Understanding competition between CO2 reduction and hydrogen evolution on a gold electrode

In this talk I will summarize our recent efforts in understanding the competition between CO2 reduction and hydrogen evolution on a gold electrode, as studied by rotating ring-disk voltammetry and online Differential Electrochemical Mass Spectrometry. I will show that in neutral media, this competition is extremely sensitive to the electrolyte composition and mass transport conditions. I will also show how by carefully adjusting the CO2 reduction rate to the proton diffusion rate, one can effectively suppress hydrogen evolution in acidic media.

- - - - - - - -

Catalysis for CO₂ electroreduction into energy-dense products, such as CO, formic acid, hydrocarbons and alcohols, represents one of the most promising strategies for CO₂ utilization. Homogeneous catalysts can be immobilized on heterogeneous conductive supports to generate cathode materials for electrolyzers: such heterogenized molecular systems thus combine the advantages of a solid material (e.g., easy recovery of products and catalysts, efficient electron transfer from the electrode support to the catalyst, high Turnover Numbers) with those of molecular complexes (e.g., synthetic control of the electronic properties and the coordination environment of the active sites). This class of hybrid catalysts for CO₂ reduction has been recently described in review articles.^[1]–[2]

In this context, [Ni(cyclam)]²⁺ is known to be a good, stable and selective molecular catalyst for CO₂ electroreduction, however, to our knowledge, there is only one precedent for electrode surface modification with [Ni(cyclam)]²⁺ for electrocatalytic CO₂ reduction ^[3] and this hybrid material proved quite inefficient and poorly selective. Additionally, there is no precedent for the non-covalent immobilization of [Ni(cyclam)]²⁺ on carbon-based nanostructured electrodes.

In the present research, a novel cyclam derivative carrying a pyrene moiety was readily synthesized for the sake of immobilizing a [Ni(cyclam)]²⁺ complex at the surface of a carbon-based electrode. The pyrene-modified complex was immobilized on a carbon nanotube-coated gas diffusion electrode using a non-covalent approach and the novel electrode was characterized electrochemically for CO₂ electroreduction. The complex proved to be much more active in the immobilized form than under homogeneous conditions, with Faradaic Yields for CO production above 90% and current densities up to 10 mA.cm^-2 in acetonitrile/water mixture. The hybrid electrode proved highly stable, leading to impressive turnover numbers (61460 after 4 h electrolysis). This is remarkable since the electrode support is carbon-based and not mercury, so far the best electrode material for CO₂ electroreduction catalysed by [Ni(cyclam)]^{2+ [4]}. The present study confirms the benefits of incorporation of molecular catalysts onto electrode surfaces using the pyrene-CNT approach for CO₂ electroreduction. Additionally, it shows that the [Ni(cyclam)]²⁺ complex provides an excellent platform on which further improvements of hybrid electrodes can be brought.

Chemical synthesis is responsible for significant emissions of carbon dioxide worldwide. These emissions arise not only due to the energy requirements of chemical synthesis, but since hydrocarbon feedstocks can be overoxidized or used as hydrogen sources. Using renewable electricity to drive chemical synthesis may provide a route to overcoming these challenges, enabling synthetic routes which operate at benign conditions and utilize sustainable inputs. We are developing an electrosynthetic toolkit in which distributed feedstocks, including carbon dioxide, dinitrogen, water, and renewable electricity, can be converted into diverse fuels, chemicals, and materials.

In this presentation, we will first share recent advances made in our laboratory on nitrogen fixation to synthesize ammonia at ambient conditions. Specifically, our lab has investigated a continuous lithium-mediated approach to ammonia synthesis and understood the reaction network that controls selectivity. We have developed non-aqueous gas-diffusion electrodes which lead to high rates of ammonia synthesis at ambient conditions. Then, we will discuss how water can be used as a sustainable oxygen-atom source for epoxidation of olefins as well as related oxygen-atom transfer reactions, providing a route to utilize oxidative equivalents in a water electrolyzer. These findings will be discussed in the context of a broader range of electrosynthetic transformations which could lead to local and on-demand production of critical chemicals and materials.

Copper catalysts are unique in enabling CO₂ reduction toward C₂₊ products, such as ethylene, ethanol and n-propanol[1]. C₂₊ product distribution varies significantly over time and depends on the catalysts’ synthesis[2]. While Cu foil produces mainly ethylene with oxygenates as side-products, oxide-derived copper have the exclusive capacity of converting CO₂ and CO to ethanol at high Faradaic efficiencies and low overpotentials[3]. Polarized copper sites, residual oxygen, defects, gran boundaries, and high electrochemical surface area have been deemed responsible for the high CO₂ reduction activity and the more favorable CO-CO dimerization step expected for these catalysts. Specific active sites toward ethylene, ethanol and 1-propanol have been identified for oxide-derived systems by isotopic labeling[4], however they have not beed characterized yet.  

