Enhancing Proton Exchange Membrane Water Electrolysis Performance through Electrospray Catalyst Deposition

Abstract

Catalysts are among the key components limiting the large-scale development of hydrogen technologies. Despite intensive research into more affordable alternatives, most catalysts used in proton exchange membrane fuel cells (PEMFCs) and proton exchange membrane water electrolyzers (PEMWEs) still rely on expensive noble metals such as platinum, iridium, and ruthenium. To reduce costs without compromising performance, new strategies must focus on optimizing catalyst utilization. Electrospray deposition is emerging as a promising technique that enhances catalyst layer morphology, thereby increasing electrochemically active surface area while reducing the amount of catalyst required. Electrospray works by applying a high voltage (in the kilovolt range) between the tip and the substrate, leading to the atomization of the catalyst ink into nanodroplets. This process enables homogeneous deposition and improved layer structure compared to conventional methods. In this work, we present the development of an electrospray system specifically designed in our laboratory for use with catalyst inks in PEMWE electrodes. Both anode and cathode catalyst layers were fabricated by electrospray deposition and compared with conventional spray-deposited electrodes. Morphological analyses confirmed that electrospray produced smoother and more uniform catalyst films, resulting in larger electrochemically active areas. Polarization curve measurements demonstrated that the improved morphology directly enhanced electrolyzer performance, particularly by reducing mass transport losses. These results show that electrospray is a valuable technique for advancing catalyst layer engineering in PEMWEs, contributing in a close future to lower noble-metal loading and reduced system costs, while maintaining high performance.

Introduction

In recent decades, renewable energy has emerged as a powerful tool in the fight against greenhouse gas emissions. However, the intermittent nature of these environmentally friendly energy sources remains a significant limitation to their widespread implementation. In this context, hydrogen is gaining increasing attention as an energy vector capable of storing excess renewable energy in the form of green hydrogen. This hydrogen can later be used for various industrial applications or serve as an energy reservoir for fuel cells to recover the stored energy. ,

Electrolyzer technology is a mature and promising solution for producing hydrogen with zero emissions using renewable energy. , Among the different types of electrolyzers, Proton Exchange Membrane Water Electrolyzers (PEMWE) are attracting the most interest due to their advantages, including high efficiency, modular assembly, low-temperature operation, and the absence of corrosive electrolytes. ,,

However, a key limitation of PEMWE technology lies in the high cost of its components, particularly the catalysts, ,− membranes, , porous transport layers (PTLs) and gas diffusion layers (GDLs) ,− and bipolar plates. , The typical catalysts used in PEMWEs are iridium oxide or ruthenium at the anode side, and platinum supported on porous carbon at the cathode side. Despite significant efforts to develop alternative catalyst formulations, these expensive and scarce metals remain the standard in commercial systems.

Catalyst deposition techniques play a decisive role in the fabrication of membrane electrode assemblies (MEAs) for proton exchange membrane water electrolyzers (PEMWEs), as they strongly influence catalyst utilization, interfacial resistance, mass transport, and long-term durability. Industrially, the most established approaches are ink-based deposition methods, where iridium-based oxygen evolution catalysts and platinum-based hydrogen evolution catalysts are deposited either as catalyst-coated membranes (CCMs) or catalyst-coated substrates (CCSs). Techniques such as spray coating (including ultrasonic spraying), doctor-blade or slot-die coating, and screen printing are widely investigated due to their scalability and compatibility with roll-to-roll processing. In PEMWE, CCM fabrication is often favored because it enhances the catalyst-membrane interface and reduces interfacial ohmic losses, enabling lower noble-metal loadings without sacrificing performance. ,

Beyond conventional coating routes, advanced thin-film deposition techniques are increasingly explored to address the critical challenge of iridium scarcity in PEMWEs. Atomic layer deposition (ALD) has emerged as a promising method to form conformal, nanometer-thin iridium oxide layers directly on porous transport layers, achieving exceptional catalyst dispersion and utilization at ultralow loadings. Electrodeposition-based approaches offer similarly precise control over catalyst thickness and morphology while maintaining good electrical contact and adhesion. Although these techniques currently face limitations in throughput and cost, recent studies highlight their potential for next-generation PEMWE MEAs aimed at drastically reducing iridium demand while maintaining high efficiency and durability. ,

