The development of sustainable strategies to produce added-value chemicals and fuels using renewable resources is particularly attractive to promote a transition towards a more sustainable energetic landscape, overcoming the dependence of fossil fuels at a global scale. One of the most promising alternatives involves the use of renewable electricity (wind, solar, hydropower, etc…) to power electrochemical conversion processes, which convert abundant molecules (e.g., water, carbon dioxide, and nitrogen) into higher-value products (e.g., hydrogen, hydrocarbons, oxygenates, and ammonia). In this context, the use of materials based on earth-abundant elements (Fe, Ni, Cu, etc…) is key to ensure the scalability and large-scale implementation. We will discuss the use of some earth-abundant (photo)electrocatalysts based on these elements for different reactions of wide technological interest, also focusing on the mechanistic understanding of the physical-chemical processes taking place during operation.

Photoelectrochemical (PEC) water splitting is an attractive method for the production of hydrogen from renewable energy sources. We have in recent years focused on both oxygen evolution reaction and hydrogen evolution reaction materials. In this talk, we want to summarize photoelectrode and surface engineering paths to high-performance electrodes for water splitting.

Firstly, we will report on BiFeO₃ (BFO) which has recently been identified as a promising photocathode material due to its light absorption and photo-electrochemical properties [1-3]. For practical applications, however, the PEC performance of BFO needs to be improved, which requires understanding of the aspects that limit its activity. We will present the effect of the ratio of Bi to Fe in the precursor solution of the sol-gel synthesis on the properties of BFO thin films [4]. Thin films with a stoichiometric ratio of Bi:Fe and thin films with 10% excess of Bi are prepared on fluorine-doped tin oxide. While bulk characterization techniques (XRD, RBS) show the formation of phase-pure BFO, surface characterization techniques (XPS, LEIS, TEM, ToF-SIMS) indicate Bi enrichment on the surface. The light absorption and band gap do not change upon adding 10% excess Bi in the precursor solution. However, the current density of the excess Bi samples is nearly two times larger than that of the stoichiometric BFO film at 0.6 V vs RHE. Electrochemical impedance spectroscopy explains this improved performance in terms of a lower recombination rate and a lower charge transfer resistance in Bi excess films. The lower recombination rate is attributed to less defects in the thin films, i.e. less Bi and O vacancies (XPS, TOF-SIMS). The low charge transfer resistance is attributed to a distinct Bi-oxide top layer that was detected on all samples (XPS, TOF-SIMS, LEIS, TEM). In summary, we can conclude that adjusting the Bi:Fe ratio is an effective manner to enhance the PEC performance of BFO thin films for hydrogen evolution reaction. 

Secondly, we will discuss a series of papers where we use WO₃ as oxygen evolution material for photo-electrochemical water splitting [5-8]. We start with discussing physical and chemical defects in WO₃ thin films and their impact on photo-electrochemical water splitting. We then transfer the thin film to scalable substrates, such as Si wafers, and discuss how the performance is boosted with WO₃/n-Si heterostructures. The role of Si and its impact on interface engineering is investigated. Then, we extend the study to the third dimension by investigating micro-pillars and nanowires and the impact of the three dimensional structure through surface area and light management on the performance.

We will conclude with sketching our further efforts to high-performance electrodes via photoelectrode and surface engineering using and developing further paths to scalable methods.

Moving from materials to devices is unthinkable without a preceding thorough assurance of the durability of all of their constituent components. Considering the crucial green energy technologies, photovoltaic cells for renewable electricity generation and water electrolyzers for hydrogen production using renewables already demonstrate the required operational stability over years, if not decades. Unfortunately, the same cannot be said for the cells used for the photoelectrochemical water splitting. As for now, with numerous efforts toward optimizing the activity, the more significant challenge of tailoring the durability of photoelectrodes to meet industrially relevant levels remains. Considerable research efforts are necessary in order to improve our understanding of the degradation processes and develop mitigation strategies. Dissolution (as part of corrosion) is one of the degradation processes. This talk will present our most recent results on the dissolution of representative photoelectrodes (photoabsorbers with or without co-catalysts) during water splitting.

Over the last years, the dissolution stability of different electrocatalysts has been investigated in our group using an inductively coupled plasma mass spectrometer (ICP-MS) directly connected to an electrochemical cell [1]. In this configuration, time- and potential-resolved dissolution analysis is possible. Recently, such a cell was equipped with a solar simulator, Air Mass 1.5 G filter, and monochromator, allowing standardized photoelectrochemical measurements [2]. The dissolution stability of representative WO₃, BiVO₄, and Fe₂O₃ photoabsorbers was investigated in different electrolytes with or without a co-catalyst overlayer [3-5]. With the help of electrolyte and interface engineering, stabilization of the photoabsorbers was demonstrated. The gained insights can then be utilized to advance synthesis and operation approaches of novel materials with improved photostability.

Besides dedicated fundamental studies on well-defined electrodes, flow cell setups are well-suited for operation in a high-throughput manner [6]. This potential was realized in a proof-of-concept study on Fe–Ti–W–O thin film materials libraries using  fully automated activity-stability measurements [7]. Such a platform would enable material discovery, which is tailored to search not only for the most active but also for the most stable material.

Photovoltaic-coupled photoelectrochemical (PV-PEC) water splitting devices offer a route to high efficiency solar-to-hydrogen (STH) conversion with low-cost and non-toxic materials, in particular given the recent advances in efficient and earth-abundant PEC (e.g. metal-oxide) and PV (e.g. organic, perovskite) materials. [1][2][3] However, achieving STH efficiencies relevant for commercialisation (~15-20%) requires careful optimisation of the electronic and optical properties of the constituent materials, as well of the device configuration to minimise parasitic optical and electrical losses.

Here, we present a optical-electronic model for simulating the current-voltage curves of PV-PEC tandem cells, and which includes descriptions of important (optical and electronic) loss mechanisms. PV operation is modelled using diode equations that incorporate loss estimates due to radiative and non-radiative recombination. For describing PEC operation, we extend the Gärtner formulism for PEC photocurrent generation to treat surface recombination due to charge trapping and correct for Fermi level pinning due to an overpotential drop over the Helmholtz layer. We then study the effect of changing semiconductor properties, including conduction and valence band levels, electron and hole masses, charge-diffusion length, Fermi level and trap densities, as well as optical considerations (e.g. parasitic light absorption due to electrolyte and bubble formation) on PV-PEC operation. This allows us to determine the PEC and PV semiconductor properties and device configurations that are required to reach STH efficiencies relevant for commercialisation, in particular using state-of-the-art metal-oxide, organic and perovskite semiconductors.

Hydrogen generated from solar-driven water-splitting process has the potential to be a clean, sustainable and abundant energy storage option [1,2]. The devices implemented in an integrated design approach, i.e., submerged in the electrolyte, pose a significant limitation when it comes to up-scaling [2]. The performance decreases due to added ohmic losses in the electrode or electrolyte and local pH gradients represents a significant challenge to overcome. The presence of the hydrogen gas bubbles at electrolyte/photoelectrode interface may assist in managing low pH gradients, if engineered properly. Beyond the state of the art, introducing a porosity to the photoelectrodes has been proposed as a shortcut to minimize the losses associated with up-scaling [2,3]. To study the optimal porous configuration of the water splitting device, mathematical modeling simulations were performed for various pore size (d_pore = 10–400 μm) and pitch of 1mm. The overall impact of tailoring the d_pore indicates that increasing porosity of the decreases the ohmic loss but increases the product bubble crossover (see Figure below). For the device to be operable with acceptable voltage loss of ~100 mV, the pore size should be less than 200 µm in diameter. These investigations are expected to contribute to advancing the development of large-scale water-splitting technology.

Figure 1 (a) Color map of total voltage losses for 10 cm long electrode with different size of pores. (b) Percentage of H₂ crossovers due to porous structures of electrodes.

Halide perovskites received significant attention in the photovoltaic community owing to its higher power conversion efficiency, high absorption coefficient, and bandgap tunability. However, their stability in aqueous medium is poor making them unsuitable to employ for direct water splitting. Various protective layers coupled with halide perovskite are utilized to prevent the exposure of the absorber to aqueous medium. However, these protective layers are susceptible to failure with prolonged operation. So, employing a water stable absorber material is found to be the key for successful photoelectrochemical (PEC) water splitting.

Vacancy ordered double perovskites with the formula A₂BX₆ are a promising alternative to conventional ABX₃ perovskites due to excellent stability, comparable absorption coefficient and bandgap tunability. This class of materials have their alternate octahedra being vacant enabling a wider range of B-site elements due to rotational freedom. Here, we have reported the most stable vacancy ordered halide perovskite Cs₂PtCl₆ and Cs₂PtBr₆ which showed extraordinary stability in air and even in extreme acidic and basic mediums. Furthermore, these materials exhibit absorption properties that encompass a substantial portion of visible light. The capability of the material to undergo anion exchange from Cs₂PtCl₆ to Cs₂PtBr₆ via core-shell mechanism to form Type-II heterostructure is demonstrated. These materials are successfully employed as photoanode for PEC water oxidation which displayed photocurrent density of >0.2 mA/cm² at 1.23 V vs. RHE.

While hydrogen production via photoelectrochemical water splitting has been demonstrated on a small scale, developing an industrial scale device is a challenge that brings together researchers from a range of disciplines, including engineers, material scientists and chemists. Development of photoelectrochemical water splitting devices requires an understanding of semiconductor physics, optics, (photo)electrochemistry (including catalysis and corrosion) and chemical engineering (e.g. fluid mechanics, heat and mass transport). These are all critical to the development of efficient and robust devices.

I will discuss our in-house developed photoelectrochemical (PEC) device, with a total photo-absorbing area in excess of 100 cm². The device operates with photoanodes fabricated by chemical vapour deposition, a prevalent and scalable method, of sequential layers of WO₃ nanorods and BiVO₄, to form a staggered heterojunction on FTO. The 2.4 to 2.5 eV bandgap of BiVO₄ enables light absorption up to 517 nm in wavelength and a theoretical solar-to-hydrogen efficiency (ɳ_STH) of up to 9.2 %. The WO₃/BiVO₄ heterojunction system is one of the most promising in terms of performance, cost and durability. Combined with a Ni mesh cathode and an externally mounted homojunction Si PV, and operated in a pH neutral phosphate buffer solution, this creates a cost-effective and scalable photoelectrochemical-photovoltaic (PV-PEC) device with a commercially viable fabrication method. In initial experiments, we have achieved a spontaneous solar-to-hydrogen efficiency (ɳ_STH) of 4.0 %.

Scale-up of PEC devices comes with numerous challenges which must be addressed by both computational and experimental approaches. I will discuss the ways in which we have combined these approaches in our research and initial progress to-date. I will also show strategies we have been developing towards mitigating performance losses within our device, including reducing severe potential and current gradients across poorly conducting photoelectrodes. I will discuss other aspects critical to the scale-up of photoelectrochemical devices, including heat and mass transfer, fluid dynamics and safety challenges associated with increasing product and precursor volumes.

Despite the urgent need for engineering progress in this area to accelerate this type of solar hydrogen technology, there are relatively few academic publications that address this topic. With a lack of demonstration prototypes, it remains difficult to envisage what an industrial-scale installation for clean hydrogen production might look like. Ultimately, availability of green hydrogen is expected to result in the hydrogen market expansion. This research seeks to elucidate engineering challenges of developing large-scale water splitting devices, to facilitate the pathway to commercially viable photoelectrochemical hydrogen production.

Key words

Photoelectrochemistry, hydrogen, scale-up, electrochemical engineering, BiVO₄

BiVO₄ is considered one of the most promising metal oxide semiconductors for photoelectrochemical water oxidation. Many studies focus on addressing the inherent material limitations, such as charge carrier mobility and slow water oxidation kinetics. For example, it has been shown that doping with some metals and non-metals increases the charge carrier mobility and the addition of a co-catalyst improves the water oxidation kinetics. These techniques are effective in such a way that ninety percent of the theoretical AM1.5 photocurrent of BiVO₄ has been achieved. This progress has shifted some focus towards scaling up, in which it has unfortunately been reported that the scale-up process significantly impacts the photocurrent and stability. Identifying and addressing the scale-up losses and optimizing the associated overpotential is therefore crucial.  In this study, we scaled-up sulfur incorporated BiVO₄ photoelectrodes using a screen printing technique. Photoelectrodes with area up to 100 cm² were fabricated uniformly, and the performance was evaluated using a continuous flow photoelectrochemical cell. Specifically, the effect of scale-up on the overall overpotential of the cell, which includes ohmic and concentration overpotentials, was considered. Strategies to minimize these losses, such as increasing the substrate conductivity and electrolyte engineering (modifying the concentration and the flow rate of electrolyte), were implemented. As a result, we successfully minimize the photocurrent loss to only ~12% when the photoelectrode area is increased by more than two orders of magnitude, from 0.24 to 100 cm². Further optimization of the ionic and concentration overpotential associated with scale-up is ongoing and will be discussed.