Here, by means of ab initio molecular dynamics simulations on seven oxygen-depleted models derived from Cu₂O and Cu pristine structures, we identified the main ensembles which control the catalytic performance of oxide-derived copper. Upon surface reconstruction and independently from the starting depletion geometry, copper can be classified in three main classes depending on its local coordination and charge: metallic Cu⁰, polarized Cu^δ+, and oxidic Cu⁺, respectively coordinated to 0, 1 and 2 oxygens. These three species form 14 ensembles, such as 4- and 6-coordinated Cu adatoms, Cu₃^δ+O₃, reconstructed (111), (110) and (100) crystalline domains and near-surface oxygens. Together with the high atomic roughness, low coordinated Cu adatoms and polarized sites are responsible for tethering CO₂ and therefore improving CO₂ reduction activity. Metastable oxygens and metallic fcc-(111) or (100)-like Cu facets promote CO–CO dimerization step via a deprotonated glyoxylate species, whose formation is theormodynamically favored and presents a negligible kinetic barrier. Characterized by vibrational fingerprints in agreement with spectroscopic features of oxide-derived copper under CO₂ reduction conditions, this new chemical species could be the elusive intermediate which enables selective C₂₊ production at low onset potentials[2]. This study[5] provides a new theoretical set of concepts for modeling complex structural rearrangements driven by the high surface polarization characteristic of CO₂ reduction conditions and suggest guidelines for new synthetic protocols for the creation of selective catalysts toward ethylene and oxygenates.

Gold and iron-doped nitrogen graphene (Fe-N-C) are among the most active and selective catalysts for electrochemical CO₂ reduction to CO. However, model DFT studies with standard GGA functionals on these materials predict CO poisoning on Fe-N-C and negligible CO coverage on Au, and the latter is contrary to observations in surface-enhanced spectroscopic experiments. Furthermore, the interfacial electric field and solvation can affect the binding strength, and these effects remain an open challenge for ab initio studies. To determine the CO adsorption energy and to assess the accuracy of DFT-GGA results, we return to a surface science experiment: temperature programmed desorption (TPD). By fitting TPD spectra to a model first order kinetic expression including the configuration entropy of the adsorbate, we determine the adsorption energy and equilibrium coverage of *CO on Au step facets [1] and Fe-N-C [2]. We find that the adsorption energies on these materials are remarkably similar, in contrast to DFT-GGA predictions. The results indicate that hybrid functionals are needed on Fe-N-C in order to accurately capture the binding strength. To investigate the impact of the electrochemical environment, we compare water adsorption energies from TPD with ab-initio molecular dynamics (AIMD) calculations. We find that the competition between water and CO adsorption can significantly affect the CO adsorption strength and result in different binding sites for *CO on Au in gas phase and electrochemical environments.

Renewably powered CO2 electrocatalysis presents an opportunity to de-carbonize chemical production. Ultimate application of CO2 reduction will require electrocatalytic systems that provide  reactants, electrons, and products at high rate and efficiency, and that are compatible with upstream and downstream processes. I will outline our progress on membrane electrode assembly based cells to meet this challenge. To accommodate O2 impurities from upstream processes we develop a hydrated ionomer catalyst coating that selectively slows O2 transport and stabilizes the copper catalyst. To increase reaction rate and energy efficiency we develop an adlayer catalyst strategy that increases local CO2 availability and tunes intermediate adsorption for ethylene production. For ethanol production we focus on minimizing product cross-over to the anode, and achieve ethanol production in excess of 10wt% - comparable to bio-ethanol production and compatible with downstream processes. Lastly I will highlight learnings, challenges and opportunities arising from our system scaling efforts in the 2020 Carbon XPRIZE competition.

The classical way to characterize reaction products in electrochemistry involves steady-state electrolysis combined with intermittent or offline product determination, for example with chromatography, nuclear magnetic resonance etc. These methods are excellent to quantify products in terms of faradaic efficiencies and reaction rates, but they offer temporal resolution in the order of several minutes. Therefore, it is not possible to determine the product formation under dynamic conditions. We recently developed an internationally unique method to characterize the products of electrochemical reactions at the time they are formed, the electrochemical real-time mass spectrometry (EC-RTMS) [1-4]. Contrary to previous approaches, EC-RTMS is not limited by the vapor pressure of analytes or the presence of non-volatile salts. I will present he basic principles of EC-RTMS, together with its capabilities exemplified for some electrochemical reactions such as the reduction of carbon dioxide or the oxidation of alcohols.

Photoelectrochemical (PEC) or photovoltaic-driven electrochemical (PV-E) water splitting is considered a promising, renewable technique for the production of the green fuel hydrogen (H₂). Nevertheless, a major bottleneck is that H₂ production through (photo)electrochemical water splitting is financially not as attractive as H₂ generation through steam methane reforming [1].

