Mobile atoms in nanoscale electrocatalysts
Our study focused on electrochemical and electrical driving forces behind structural transformations in geometrically complex nanoelectrocatalysts used in carbon dioxide electroreduction and other cathodic processes of intense interest.
Structural complexity in nanoscale electrocatalysts often correlates with their superior catalytic performance in useful cathodic reactions such as electroreduction of water, CO2 or N2 to fuels and valuable chemicals. Such structural design is highly promising for the industrial implementation of these materials, which brings up the question: how stable are these morphologies under bias and how can we predict it? To date, only a fraction of the reports on nanocatalysts with complex shapes includes notes on their structural (in)stability during electrolysis (e.g., see Supplementary Tables 1–3 here). The observed general trends are somewhat contradictory and inconclusive. To fill this gap, we set out on a journey to closely study the structural behaviour of shaped metal nanoparticles under bias and mechanistically understand the structural transformation phenomenon.
Our study subjects were Au and Pd nanocages and branched nanoparticles, which we subjected to electrolysis under typical CO2 reduction and hydrogen evolution reaction conditions. By examining these multiple structures after different electrolysis times using high-resolution scanning electron microscopy, we found that they can restructure, depending on the material and the reaction taking place on their surface. Geometric definition can even completely degrade, as in the case of Au nanocages in CO2 reduction electrolysis.
|Our observations suggested that the metal atoms in these structures can move, and the reaction parameters play a role in this process. Using DFT modelling, we indeed found that adsorbed reaction intermediates can decrease the energetic barrier to surface migration (see cartoon on the right). Other factors affecting the surface mobility include the catalyst material, exposed facets, and likely other species beyond reaction intermediates that can interact with the surface.|
Besides the fact that the particles could structurally change, we noticed that these transformations were not quite spatially uniform. Specifically, more pronounced changes occurred between the particles and at the interface with the substrate: these areas are less accessible to the reaction intermediates and other species. Thus, we needed to consider other driving forces at play to explain this behaviour. To this end, we performed Multiphysics calculations to map out how the charge carriers flow through the metal structures towards the interface with the electrolyte. In the resulting 3D simulations, we found that small constrictions in nanoscale geometries incurred very high current densities during normal electrode operation. Note the current density within the metal electrode where a branched nanoparticle interfaces with the substrate:
A similar effect in microelectronics is referred to as current crowding, and is known to lead to electromigration, i.e., mass transfer of the conductor material induced by so called electron wind force (see Supplementary Note 2 here). This process is generally more pronounced at the surfaces and along grain boundaries and other defects. The surface electromigration of gold atoms can be illustrated with this simplified cartoon:
In microelectronics, we generally want conductor features to be dry because in a wet conductive environment electromigration can be further accelerated. Consequently, in the case of electrolysis, this process can happen within the electrodes comprised of geometrically complex nanoparticles. Specifically, areas experiencing current crowding and containing grain boundaries, interfaces between multiple particles, and particle surfaces with various adsorbed species could be prone to experience atomic mobility. Electromigration in this context has been previously overlooked.
Overall, our results showed that the surfaces can be mobilized by reaction intermediates, while shapes with geometric constrictions can lead to current crowding and subsequent electromigration of the electrode material. The presented ab initio and numerical methods can be used for stability assessment, and potentially for creating structures that can be activated in situ in ways we usually don’t think of. In certain circumstances, the mechanisms of the reactions themselves may be affected by the presence of mobile surface atoms of the heterogeneous catalyst material, which jiggle on this surface without falling into potential minima.