Localized Chemistry Department of Chemistry, University of Waterloo

Research

Electrochemistry & Energy Conversion

Understanding the materials we use at electrified interfaces and the mechanisms by which they drive catalysis.

increasing long-range orderaThree regimes of host ordernetworks are schematic — not measured dataamorphousActivity saturatesA few per cent Fe capturesnearly the whole effect.Science 2013 · JACS 2013 · EES 2018 · Chem. Mater. 2019disorderedActivity tracks the averageThe pre-catalytic potentialfollows the mean M–Mdistance.Energy Environ. Sci. 2022crystallineActivity tracks local strainEvery site is defined, buteach dopant strains itsneighbours.ACS Catal. 2022 · Chem. Sci. 2024 →Co K-edgeFe K-edge-20-100102030Relative energy (eV)bNormalised XANESFe and Co K-edges nested, 0–30 % Fe030 % FecDifferencesubtractive analysis−25 to 0 eVasymmetry → lower Tafel slopedStrain vs. activitymeasured hereAlsac & Smith, Chem. Sci. 2024, 15, 15339

Iron substitution has less effect on catalytic activity as the host becomes more crystalline. (a) Three regimes of host order, drawn as pieces of one hexagonal metal-hydroxide sheet. The networks are schematic. Metal–oxygen distances are comparable in all three, and the regimes differ in how far a coherent domain extends. In an amorphous network, a few per cent Fe captures nearly the whole change in catalytic activity, and adding more does little. In a partly ordered material, the response follows an averaged structural parameter. In a crystalline host, every substituted site sits in a defined environment, the lattice limits how far each dopant can strain the sites around it, and the activity tracks that localized strain. (b–d) Normalised XANES at the Fe and Co K-edges across a highly crystalline series of Co(OH)2 containing 0–30 % Fe. A subtractive analysis developed for this work relates local O–M–O bond-angle asymmetry to electronic structure and to the oxygen-evolution Tafel slope. Data from Alsac & Smith, Chem. Sci. 2024, 15, 15339–15351.

Electrochemical energy conversion relies on materials that are rarely perfect crystals. Catalysts and electrodes contain defects, local distortions, and disordered regions, and these features often affect activity and stability more than the average structure does.

We identify these defects and distortions, quantify them, and relate them to catalytic behaviour. Our main tools are X-ray absorption spectroscopy, potential-resolved vibrational spectroscopy, and Fourier-transformed alternating-current voltammetry. With them we follow changes in oxidation state, coordination environment, and lattice distortion as composition, crystallinity, and applied potential are varied, and use what we find to improve models of structure and reactivity.

What we work on

  • Oxygen evolution. The local structure of layered double hydroxides, how that structure changes over long periods of use, non-traditional coordination environments in their lattices, and how lattice strain develops under operating conditions and affects electron-transfer kinetics.
  • Oxygen reduction. Layered perovskites, the Jahn-Teller effect as a way to tune reactivity, and how local pH gradients and other environmental effects shift product selectivity.
  • CO2 reduction. Spontaneous reactions between CO2 and catalyst surfaces, surface reconstruction during operation, and the identification of active sites within grain-boundary assemblies.
  • Electrochemical sensing. Applying what we learn about electrode mechanisms to the design of electrochemical sensors, together with instrumentation companies.

How crystallinity limits a dopant’s effect

The effect of iron substitution on a catalyst depends on how ordered the host material is. The more crystalline the host, the smaller the change in activity that a given amount of iron produces. In an amorphous oxide, a few per cent iron captures nearly the whole change in Tafel slope, and adding more barely moves it. Every site already differs from every other, so the distortion from additional iron adds little to a distribution that was broad to begin with. In a partly ordered material, the response tracks an average structural parameter instead. In a crystalline host, every substituted site sits in the same defined environment, and the lattice energy makes it hard for a dopant to distort the sites around it. The strain that does develop is small and local, and the catalysis tracks it. In our layered hydroxides, this strain can be measured as the difference in O–M–O bond angles between the constituent metals. In Fe-doped materials, the metal atoms adjacent to the iron are activated for catalysis.

As a result, a structure–activity correlation holds only within one crystallinity regime. A descriptor fitted to crystalline samples will not describe amorphous ones. A response that saturates in an amorphous film also says nothing about whether the underlying interaction is weak. This is why we study samples from fully amorphous to highly crystalline.

Why it matters

Once the structural features responsible for activity have been identified and measured, catalysts can be designed around them. That knowledge carries over to related materials, which provides a route to improving catalyst performance.