Phase shuttling-enhanced electrochemical ozone production

Abstract

Ozone can be produced by the electrochemical oxidation of water, which provides a technical solution to on-demand ozone production for disinfection and sterilization. Lead oxides have been found to be unique in catalyzing such a process. However, the fundamental understanding of these catalysts’ mechanisms remains limited, hindering the development of high-performance catalysts for electrochemical ozone production (EOP). Herein, the effect of phase shuttling on the reactivity of Pb₃O₄ was systematically investigated during the EOP process by in situ/ex situ characterizations. It was found that Pb₃O₄ undergoes a phase shuttle towards β-PbO₂via the lattice oxygen oxidation mechanism (LOM) pathway, and the reconstructed β-PbO₂ shows enhanced EOP activity and stability compared to commercial β-PbO₂. The ex situ characterization of materials combined with theoretical calculations reveals that the performance enhancement is mainly attributed to the stable presence of (101) and (110) surfaces in the reconstructed β-PbO₂ with undercoordinated Pb–O. Pourbaix diagrams of lead oxides calculated by DFT demonstrate that the phase shuttling to β-PbO₂ is thermodynamically favorable under EOP conditions. Surface Pourbaix diagrams of β-PbO₂(101) and Pb₃O₄(110) further reveal the adsorption behavior of O*/OH* intermediates and explain the observed change of EOP kinetics at ∼1.6 V vs. RHE. The catalyst is integrated and assembled in a membrane electrode assembly (MEA) electrolyzer, and the produced ozonated water successfully inactivated severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). This work provides a new insight into EOP catalysts and demonstrates the possibilities of further optimization of electrochemical approaches for on-demand ozone generation.

---

Broader context

Electrification is one of the key trends shaping the current global energy transition in the context of climate change. Due to recent geopolitical developments, favorable economics, and the urgent demand for decarbonization, this trend is accelerating steadily. Alongside climate targets, the COVID-19 pandemic has demanded more radical change but also new opportunities with a sense of urgency. One of the immediate impacts has been elevating emphasis on sanitizing and disinfecting chemicals. Surging demand for these products requires developing advanced, climate-neutral technologies to support this sector's growth. Here, we report a breakthrough in fundamental and applied research for electrochemical ozone production (EOP). Ozone is the most effective, safe, and cost-effective solution for disinfection and sanitization that breaks down into benign components. EOP is a promising way to produce it in a sustainable manner. To date, achieving long-term operation of EOP at industrial current density in ultrapure water remains one of the biggest challenges. In this paper, we report that using a pre-catalyst can significantly promote the EOP performance. We found that the pre-catalyst Pb₃O₄ undergoes phase shuttling to β-PbO₂ during the EOP process. The reconstructed β-PbO₂ exhibited improved EOP activity and stability compared to commercial β-PbO₂. This work provides fundamental insights into the lead oxides under EOP conditions and paves a new way to design high-performance EOP catalysts for potential industrial applications.

Introduction

Electrification is one of the key trends shaping the current global energy transition in the context of climate change. Due to recent geopolitical developments, favorable economics, and the urgent demand for decarbonization, this trend is accelerating steadily. Alongside climate targets, the pandemic has demanded more radical change but also new opportunities with a sense of urgency. One of the immediate impacts has been elevating emphasis on sanitizing and disinfecting chemicals. Surging demand for these products requires developing advanced, climate-neutral technologies to support this sector's growth. New green technologies could emerge, considering breakthroughs in fundamental and applied research in electrochemistry.

In this context, ozone has proven to be one of the most effective, safe, and cost-effective solutions for disinfection and sanitization that breaks down into benign components.¹ Its main applications range from water purification to industrial exhaust treatment, public health management, food processing, aquaculture disinfection, and value-added compound production (Scheme 1a). Compared with the standard ozone production methods (ultraviolet excitation and electric discharge (corona)) (Scheme 1b), electrochemical ozone production (EOP) from water electrolysis is promising in both high-end applications of ozonation^2,3 and advanced oxidation processes (AOPs).⁴ Furthermore, considering the instability of ozone, EOP is highly suitable for decentralized, on-demand, and on-site ozone generation, thereby eliminating transportation and safety risks as well as storage costs (Scheme 1c).

Scheme 1 Applications and preparation methods of ozone. (a) Applications of ozone in various industries. (b) Ozone preparation through the ionization discharge process. (c) Ozone preparation by the electrochemical oxidation of water.

EOP is a six-electron reaction process competing with the oxygen evolution reaction (OER). These two anodic reactions can be expressed in eqn (1) and (2).

2H₂O → O₂ + 4H⁺ + 4e⁻, E^θ = 1.23 V vs. RHE (1)

3H₂O → O₃ + 6H⁺ + 6e⁻, E^θ = 1.51 V vs. RHE (2)

Thermodynamically, the EOP takes place at a potential higher than 1.51 V vs. RHE, while the OER starts from 1.23 V vs. RHE, which clearly indicates that the distinguishability between these two reactions is the key challenge.

To date, lead oxides have been found to be unique in catalyzing the EOP in water oxidation, and β-PbO₂ has been the state-of-the-art catalyst in commercial devices. However, challenges remain due to the unsatisfactory stability and activity at high current density.⁵ Although many strategies have been applied to use lead oxides in some works, significant gaps still exist, especially in the underlying mechanisms that are not well understood. The experimental conditions used to perform water oxidation are harsh for the catalysts for both OER and EOP. As a result, most catalysts change their surface chemistry to adapt to the working environment. It has been widely reported that most OER electrocatalysts change their surface phase under the OER process, in which more active surfaces are formed.⁶ The inherent reason for this change can be attributed to the relatively unstable bulk chemistry and the strong interactions between the surface oxygen-containing adsorbate and the exposed surface. In addition, the structural phase change of the OER catalysts depends on the applied potential,^7,8 which electrically drives the surface oxidation/reduction of pre-catalysts to switch to the new surfaces.^9,10 Despite many insights revealed for the OER, little has been known regarding the phase and surface stability of EOP catalysts, which is crucial in understanding the active phase and developing efficient EOP catalysts.

Here, we report a systematic study of lead oxides during the EOP process. We proved that using the pre-catalyst Pb₃O₄ as a catalyst provides a remarkable enhancement of the EOP compared to the commercial β-PbO₂. The phase shuttling of Pb₃O₄ under EOP conditions leads to in situ formation of β-PbO₂, which is rich in (101) and (110) surfaces with undercoordinated Pb–O. The Pb₃O₄-reconstructed β-PbO₂ shows better stability than commercial β-PbO₂ at the industry-level current density of 1 A cm⁻² for more than 900 h. The significance of these findings is strengthened by in situ differential electrochemical mass spectrometry (DEMS) experiments, which also show that Pb₃O₄ gradually reconstructs to β-PbO₂ by exchange of lattice oxygen with water, and the EOP happens through a lattice oxygen-involved mechanism (LOM).

Results and discussion

EOP performance of the catalysts

The electrochemical water oxidation properties of the commercial Pb₃O₄ and β-PbO₂ were first evaluated and compared using linear sweep voltammetry (LSV) tests in a saturated K₂SO₄ solution (pH = 7.0) at room temperature (Fig. S1, ESI†). The LSV curves in Fig. 1a show that a higher overpotential is required for Pb₃O₄ to achieve the same current density of 10 mA cm⁻² compared to β-PbO₂ (2.73 and 2.51 V vs. RHE for Pb₃O₄ and β-PbO₂, respectively). Taking into account the electrochemical active surface area (ESCA) values (shown in Fig. S2 and S3, ESI†), these results suggest the superior overall water oxidation activity of β-PbO₂. The electrochemical water oxidation kinetics were further investigated by calculating the Tafel slopes. Fig. 1b compares the Tafel slopes of the Pb₃O₄ and β-PbO₂ electrocatalysts, where two apparent linear segments can be found for both cases. The Tafel slopes of β-PbO₂ and Pb₃O₄ were calculated to be 441 mV dec⁻¹ and 434 mV dec⁻¹, respectively, at the potential region below ca. 1.6 V vs. RHE. At the higher potential range, a similar trend of Tafel slope increase was observed for both electrocatalysts, where the Tafel slopes of β-PbO₂ and Pb₃O₄ are 825 mV dec⁻¹ and 681 mV dec⁻¹, respectively. Similar findings have been reported in EOP-related literature (Table S1, ESI†). The inflection in the Tafel slopes implies a change in the water oxidation kinetics, which could be originated from continual alteration of electrode surfaces and interface structures in different potential regions.^11,12

Fig. 1 Electrochemical ozone production (EOP) performance of commercial Pb₃O₄ in the neutral electrolytes. (a and b) The linear sweep voltammetry curves and corresponding Tafel slopes of β-PbO₂ and Pb₃O₄ in saturated K₂SO₄ solution (pH = 7.0). (c) The faradaic efficiency towards ozone generation for the Pb₃O₄ at different applied current densities using the EOP MEA electrolyzer. (d) Tafel plot of the partial O₃ current densities normalized by electrochemically active surface area of the Pb₃O₄. (e) Pourbaix diagram of the lead oxide phases at the different electric fields and pH conditions. (f) DFT calculated pH-potential phase diagram of the β-PbO₂(101) surface. The phase diagram is constructed for the lead oxide surfaces with total OH*/O* occupation. (g) Comparison of the EOP performances of Pb₃O₄ and β-PbO₂ for more than 900 h at the constant current condition of 1 A cm⁻² (ultrapure water is used in the EOP MEA electrolyzer as the electrolyte (c–e)).

The EOP performance of the Pb₃O₄ electrocatalyst was then evaluated by incorporating and assembling the catalyst into a membrane electrode-assembly (MEA) based EOP electrolyzer (Fig. S4, ESI†). The MEA consists of a Nafion 117 membrane coated with Pb₃O₄ and 20% Pt/C on each side as the anode and cathode catalysts, respectively. Fig. 1c and Fig. S5, S6 (ESI†) show the Faraday efficiencies (FE) and yield rate towards ozone production using Pb₃O₄ as a catalyst at various current densities. From Fig. S5 (ESI†), it is apparent that an increase in the current density increases the FE(O₃) of the Pb₃O₄ electrocatalyst and reaches a maximum value of 13.4% at 1.0 A cm⁻², where the responsive potential is ∼2.6 V vs. RHE. However, the FE(O₃) slightly drops with further current density increment. A similar trend can also be observed for β-PbO₂ (Fig. S7 and S8, ESI†), where the maximum FE(O₃) of 16.4% is also obtained at the current density of 1.0 A cm⁻². Additionally, when the current density was lower than 0.5 A cm⁻², the FE(O₃) rises rapidly with the current density increment for both cases, where the anode potential corresponded to ∼1.5–1.9 V vs. RHE. Moreover, the corresponding Tafel plot obtained from the FE(O₃) and the LSV current densities (Fig. 1d) exhibits an inflection at ∼1.9 V vs. RHE, where the Tafel slope value increases from 367 to 1065 mV dec⁻¹. These results indicate that the EOP process's kinetics becomes slower when the current density reaches 0.5 mA cm⁻². Similarly, the Tafel plot of OER current density also shows a slope increment, but at 1.6 V vs. RHE, the same as the overall water oxidation reaction mentioned above (Fig. S9, ESI†).

To further investigate the mechanisms responsible for the kinetic change, the Pourbaix diagram of lead oxide phases and the pH-potential phase diagram of lead oxide surfaces are calculated. Fig. 1e shows that the PbO₂ phase exhibits higher stability under potential higher than 1.5 V vs. RHE in pH 7 electrolyte. These results suggest that the Pb₃O₄ surface is thermodynamically favored to be reconstructed to PbO₂ under the tested potential range, which explains the similarity of the Tafel plots obtained for the Pb₃O₄ and PbO₂ catalysts. Meanwhile, Fig. 1f suggests that the process 1/2 ML O*/OH* phase is transferred to 1ML O* at the potential of ∼1.6 V vs. RHE at pH 7, which is very close to the potential of the Tafel plot inflection points for overall water oxidation and OER. This observation indicates that the 1/2 ML O*/OH* surface adsorbates are kinetically more favorable for the OER than 1ML O*. Moreover, the EOP starts with the transformation into 1ML O*, implying the important role of 1ML O* surface adsorbates in facilitating EOP and hindering the OER.

On the basis of this analysis, the long-term operation of Pb₃O₄ in the EOP electrolyzer was evaluated at high current densities in ultrapure water instead of K₂SO₄ electrolyte to investigate the feasibility of potential industrial applications (Fig. 1g) and to avoid serious corrosion problems.⁷ Impressively, the Pb₃O₄ anode-based electrolyzer could stably drive EOP at a high current density of 1 A cm⁻² for more than 900 hours of continuous operation. Our strategy substantially outperforms other reported EOP electrocatalysts in current density, product generation rate, and reported operation time while maintaining high faradaic efficiency (Fig. S10 and Table S2, ESI†). Furthermore, during the entire test, the corresponding anode potential is stabilized in the range of 2.40 to 2.45 V vs. RHE, and FE was maintained at 12.0%. For the sake of comparison, the β-PbO₂ anode exhibits reduced activity during the over-900 hour-test under similar conditions (Fig. 1e and Fig. S11, ESI†). In detail, the optimal EOP activity by β-PbO₂ is reached with an ozone yield rate of ∼490 mg h⁻¹ g⁻¹_catalyst and a corresponding FE of 16.4% after 100 h of EOP operation. After that, the ozone production rate of β-PbO₂ drops to ∼218 mg h⁻¹ g⁻¹_catalyst after operating over 900 h, while the FE falls to 7.5%. Although β-PbO₂ exhibits higher EOP activity at the initial stage, these results suggest lower overall stability of β-PbO₂ compared to Pb₃O₄. Therefore, the inherent reason for the excellent stability of Pb₃O₄ during the long-term EOP process was further analyzed.

Structural changes during EOP

Extensive characterization techniques were employed to study the change in the catalyst's structure during the EOP process, aiming to elucidate activity and stability enhancement mechanisms. To confirm the structural changes of Pb₃O₄ and β-PbO₂ during the EOP process, the microstructure of the electrocatalyst was monitored at different activation times. It is interesting to note that after 100 h of continuous operation at 1 A cm⁻² (∼2.4–2.45 V vs. RHE), the EOP electrocatalyst turns dark brown, indicating possible structural changes of Pb₃O₄ during the EOP process (Fig. S12 and S13, ESI†). Furthermore, ex situ HRTEM was employed to clarify the time-dependent structure transformation with respect to reaction time (Fig. 2a and Fig. S14, ESI†). As shown in Fig. S15 (ESI†), a newly formed species with a lattice distance of 0.35 nm was detected after 15 h EOP reactions, corresponding to the (110) plane of β-PbO₂ (PDF#41-1492. Moreover, the exposed interplanar spacings of 0.185 and 0.279 can be assigned to the (211) and (101) surfaces of β-PbO₂ after a 50 h EOP process,^13,14 indicating that the structural reconstruction of Pb₃O₄ to β-PbO₂ was essentially completed after 50 h of EOP operation. Furthermore, Fig. S16 and S17 (ESI†) show the decrease in the Pb₃O₄ particle size, reflecting the transformation that occurred during the EOP process. In contrast, no phase or morphological change was observed for β-PbO₂ (Fig. 2b and Fig. S18, ESI†). These observations agree well with the Pourbaix diagram calculated previously, which reveals that the β-PbO₂ is more thermodynamically stable under the EOP operating conditions.

Fig. 2 Structural characterization of Pb₃O₄ and β-PbO₂ in MEA under different EOP durations. (a and b) HRTEM images and (c and d) ex situ XRD patterns of Pb₃O₄ and β-PbO₂ before and after EOP treatment in EOP electrolyzer at a high current of 1 A cm⁻².

