Enhanced localized electron density from PdCu nanoparticle loading on a defective TiO₂ support for selective nitrate electroreduction to ammonia

Abstract

The electrocatalytic nitrate (NO₃⁻) reduction reaction (NITRR) to synthesize ammonia (NH₃) offers a promising method for both pollution abatement and chemical production. However, the sluggish multi-electron/proton-involved steps limit ammonia selectivity and yield. Herein, we loaded PdCu nanoparticles (PdCu NPs) onto defective TiO₂ to regulate the localized electron density for selective NH₃ synthesis. Our results show a significant increase in NH₃ yield and selectivity (112.0 to 322.7 μmol cm⁻² h⁻¹ and 34.6 to 80.1%, respectively), induced by the introduction of PdCu NPs. Importantly, in situ and operando investigations combined with density functional theory (DFT) calculations reveal that PdCu NPs enhance the localized electron density at oxygen vacancies. This regulated electron density suppresses N–N coupling and accelerates N–H bonding processes, thus modulating NH₃ selectivity. This unique insight based on localized electron density control provides a rational design strategy for metal-defect support catalysts in ammonia production using electrochemical processes.

---

1. Introduction

Ammonia (NH₃) is essential in industry, the military, and the economy.^1–3 However, the industrial NH₃ synthesis pathway, known as the Haber–Bosh process, is operated on a centralized and large scale at high temperatures and pressures, contributing to 2% of global energy consumption and 1.5–2% of global CO₂ emissions. Therefore, there is a high demand to develop cost-effective and sustainable technologies for decentralized NH₃ synthesis under mild conditions.^4,5 The electrochemical nitrate reduction reaction (NITRR) to ammonia provides a promising way to turn waste into wealth due to the high solubility and weak N–O bond of nitrate (204 kJ mol⁻¹ compared to 941 kJ mol⁻¹ for the N≡N bond).^6–15 However, the low ammonia selectivity and limited yield, alone with several competing reactions, hinder the practical applications of NH₃ synthesis via the NITRR. For example, the hydrogen evolution reaction (HER), which starts at 0 V vs. RHE (all potential mentioned below is converted to RHE), compared to NH₃ at 0.88 V, intensifies with increasing overpotential and suppresses the NITRR to some extent.¹⁶ Furthermore, the N₂ formation reaction (1.25 V) involves the transformation of five electrons and six protons, making it more favourable in terms of reaction kinetics than NH₃, for which the transformation of eight electrons and nine protons is involved.^17–19 Therefore, it is critical to reveal the promoting mechanism for the selective NH₃ synthesis and to understand the underlying mechanism for N₂ generation in order to design efficient catalysts for the NITRR.

The reduction of NO₃⁻ to N₂ involves an N–N coupling step, which can be prevented to hinder N₂ formation. However, the unclear hindering mechanism limits the selectivity of the NITRR. For instance, Peng et al.²⁰ reported that PdCu/Cu₂O exhibits efficient inhibition on N₂ formation, while Shi et al.²¹ suggested that PdCu/Al₂O₃ presents 80.37% N₂ selectivity during the NITRR. Nevertheless, the mechanism of the inhibition and selectivity of N₂ remains unclear. Some studies attribute the variation of selectivity during the NITRR to the metal-carrier-induced interface interactions and alloying effect,^22,23 while others ascribe it to defects on the support.^24,25 PdCu/TiO_2−x exhibits the highest NH₃ yield on PdCu-loaded catalysts via the NRR,^26–32 (Table S1†) suggesting the potential for high NH₃ yield via the NITRR. Additionally, PdCu/TiO₂ has been used in urea synthesis,³³ the NRR,³² and the NITRR,³⁴ showing the universality of PdCu/TiO₂ in nitrogenous catalytic reactions. Therefore, we constructed PdCu/TiO_2−x to understand the effect of interaction between metal nanoparticles (NPs) and support on the selectivity of the NITRR.

In this study, oxygen vacancy (OV) rich-TiO₂ (TiO_2−x) was successfully synthesized using commercial TiO₂via a facile reduction method and PdCu NPs were supported on TiO_2−xvia the impregnation method. It was found that the interface is the main active site, and PdCu NPs mainly served as electron donors, based on density function theory (DFT) calculations. Electrochemical characterization, X-ray photoelectron spectroscopy (XPS), and further DFT calculations revealed the mechanism of electron donor effects on the NH₃ synthesis via the NITRR, which enhanced the *H formation and adsorption, improved the local electron density, and hence enhanced the proton supply and electron transfer, resulting in accelerated selective NH₃ formation. As a result, PdCu NPs/TiO_2−x achieved the optimized NH₃ selectivity (80.1%) and yield rate at −1.4 V (322.7 μmol cm⁻² h⁻¹), in contrast with TiO_2−x (34.6% and 112.0 μmol cm⁻² h⁻¹). The findings in this work would shed light on the effect of metal-defect interaction on NH₃ synthesis via the NITRR, and provide unique insights into the rational design of metal-defect support catalysts.

2. Experimental section

The DFT calculations were performed using the Vienna ab initio simulation package (VASP) with projector augmented wave (PAW) pseudopotential. The exchange-correlation energy was described using the generalized gradient approximation of the Perdew–Burke–Ernzerhof (GGA-PBE) functional. Considering the strong correlation effect among the localized Ti 3d electrons, the GGA + U method with an effective U value of 3.5 eV was used. The vacuum layer, cut-off energy, convergence threshold for the force and total energy were set to 15 Å, 380 eV, 0.02 eV Å⁻¹ and 10⁻⁵ eV, respectively. Detailed characterization, analysis, calculation and preparation process can be obtained in the ESI.†

3. Results and discussion

3.1 Theoretical simulations and material characterization

Considering that catalytic activity is influenced by the Pd/Cu atomic ratios, a structural model of PdCu NPs was designed based on previous experimental results, as shown in Fig. S1.† The PdCu NPs were simulated as a model consisting of 4 Pd and 8 Cu atoms, and the geometry-optimized structure is shown in Fig. S2.† Six possible NITRR active sites on (101) edge TiO_2−x and PdCu/TiO_2−x were studied: an oxygen vacancy site on TiO_2−x (I) and PdCu/TiO_2−x (II), the interface site between PdCu NPs and TiO_2−x, *NO₃ mainly bonded to Ti³⁺ (III), the interface site between PdCu NPs and TiO_2−x, *NO₃ mainly bonded to Cu (IV), and Pd (V) and Cu (VI) sites on PdCu NPs, as shown in Fig. 1a. It can be found that PdCu NP loading induces an increase in the N–O bond length, indicating that the PdCu NP loading promotes *NO₃ activation.

Fig. 1 Theoretical prediction of the NITRR using DFT calculations. (a) The structural model and N–O bond length (Å) of *NO₃ adsorbed at different sites, (b) the Gibbs free energy (ΔG) of the adsorption of *NO₃ at different sites, (c) the Gibbs free energy (ΔG) of the adsorption of *H at different sites, and (d) the ΔG of the entire NITRR process of two possible reaction pathways at the main active sites.

To investigate the effect of PdCu NPs on NO₃⁻ adsorbability, we calculated the Gibbs free energy (ΔG) and adsorption energy (E_ad) of the adsorbed *NO₃ (Fig. 1b and Table S2†). The results show that the introduction of PdCu NPs significantly enhances the adsorbability of NO₃⁻, even though the PdCu NPs are not the main adsorption sites for *NO₃. The interface sites between PdCu NPs and TiO_2−x (III and IV) exhibit the strongest *NO₃ adsorption, where *NO₃ is bonded to both Ti³⁺ and Cu. Specifically, the adsorption configuration (III) is more stable, indicating that Ti³⁺ at the interface is the primary *NO₃ adsorption site. Moreover, we found that the adsorption sites (I, II, and III) related to Ti³⁺ are the preferred *H adsorption sites over PdCu NPs (V and VI), suggesting that Ti³⁺ at the interface is the primary *H adsorption site, as shown in Fig. 1c. Therefore, the Ti³⁺ at the interface is the main active site due to its strongest adsorption of *NO₃ and *H. Furthermore, we established a structural model for the two sites (I and III) to investigate the interaction between OVs and PdCu NPs (Fig. S2†). We studied two possible reaction pathways,^35,36 for electrocatalytic NH₃ synthesis via the NITRR. The structure models of intermediates and reaction pathways are shown in Fig. S3,† and the E_ad values of intermediates are provided in Table S3.†

The introduction of PdCu NPs changes the preferred reaction pathway and reduces the reaction barrier in the rate-determining step (RDS). Fig. 1d summarizes the entire NITRR process in two possible reaction pathways at the main active sites of TiO_2−x and PdCu NPs/TiO_2−x. Prior to the loading of PdCu NPs, path 2 (*NOH → *NHOH → *NH₂OH → *NH₂) is the preferred reaction pathway due to the lower Gibbs free energy for the formation steps of *NHOH (ΔG = 0.023 eV) and *NH₂OH (ΔG = −2.68 eV), compared to the formation steps of *N (ΔG = 0.024 eV) and *NH (ΔG = −2.22 eV), which is consistent with a previous study.³⁶ Following the introduction of PdCu NPs, the hydrogenation reaction of *NOH is significantly enhanced, and the formation of *N becomes more favourable than that of *NHOH. The hydrogenation reactions of *NOH, *N and *NH are spontaneous, but the hydrogenation reaction of the *NH₂OH intermediate is significantly suppressed. A similar trend was observed in a previous study, where the formation of *N was significantly promoted and the formation of *NOH and *NH₂OH was inhibited during the NITRR over PdCu/Cu₂O, in comparison to Cu₂O.²⁰ As a result, Path 1 becomes the preferred reaction pathway during the NITRR on PdCu NPs/TiO_2−x. Furthermore, the loaded PdCu NPs significantly enhance the *N adsorption (2.28 eV) and accelerate the rate-determining step, from ΔG_RDS = 1.49 eV for TiO_2−x to ΔG_RDS = 1.30 eV for PdCu NPs/TiO_2−x (Fig. S4†). Therefore, the introduction of PdCu NPs could enhance the *NO₃ and *H adsorption, reduce the energy barrier of the rate-determining step, and promote the electrocatalytic synthesis of NH₃via the NITRR.

Based on theoretical predictions, PdCu NPs loaded onto defective TiO_2−x with oxygen vacancies are constructed using a facile reduction and solution impregnation method with commercial TiO₂. The microscopic morphology of PdCu NPs/TiO_2−x (Fig. S5a†) shows that numerous NPs (particle average size is 3.1 ± 0.9 nm) are uniformly distributed on the surface of TiO_2−x particles (average size is 34.0 ± 0.8 nm, as shown in Fig. S5b†), The loaded NPs and support exhibit characteristic spacings of 0.199 and 0.234 nm for the (200) lattice planes of the PdCu alloy (JCPDS 48-1551)²⁸ and (112) lattice planes of anatase TiO₂ (ICDD 00-004-0477), respectively, indicating that the PdCu alloy NPs are loaded on the surface of TiO_2−x (Fig. 2a). The EDS elemental maps (Fig. 2b) reveal that the compositional distributions of both elements (Pd and Cu) in PdCu NPs/TiO_2−x are uniform, suggesting an alloy structure. The compositional distributions of Pd, Cu, Ti, and O at the energy dispersive spectroscopy (EDS) mapping surface are presented in Table S4.† The amount of Cu and Pd in PdCu NPs/TiO_2−x is quantified via ICP-MS, and the results are shown in Table S5,† which indicates that the chemical formula of the catalyst can be specified as Pd_0.026Cu_0.051 NPs/TiO_2−x. Furthermore, the selected-area electron diffraction (SAED) image of PdCu NPs/TiO_2−x (Fig. S6†) reveals the existence of PdCu alloy NPs. The transmission electron microscopy (TEM), high-angle annular dark-field – energy dispersive spectroscopy (HAADF-EDS), and SAED images confirm the successful loading of PdCu NPs and the presence of the PdCu alloy in the PdCu NPs.

Fig. 2 Microscopic morphology of as-prepared PdCu NPs/TiO_2−x. (a) TEM image of PdCu NPs on the TiO_2−x support, and (b) HAADF-EDS elemental mapping images. XPS spectra of (c) Pd 3d and (d) Cu 2p of Pd, Cu and PdCu NPs supported on TiO_2−x. The peak area ratio for Pd²⁺ increases from 9.8% to 32.3% with the introduction of Cu, while the peak area ratio for Cu²⁺ decreases from 25.8% to 11.1% with the introduction of Pd.

To further confirm the crystalline phases and oxygen vacancies of the samples, we conducted X-ray diffraction (XRD) and electron paramagnetic resonance (EPR) analyses on the catalysts. As shown in Fig. S7a,† the peaks in TiO₂ can be indexed to anatase phase TiO₂ (ICDD 00-004-0477). However, the diffraction peak related to Pd or Cu cannot be observed in the PdCu NPs/TiO_2−x spectra, indicating that PdCu is highly dispersed on TiO_2−x, which also agrees with the results of HAADF-EDS. As shown in Fig. S7b,† the TiO_2−x and PdCu NPs/TiO_2−x exhibit symmetrical EPR signals at g = 2.003, indicating that the electrons are trapped on oxygen defects.^37–39 No oxygen vacancies were found in the TiO₂ sample, whereas TiO_2−x possesses the highest concentration of oxygen vacancies. The oxygen vacancy concentration decreases after the PdCu NP loading process, indicating that some oxygen vacancies were refilled after PdCu NP loading.^40,41

To reveal the chemical states and electron transfer of TiO_2−x-based catalysts, we conducted XPS and atomic emission spectroscopy (AES) characterization. As shown in Fig. 2c, the Pd 3d spectra of Pd NPs/TiO_2−x exhibit four peaks of Pd⁰ 3d_5/2, Pd²⁺ 3d_5/2, Pd⁰ 3d_3/2, and Pd²⁺ 3d_3/2.⁴² The introduction of Cu positively shifts the binding energies (BEs) of Pd 3d, indicating that electrons are transferred from Pd to Cu.⁴³ Furthermore, the peaks of Cu 2p are assigned to Cu⁰ (or/and Cu⁺) 2p_3/2, Cu²⁺ 2p_3/2, Cu⁰ (or/and Cu⁺) 2p_1/2, and Cu²⁺ 2p_1/2,^44,45 which are negatively shifted with the introduction of Pd, confirming the electron transfer from Pd to Cu, as shown in Fig. 2d. Moreover, it is apparent from Fig. 2c and d that the peak area ratio for Pd²⁺ increases with the introduction of Cu and the peak area ratio for Cu²⁺ decreases with the addition of Pd, indicating electron transfer from Pd to Cu and the presence of the PdCu alloy. The Cu⁰ and Cu⁺ were distinguished using AES, as shown in Fig. S8.† The peaks of Cu NPs/TiO_2−x are assigned to Cu⁰, Cu⁺ and Cu²⁺.⁴⁶ The main component of Cu in the PdCu NPs is Cu⁰. For PdCu NPs/TiO_2−x, the kinetic energy of the Cu LMM Auger electrons positively shifted, reflecting the transformation from Cu²⁺ into Cu⁰ with the introduction of Pd. This indicates that electron transfer from Pd to Cu improves the stability of Cu⁰.²⁶ As one of the NITRR active elements, the stabilization of Cu⁰ leads to the promotion of the NITRR.

To understand the interaction between metal NPs and the TiO_2−x support, XPS was conducted for Ti 2p and O 1s of the catalysts, as shown in Fig. S9.† With the loading of Pd and Cu NPs, the characteristic Ti 2p and lattice O 1s peaks shift toward higher BEs, indicating electron transfer from the TiO_2−x support to Pd and Cu NPs due to the strong metal-support interaction (SMSI).^47,48 Presumably, the high concentration of Ti³⁺ ions in the support transfers electrons to Pd²⁺ and Cu²⁺ ions and maintains more Pd and Cu species with low oxidation states. In contrast, the Ti 2p and lattice O 1s peaks shift towards lower BEs with the loading of PdCu NPs, indicating that the PdCu NPs transfer electrons to the TiO_2−x support and possess different properties than Pd and Cu NPs, confirming the presence of an alloy state for PdCu NPs. The bimetal synergistic effect of the PdCu alloy promotes the breaking of the N–O bond,⁴⁹ and the effective d–d coupling of Cu 3d and Pd 4d improves the d-electronic exchange and transfer ability, hence improving the electrocatalytic activity.²⁸ As shown in Fig. S9b,† the peaks of O 1s centred between 530.7 and 531.4 eV are assigned to the O atom correlated with OVs,⁵⁰ and the negative shifting of the peaks indicates the electron transfer from metal NPs to OVs. It is apparent from the degree of shift that the formation of the PdCu alloy inhibits the electron transfer from Cu to OVs. Furthermore, the peak area ratios for O vacancies are 19.5%, 17.2%, 29.0%, and 22.6% in TiO_2−x, Pd NPs/TiO_2−x, Cu NPs/TiO_2−x, and PdCu NPs/TiO_2−x, respectively. The results indicate that introduction of Cu NPs led to an increase in the number of OVs in the TiO_2−x support,⁵¹ while the presence of Pd inhibited the process. The EPR and XPS analyses confirm the presence of OVs and the PdCu alloy in PdCu NPs/TiO_2−x and the enhanced localized electron density at OVs induced by PdCu NPs.

3.2 Electrocatalytic performance

The experiments on NH₃ yield were conducted to confirm that the loading of PdCu NPs improves the yield and selectivity of NH₃. The linear sweep voltammetry (LSV) curves (Fig. S10†) show that the current density of PdCu NPs/TiO_2−x is higher than that of TiO_2−x under the same conditions, indicating that the loading of PdCu NPs improves the electron transfer ability on the surface of the catalysts.²⁸ Moreover, the current density in the electrolyte with nitrate is higher than that in the electrolyte without nitrate at a potential higher than −1.0 V. Therefore, the NITRR was studied at different potentials from −1.0 to −1.8 V, as shown in Fig. 3a. PdCu NPs/TiO_2−x exhibits a higher NH₃ yield rate and faradaic efficiency compared to PdCu NPs and TiO_2−x, indicating improved catalytic performance with the loading of PdCu NPs. A higher NH₃ yield was achieved at higher potential, but the gradually intensified HER led to a decrease in FE. To verify the inflection point of FE at −1.4 V, the NITRR of PdCu NPs/TiO_2−x was studied at lower potentials from −0.1 to −0.6 V (Fig. S11†); it shows that the NH₃ yield rate and FE increase with the increase of potential. For PdCu NPs/TiO_2−x, the highest FE (38.5%) was achieved at −1.4 V and the NH₃ yield was 322.7 μmol cm⁻² h⁻¹ under these conditions (compared to 22.6% EF and 112.0 μmol cm⁻² h⁻¹ NH₃ yield for TiO_2−x).

Fig. 3 (a) The NH₃ yield rate and faradaic efficiency over TiO_2−x, PdCu NPs and PdCu NPs/TiO_2−x at given potentials. (b) The nitrogenous product selectivity of TiO_2−x, PdCu NPs and PdCu NPs/TiO_2−x at −1.4 V.

More importantly, the loading of PdCu NPs significantly inhibited the production of N₂ (Fig. S12†), and hence the selectivity of NH₃ in mainly nitrogenous products (NH₃, NO₂⁻ and N₂) was increased from 34.6% to 80.1% at −1.4 V, as shown in Fig. 3b. The formation of N₂ during the NITRR was significantly inhibited after PdCu NP loading, which is consistent with the DFT result (Fig. S14†). While the formation and desorption of *N₂ are spontaneous processes on TiO_2−x, the energy barrier inhibits these processes after PdCu NP loading (ΔG_RDS = 1.07 eV for interface sites of PdCu NPs/TiO_2−x). The current densities of PdCu NPs/TiO_2−x remained relatively stable during the 48 h NITRR process, and the NH₃ yield was relatively stable during 10 cycle tests, as presented in Fig. S15,† confirming the good stability of the catalyst.

We carried out experiments using different types of catalysts, including commercial TiO₂, defective TiO_2−x, Pd NPs/TiO_2−x, Cu NPs/TiO_2−x and PdCu NPs/TiO_2−x. We found that the loading of PdCu NPs significantly promotes NH₃ synthesis, as depicted in Fig. S16.† The results clearly show that PdCu NPs/TiO_2−x exhibits the highest NH₃ yield rate and current density among the catalysts. The improvement in NH₃ yield is more significant at relatively low potential, especially at −1.4 V. This finding indicates that the coexistence of Pd, Cu and partial PdCu alloy in the NPs promotes the synthesis of NH₃via the NITRR.²⁰ Pd NPs/TiO_2−x and Cu NPs/TiO_2−x present similar NH₃ yields at low potential (from −1.0 to −1.4 V). In contrast, Cu NPs/TiO_2−x significantly improves NH₃ yield at higher potential. This suggests that the competitive HER is more significant over Pd compared with Cu species, hence inhibiting the NITRR at high potential.⁵² The yield rate of NH₃ and NO₂⁻ was determined to examine the catalytic performance in detail, as shown in Fig. S17 and 18.† The results show that the main liquid product over carbon paper (CP) and PdCu NP loaded CP is NO₂⁻, while the main product over TiO_2−x and PdCu NPs/TiO_2−x loaded CP is NH₃, indicating that the loading of PdCu NPs/TiO_2−x promotes the hydrogenation processes from *NO₂ to NH₃. The promotion effect over PdCu NPs/TiO_2−x is more significant than that over TiO_2−x. The comparison of NITRR performance between TiO_2−x, PdCu NPs and PdCu NPs/TiO_2−x confirms that PdCu NP loading on the defective TiO₂ support promotes the hydrogenation processes and accelerates nitrate electroreduction to ammonia.

To further reveal the enhancement mechanism of electrocatalytic activity via PdCu NP loading, operando electrochemical impedance spectroscopy (EIS), cyclic voltammetry (CV) measurements, and open circuit potential (OCP) tests were performed. As shown in Fig. 4a, a characteristic peak centred at 3.2 Hz (start at −1.2 V) is attributed to charge storage via adsorption/desorption of *H intermediates.^53,54 The peak shifts to higher frequencies (3.2–39.8 Hz) and lower phase degrees (10.5–4.9°) with the increase in potentials (from −1.2 to −1.5 V), indicating enhanced intermediate adsorption and accelerated charge transfer at higher potential.⁵⁵ In contrast, the Bode plots for TiO_2−x in electrolyte with 0.1 M NaNO₃ (Fig. 4b) show an extra peak at 1.3 Hz from −0.8 V, which is attributed to the adsorption of *NO₃ intermediates. PdCu NPs enhance *H adsorption as evidenced by the characteristic peak of *H adsorption appearing at a lower potential (−0.7 V) compared to that for TiO_2−x (−1.2 V), as shown in Fig. 4c. Similarly, the comparison between Fig. 4b, d and f demonstrates that the promotion of *NO₃ adsorption is related to the interaction between PdCu NPs and the TiO_2−x support. By comprehensively analyzing the results from Fig. 4, it can be concluded that the adsorption of intermediates (*NO₃ and *H) is enhanced via PdCu NP loading on TiO_2−x, and the equilibrium potential for *NO₃ adsorption on PdCu NPs/TiO_2−x is lower than that for *H adsorption, as confirmed by the DFT calculations. ΔG of *NO₃ adsorption is −1.23 and −2.89 eV at OV and interface sites, respectively, and that for *H is −0.55 and −0.62 eV (Fig. S21†). This indicates that the major adsorption sites for *H are the interface and OV sites of PdCu NPs/TiO_2−x rather than metal NPs. The enhanced adsorption of *NO₃ and *H intermediates accelerates the hydrogenation processes hence promoting the NITRR process.

Fig. 4 Operando EIS analysis. Bode plots of TiO_2−x in 0.5 M Na₂SO₄ supporting electrolyte (a) without NaNO₃ and (b) with 0.1 M NaNO₃. Bode plots of PdCu NPs in 0.5 M Na₂SO₄ supporting electrolyte (c) without NaNO₃ and (d) with 0.1 M NaNO₃. Bode plots of PdCu NPs/TiO_2−x in 0.5 M Na₂SO₄ supporting electrolyte (e) without NaNO₃ and (f) with 0.1 M NaNO₃. An equivalent circuit model is shown in Fig. S20.†

The unfitted Nyquist plots at open circuit potential show the intrinsic impedance of TiO_2−x, PdCu NPs and PdCu NPs/TiO_2−x (Fig. S22a†). This demonstrates that the charge transfer of the catalyst is promoted with the loading of PdCu NPs on TiO_2−x, which is consistent with the LSV results. To evaluate the influence of PdCu NPs on the adsorption behaviour of NO₃⁻, the open circuit potential (OCP), which is affected by the variation of absorbates in the Helmholtz layer,⁵⁶ was analyzed, as shown in Fig. S22b.† With the injection of 0.1 M NaNO₃, the electrochemical cell with the PdCu NPs/TiO_2−x electrode exhibits a more significant increment in OCP (30.6 mV) than that with the TiO_2−x electrode (24.7 mV) and PdCu NP electrode (2.5 mV), indicating that the *NO₃ adsorption sites are strongly related to TiO_2−x and the *NO₃ adsorption is enhanced after PdCu NP loading. Furthermore, the double layer capacitance (C_dl), which was fitted based on the CV result, is considered an indicator of the electrochemically active surface area (ECSA) on different catalysts (Fig. S23†). The C_dl of catalysts decreases from 5.96 to 3.66 mF cm⁻² with the loading of PdCu NPs, indicating the agglomeration of TiO₂ NPs during the catalyst synthesis process and suggesting that the pore structure of TiO_2−x was partially filled by the PdCu NPs. Despite the reduction in ECSA after PdCu NP loading, the PdCu NPs/TiO_2−x catalyst shows enhanced catalytic performance. These analyses collectively demonstrate that the enhanced adsorption of *NO₃ and *H intermediates and accelerated electron transmission promote the hydrogenation processes, improving ammonia yield and selectivity.

3.3 Mechanism investigation

To dynamically monitor the NITRR process and reveal the NITRR mechanism over PdCu NPs/TiO_2−x, in situ electrochemical attenuated total reflectance surface-enhanced infrared absorption spectroscopy (ATR-SEIRAS) was performed (Fig. S24†). A series of vibrational bands can be identified, with peaks centered at 2963, 1648, 1456, 1238, 1180, and 1117 cm⁻¹ attributed to NH₄⁺, H₂O, NH₄⁺, –NO₃, –NH₃, and –NH₂, respectively.^57–60 The peak intensities for NH₄⁺, –NO₃, –NH₃ and –NH₂ change significantly with an increase in electroreduction time from 0 to 14 min, convincingly confirming that NH₃ originates from the NITRR process. However, some peaks from NH_x intermediates are covered by –OH. To distinguish Fourier transform infrared spectroscopy (FTIR) peaks from NH_x and –OH, comparison experiments based on ATR-SEIRAS characterization using three different electrolytes, including 0.1 M Na¹⁵NO₃ + 0.5 M Na₂SO₄, 0.1 M NaNO₃ + 0.5 M Na₂SO₄, and 0.5 M Na₂SO₄ were conducted. Fig. 5a and b show spectra of ¹⁵N isotope labeling experiments. The results clearly show that the center of the –OH peak is located at 3309 cm⁻¹ and the intensity of peaks (3500–3100 and 1800–1500 cm⁻¹) decreases in the electrolyte without NaNO₃, which are related to the HER. In contrast, for the electrocatalytic reaction in the electrolyte with 0.1 M NaNO₃, the increases in peak intensity are related to the NITRR, with the peaks centered at 3391, 1109 and 1025 cm⁻¹ assigned to –NH₂ and the peaks centered at 1652 and 1243 cm⁻¹ assigned to –NO₂ and –NO₃, respectively.⁶⁰ Furthermore, the -¹⁵NH₂ peak center shifts to a lower wavenumber as shown in Fig. 5b which may be related to the isotopic effect or the overlap of adjacent peaks.⁶⁰ Repeated experiments have consistently shown that the peaks shift to higher wavenumbers, suggesting that the shift is caused by the overlapping of adjacent NH₄⁺ peaks,⁶⁰ as illustrated in Fig. S25.† Contrast experiments conducted using TiO_2−x and PdCu NPs/TiO_2−x in ultrapure water and NaNO₃ electrolyte indicate that the introduction of PdCu NPs promotes H₂O dissociation and *H supply, thereby accelerating hydrogenation and *NH₂ formation. The –NH₂ peak for TiO_2−x moves to a lower wavenumber and lower intensity than that for PdCu NPs/TiO_2−x, due to the strong HER (Fig. S26†). However, the spectrum of TiO_2−x shows a more intense negative –NO₃ peak at 1238 cm⁻¹. Considering that TiO_2−x exhibits a much higher N₂ yield than PdCu NPs/TiO_2−x, the difference in –NO₃ behaviour suggests that the N₂ formation reaction on PdCu NPs/TiO_2−x is inhibited.

Fig. 5 (a) Time-resolved in situ electrochemical ATR-SEIRAS spectra over PdCu NPs/TiO_2−x at −1.8 V, and (b) potential-resolved in situ electrochemical ATR-SEIRAS spectra over PdCu NPs/TiO_2−x (30 s reaction time at each potential) in 0.5 M Na₂SO₄ supporting electrolyte with isotopically labelled 0.1 M Na¹⁵NO₃ electrolytes, unlabeled NaNO₃ electrolytes, and electrolytes without NaNO₃. The ATR-SEIRAS counter spectrum of (c) TiO_2−x in ultrapure water, (d) PdCu NPs/TiO_2−x in ultrapure water, (e) TiO_2−x in 0.1 M NaNO₃, and (f) PdCu NPs/TiO_2−x in 0.1 M NaNO₃.

The spectra in Fig. 5c–f indicate that PdCu NPs/TiO_2−x exhibits a repressed –OH desorption peak in ultrapure water-electrolyte, but an increased formation of the *NH₂ intermediate in 0.1 M NaNO₃ electrolyte, in comparison to TiO_2−x. During the NITRR process, H₂O dissociates on the surface of the PdCu NPs/TiO_2−x cathode and H⁺ adsorbs on the surface and participates in the NITRR and HER, while OH⁻ desorbs from the surface and diffuses to the anode, resulting in the observed –OH desorption peak.⁶¹ Considering the stronger HER on the PdCu NPs/TiO_2−x electrode, the suppressed –OH desorption peak at the same potential indicates stronger H₂O dissociation on its surface. The FTIR results are in accordance with the DFT results; as shown in Fig. S27,† the ΔG of the rate-determining step (*H₂O + e⁻ = *H + OH⁻) is reduced from 6.86 to 3.93 eV after PdCu NP loading, which promotes H₂O adsorption and dissociation, further leading to accelerated OH⁻ and *H formation, resulting in a suppressed –OH desorption peak and enhanced HER and NITRR. Fig. 5e and f clearly show that the loading of PdCu NPs enhances the formation of NH₂ and NO₂. The promotion of *NH₂ formation is more significant than that of *NO₂, suggesting that PdCu NPs have a greater effect on the hydrogenation of *NH_x than the reduction of *NO₃. These results demonstrate that the PdCu NPs accelerate H₂O dissociation, intensify the proton supply, and promote hydrogenation processes, thereby enhancing NH₃ formation.

We further performed DFT calculations to gain insights into the reaction mechanism and to reveal the enhanced activity and selectivity of NH₃. Fig. 6 depicts the bonding interactions between *NO₃ and the catalyst. We calculated the integrated –pCOHP (–ICOHP) to provide quantitative information on bonding strength (by calculating the energy integral up to the Fermi energy).⁶² Our results indicate that the electron transfer from the catalyst to *NO₃ is improved with the introduction of PdCu NPs, and the injected extra electrons fill the Ti–O bonding orbits, as evidenced by the improvement in the -ICOHP of the Ti–O bond from 0.82 to 1.42. The loading of PdCu NPs strengthens the Ti–O bond and weakens the N–O bond, thereby promoting the adsorption and activation of NO₃⁻. Moreover, we observed that the introduction of PdCu NPs increases the d band centre of Ti 3d (Fig. 6e). The d band centre is an indicator of the average energy of d states, and the higher the d states, the more empty the antibonding states, resulting in a stronger adsorption bond, as shown in Fig. 6f.⁶³ The Ti 3d state increases with PdCu loading, indicating electron transfer from PdCu NPs to Ti³⁺, which enhances the Ti–O bond and promotes NO₃⁻ adsorption. This is in line with the findings of the electrochemical experiments and active site study. In addition, PdCu NP loading reduces the selectivity of N₂ during the NITRR. Since CNTs@mesoC@CuPd exhibited 98% selectivity for N₂ during the NITRR,⁶⁴ the N₂ selectivity is mainly affected by the interaction between Ti³⁺ and PdCu NPs (Fig. S31 and 32†). The asymmetrical Ti and N pDOS for different electron spin states suggest the presence of unpaired electrons in *N and *N₂ before PdCu NP loading. For the *N intermediate, the significantly enhanced bonding strength (from 4.50 to 7.17) after loading PdCu NPs suggests enhanced *N adsorption (consistent with E_ad, as shown in Table S3†). The improved Bader charge shows that PdCu loading promotes electron transfer to *N and enhances the local electron density of *N. The enhanced local electron density and facilitated H⁺ formation and adsorption together promote the hydrogenation process, leading to the promotion of *NH formation and inhibition of *N desorption and *N₂ formation. For the *N₂ intermediate, the enhanced adsorption of *N₂ is evidenced in Fig. S30,† due to the negligible bond strength loss of the N–Ti bond and the formation of an extra N–Cu bond. The N≡N bond is weakened (reduced from 21.98 to 17.25), and the positive shifting of pDOS indicates the activation of N₂. As a result, the desorption of *N₂ is inhibited and the formation of *NNH is promoted, consistent with the NRR study.³²

Fig. 6 The differential charge density map and electron transfer between PdCu NPs, TiO_2−x, and the *NO₃ intermediate adsorbed at (a) TiO_2−x OVs and (b) PdCu NPs/TiO_2−x interface, respectively. The partial density of states (PDOS) and partial crystal orbital Hamilton population (COHP) of the *NO₃ intermediate adsorbed at (c) TiO_2−x OVs and (d) PdCu NPs/TiO_2−x interface. (e) The d band centre of Ti 3d of TiO_2−x and PdCu NPs/TiO_2−x. (f) The schematic illustration of the higher d-band centre promoting catalytic kinetics. (With an isosurface value of 0.005 electrons per ų, the green areas represent electron depletion and the yellow areas represent electron accumulation).

The pDOS of other intermediates, including *NO₂, *NO, *N, *NH, *NH₂ and *NH₃ were studied to reveal the effect of PdCu NPs on the hydrogenation process, as shown in Fig. S33 to S38.† For *NO₂ intermediates (Fig. S33†), the peaks of N 2p and O 2p orbits are co-located between −9.3 and −7.7 eV. The N 2p peaks in this region can be regarded as an indicator of the N–O bond since the major contributions of N 2p and O 2p are present. The results suggest that the electron density on Ti³⁺ is increased, and more electrons fill in the N–Ti orbits, enhancing the N–Ti bonding state and weakening the N–O bond due to the introduction of PdCu NPs, consistent with the results for *NO intermediates (Fig. S34†). Therefore, the adsorption and activation of *NO_x intermediates are enhanced with PdCu NP loading, as supported by the results from electrochemical characterization and DFT calculations of E_ad. In the case of the *N intermediate (Fig. S35†), the N–Ti bond is strengthened and *N is stabilized due to the presence of PdCu NPs. Furthermore, negative shifts (1.35 eV) in the N 2p and Ti 3d peaks indicate that more electrons fill the vacant N 2p and Ti 3d orbits due to the electrons provided by PdCu NPs. Similar negative shifts (0.84 eV) can be found in *NH (Fig. S36†), and the shifts of pDOS reduce with the hydrogenation process, as shown in the pDOS of *NH₂ (Fig. S37†), due to the occupation of N 2p orbits during hydrogenation. This finding suggests that the injection of electrons from PdCu NPs promotes the hydrogenation processes. For the *NH₃ intermediate (Fig. S38†), the PDOS of Ti 3d and N 2p for TiO_2−x and PdCu NPs/TiO_2−x are studied to investigate the effect of PdCu NPs on the *NH₃ desorption process. After PdCu NP loading, the peak intensity of Ti 3d and N 2p decreases, indicating that the interaction between Ti³⁺ and *NH₃ weakens and the desorption of *NH₃ is enhanced. In contrast to the theoretical prediction of the NITRR at transition metal surfaces,⁶⁵ PdCu NPs promote nitrate electroreduction to ammonia instead of NO, due to the interaction between PdCu NPs and OVs. The comprehensive study of charge density, Bader charge, pDOS and COHP of intermediates reveals that PdCu NPs enhance nitrate electroreduction to ammonia by promoting hydrogenation and inhibiting the formation of N₂.

4. Conclusions

In conclusion, we demonstrated enhanced NH₃ selectivity and yield in the NITRR via introducing PdCu NPs to modulate the local electrons of OVs on a defective TiO_2−x support. Our in situ characterization and DFT calculation results clearly show that electron injection from PdCu NPs effectively regulates the localized electron density of OVs, boosting the hydrogenation processes and inhibiting N–N coupling, ultimately enhancing the NH₃ selectivity and yield. This work sheds light on the mechanism behind the enhancement of alloy loading on the activity and selectivity of NH₃ synthesis via the NITRR and provides a feasible approach to design alloying catalysts for selective NH₃ synthesis via the NITRR.

Author contributions

Haoran Wu: investigation, formal analysis, writing the original draft. Heng Guo: conceptualization, review, funding acquisition. Ying Zhou: conceptualization, review, supervision, funding acquisition. Fengying Zhang: methodology, review. Peng Yang: methodology. Jiaxin Liu: methodology. Yuantao Yang: methodology. Zhenfeng Huang: resources. Guidong Yang: resources. Xin Tu: resources, review. Chenyuan Zhu: review. Weitao Wang: review.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported by the National Key R&D Project of China (Grant No. 2020YFA0710000), the National Natural Science Foundation of China (22109132), the Provincial Key Research and Development Project of Sichuan (22SYSX0142), the International Science and Technology Cooperation Project of Chengdu (2021-GH02-00052-HZ), the Technology Innovation R&D Project of Chengdu (2022-YF05-00978-SN) and the Scientific Research Starting Project of SWPU (No.2021QHZ014 and 2021QHZ028). X. T. acknowledges the support of the Engineering and Physical Sciences Research Council (EP/X002713/1).

References

1. S. L. Foster, S. I. P. Bakovic, R. D. Duda, S. Maheshwari, R. D. Milton, S. D. Minteer, M. J. Janik, J. N. Renner and L. F. Greenlee, Nat. Catal., 2018, 1, 490–500.
2. C. Tang and S. Z. Qiao, Chem. Soc. Rev., 2019, 48, 3166–3180.
3. Y. Ren, C. Yu, X. Tan, H. Huang, Q. Wei and J. Qiu, Energy Environ. Sci., 2021, 14, 1176–1193.
4. D. R. MacFarlane, J. Choi, B. H. R. Suryanto, R. Jalili, M. Chatti, L. M. Azofra and A. N. Simonov, Adv. Mater., 2020, 32, 1904804.
5. L. Li, C. Tang, B. Xia, H. Jin, Y. Zheng and S. Z. Qiao, ACS Catal., 2019, 9, 2902–2908.
6. P. H. van Langevelde, I. Katsounaros and M. T. M. Koper, Joule, 2021, 5, 290–294.
7. L. Li, C. Tang, X. Cui, Y. Zheng, X. Wang, H. Xu, S. Zhang, T. Shao, K. Davey and S. Qiao, Angew. Chem., Int. Ed., 2021, 60, 14131–14137.
8. X. Zhang, T. Wu, H. Wang, R. Zhao, H. Chen, T. Wang, P. Wei, Y. Luo, Y. Zhang and X. Sun, ACS Catal., 2019, 9, 4609–4615.
9. L. Yang, S. Feng and W. Zhu, J. Phys. Chem. Lett., 2022, 13, 1726–1733.
10. T. Mou, Y. Wang, P. Deák, H. Li, J. Long, X. Fu, B. Zhang, T. Frauenheim and J. Xiao, J. Phys. Chem. Lett., 2022, 13, 9919–9927.
11. P. Yang, H. Guo, H. Wu, F. Zhang, J. Liu, M. Li, Y. Yang, Y. Cao, G. Yang and Y. Zhou, J. Colloid Interface Sci., 2023, 636, 184–193.
12. H. Guo, M. Li, Y. Yang, R. Luo, W. Liu, F. Zhang, C. Tang, G. Yang and Y. Zhou, Small, 2023, 19, 2207743.
13. S. B. Patil and D. Y. Wang, Small, 2020, 16, 2002885.
14. S. B. Patil, H. Chou, Y. Chen, S. Hsieh, C. Chen, C. Chang, S. Li, Y. Lee, Y. Lin, H. Li, Y. J. Chang, Y. Lai and D. Wang, J. Mater. Chem. A, 2021, 9, 1230–1239.
15. S. B. Patil, T. Liu, H. Chou, Y. Huang, C. Chang, Y. Chen, Y. Lin, H. Li, Y. Lee, Y. J. Chang, Y. Lai, C. Wen and D. Wang, J. Phys. Chem. Lett., 2021, 12, 8121–8128.
16. T. Hu, M. Wang, C. Guo and C. M. Li, J. Mater. Chem. A, 2022, 10, 8923–8931.
17. J. M. McEnaney, S. J. Blair, A. C. Nielander, J. A. Schwalbe, D. M. Koshy, M. Cargnello and T. F. Jaramillo, ACS Sustain. Chem. Eng., 2020, 8, 2672–2681.
18. Q. Li, Y. Liu, Z. Wan, H. Cao, S. Zhang, Y. Zhou, X. Ye, X. Liu and D. Zhang, Chin. Chem. Lett., 2022, 33, 3835–3841.
19. F. Ni, Y. Ma, J. Chen, W. Luo and J. Yang, Chin. Chem. Lett., 2021, 32, 2073–2078.
20. H. Yin, Z. Chen, S. Xiong, J. Chen, C. Wang, R. Wang, Y. Kuwahara, J. Luo, H. Yamashita, Y. Peng and J. Li, Chem Catal., 2021, 1, 1088–1103.
21. Z. Zhang, Y. Xu, W. Shi, W. Wang, R. Zhang, X. Bao, B. Zhang, L. Li and F. Cui, Chem. Eng. J., 2016, 290, 201–208.
22. Z. Y. Wu, M. Karamad, X. Yong, Q. Huang, D. A. Cullen, P. Zhu, C. Xia, Q. Xiao, M. Shakouri, F. Y. Chen, J. Y. T. Kim, Y. Xia, K. Heck, Y. Hu, M. S. Wong, Q. Li, I. Gates, S. Siahrostami and H. Wang, Nat. Commun., 2021, 12, 2870.
23. G. F. Chen, Y. Yuan, H. Jiang, S. Y. Ren, L. X. Ding, L. Ma, T. Wu, J. Lu and H. Wang, Nat. Energy, 2020, 5, 605–613.
24. R. Jia, Y. Wang, C. Wang, Y. Ling, Y. Yu and B. Zhang, ACS Catal., 2020, 10, 3533–3540.
25. X. Wan, W. Guo, X. Dong, H. Wu, X. Sun, M. Chu, S. Han, J. Zhai, W. Xia, S. Jia, M. He and B. Han, Green Chem., 2022, 24, 1090–1095.
26. M.-M. Shi, D. Bao, S.-J. Li, B.-R. Wulan, J.-M. Yan and Q. Jiang, Adv. Energy Mater., 2018, 8, 1800124.
27. X. Zhang, Y. Cao, Z. F. Huang, S. Zhang, C. Liu, L. Pan, C. Shi, X. Zhang, Y. Zhou, G. Yang and J. J. Zou, Carbon Energy, 2023, 5, e266.
28. W. Tong, B. Huang, P. Wang, L. Li, Q. Shao and X. Huang, Angew Chem. Int. Ed. Engl., 2020, 59, 2649–2653.
29. F. Pang, Z. Wang, K. Zhang, J. He, W. Zhang, C. Guo and Y. Ding, Nano Energy, 2019, 58, 834–841.
30. Z. Wang, X. Wu, J. Liu, D. Zhang, H. Zhao, X. Zhang, Y. Qin, N. Nie, D. Wang, J. Lai and L. Wang, Nano Lett., 2021, 21, 9580–9586.
31. L. Han, Z. Ren, P. Ou, H. Cheng, N. Rui, L. Lin, X. Liu, L. Zhuo, J. Song, J. Sun, J. Luo and H. L. Xin, Angew. Chem., 2021, 133, 349–354.
32. C. Liu, X. Guo, