational design. Instead of treating grain boundaries as a liability or as a fixed feature, this work treats them as tunable parts of a system, ones that can be engineered for stability, activity, and longevity at the same time.
For authors and communications departmentsclick to open
Lay summary
Prefilled posts
Nanowerk Newsletter
Get our Nanotechnology Spotlight updates to your inbox!
Thank you!
You have successfully joined our subscriber list.
Become a Spotlight guest author! Join our large and growing group of guest contributors. Have you just published a scientific paper or have other exciting developments to share with the nanotechnology community? Here is how to publish on nanowerk.com.