The X-ray diffraction (XRD) technique was also used to get more insight into structural changes of Pb₃O₄ and β-PbO₂ during the EOP operation. Fig. 2c and Fig. S19a (ESI†) demonstrate that the XRD pattern of the Pb₃O₄ structure showed a pronounced sharp peak at around 26.3°, which is attributed to the (211) plane of Pb₃O₄ (PDF#41-1493, Fig. S20, ESI†). Other characteristic peaks of (110), (220), (112), (310), (202), and (332), appeared at 2θ = 14.2°,28.6°, 30.7°, 32.0°, 34.0°, and 52.0°, respectively, which are well consistent with standard peaks of Pb₃O₄. A notable peak at 2θ = 49.8° attributed to the (402) plane significantly becomes weak after 10 h of EOP operation, and no notable peak shift is observed, which indicates an inconspicuous phase change degree at this stage. A new peak at ∼25.4° assigned to the (110) of β-PbO₂ starts to appear after 15 h, which is consistent with the HRTEM results mentioned above. Meanwhile, the characteristic peaks of β-PbO₂ at 31.9° (101), and 49.0° (211) appear and gradually intensify as the EOP test continues. Meanwhile, the diffraction peak at 26.3° becomes broadened and weakened, which can be attributed to the phase change of the Pb₃O₄ samples after 25 h. After 50 hours, the peak assigned to the Pb₃O₄ phase disappeared. At the same time, new diffraction peaks located at 2θ = 25.4°, 31.9°, 36.2°, 49.0°, and 52.1° appeared, which can be assigned as the (110), (101), (200), (211), and (220) facets of β-PbO₂, respectively. Moreover, these diffraction peaks remain steady even after 100 h, indicating that Pb₃O₄ has been completely converted to β-PbO₂.

To further probe the time-resolved evolution of Pb₃O₄ during EOP at anodic potentials at around 2.4 to 2.45 V vs. RHE, the ex situ Raman characterization was performed (Fig. S19b, ESI†). The Raman spectrum of the initial sample shows peaks at 122, 152, 224, 314, 390, 480, and 550 cm⁻¹, which are attributed to the feature of Pb₃O₄ (Fig. S21, ESI†).^15,16 Note that at the initial stage (<10 h), all the main peaks of Pb₃O₄ remain unchanged. The peak at 142 cm⁻¹ appears after 15 h of operation, demonstrating the formation of new species. Two new peaks at 140 and 277 cm⁻¹ are detected after 25 h of EOP operation. Then, another two peaks at 136 and 270 cm⁻¹ are gradually formed after 50 h, which demonstrates the complete transformation of Pb₃O₄ to β-PbO₂¹⁶ (Fig. S22, ESI†). These results confirm that the electrocatalyst reaches a stable state after an activation stage of 50 h, which agrees with the XRD results. In contrast, the commercial β-PbO₂ does not undergo phase change during long-term EOP operation, which is confirmed by the ex situ XRD and ex situ Raman results (Fig. 2d and Fig. S23 and S24, ESI†).

As β-PbO₂ is thermodynamically more stable, the phase shuttling from Pb₃O₄ to β-PbO₂ observed in the XRD and Raman characterizations can be explained by the Pourbaix diagram mentioned above. However, the mechanism underneath the fact that the Pb₃O₄-reconstructed β-PbO₂ is more active and stable during the long-term test is still unclear. In our previous work, the (101) and (110) facets of β-PbO₂ are confirmed to contribute to the high activity of EOP. Notably, the XRD patterns show that the Pb₃O₄-reconstructed β-PbO₂ has significantly different crystalline plane ratios than the commercial β-PbO₂ during the EOP period of 500 h to 950 h. Nevertheless, the Pb₃O₄-reconstructed β-PbO₂ remains stable in the ratio of (101) to (110) facet, as shown in Fig. 2c, d and Fig. S25 (ESI†), while the commercial β-PbO₂ exhibited a remarkable decrease in the ratio of (101) to (110) facet, which can explain its difficulty in maintaining EOP activity.

Electronic property changes during EOP

The O 1s X-ray photoelectron spectra (XPS) were employed to analyze the surface oxygen vacancies of the Pb₃O₄ and β-PbO₂ electrocatalysts during the long-term EOP tests (Fig. 3a–c). The formation of adsorbed oxygen species at about 531.0 eV is closely related to the existence of surface oxygen vacancies (V_O).^17,18 The content of V_O on the surface of the β-PbO₂ catalyst increases by ca. 50% after 950 h of EOP operation, while slight increases in V_O content are observed for the Pb₃O₄ catalyst during the reconstitution process. Besides, the surface lattice oxygen ratio of Pb₃O₄ declines slightly during the reconstruction process (Fig. S26 and Tables S3 and S4, ESI†), which indicates that the Pb₃O₄-reconstructed possesses a more stable surface crystal structure during the EOP process. Furthermore, the XPS on Pb element valence state was also conducted. The valence states of Pb elements in the Pb₃O₄ exhibit slight variation during the reconstruction, while there is a significant decrease in the Pb⁴⁺ proportion in β-PbO₂ during the same period, which is consistent with the decrease in the proportion of its lattice oxygen (Fig. S27 and Tables S5 and S6, ESI†). These results declare that the Pb₃O₄-reconstructed β-PbO₂ has more stable surface structures during the EOP process.

Fig. 3 Structural characterization and stability investigations of Pb₃O₄ and β-PbO₂ during the EOP process. (a and b) O1s spectra of Pb₃O₄ and β-PbO₂ before and after EOP treatment in EOP electrolyzer at a high current of 1.0 A cm⁻². (c) Changes in surface oxygen vacancy ratios (V_O%) of Pb₃O₄ and β-PbO₂ during the EOP process corresponding to a and b. (d–f) XAFS analysis of Pb₃O₄ and β-PbO₂ under different EOP durations. The XANES Pb L₃-edge spectra (d and e), and Pb L₃-edge curve fits (lines) of Pb₃O₄ and β-PbO₂, shown in R-space (f).

To further elucidate the stability of the exposed lead oxide surfaces, COHP for the exposed Pb–O covalent bonds over the β-PbO₂101), β-PbO₂(110), Pb₃O₄(110), and Pb₃O₄(211) surface was calculated (Fig. S28, ESI†). The COHP results show that the Pb–O covalent interaction is mainly dominated by the interaction between Pb-2s/2p and O-2p orbital; the Pb-5d orbital is fulfilled and not responsible for the Pb–O covalent interaction. Such Pb–O covalent interaction is different from most transition metal oxides (TiO₂, Al₂O₃, etc.), in which the interaction is dominated by the d orbital and O-2p orbital.¹⁹ The weak Pb–O bond strength could result in the migration of lattice oxygen, which also explains the reconstruction of lead oxides during the EOP process. The integrated COHP (ICOHP) values are calculated to be 4.26 and 3.54 for β-PbO₂(101), and Pb₃O₄(110), respectively. By comparing the ICOHP values of the Pb–O bond over different surfaces, it can be concluded that the Pb–O covalent interaction on different surfaces follows the order of PbO₂(101) > Pb₃O₄(110). These results again prove that the β-PbO₂(101) possesses higher surface stability, which also explains why the β-PbO₂ with a high ratio of (101) facets obtained by Pb₃O₄ phase shuttling represents excellent long-term EOP stability compared to that of commercial β-PbO₂.

Moreover, the X-ray absorption spectroscopy (XAS) tests of the Pb L₃-edge were carried out to track the evolution process of the atomic structures of Pb₃O₄ and β-PbO₂ during the long-term EOP operation.^20,21 As depicted in the X-ray absorption near edge structure (XANES) spectra (Fig. 3d), the positive shift in energy absorption of the white line by the Pb₃O₄ pre-catalyst can be found (red arrow direction) after the EOP operation (Fig. 3d), which demonstrates that the valence states of Pb gradually rise during the whole process. Moreover, the negative shift in energy absorption of the white line of the β-PbO₂ pre-catalyst is displayed with time (Fig. 3e), which implies the continuous decrease in the valence state of Pb during the EOP process (Table S7, ESI†). It should be noted that the curves have a bond-length compensation in Fig. 3f. The curve-fitting analysis of the extended X-ray absorption fine structure (EXAFS) spectrum (Fig. S29, S30 and Table S8, ESI†) indicates that the coordination number of the Pb₃O₄ sample is 4.0 at 0 h, and then increases to 4.1 at 500 h and 4.3 at 950 h. The coordination number of the Pb–O bond keeps increasing during the EOP process, revealing that O atoms are continuously filled into the lattice structure during the reconstruction of Pb₃O₄, which can lead to the decrease of the Pb–O bond length and the increase of the Pb oxidation state. Such interpretation is also consistent with the XANES results. In addition, the fitting results also exhibit that the coordination number of the Pb–O bond of β-PbO₂ decreases from 6 at 0 h to 5.0 at 950 h, which are higher than the Pb–O coordination numbers of Pb₃O₄-reconstructed β-PbO₂ at the same time. The lower coordination number of Pb–O bonds signifies that the structure of Pb₃O₄-reconstituted β-PbO₂ contains abundant Vo, which can affect the electronic structure of the catalyst hosts.²² This electronic regulation facilitates lattice oxygen release during EOP, which can eventually lead to activity enhancement.²³

Phase shuttle mechanism

To achieve an in-depth understanding of the origin of the phase shuttling of Pb₃O₄ during the EOP process, the mechanism of the Pb₃O₄ electrocatalyst was investigated in neutral media using ¹⁸O-labeling combined with DEMS (Fig. S31 and S32, ESI†).^24,25 The controlled potential electrolysis was carried out at 2.4 V vs. RHE in saturated K₂SO₄ prepared with 98% H₂¹⁸O. Fig. 4a and b show the occurrence of both OER and EOP processes simultaneously, and the EOP FE reaches ∼17.4%. Meanwhile, the signal strength of O₃ (¹⁶O¹⁶O¹⁶O) with a mass-to-charge ratio of 48 is the greatest (Fig. 4c), which is approximately 500 times higher than that of other O₃ signals. This result suggests that ∼99.8% of the O₃ signal comes from the ⁴⁸O₃ (¹⁶O¹⁶O¹⁶O) signal (Table S9 and Fig. S33, ESI†). The above results indicate that the product of the EOP process using Pb₃O₄ is mainly composed of three lattice oxygen atoms coupling with each other to generate ⁴⁸O₃, associated with the participation of lattice oxygen.^26–28 Meanwhile, the dominating signal (∼80.6%) of the oxygen generated during the electrocatalytic process comes from ³⁶O₂ (¹⁸O¹⁸O) (Fig. S33, ESI†). Unlike the EOP process, the OER process undergoes both the adsorbate evolution mechanism (AEM) and LOM pathways,^29,30 and the AEM pathway played a significant role in the OER process.

Fig. 4 DEMS and theoretical calculation results. (a) DEMS signals of O₂ and O₃ from the reaction products cycled in saturated K₂SO₄ (H₂¹⁸O) by in situ DEMS of Pb₃O₄ at 2.4 V vs. RHE. (b) DEMS signals of ³²O₂, ³⁴O₂, ³⁶O₂, ⁴⁸O, ⁵⁰O₃, ⁵²O₃ and ⁵⁴O₃ on Pb₃O₄ during the OER and EOP process at 2.4 V vs. RHE. (c) Constituent content of ¹⁶O¹⁶O¹⁶O (m/z = 48), ¹⁶O¹⁸O¹⁶O (m/z = 50), ¹⁸O¹⁸O¹⁶O (m/z = 52) and ¹⁸O¹⁸O¹⁸O (m/z = 54) through DEMS measurements. (d) DFT calculated potential energy diagrams of AEM and LOM reaction pathways under the conditions of U = 0.0 V,³¹ (e) DFT calculated potential energy diagrams of AEM and LOM reaction pathways under the equilibrium potential condition (U = 1.51 V vs. SHE).³¹ (f) Applied potential effect on O₂/O₃ adsorption energies of the PbO₂(101) surface.

To investigate the underlying mechanism for the phase shuttling of Pb₃O₄ to β-PbO₂ during the EOP process, two sets of ¹⁸O isotope labeling experiments (series A and B) were performed. In experiment set A, the oxygen atoms of Pb₃O₄ were labeled with ¹⁸O in K₂SO₄ electrolyte prepared with H₂¹⁸O at constant voltage for 20 min (Fig. S34 and S35, ESI†). Since the isotope content of the Pb₃¹⁸O₄ is identical to the isotope content of the electrolyte, no change in the isotope content has been observed. Afterward, ¹⁸O-labeled electrodes were rinsed with H₂¹⁶O five times to remove the remaining H₂¹⁸O. In experiment set B, the prelabeled oxide was then used to evolve O₂ and O₃ in H₂¹⁶O-containing electrolytes. Even though the measurement was conducted in H₂¹⁶O-containing electrolytes, the spectra showed that ³⁴O₂, ³⁶O₂, ⁵⁰O₃, ⁵²O₃, and ⁵⁴O₃ were formed at potentials for EOP (Fig. S36 and S37, ESI†). Therefore, the Pb₃O₄ can be gradually reconstructed into β-PbO₂ during the EOP process through oxygen exchange between the electrocatalyst and locally adsorbed water molecules on the electrocatalyst surface. Furthermore, our previous work has comprehensively investigated the possible reaction pathways of β-PbO₂ during the EOP process, of which the results reveal that the LOM pathway dominated the O₃ formation on the anode, as shown in Fig. 4d and e.³¹ Meanwhile, the LOM-2 reaction path (three oxygen atoms originating from the lattice oxygen) is more favorable and consistent with the above DEMS results. Based on the lattice oxygen mechanism (LOM), the O₃* is generated via the coupling of the adsorbed O₂* intermediate and one surface lattice oxygen atom. Thus, the key descriptor for O₂/O₃ selectivity could be attributed to the adsorption energies of O₂ and O₃, and Fig. 4f shows the calculated O₂/O₃ adsorption energies of the PbO₂(101) surface. With the increase of the applied potential, both the O₂ + 2O_v (Oxygen vacancy) and O₃ + 3O_v formation energies on the β-PbO₂(101) surface become more negative, indicating the facile migration and coupling of surface lattice oxygens to O₂/O₃. Additionally, the energies of O₃ + 3O_v drop more dramatically, indicating the increasing selectivity of O₃ on the β-PbO₂(101) surface. These results further explain why the Pb₃O₄-reconstructed β-PbO₂ with the stable presence of (101) active surfaces exhibits a more stable EOP activity than the β-PbO₂.

From the above experimental and theoretical results, the concept of “phase shuttling” during the EOP process is proposed. As shown in Scheme 2, the phase shuttling originating from the external electric field can promote the process of surface lattice oxygen diffusion and reconstruction of the Pb₃O₄ phase to the more stable β-PbO₂ phase (Fig. S38, ESI†), in which the weak covalent interaction of Pb–O most likely results in the phase shuttling of Pb₃O₄. In spite of the potential enhancement of the reaction activity through the LOM process, the participation of the lattice oxygen can also lead to catalyst instability from thermodynamic considerations.³² In this work, the Pb₃O₄-reconstructed β-PbO₂ maintains excellent EOP activity and stability during the long-term EOP test, which can be attributed to its lower coordination structure of Pb–O bonds and the stable presence of (101) and (110) active surfaces.

Scheme 2 Schematic illustration of phase shuttling of Pb₃O₄ by the applied electric field.

