Elettra | BACH Probing nanoporous copper catalyst layers under bias The development of electrochemical processes based on renewable electricity and feedstocks could help reduce the carbon footprint of the chemical and power sectors. In particular, the prospect of electrochemically converting carbon dioxide (CO2) and carbon monoxide (CO) to value added chemicals such as ethylene, ethanol and acetate has gained traction over the past decade. To accelerate the commercialization of these carbon utilization technologies, energy efficiency has to be maximized and this entails lowering power consumption and increasing current efficiency. To achieve the former, the overpotentials at the anode and cathode plus the overall cell electrical resistance must be minimized by using high surface area catalyst layers and short inter-electrode distances. To maximize selectivity, it is necessary to tailor the electronic and the porous properties of the catalyst layer. In line with Sebatier’s principle, Cu possesses unique selectivity in the aforementioned reactions because it has a quasi optimal binding energy towards *CO and *H adsorbates. In contrast, other metals such as Ag or Pt have a weaker and stronger interaction with *CO, leading to poor *CO adsorption or poisoning, respectively. In contrast, ensembles of Cu atoms allow C-C coupling and formation of hydrocarbons and alcohols. Our previous study has shown that assemblies of copper nanoparticles and nonporous magnetron-sputtered thin-films selectively yield methane and ethylene depending on catalyst loading and thickness. Nevertheless, it remained unclear to what extent surface area and porosity affected that reaction outcome and if the initial structure was maintained when in contact with the electrolyte at open-circuit or under negative electrical bias. To fully understand the dynamic processes taking place inside the catalyst layer under such conditions, spectroscopic analysis was needed. Therefore, we employed in-situ soft x-ray Absorption Spectroscopy (XAS) at the beamline for advanced dichroism (BACH) at Elettra to probe the Cu L-edge region and identify the specific electronic transitions characteristic of Cu and its oxides under electrical bias. Using a suite of complementary techniques and CO reduction monitoring, our goal was to quantify structural and electronic properties and provide insights that would allow improved design and modeling of future catalyst layers. In this work, we employed the stardust machine (a multiple ion cluster source with an aggregation zone) at the institute of materials science in Madrid (ICMM) to fabricate Cu catalyst layers. By varying the source conditions and oxygen flow into the aggregation chamber, 0.35-2 µm-thick films consisting of sub 10-nm copper oxide nanocrystals were formed on gas diffusion electrodes. Their porosity was found to be 66% based on accurate measurements of loading and thickness. In addition, electron tomography and segmentation of the reconstructed volume yielded similar porosity values and a specific surface area of 94 m2/g on average. Analogous catalyst layers were similarly deposited onto Au-coated SiNx membranes for Cu L-edge XAS analysis (Fig. 1 a) in realistic conditions. The membranes were mounted in air onto the microfluidic electrochemical cell for in-situ XAS experiments for studying the stability of the Cu oxides before and after initial exposure to alkaline media. Such environment is produced locally inside a real electrolyzer if an OH-containing electrolyte or membrane is placed in contact with the catalyst layer. Using the porosity values previously obtained, the effective medium density of the film was calculated (3.7 g/cm3) so that the probing depth at the attenuation depth was in the range of 100-200 nm (Fig. 1 b). That meant that only the bottom part of the film was probed by the x-ray beam. In the dry state, the composition of the nanocrystal film consisted of 76% Cu2O and 24% CuO, judged by the magnitude of the peaks at 933.7 eV and 931.3 eV and by their respective absorption cross-sections (Fig. 1 (d)). Upon bias, the peaks attributed to CuO and Cu2O disappeared, and the typical fingerprint spectrum of metallic copper appeared. At the same time, the transition from open-circuit to negative potentials was accompanied by a clear reduction wave and negative current in the recorded voltammogram (Fig. 1 c). Considering the different absorption cross sections of CuO, Cu2O and Cu, we found that there was no loss of material inside the probed region. This indicated that there was no significant contraction of the film (as metallic Cu is denser than Cu2O) in the direction perpendicular to the electrode upon reduction, nor dissolution. Upon cycling to more oxidative potentials, the catalyst layer first oxidized to Cu(I), then partially to Cu(II), in agreement with the observed oxidation wave. When potential control was released, the measured oxidation state of the nanocrystals was found again to be predominantly Cu(I), as with the initial measurement upon immersion. These measurements proved that this experimental protocol caused little to no structural modification inside the thick catalyst layer. This motivated us to probe the electrochemically active surface area of similar films which were deposited on gas diffusion electrodes using Pb underpotential deposition, which revealed significant differences between the nonporous and nanoporous Cu samples (Fig. 1 e). Combined with the initial surface area derived from tomography, we have concluded that the nanoporous catalyst layers effectively utilized up to 45% of their initial surface area during reaction; their roughness (Fig. 1 f) closely aligning with trends in CO reduction product distribution and alcohol crossover, which are both a function of accessible reaction volume and gas-liquid contact area. Elettra Highlights 2024–2025 110
RkJQdWJsaXNoZXIy MjQ4NzI=