In this work, we discuss the partial oxidation of water to the commodity chemical hydrogen peroxide (H₂O₂) (E⁰ (H₂O₂/H₂O) = +1.78 V vs RHE) as a financially promising substitute for anodic water oxidation to O₂ [2,3]. Particularly, we will consider two scenarios: PEC and PV-E H₂O₂ production.
First, we demonstrate that the H₂ price can be lowered significantly compared to ‘classic’ water splitting for a PEC system using a techno-economic analysis. Here we consider (i) a near-optimal scenario and (ii) a literature-based state-of-the-art scenario. It will be shown that H₂ production through steam methane reforming is even financially outcompeted.
Second, we will focus on PV-E systems and discuss on an experimental basis the suitability of various electrode materials for anodic H₂O₂ production. We demonstrate that boron-doped diamond yields promising results, reaching Faradaic efficiencies to H₂O₂ over 30% and H₂O₂ production rates slightly over 4.5 μmol min^-1 cm^-2. Finally, implementing the obtained experimental results in (a slightly altered version of) our techno-economic model, we will discuss the most important areas where research should be pursued in to make anodic H₂O₂ production even more attractive.

The development of green industrial processes has been attracting more attention in the last years and it requires renewable energy sources. Large-scale electrification of the energy and chemistry sectors is a crucial condition for the transition from a society based on fossil resources to a sustainable society based on renewable building blocks. Transformation of CO₂ into chemicals has been intensively studied over the past decades because this molecule can be used as C₁ feedstock in electro-organic synthesis due to its abundant, cheap, renewable and non-toxic nature1. Of particular interest of this molecule is the synthesis of two- to four-carbon atom products (C₂-C₄). Concerning the synthesis of C₂-C₄ products, tartaric acid (TTA) attracts attention because it is an important product used in the textile printing, dyeing, pharmaceutical and food industries². In the past years, the price of TTA and its salts has varied as much as 50 %, as might be expected when the source of the raw material is a by-product of another industry.

Herein we describe a study of the electrochemical reduction of oxalic and glyoxylic acids towards a feasible green and sustainable production of tartaric acid in aqueous and/or acetonitrile solvent using silver and lead electrodes.

Our results show that on the silver electrode, for both oxalic acid and glyoxylic acid, the reduction reaction is more favorable towards the dimerization step, leading to tartaric acid, due to the increase in the local pH, while on the lead electrode the step involving the protonation of the intermediate is more favorable, leading to the formation of glycolate.

Techno-Economic Analysis shows that tartaric acid production from glyoxylic acid and from oxalic acid via electrochemical synthesis can be a potential process at an industrial scale. In the present case, the oxygen evolution reaction was chosen as the reaction at the other electrode for practical reasons, but oxygen is a low-value product. Another anodic reaction with a more valuable oxidation product can be selected to increase the profitability of the overall electrochemical process, andthereby decrease the total production costs of tartaric acid.

Fuel cells are the upcoming alternative for the combustion engine. Using hydrogen and oxygen to produce water and electricity, they promise higher efficiencies and no pollution. These green electrochemical reactions, hydrogen oxidation reaction and oxygen reduction reaction (ORR) are excellently catalyzed by noble metals. However, these elements are costly and scarce, limiting the use of fuel cells to only novelty projects. For large scale application, affordable and abundant electrocatalysts are a must.

As an alternative, we and others developed N-doped carbons, promising catalysts at a fraction of the price of noble metals and abundant enough to use in every imaginable application.^[1,2] Inherently, N-doped carbons consist of defect rich disordered amorphous domains and ordered graphitic domains, both active in ORR. But which domain is more active for ORR, amorphous or graphitic? An important question, as decreasing inactive material and thus decreasing electrode area is the difference between hydrogen fuel cells fitting only Hummers or also fitting Smarts.

To answer the question, we changed the ratios between these phases in our well characterized carbon. Using this method, we got three valuable insights. First, as promised we showed which phase N-doped carbon phase is more ORR active. But we did not stop there. We also aimed to understand the post-synthesis modifications and how these affected the N-doped carbon morphology and surface structure. Lastly, we showed that, just like numbers without context, synthesis/modifications without understanding provide wrong conclusions.

DMC is a valuable chemical used on industrial scale for both production (e.g. carbonylation reactions) and in end products (e.g. paints and batteries). DMC is currently formed by thermal catalysis but can be formed electrocatalytically, it is a carbonylation product of methanol¹, formed electrochemically by the reaction 2CH₃OH + CO à (CH₃O)₂CO + 2H⁺ + 2e^-.

The purpose of the work is to understand the fundamental aspects of the carbonylation reaction over an electrode surface, i.e. by heterogeneous electrocatalysis, rather than homogeneous or indirection electrosynthesis. Knowing the steps of formation and the reaction intermediates involved will allow us to develop the catalytic system in a rational manner as well as apply the knowledge to later reactions of interest (e.g. carbonylation of other alcohols, such as phenol, to form valuable products). Understanding the active site and mechanism of the reaction is paramount for rational optimisation strategies.