EOP electrolyzer for SARS-CoV-2 disinfection

As a representative demonstration of on-site on-demand application, we used the as-synthesized ozone from our device for disinfection against SARS-CoV-2. Ozone has been demonstrated to be effective in diminishing bacteria, spores, and viruses. However, few reports have focused on using ozonated water for virus inactivation.^33,34 The Pb₃O₄-based EOP electrolyzer was used to generate ozonated water for inactivating SARS-CoV-2 (Fig. 5a). Fig. 5b shows that after injecting ozone gas into quantitative normal saline for 40 min, the concentration of ozonated water reaches the peak and remains stable for 60 min at ca. 2.12 mg L⁻¹ (Fig. S39 and S40, ESI†). The results show that the virus titer is significantly reduced along with a contact time over 30 s to 10 min by the ozonated water. Within 30 s, more than 99.99% of the viruses can be inactivated, and the inactivation efficiency can rise to 100.00% when the contact time increases to 5 min (Fig. 5c). Fig. 5d shows Vero cells as a negative control (with ozonated water treatment) and infected Vero cells (without ozonated water treatment) under a 400× microscope, which clearly demonstrates the excellent inactivation effect of ozonated water on SARS-CoV-2. It is generally considered that ozone could inactivate viruses by diffusing through the protein coat into the nucleic acid core, damaging the viral RNA, and disrupting the exterior protein layer by oxidation.³⁵ All of these results demonstrate the efficiency of EOP-generated ozone on SARS-CoV-2 inactivation, making our design economically and environmentally appealing for practical use when scaled up, which provides a solution to help contain the spread of SARS-CoV-2.

Fig. 5 Inactivation of SARS-CoV-2 by ozonated water. (a) Schematic diagram of the electrochemical inactivation process. (b) The concentration of ozonated water collected in 15 mL of normal saline under normal temperature and pressure (Inset: The qualitative result of ozonated water concentration). (c) SARS-CoV-2 inactivation ratio by ozonated water. (d) Microscopic pictures of negative control and infected cells by SARS-CoV-2 (400× magnification).

Conclusions

In summary, the phase shuttle effect on the EOP performance of Pb₃O₄ has been comprehensively investigated by in situ/ex situ approaches. The transformation from Pb₃O₄ to β-PbO₂ was observed during the EOP test, and the as-reconstructed β-PbO₂ exhibits enhanced EOP activity and stability compared to the commercial β-PbO₂. The calculated Pourbaix diagrams of lead oxides prove that the phase shuttling to β-PbO₂ is thermodynamically favorable under the EOP operating conditions. The ex situ XRD and XAS characterizations suggest the excellent performance mainly attributed to the stable presence of (101) and (110) surfaces and the low Pb–O coordination number of the reconstructed β-PbO₂. The in situ DEMS tests reveal that the phase shuttling from Pb₃O₄ to β-PbO₂ undergoes the LOM pathway. In addition, the ozonated water prepared by the Pb₃O₄-based EOP MEA electrolyzer can inactivate SARS-CoV-2 efficiently and rapidly. This work provides fundamental insights into the lead oxides under EOP conditions and paves a new way to design high-performance EOP catalysts for potential industrial applications.

Author contributions

X. Z., Z. W., Z. J. X., and J. G. W. conceived and supervised the study. J. L. performed most of the experiments, carried out the analysis, and produced the figures with the support of X. Z. L. S. and X. H. helped to characterize the samples. G. F. and W. W. L. contributed to the theoretical calculation with the support of J. G. W. and S. B. W. Z. N. Y. and B. B. W. and H. W. performed experiments for the inactivation of SARS-CoV-2. L. J. and S. B. W. analyzed the data and constructed the discussion. J. L., S. B. W., C. C. D. and Kamal wrote the manuscript. All authors participated in manuscript editing and approved the manuscript.

Conflicts of interest

The authors declare no competing interests.

Acknowledgements

The authors acknowledge the financial support from Zhejiang Provincial Natural Science Foundation (No. LR22B060003), Key Research&Development Program of Zhejiang (2022C03120), National Health Commission Scientific Research Projects (WKJ-ZJ-2113, WKJ-ZJ-2220) and the National Natural Science Foundation of China (NSFC-22078293, 21625604, 2017R5203, and 91934302). We greatly thank the Research Center of Analysis and Measurement, Zhejiang University of Technology, for using their XRD equipment. We also thank Prof. R. R. Li. and Dr X. L. Chen. (Engineering Research Centre of Recycling & Comprehensive Utilization of Pharmaceutical and Chemical Waste of Zhejiang Province, Taizhou University) for help with the ex situ Raman characterizations. This work is partially supported by the Singapore MOE Tier 1 grant (RG62/21) and the Singapore National Research Foundation under its Campus for Research Excellence and Technological Enterprise (CREATE) programme.

References

1. J. Wang and H. Chen, Sci. Total Environ., 2020, 704, 135249.
2. K. Lee, H. Lin and K. F. Jensen, React. Chem. Eng., 2017, 2, 696–702.
3. S. G. Van Ornum, R. M. Champeau and R. Pariza, Chem. Rev., 2006, 106, 2990–3001.
4. C. G. Joseph, Y. Y. Farm, Y. H. Taufiq-Yap, C. K. Pang, J. L. H. Nga and G. Li Puma, J. Environ. Chem. Eng., 2021, 9, 106099.
5. V. Naresh, L. Elias, S. A. Gaffoor and S. K. Martha, J. Electrochem. Soc., 2019, 166, A74–A81.
6. Y. Zeng, M. Zhao, Z. Huang, W. Zhu, J. Zheng, Q. Jiang, Z. Wang and H. Liang, Adv. Energy Mater., 2022, 12, 2201713.
7. F.-Y. Chen, Z.-Y. Wu, Z. Adler and H. Wang, Joule, 2021, 5, 1704–1731.
8. X. Liu, J. Meng, J. Zhu, M. Huang, B. Wen, R. Guo and L. Mai, Adv. Mater., 2021, 33, e2007344.
9. S. Jin, ACS Energy Lett., 2017, 2, 1937–1938.
10. S. Zhao, J. Huang, Y. Liu, J. Shen, H. Wang, X. Yang, Y. Zhu and C. Li, J. Mater. Chem. A, 2017, 5, 4207–4214.
11. R. Amadelli, L. Armelao, A. Velichenko, N. Nikolenko, D. Girenko, S. Kovalyov and F. Danilov, Electrochim. Acta, 1999, 45, 713–720.
12. A. Babak, R. Amadelli, A. De Battisti and V. Fateev, Electrochim. Acta, 1994, 39, 1597–1602.
13. Y. Yan, Y. Gao, H. Zheng, B. Yuan, Q. Zhang, Y. Gu, G. Zhuang, Z. Wei, Z. Yao, X. Zhong, X. Li and J. Wang, Appl. Catal., B, 2020, 266, 118632.
14. J. D. McGettrick, K. Hooper, A. Pockett, J. Baker, J. Troughton, M. Carnie and T. Watson, Mater. Lett., 2019, 251, 98–101.
15. I. Costantini, P. P. Lottici, K. Castro and J. M. Madariaga, Minerals, 2020, 10(5), 468.
16. D. Guo, C. Robinson and J. E. Herrera, Corros. Sci., 2016, 103, 42–49.
17. C. Zhao, N. Li, R. Zhang, Z. Zhu, J. Lin, K. Zhang and C. Zhao, ACS Appl. Mater. Interfaces, 2019, 11, 47858–47867.
18. X. Duan, W. Wang, Q. Wang, X. Sui, N. Li and L. Chang, Chemosphere, 2020, 260, 127587.
19. N. J. O’Connor, A. S. M. Jonayat, M. J. Janik and T. P. Senftle, Nat. Catal., 2018, 1, 531–539.
20. W. Qu, S. Niu, D. Sun, H. Gao, Y. Wu, Z. Yuan, X. Chen, Y. Wang, T. An, G. Wang and F. Zhao, Adv. Sci., 2021, 8, 2002889.
21. D. G. Nevidomskaya, T. M. Minkina, A. V. Soldatov, V. A. Shuvaeva, Y. V. Zubavichus and Y. S. Podkovyrina, J. Soils Sediments, 2015, 16, 1183–1192.
22. Y. Tian, S. Wang, E. Velasco, Y. Yang, L. Cao, L. Zhang, X. Li, Y. Lin, Q. Zhang and L. Chen, iScience, 2020, 23, 100756.
23. H. Liu, X. Li, C. Peng, L. Zhu, Y. Zhang, H. Cheng, J. Cui, Q. Wu, Y. Zhang, Z. Chen, W. Zou, W. Gu, H. Huang, J. Wang, B. Ye, Z. Fu and Y. Lu, J. Mater. Chem. A, 2020, 8, 13150–13159.
24. A. Grimaud, O. Diaz-Morales, B. Han, W. T. Hong, Y. L. Lee, L. Giordano, K. A. Stoerzinger, M. T. M. Koper and Y. Shao-Horn, Nat. Chem., 2017, 9, 457–465.
25. N. Zhang and Y. Chai, Energy Environ. Sci., 2021, 14, 4647–4671.
26. A. Grimaud, W. T. Hong, Y. Shao-Horn and J. M. Tarascon, Nat. Mater., 2016, 15, 121–126.
27. N. Zhang, X. Feng, D. Rao, X. Deng, L. Cai, B. Qiu, R. Long, Y. Xiong, Y. Lu and Y. Chai, Nat. Commun., 2020, 11, 4066.
28. Z.-F. Huang, J. Song, Y. Du, S. Xi, S. Dou, J. M. V. Nsanzimana, C. Wang, Z. J. Xu and X. Wang, Nat. Energy, 2019, 4, 329–338.
29. J. T. Mefford, X. Rong, A. M. Abakumov, W. G. Hardin, S. Dai, A. M. Kolpak, K. P. Johnston and K. J. Stevenson, Nat. Commun., 2016, 7, 11053.
30. Y. Pan, X. Xu, Y. Zhong, L. Ge, Y. Chen, J. M. Veder, D. Guan, R. O’Hayre, M. Li, G. Wang, H. Wang, W. Zhou and Z. Shao, Nat. Commun., 2020, 11, 2002.
31. W. Li, G. Feng, S. Wang, J. Liu, X. Zhong, Z. Yao, S. Deng and J. Wang, J. Phys. Chem. C, 2022, 126, 8627–8636.
32. T. Binninger, R. Mohamed, K. Waltar, E. Fabbri, P. Levecque, R. Kotz and T. J. Schmidt, Sci. Rep., 2015, 5, 12167.
33. A. Blanco, F. B. Ojembarrena, B. Clavo and C. Negro, Environ. Sci. Pollut. Res. Int., 2021, 28, 16517–16531.
34. C. Tizaoui, R. Stanton, E. Statkute, A. Rubina, E. Lester-Card, A. Lewis, P. Holliman and D. Worsley, J. Hazard. Mater., 2022, 428, 128251.
35. C. Tizaoui, Ozone: Sci. Eng., 2020, 42, 378–385.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d3ey00015j

‡ Equal contribution.

---