Direct connection of PV devices to electrochemical (EC) water-splitting cells or batteries (B) is a highly efficient and material-saving solution to stabilize intermittent PV generation on long- and short-time scales. In our study, we address a hybrid device with EC water splitting cell in parallel connection to a battery. The first obvious advantage of this PV-EC-B system is the much more stable operating conditions of the electrolyzer. A sufficiently scaled battery keeps the EC cell running overnight and shaves peak load during the daytime. In our works, we have shown the feasibility of the self-sustained operation of the PV-EC-B system [1, 2]. Besides the foreseeable stabilizing effects, we have found theoretically [1] and confirmed experimentally [2] that batteries can improve solar-to-hydrogen efficiency (STH) in PV-EC-B devices as compared to the PV-EC reference. Batteries can boost STH in PV-EC-B systems because, under periodic irradiance conditions, the battery transfers part of the daytime PV energy to the night, reducing EC cell power and suppressing related overpotential (kinetics) losses. The gain is synergistic – even despite additional potential losses in the battery, the EC-B combination has lower losses than the EC cell operating alone [1, 2]. From the point of view of the system design the reduction of operating power and its stability allows for the use of a smaller electrolyzer at the same nominal PV capacity.

In our report we will present details of the PV-EC-B operation, principles behind the gain in STH attained with battery, and our recent experimental study. An increased average STH efficiency per cycle (11.4 % vs. 10.5 % without the battery) has been observed in the PV-EC-B system of a Si heterojunction PV module, bifunctional NiFeMo electrolyzer, and a commercial Li-ion NMC battery. The results are discussed in the context of the generalized STH limit analysis developed earlier for PV-EC combinations [3].

Applying our analysis to different literature sources using PV-EC without batteries we present an analysis of what could be achieved if a battery is applied to these systems in the figure below. We calculate the STH limit with and without a battery for each experimentally reported system in literature as well as the STH gain with a battery.

Each system reported in the literature is represented by four symbols: filled blue circles show the reported experimental STH of each PV-EC system; open blue circles indicate the STH limit calculated for each system using the reverse analysis [3]; horizontal dashes represent the estimate for the STH limit in this system if the battery is included; finally, the red circles show the estimate of the STH gain for each system.

A separate set of points denoted “this work” represents the experiment on the PV-EC-B system in this work.

For approximately half of the reported systems STH gain of 0.5%_abs – 1%_abs can be expected, several systems can gain 1%_abs – 2%_abs, while several systems can gain 2%_abs – 4%_abs. There is a general trend for higher STH gain at higher PV efficiencies, but the data presented in Figure 4 does not constitute any dependence.

Suitable photo-electrode materials are known to be the main bottleneck for large scale deployment of devices for solar fuel production. These semiconducting materials must be efficient, stable and scalable, although, usually two out three of these qualities are met. Hematite is an extensively researched photoelectrode material, known for its long-lasting properties and facile synthesis. However, the position of its band edges and high recombination rates, resulting in low efficiencies, preclude hematite to be used in real applications. On the other hand, complex photo-electrodes with laborious synthesis, e.g. III-V tandems absorbers (GaInP/GaAs), coupled with RuO_x and PtRu catalysts, have proven to be highly efficient for water splitting [1].

With the aim to improve scalability, concentrated light sources allow the use of smaller photoelectrodes, provided they are stable and efficient at operating conditions. However, the use of higher irradiation (>100 kW m^-2) unlocks additional constraints related to heat management, ohmic drop due to inordinate current densities, and unknown electron/hole transfer kinetics under high photon flux and at higher temperatures. It is expected that these effects could be alleviated by careful reactor design, and choosing suitable substrates and operating conditions. However, these issues have been seldom investigated and only a few studies on photoelectrochemical cells for irradiations typically below 30 suns have been reported [2].