Different transition metals (Cu, Au, Pt) have been tested for their activity towards DMC production^1,2,3. Au is more active than Pd in terms of current efficiency as well as current density (Figure 1), whereas Cu catalysis DMC synthesis via an inefficient solution phase mechanism. We have used different techniques to understand catalyst activity: in situ infrared spectroelectrochemistry (in situ FTIR) and headspace gas chromatography mass spectrometry (HS-GC-MS). The results that we have obtained so far have been discussed in combination with a theoretical analysis which has helped us understand the mechanisms involved. Further development is underway for optimising the catalyst as well as reactor system.

[1] Šarić, M., Davies, B.J.V. et al, Green Chem., 2019, 21, 6200-6209
[2] Davies, B.J.V. et al, ACS Catal., 2019, 9, 2, 859-866
[3] Davies, B.J.V. et al, J. Phys. Chem. C, 2019, 123, 20, 12762-12772

I will review recent progress in CO2 reduction electrocatalysis toward sustainable chemicals. This will include a discussion of systems that enable the needed high rate of electroproduction in order to render capital costs acceptable. It will also include work on the catalysts and their electrolyte that are capable of producing a relatively high selectivity towards a desired product, such as for example ethylene. Overall I will look at further progress needed to achieve high CO2 utilization, and - crucially - to continue to increase energy efficiency on the path towards prospective technoeconomic viability. The approaches used will include computational studies as well as nanomaterials synthesis and in situ spectroscopies.

Electrochemical CO2R is a particularly challenging process from both modelling and experimental perspectives.  The challenge arises from the complexity of its reaction network and the sensitivity of its activity and selectivity to the electrolyte composition and the associated interfacial field effects. In this talk, I will first discuss developments in methods to model the electrochemical interface - the accuracy of widely applied continuum approximations of the electrolyte, as well as a new unified approach to obtain electrochemical barriers from both explicit and implicit solvent simulation approaches.  In particular I emphasize that the surface charge gives the most appropriate proxy of the local potential (vs. the traditionally considered work function).  Then, I will discuss the impact of the electrolyte composition - pH and ions - on electrochemical CO2R activity of both transition metals as well as supported single atom catalysts.  Finally, I discuss the implications of our findings for electrocatalyst design and electrolyte engineering.

We present a novel multiscale numerical approach to estimate in an efficient, accurate, and high-throughput fashion the current density and mass activity of individual Pt nanoparticles as well as morphologically diverse but size-selected samples for Oxygen Reduction. In this newly developed framework we adapt the computational hydrogen electrode model to forecast currents from reactions taking place at any active site, by means of a statistical learning approach and a geometrical descriptor that bridges the active site topological and catalytic properties.

By exploiting the above framework we then propose specific design rules of Pt-nanoparticles for the electrochemical reduction of molecular oxygen identifying the size-range up to 5.5 nm as the one where structural effects are fundamental and can not be neglected. We confirm the peak of the activity of defected and concave polyhedra at 2-3 nm whilst spherical but amorphous isomers are the most active between 3-5 nm, with a large mass activity, 2.7 A/mg. Finally, we discuss possible discrepancies in the experimentally measured mass activity of size-selected samples in terms of the different distributions of Pt-isomers in each specimen.

The extension of our model to other electrochemical reactions will be also discussed if time allows.

Electrochemical technologies are considered crucial to future sustainable energy scenarios. However, the complexity of these systems has hindered the first-principles understanding and design of materials for such energy conversions. I'll describe our approaches to understand electrochemical reactions based on a grand-canonical---or constant-voltage---approach to atomistic simulations, and how these calculations can be directly linked to experimental observations. I will cover the methodology we have developed, which we refer to as the "solvated jellium" approach, which enables electronically grand-canonical calculations to run at comparable cost to canonical calculations. A key ramification of processes at constant potential is that the number of electrons involved in a reaction is not known a priori, and is not in general an integer quantity. Examples will be provided on how we use these approaches to inform and guide experiments.

This work was supported by Grant 9455 from VILLUM FONDEN and the European Union’s Horizon 2020 research and innovation programme under Marie Sklodowska-Curie Grant Agreement 713683.A persistent challenge in electrocatalysis is the limited understanding of the microscopic processes occurring at electrified solid/electrolyte interfaces. Even single-crystal studies which treat well-defined surfaces are often challenged to provide a deeper insight due to a lack of sensitive analytical methods. Atomistic simulations potentially fill this gap of missing analytics but may not align with experimental observations. Possible discrepancies may arise not only from the absence of the necessary complexity in computational model systems but also due to the limitations in a thermodynamic description. Here we present a rigorous joint experimental–theoretical study on the single-crystal (SC) Cu/aqueous interface based on sufficiently clean experiments and a kinetic description of the electrochemical processes in our theoretical model. In the direct comparison between simulated and experimentally measured voltammograms, we achieve agreement within typical computational uncertainty. The experimental and computational consensus allows us to unequivocally identify the *OH adsorption feature in the fingerprint region of Cu(110), Cu(100), and Cu(111) SCs under alkaline conditions. In our computational treatment we find the inclusion of hydrogen evolution reaction kinetics to be crucial for an accurate steady-state description that gives rise to a negligible H* coverage on the Cu surface [1]. In contrast, a purely thermodynamic description of the H* coverage through a Pourbaix analysis would incorrectly lead to a H* adsorption peak. We further demonstrate how we can achieve near quantitative agreement of peak positions between theory and experiment via sensitive improvements of the solvation energies using corrections predicted by molecular dynamics simulations [2]. Our work demonstrates the importance of reaction kinetics to elaborate on the electrocatalysts’ surface composition and how a refined description of the electrochemical interface, particularly through solvation, improves accuracy.

Using the electrode potential and electrolyte to manipulate reaction thermodynamics and kinetics forms the backbone of all electrochemistry and electrocatalysis. Especially important reactions to understand are proton-coupled electron transfer (PCET) reactions forming the mechanistic basis of e.g. oxygen, CO2, and N2 reduction and hydrogen evolution reactions. While clever and cheap schemes for evaluating electrochemical thermodynamics and kinetics have been developed, a rigorous treatment is needed to test the accuracy and to define well-controlled computational models.

In my contribution I will present a rigorous theory and numerical techniques to simulate electrochemical solid-liquid interfaces using grand canonical ensemble density functional theory (GCE-DFT) [1] – this approach provides an exact theory and well-defined approximations to compute thermodynamics as a function of the electrode potential and electrolyte concentration. Besides thermodynamics, I will present the newly established generally valid GCE rate theory (GCE-RT) to address PCET reaction kinetics as a function of the electrode potential.[2,3] The GCE-RT can account for (non-adiabatic) proton and electron tunneling which may significantly contribute to PCET kinetics and long-range electron transfer, respectively.[3] Besides theory, I will present how the thermodynamics, kinetics, and nuclear quantum effects of a gold-catalyzed Volmer reaction can be addressed and understood from the atomic scale with GCE-DFT and GCE-RT[2-3]. I will also show how GCE-DFT can be used for parametrizing effective Hamiltonians for predicting and explaining electron transfer kinetics and experimental measurements (in preparation).

The presentation provides an account on the recent theoretical and methodological developments in using constant potential, grand canonical ensemble DFT for simulating electrochemical interfaces and reactions. These developments enable simulating both the thermodynamics and kinetics of electrochemical reactions, including both classical adiabatic inner-sphere reactions as well as e.g. outer-sphere reactions, non-adiabaticity, and nuclear tunneling at the DFT level.

The Open Catalyst Project aims to develop new ML methods and models to accelerate the catalyst simulation process for renewable energy technologies and improve our ability to predict activity/selectivity across catalyst composition. To achieve that in the short term we need participation from the ML community in solving key challenges in catalysis. One path to interaction is the development of grand challenge datasets that are representative of common challenges in catalysis, large enough to excite the ML community, and large enough to take advantage of and encourage advances in deep learning models. Similar datasets have had a large impact in small molecule drug discovery, organic photovoltaics, and inorganic crystal structure prediction. We present the first open dataset from this effort on thermochemical intermediates across stable multi-metallic and p-block doped surfaces. This dataset includes full-accuracy DFT calculations across 53 elements and their binary/ternary materials, various low-index facets. Adsorbates span 56 common reaction intermediates with relevance to carbon, oxygen, and nitrogen thermal and electrochemical reactions. Off-equilibrium structures are also generated and included to aid in machine learning force field design and fitting. Collectively, this dataset represents the largest systematic dataset that bridges organic and inorganic chemistry and will enable a new generation of catalyst structure/property relationships. Fixed train/test splits that represent common chemical challenges and an open challenge website will be discussed to encourage competition and buy-in from the ML community.

Bimetallic M_A(II)M_B(III) layered double hydroxides (LDHs), in particular Fe(III) containing ones, are among the most active and widely studied catalysts for the oxygen evolution reaction (OER) in alkaline electrolytes. However, little is known about the dynamic structural transformation of the inactive alpha-phase to the catalytically active gamma-phase under concomitant ion exchange.

I will highlight recent advances of our understanding of the active catalyst structure, the reactivity, and the likely molecular reaction mechanism of the electrocatalytic evolution of molecular of oxygen on M(II)M(III) layered double hydroxide materials. Focus is placed on NiFe and CoFe catalyst using operando X-ray and DEMS analysis, coupled to ab Initio Molecular Dynamics (AIMD) modeling [1].

Exploration of precious-metal-free catalysts for water splitting is of great importance in developing renewable energy conversion and storage technologies. Here, we investigate the link between the oxygen evolution reaction (OER) activities and the electronic properties of pure and first-row transition-metal-doped AlN and GaN monolayers. We find that Ni-doped layers are singularly appealing because they lead to a low overpotential (0.4 V). Early transition-metal dopants bind the intermediate species OH or O too strongly. The late transition-metal dopants Cu and Zn bind intermediate species too weakly which leads to large overpotentials or to no OER activity at all. Ni dopant breaks this trend and stabilizes instead the OOH adsorbant which can be correlated with a switch from a high- to a low-spin state of the dopant atom. This ability to change spin states offers an exciting ingredient for the design of OER catalysts [1].

The morphology of semiconductor photoelectrodes and metallic electrodes significantly affects the performance of (photo)electrochemical devices. Complex anisotropic morphologies are important for overcoming performance limiting bulk transport properties of semiconductor materials or enhancing the selectivity of CO₂ reduction electrodes, but are often also an unintended outcome of the fabrication process. A better understanding of morphology-induced transport limitations of (photo)electrodes is needed to better understand the kinetics, degradation, and transport limitations, and to subsequently guide mesostructured design and performance optimization. We use direct pore-level simulations for the coupled transport characterization of mesostructured (photo)electrodes. I will introduce and discuss four short examples to show the importance of local heterogeneities and reaction environment on the performance and degradation: i) stochastic, particle-based lanthanum titanium oxynitride (LTON) photoanodes for water oxidation, ii) degradation in porous photoanodes made of compound semiconductors, iii) structured, inverse opal silver electrodes for CO₂ reduction, and iv) CO₂ reduction in gas diffusion electrodes.

In this work, we study the role of alkali metal cation concentration and electrolyte pH on the kinetics of hydrogen evolution reaction (HER) at Au electrodes in order to discern a clear activity descriptor for this reaction under alkaline conditions. We find that at moderately alkaline pH (pH =11), increasing cation concentration significantly enhances the HER activity on Au electrodes (reaction order » 0.5). We propose that the cations near the interface interact favorably with the transition state of the rate-determining Volmer step by stabilizing the (partially) negative hydroxide which is being split off from the reacting water molecule (*H--OH^δ---cat⁺). Remarkably, at higher pH, the effect of the concentration of alkali cations is diminished, and it is even negative at pH=13. Furthermore, capacitance curves obtained via impedance spectroscopy suggest that the electrolyte pH also influences the near-surface composition of the electrolyte such that an increasing electrolyte pH leads to a corresponding increase in the near-surface cation concentration. This results in an apparent pH dependence for the HER activity on the Au electrodes where synonymous to the cation concentration effect, saturation is observed at extreme pH values (pH 13 to pH 14). We attribute the saturation and even inhibitive effects observed at high pH and at high cation concentration to a blockage of the surface by cations beyond a threshold concentration.

This work shows that the electrolyte pH and the near-surface cation concentration are inter-dependent parameters, which cannot be de-coupled while probing the metal-electrolyte interface in the alkaline media. Hence, our work provides foundational insights on the complex molecular origins behind the pH dependence of HER, and we believe that these insights will be instrumental in guiding the design of optimized catalyst-electrolyte conditions for HER in the alkaline media. Finally, we also extend the understanding gained from these studies to rationalize the unusual role of mass transport in determining the HER activity by controlling the local pH using rotating disk electrode voltammetry (1).

Novel and efficient devices taking advantage of renewable energy sources, supercapacitors, water-splitting devices and high-capacity batteries, as well as various sensors, are hot topics within the society, and at the same time are challenges for scientific community. The latest approaches and dynamic development in those fields benefit from materials research at nanometre-size level. Significant efforts are being aimed especially at fabrication of functional nanomaterials where two materials are combined and synergistic effect is observed which is atypical for the single elements [1, 2]. To fit in with the latest trends, we combined titania nanotubes (TiO₂NTs) with gold species in order to obtain photoactive material. Moreover, we utilized laser irradiation during fabrication route that apart from being rapid and easily scaled up to industrial level also allows for modification of the sample in the specified area of any shape. Base material, i.e. TiO₂NTs with deposited Au layers, underwent treatment by means of UV laser (355 nm, 12 ns, 2 Hz, 30-240 mJ/cm²). SEM inspection revealed the formation of Au nanoparticles on the rims of the tubes for mild irradiation conditions while increasing of laser beam energy led to melting of the top layer of the nanotubes. Nevertheless, underneath, the initial architecture of the TiO₂NTs remained intact. Structural measurements confirmed the anatase phase of the material and no shift of the main anatase mode peak was observed. The optical studies indicated that laser processing is causing the narrowing of the energy bandgap, which is of key importance for future applications in solar-driven processes. For the optimised, laser-modified 5 nm Au film deposited onto titania nanotubes, the current density registered in dark exceeds 2 mA/cm², while for the 10 nm Au coating – over 4 mA/cm² at 1.5 V. When the materials were illuminated by simulated solar light, the current densities reach over 3.5 and almost 6 mA/cm², respectively. It was also proven that the concentration of donor density increases with the laser fluence, and for TiO₂NTs covered with a 10 nm Au layer and irradiated with 240 mJ/cm² laser fluence, it drastically increases in comparison to the other configurations. This indicates the superior unique effect arising when titania covered by gold film is treated by laser beam under optimised conditions. We strongly believe that Au-TiO₂ nanomaterial with boosted photoactivity can find application in water splitting simulated by the Sun, as water photo-assisted decomposition was observed in the studied case.

The oxygen evolution reaction (OER) is a key process in the general development of sustainable solar fuels, and essential in the progress/growth of an alternative hydrogen economy. This is a very demanding reaction, since a OER catalysts must be fast and efficient, but also stable and robust to high oxidation potentials, water, oxygen, radicals, acids or bases, etc. Several earth abundant metal oxides have appeared as excellent OER electrocatalysts in alkaline conditions, thanks to the robust stability of metal oxides in high pH.¹ The most common agreement suggests that FeNi oxo-hydroxides are arguably the best OER catalyst at pH > 13,² although many other binary and ternary candidates appear close in activity.³ The complexity of this reaction makes difficult to understand and optimize the performance of these catalysts, with the remaining goal to minimize overpotentials, and to maximize current densities. Indeed, after thousands of experiments with multiple phases, the latest progress reports deal with further nanostructuration and smart processing of the materials to increase active surface area.⁴

In this presentation we will introduce some novel strategies to enhance OER activity in alkaline conditions. On one end, we will demonstrate how the use of magnetic fields may boost the performance of classic magnetic metal oxides.⁵ On the other end, we will show how a non-redox doping approach opens alternative (and faster) mechanistic pathways.⁶ Both strategies were developed from the combination of experiment and theory, as an example of the powerful contribution that computational chemistry can bring to the field, opening new alternatives difficult to disclose or understand exclusively from experimental data analysis.

Electrocatalytic conversion of water and CO2 into fuels and chemicals has obtained large attention as potential routes for large scale energy storage and renewable feedstock. However, the scaling and intensifying electrocatalytic CO₂ conversion faces many challenges. My work targets electrocatalytic reduction processes for upscaled reactors, operation at a typical current density > 200 mA/cm².

One of the well-known issues at these high current density is the mass transport limitation of e.g. dissolved CO₂ in aqueous phase. Gas-phase CO₂ conversion has been extensively explored in literature to mitigate the CO₂ diffusion limitation. At the same time, electrochemical reduction in vapour phase introduces challenges regarding water management and product cross-over. Therefore, new strategies that can boost the mass transport in water-based electrochemical systems do have potential for large scale CO₂ reduction. Moreover, other electrochemical systems, such as flow batteries and membrane technologies, benefit from the same mass transfer enhancement in water-based systems. For enhancing mass transfer in aqueous electrochemical systems, gas bubbles play a central role. For many processes, gas bubbles decelerate the process via electrical resistance and shielding catalytic area. However, when controlling the gas bubble movements, we can leverage the convective properties of gas bubbles to improve the mass transfer. In this talk, I will highlight the perspectives for CO₂ reduction in gas phase, but also possibilities for CO₂ reduction at high current density in aqueous phase.

The development of sustainable, fossil-free strategies to synthesize fuels and added-value chemicals has raised enormous interest in the last years, in order to provide reliable energy vectors as well as the feedstocks needed for the chemical industry at a global scale. In this context, water oxidation stands out as one of the preferred reaction to provide the protons and electrons needed for electrochemical conversion processes, although this reaction is considered a kinetic bottleneck and consequently, the development of electrocatalytic materials that effectively oxidize water is essential for improving the efficiency of the overall electrochemical conversion process. Currently, the most efficient water oxidation catalysts (WOC) are based on iridium and ruthenium oxides, IrO₂ and RuO₂. However, due to the scarcity and high-costs of these precious metal-based oxides, other alternatives need to be explored. Ni-based materials constitute one of the best alternatives due to their high electrocatalytic activity and stability under alkaline conditions, as a consequence of their high electrical conductivity and corrosion resistance. Indeed, large scale commercial Liquid Alkaline (LA) electrolyzers, preferentially use nickel-based anodes. Herein, an “Earth-abundant” undoped Nickel oxide (NiO_x) electrocatalyst for water oxidation was investigated. An up-scalable and low-cost solution-based sol-gel synthetic route was employed to obtain non-stoichiometric NiO_x electrocatalyst baked at mild temperatures from 100 to 500 ºC. The catalytic activity towards water oxidation was found to be inversely proportional to the baking temperature. The defective and amorphous nature of the NiO_x electrocatalysts baked at the lowest temperatures (< 200 ºC) was assessed by a wide range of different structural, optical and spectroscopic techniques, leading to a higher acceptor density, attributed to a higher density of structural/electronic defects. Our champion NiO_x catalyst grants 358 mV of overpotential at 10 mA cm^-2 and more than 60h of continuous operation without significant losses.

Electrochemical reduction of CO2 offers to decrease the atmospheric emission of a greenhouse gas, meanwhile producing valuable chemicals – hence closing the “artificial carbon-cycle”. One of these potential products is CO, which combined with H2 could (partly) replace the fossil fuel based organic synthesis methods, and could be a precursor for the formation of a large variety of chemicals including dimethyl ether or specialty wax.