Among the conventional methods used for electrode preparation such as drop-casting, airbrush spraying, and brush painting, the electrospray technique stands out as a valuable approach for optimizing catalyst deposition. − Electrospray is a physical process that involves applying a high voltage (on the order of kilovolts) to a catalyst ink, leading to its atomization into nanodroplets. , This results in improved catalyst distribution and increased electrochemically active surface area, which enhances overall electrode performance.

In this study, we evaluate the preparation of both electrodes of an electrolyzer (anode and cathode) using an electrospray system specifically developed in our laboratory, designed for compatibility with catalyst inks used in PEMWE electrodes. After introducing the basic principles of the electrospray technique, we investigate the morphological differences between electrodes produced by that technique in comparison with the conventional spray methods. Finally, we assess how these morphological differences generated that is related to the methods of deposition employed stright forward affects to the performance of the electrolyzer.

Experimental

Iridium-ATO Preparation

The iridium-based electrocatalyst supported on Antimony Tin Oxide (ATO) was prepared following a modified version of the procedure described by Lettenmeier et al. , All steps were carried out under an inert nitrogen atmosphere, using anhydrous ethanol as the solvent. First, iridium chloride (IrCl ₃) was dissolved in anhydrous ethanol at a concentration of 0.05 M using an ultrasonic bath. In parallel, a 0.5 M cetyltrimethylammonium bromide (CTAB) solution was prepared in anhydrous ethanol under the same conditions. The two solutions were then combined, and the resulting IrCl ₃/CTBA mixture was added dropwise to a suspension of 10 g/L of ATO in anhydrous ethanol under continuous stirring.

Subsequently, a NaBH ₄ solution in anhydrous ethanol (1.5 M) was added dropwise as the reducing agent. The mixture was maintained under stirring in an inert atmosphere until the complete consumption of the reducing agent, typically after 10 h.

The solid product was recovered by ultrafiltration under 1–2 bar nitrogen pressure, followed by centrifugation and filtration. The material was washed thoroughly with hot water and ethanol to remove residual surfactant. Characterization by SEM/EDX, XRD, and TEM confirmed a final iridium purity of approximately 75%.

MEA Preparation

The membrane electrode assembly (MEA) represents the core component of the electrolyzer. Catalyst layers are typically fabricated from catalyst inks composed of an ionomer, catalyst, and solvent. In this work, separate catalyst inks were prepared for the cathode and anode.

For the anode, the ink was prepared by mixing 1.03 g of Ir-ATO, 0.54 mL of water, 1.22 mL of 2-propanol, and 0.054 mL of Nafion solution (Alfa Aesar, 5 wt %). Concerning the cathode, the ink composition was 0.021 g Pt/C 60% (Thermo Scientific), 0.525 mL of water, 1.1725 mL of 2-isopropanol and 0.0525 mL of Nafion solution (Alfa Aesar, 5 wt %). Then both solutions were sonicated for 40 s using an ultrasonic processor (UP200HT) to obtain a homogeneous suspension and to prevent detachment of platinum particles from the support. The ink was kept under stirring until use. Both solutions were used for the deposition of the catalyst using both deposition techniques studied in this work.

The catalyst layers were fabricated by electrospray or aerography onto a Nafion-115 (DuPont) electrolyte membrane, forming thin catalyst films. The final loadings were 0.8 mgPt cm^–2 at the cathode and 2.4 mgIr cm^–2 at the anode. For the cathode, a carbon-based eCoCell-NN40-e gas diffusion layer (GDL) was employed, while a titanium felt porous transport layer (PTL) eCoCell-TIFF-40, was used at the anode (eCoCell is commercilized by Hydrogreen Energy, hydrogreenergy.com).

To evaluate the performance of the fabricated MEAs, a 3 × 3 cm PEM electrolyzer single cell was assembled, as depicted in Figure .