Here, we present the model and preliminary experimental validation of a photo-electrochemical cell operating under concentrated irradiation (>100 kW m^-2), using metal oxides (Fe₂O₃ and BiVO₄) as photoelectrode materials, which were selected due to their well-known and predictable photoelectrochemical behavior [3,4]. The effects of the substrate material (Ti or transparent conductive oxides deposited on glass), electrode configuration, electrolyte flow and spectrally resolved photon intensities on the performance of the cell will be presented. Initial predictions indicate that conductive glass is not a suitable substrate due to its low thermal conductivity, resulting in undesirable high temperatures (>80 °C) when operating under irradiations of 600 kW m^-2.

References:

[1] Young, J., Steiner, M., Döscher, H. et al. Nat Energy 2017, 2, 17028.

[2] Vilanova, Mendes, A. Journal of Power Sources 2020, 454, 227890.

[3] Bedoya-Lora, F. Hankin, A. Kelsall, G. J. Mater. Chem. A, 2017, 5, 22683-22696

[4] Gaudy, Y. Haussener, S. et al. J. Mater. Chem. A, 2018, 6, 17337-17352

In my talk I will revise the recent efforts done in my group regarding the integration of first principles studies into more advanced multiscale simulations. I will show the main difficulties and bottlenecks for the implementation of a more robust computational framework that can be used in a more general manner and that should give us access to more adequate models in the field. For instance I will address the main difficulties in the modeling of excited states and how they are more ubiquotous and diverse than previously considered and how modeling can address some of the challenges ahead of us to obtain proper representations. Also I will revise the types of approaximations when attempting multiscale approaches and how they need to be tuned and supported by machine learning techniques in some of the cases to reduce the dimensionality problems and the low energy events that typically consume the simulation times.

Green H₂ has been recognized as an important element in efforts to decarbonize our fossil fuel-dependent society. One approach to produce green H₂ is solar water splitting in a photoelectrochemical (PEC) device. Solar-to-hydrogen (STH) efficiencies of up to 30% have been demonstrated[1] but studies have shown that this approach still results in a relatively high levelized cost of hydrogen (LCOH) of ~10 US$ per kg of H₂.[2-3] This is ca. one order of magnitude higher than that of hydrogen produced via steam methane reforming (SMR), which forms the bulk of the currently produced H₂. A possible solution is to incorporate an upgrading process of biomass feedstock that generates valuable chemicals into the solar water splitting device. This is expected to not only decrease the overall LCOH but also introduce an alternative renewable pathway in chemical manufacturing. In this study, we propose the concept of solar-driven hydrogenation of biomass-derived feedstock. Photoelectrochemically generated H₂ in our solar water splitting device is coupled in situ with the homogenously catalyzed hydrogenation of itaconic acid (IA) to methyl succinic acid (MSA). IA has been identified by the US Department of Energy as one of the twelve building blocks that possess the potential to be transformed subsequently into several high-value bio-based chemicals or materials. MSA is a valuable chemical compound with an estimated global market size of up to ~15,000 tonnes, whose derivatives are ubiquitously used as solvents in cosmetics, polymer synthesis, binders in powder coatings, and organic synthesis, especially for pharmaceutical synthesis.[4-5] Our coupled hydrogenation approach—performed in the PV-electrolyzer and PEC configurations using III-V PV cells and BiVO₄-based photoelectrode, respectively—successfully demonstrate solar-driven H₂-to-MSA conversion as high as 60%. In comparison to the non-coupled approach (i.e., direct hydrogenation), our coupled system offers synergistic benefits in terms of prolonged durability and a higher degree of flexibility toward other important chemical transformation reactions. In addition, life-cycle net energy assessment and technoeconomic analysis results show that adding the coupled hydrogenation process significantly lowers the energy payback time and the LCOH, respectively, to a point that is competitive even with SMR-produced H₂.[6] Further implications and optimization potentials of the coupled PEC hydrogenation approach, also beyond the demonstrated hydrogenation of IA to MSA, will be discussed.