We present a zero gap electrolyzer cell for the continuous-flow electrochemical reduction of CO2, and its possible scale-up avenues in terms of lateral size increase and the construction of multi-layer electrolyzer stacks.[1] Notably, all these structural changes involve the fine-tuning of the operation conditions, achieved by using a custom-designed, almost fully automatized test station. Using commercially available silver catalyst in a gas diffusion electrode configuration we show the selective formation of CO (with H2 being the only cathodic by-product) with similar efficiency metrics in the different sized electrolyzer cells. Further, employing a poly(aryl piperidinium)-based anion exchange membrane (PiperION) with high carbonate conductance in conjunction with a tailored electrolyzer cell structure, we present high partial current densities for CO production, while maintaining high conversion, selectivity at a reasonable cell voltage.[2]

Microfluidic and polymer electrolyte membrane (PEM) reactor designs with different configurations were used for studying electrochemical CO₂ reduction at room temperature, each having particular advantages. PEM based reactors with zero-gap configuration containing a membrane electrode assembly (MEA) might offer several advantages over other device architectures such as having lower ohmic drops at high current density, higher volumetric energy density, making them more suitable to scale-up.^1,2 However, PEM based and microfluidic reactors with flowing liquid catholyte are the most numerous configuration in electrochemical CO₂ reduction so far.^3,4 The studies in gas-fed electrochemical cells with flowing liquid catholyte provided crucial information on the effect of process conditions and material parameters to the selectivity and activity of the electrocatalytic process. Although this configuration might suffer from huge ohmic loses at high current density and require more technical control for scaling up ,from both a fundamental and applied perspective, it is important understand the effect of mass transport and process conditions on the performance and catalyst screening in these systems.

We will present results of a 2-D transport model for a gas diffusion electrode performing CO₂ reduction to CO with a flowing catholyte, including the concentration gradients along the flow cell, spatial distribution of the current density and local pH in the catalyst layer. The model predicts that both the concentration of CO₂ and the buffer electrolyte gradually diminish along the channel for a parallel flow of gas and electrolyte as a result of electrochemical conversion and non-electrochemical consumption. The effect of concentration gradients along the flow channel on the current density distribution becomes prominent at high conversions (e.g. current density) when compared to the ohmic drops across the electrochemical cell (Figure 1a), and a strong variation of the electrochemical performance is observed along the flow path (Figure 1b). In addition, the contribution of concentration overpotentials to overall potential losses dramatically changes with the CO₂ gas inlet feed flow rate, which results in differences in outlet concentrations at high conversions. Fundamental and practical implications of our findings to electrochemical CO₂ reduction are discussed particularly at high single-pass conversions.

The longevity of catalysts is crucial for making industrial scale applications economically feasible. The high activity of platinum favors its applications in electrochemical energy conversion systems, although platinum clearly suffers from degradation during fuel cell operation. The degradation is caused by the nucleation and growth of nanoislands, which undesirably roughens the surface [1, 2, 3, 4]. It has been shown that the extent of this roughening can be tuned by including additives, varying the pH of the electrolyte, or by applying different potential windows [5, 6, 7].
Considering fundamental surface science background knowledge, in particular, adatom density, Gibbs-Thomson pressures, and critical nuclei sizes, we set out to explore the roughness evolution on Pt(111)-like surfaces with different step densities. Intending to decrease the roughness evolution, we show that it is possible to smartly tune the step density with respect to the growth speed: this results in a significantly decreased interface width (roughness) and, when pushing the limits, to an even completely flat single crystalline surface exhibiting only the preexisting steps without any additional roughness build-up.

 

CO₂ emissions are significantly contributing to anthropogenic climate change, but at the same time the costs of sustainably based electricity has greatly been reduced in the last decade.  This has provided a perfect economic and societal opportunity for electrochemical CO₂ reduction into chemicals. However, this field is not technologically mature yet, and there are many engineering details that needs to be investigated.  This talk will focus on understanding the complete carbon balance on low temperature electrochemical CO₂ reduction reactors operating at industrial relevant current densities (100-500 mA/cm²). The focus will be primarily be on using a gas diffusion approach with a liquid catholyte using sputtered Cu as a catalyst.

By employing a post-reactor volumetric flow rate to measure outlet flow rates, highly accurate faradaic efficiency measurements were achieved. Furthermore, these results show that if one instead assumes inlet gas flowrates, this can overestimate faradaic efficiencies of gas products by up to 25%. By careful analysis of both gas and liquid products we show a near 100% faradaic efficiency as well as completely accounting for all the carbon in our reactor.  Our liquid analysis showed glycoaldehyde and ethylene glycol, two products never seen before in high current density electrochemical CO₂ reduction reactors

Using an anion exchange membrane, we show significant crossover to the anode of anionic species such as acetate, and formate, as well as carbonate crossover and subsequent CO₂ evolution on the anode.  We show that the ratio of CO₂ to oxygen evolution on the anode is 2:1 throughout a current density range of 100-300 mA/cm², however the time it takes to equilibrate at this ratio is both current dependent and dependent upon the catholyte and anolyte reservoirs.  This talk will also discuss variations in pH at both the anode and cathode as a function of operating parameters and give insight into the potential advantages and pitfalls of various reactor designs.