1.

PEM cell electrolyzer used to test the MEAs fabricated in this study.

In both cases corresponding to MEAs prepared by electrospray and by aerography the catalyst loading was kept identical.

Electrospray Optimization

Electrospray is based on the application of a high voltage in the kilovolt range between a substrate (collector) and the tip of a syringe needle. , The resulting electric field destabilizes the liquid surface tension, generating an aerosol of nanodroplets. Figure shows a schematic and a photograph of the electrospray equipment developed in our laboratory. By applying a potential difference of 9–10 kV between the tip and the substrate, catalyst nanodroplets were successfully produced, as illustrated in Figure . To prevent membrane swelling during the electrospray, the substrate was thermostatted at 40 °C.

2.

(a) Esqueme of the electrospray equipment used in this work. (b) Photo of the real system developed for generating the electrospray deposition.

3.

Catalyst nanodepositions generated with the electrospray technique.

The electrospray setup was specifically designed to handle different catalytic inks, including those in which the catalyst particles tended to precipitate during the long deposition time required for electrode fabrication. This instability was a serious issue we encountered, since deposition typically required several hours, during which the catalyst gradually settled out of suspension. Initially, we observed that Ir-ATO precipitated along the tubing connecting the catalyst reservoir to the electrospray tip, severely affecting the uniformity of the deposition and producing the interruption of the electrospray.

To circumvent this problem, the ink solutions were continuously stirred throughout during the whole electrospray process. In addition, two external vibrators were coupled to the system to prevent particle sedimentation and to maintain the catalyst in suspension during operation, as illustrated in Figure .

This solution, however, introduced a technical complication: the system required careful electrical isolation to prevent current leakage, given the high electrical conductivity of the catalyst inks. After optimization, the critical parameters for electrospray deposition of both iridium- and platinum-based inks were as follows: a tip-to-substrate distance of 35 mm, an applied voltage of 9.5 kV, and a needle of 0.36 mm outer diameter and 0.18 mm inner diameter.

To enable large-scale production, this technique can be scaled by extrapolating the performance of a single tip to multiple tips operating simultaneously, as illustrated in Figure , where ten tips are used in parallel. This approach significantly reduces the time required to fabricate MEAs to a level compatible with stack-scale production.

4.

Electrospray equipment designed for large-scale MEA production. The figure shows a prototype system under development in our lab to scale up the technique, with ten electrospray tips operating simultaneously for parallel MEA fabrication.

Catalyst Morphological Factor, φ

To evaluate the catalytic activity of the films prepared by electrospray compared with those obtained by standard aerography, the electrochemical response of the corresponding electrodes was analyzed. Ir-ATO and Pt/C electrodes were deposited on stainless steel strips with a geometric surface area of 1.5 cm².

Cyclic voltammograms used to probe the morphology of the catalyst layers were recorded using an AUTOLAB PGSTAT-302 potentiostat/galvanostat in a standard three-electrode cell.

To establish a correlation between electrode morphology and electrochemical response, the approach proposed by Da Silva et al. was adopted. This method is based on cyclic voltammetry (CV) performed in the capacitive region at different scan rates (typically 5–200 mV s^–1), where morphological differences can be inferred from variations in capacitive current.

Catalyst layers prepared by both deposition techniques (electrospray and aerography) were immersed in 0.5 M H₂SO₄ solution, and CV experiments were carried out at different scan rates. The potential window was restricted to 0.3–0.5 V to ensure that the CVs remained within the capacitive region. Supporting Information (Supporting Information correponding to FigS 1–4) provides the cyclic voltammetries performed for each electrode generated with aerography and electrospray in the capacitive region.

Variations in the capacitive current were correlated with morphological differences in the catalyst layers, such as pore density, crack formation, or compaction of the film. This procedure assumes that electrochemical activity in the capacitive region is proportional to the accessible surface area of the electrode. At higher scan rates ν, inhomogeneous regions of the electrode surface are excluded from participation in the charge–discharge process, reducing the measured capacitive current.

Accordingly, the linear dependence of the capacitive current (i _c ), taken at 0.4 V (the midpoint of the capacitive region), on the scan rate ν yields two distinct linear regimes with different slopes. These slopes are associated with porosity and surface roughness in the catalyst layer and correspond to different double-layer capacitances (C _d ) at the electrode/electrolyte interface. The capacitance, C _d , is obtained from the slope of i _c versus ν, at low scan rates (ν < 50 mV s^–1, see Figure ), where i _c is the anodic current measured at 0.4 V in the capacitive region.

5.

Anodic current versus scan rate corresponding to Ir-ATO film generated by aerography (a) and electrospray (b), and Pt/C film generated by aerography (c) and electrospray (d). From the slopes of both stright lines, we were able to calculate the corresponding morphology factors, φ.

Thus, for this first linear segment, i _c can be expressed as follows:

$$i_c=C_d.\nu$$ 1

Similarly, for the domain corresponding to high ν, a second linear segment can be identified

$$i_{c,e}=i_{c,r}+C_{d,e}\nu$$ 2

where i _c,r is the origin coordinate of this second straight line when ν→ 0.

Thus, the total capacitive current, i _c , is the sum of the capacitive current of the inner zone i _c,i , and the outer zones i _c,e , in the catalyst layer.

$$i_c=i_{c,i}+i_{c,e}$$ 3

By combining the above equations, we can obtain the following expression,

$$i_{c,i}=C_d\nu -(i_{c,r}+C_{d,e}\nu )$$ 4

Finally, the differential capacity between the inner and outer region can be determined as follows:

$$C_{d,i}=C_d-C_{d,e}$$ 5

Spinolo et al. defined the morphological factor, φ, as the ratio between the accessible internal surface area and the total electrochemically active area of the electrode. Accordingly, the morphological factor can be expressed as:

$$\phi =\frac{C_{d,i}}{C_d}$$ 6

where φ is the morphological factor, and C _d,i and C _d represent the capacitances associated with the internal and total accessible surface area of the catalyst layer, respectively. Therefore, φ can vary between 0 and 1. When φ → 1, the electrode exhibits a higher degree of porosity, indicating a greater contribution from the internal surface area. Figure shows the anodic current i _c versus ν for the catalyst layer generated by electrospray at 10 kV, considering the four cases studied.

By repeating this procedure for electrodes prepared using both deposition techniques, the estimated morphological factor, φ, for the catalyst layers (CLs) generated with each method is presented in Table .

1. Morphological Factor φ of the Catalyst Layers Corresponding to Ir-ATO and Pt/C Generated with Electrospray and Aerography, Respectively.

electrode	electrospray	aerography
Pt/C	0.98	0.93
Ir-ATO	0.46	0.33

Thus, Table presents the morphological factor, φ, of the catalyst layers generated using different deposition techniques. The results indicate that electrospray deposition increases the morphological factor for both catalyst films. This improvement reflects a larger accessible surface area within the catalyst layers, which in turn should translate into higher electrochemical activity compared to layers produced with the standard aerography technique.

Morphology of the Catalyst Films

Figure shows two images of the electrodes fabricated using aerography and electrospray. A simple visual inspection reveals that the electrospray technique produces a significantly more uniform and homogeneous catalyst film than the one obtained by aerographic deposition.

6.

Images of the electrodes of 3 cm × 3 cm depositions: (a) and (b) correspond to iridium and platinum layers deposited using electrospray, while (c) and (d) correspond to iridium and platinum layers deposited using aerography.

In this context, to analyze the morphological differences in the catalyst layers as a function of the deposition technique employed, scanning electron microscopy (SEM) was performed. Figure presents the SEM images of the Iridium and Pt/C catalyst layers fabricated using both deposition methods.

7.

SEM images of the catalyst layers generated with electro Electrospray and aerography. (a) Ir-ATO catalyst layer generated with aerography. (b) Ir-ATO catalyst layer generated with electrospray. (c) Pt/C catalyst layer generated with aerography. (d) Pt/C catalyst layer generated with electrospray.

Figure illustrates how the deposition technique used for fabricating the catalyst film strongly influences the morphology of the catalyst layer. For the Ir-ATO layer, the electrospray method produced a smoother and less rough surface compared to the layer obtained by aerography. A similar effect was also observed for the Pt/C catalyst layer.

Chemical Stability of the Catalysts after its Deposition

The stability of the catalysts during the electrospray process was carefully evaluated to ensure that the chemical composition of the active phase was not significantly altered by the high voltage and long deposition times required for layer fabrication.

As explained above, the electrospray technique relies on the application of a very high voltage to the catalyst ink suspension, which induces ionization of the solution and destabilization of its surface tension, thereby generating nanodroplets. However, one aspect that has not been widely reported is the stability of the metallic groups within the electrocatalysts under such high-voltage conditions. To address this, an XPS analysis was performed on catalyst films deposited by both electrospray and aerography, in order to verify the absence of alterations in the oxidation states of the electrocatalysts. XPS measurements were performed on a SPECS FlexPS-E system.

Figure a,b show the Pt 4f and Ir 4f core levels, respectively, for the samples deposited by aerography (black line) and electrospray (red line). In both spectra, the Pt 4f_7/2 core level appears at 71.5 eV with a spin–orbit splitting of 3.3 eV, clearly indicating that the Pt chemical state remains metallic in both cases. , Moreover, the lineshapes of both spectra are very similar, regardless of whether the deposition technique used was aerography or electrospray.

8.

Pt (a) and Ir (b) XPS spectra obtained for films generated by electrospray and aerography.

Figure b shows the Ir 4f core levels of the samples deposited by aerography and electrospray. In both cases, the Ir 4f_7/2 peak is located at 60.9 eV with a spin–orbit splitting of 3.0 eV, indicating that the dominant Ir species is metallic Ir. However, for the aerograph sample, the presence of a small Ir oxide contribution cannot be ruled out, as suggested by the appearance of a shoulder at this binding energy.

PEMWE Performance

To evaluate the practical impact of the catalyst deposition technique, the membrane electrode assemblies (MEAs) prepared by electrospray and aerography were tested in a single-cell PEM water electrolyzer (PEMWE).

The stability of the MEAs fabricated by aerography and electrospray was evaluated by monitoring the high-frequency resistance (HFR) of the electrolyzer cell. This parameter was determined through HFR measurements performed using a BK Precision BA6010 mini-ohmmeter at a constant frequency of 1 kHz. Figure shows the evolution of the internal resistance over 130 h of operation. The results indicate stable behavior for both MEAs; however, the MEA prepared by aerography consistently exhibits a higher internal resistance than the MEA fabricated by electrospray.

9.

High-frequency resistance (HFR) of the electrolyzer cell at 1 kHz for MEAs fabricated by aerography and electrospray.

Polarization curves were also performed to compare the electrochemical performance and efficiency of both approaches. Figure depicts the polarization curve of a single cell PEMWE, where the catalyst layers were deposited using both techniques.

10.

Polarization curves of the PEMWE with catalyst layers generated with electrospray and aerography. Strigh lines correspond to the fit of the of both CP to the eq .

Thus, Figure presents the polarization curves obtained in a PEMWE using catalyst layers fabricated by electrospray and aerography. The results clearly demonstrate that the improved morphology of the catalyst layer achieved with electrospray deposition enhances the electrochemical performance of the electrolyzer compared to the standard aerography technique. From the analysis of the polarization curves, relevant information can be extracted regarding the performance parameters of the catalyst films.

The electrochemical behavior of a PEMWE can be described by a current–voltage (i–V) model. The cell voltage, E _cell, is defined as follows: ,

$$E_{\mathrm{cell}}=E_{\mathrm{rev}}+{\eta }_{\mathrm{act}}+V_{\mathrm{ohm}}+{\eta }_{\mathrm{diff}}$$ 7

where E _rev is the reversible voltage required for the electrolysis reaction, η_act is the activation overpotential, and V _ohm represents the ohmic voltage loss, which arises from the electrical resistances of several components, including current collectors, electrodes, electrolyte, and the membrane. In addition, η_diff accounts for mass transport limitations, particularly diffusion processes (concentration overpotentials), occurring at different interfaces within the cell.

Thus eq can be rewritten as follows:

$$E_{\mathrm{cell}}=E_{\mathrm{rev}}+b·\log (i)+R·i+{\eta }_{\mathrm{diff}}$$ 8

where E _cell corresponds to the open-circuit potential of the electrolyzer, b is the Tafel slope, R represents the ohmic resistance, and η_diff denotes the overpotential associated with diffusive processes.

Table presents the fitting parameters extracted from the polarization curves in Figure , corresponding to the MEAs fabricated by aerography and electrospray, respectively.

2. Parameters Corresponding to the Fit of Both Polarization Curves of Figure to the Eq .

	E 0 (V)	b (V/dec)	R (Ohms)	η (V)
aerography	1.5	0.034	0.331	0.0098
electrospray	1.47	0.032	0.285	0.0045

From the results in Table , it can be concluded that diffusion-related processes are the main factor responsible for the differences between both polarization curves. Specifically, the contribution of diffusion processes in the aerography-based MEA is approximately twice as high as that observed in the MEA fabricated by electrospray.

Discussion

Electrospray is a promising technique for fabricating homogeneous catalyst layers and optimizing catalyst deposition at the nanometric scale. However, several technical challenges must be addressed when applying this method to catalyst suspensions, particularly due to the poor stability of iridium-based inks compared with Pt/C suspensions. Initially, we encountered the critical issue of maintaining the suspension stable until the aerosol was generated at the syringe tip. Iridium tended to precipitate along the tubes connecting the catalyst reservoir to the tip, making it impossible to apply the technique under these conditions. To overcome this limitation, we introduced three key modifications: shortened the length of all the tubes in the instalation, implemented continuous stirring in the reservoir, and coupled external vibrators to the system. These improvements allowed us to maintain suspension stability throughout the electrospray process. Importantly, this problem straightforward in other context had to be solved while carefully preventing current leakages caused by the high voltages (on the order of kilovolts) required for electrospray operation.

Regarding the electrodes produced, the study of the morphological factor for both Ir-ATO and Pt/C catalyst layers revealed that the electrospray technique generates films with a higher morphological factor compared to those obtained by standard aerography. This translates into an increase in the available catalyst surface area and, consequently, an enhancement of catalytic activity.

This increase in the morphological factor is more pronounced in the Pt/C catalytic layer than in the Ir-based one, due to differences in their specific densities. The high compactness of Ir-ATO film, associated with its greater density, resulted in morphological factors of 0.46 and 0.33 for layers generated by electrospray and aerography, respectively. In contrast, the Pt/C layers showed much higher values of 0.98 and 0.93 for electrospray and aerography, respectively.

In this context, the analysis of the electrochemical response of MEAs fabricated using the two deposition techniques revealed that electrospray-deposited MEAs exhibit lower diffusion resistance compared to those produced by aerography, in excellent agreement with the morphological factor values discussed above

Conclusions

This work focused on the electrospray deposition technique as a valuable approach for optimizing catalyst utilization during the fabrication of MEAs for PEMWEs. A technical solution was developed in the design of the electrospray equipment to maintain catalyst suspensions stable over the deposition time, while simultaneously preventing current leakages during high-voltage operation.

We demonstrated that electrospray enables catalyst deposition at the nanometric scale, leading to morphologies that enhance catalytic activity. From the analysis of the morphological factor, it was shown that electrospray increases the accessible catalytic surface, thereby improving the electrochemical response and overall PEMWE performance. Polarization curve analysis confirmed that diffusion processes play a dominant role in electrodes fabricated by aerography, whereas electrospray-produced layers exhibit reduced diffusion resistance. Finally, XPS analysis confirmed that the oxidation states of the catalysts remained unaltered during electrospray deposition, ensuring the chemical stability of the catalyst.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.5c11848.

• CV measurements performed in the potential range of 0.3–0.5 V confirming that the experiments were conducted within the capacitive region (PDF)

The authors declare no competing financial interest.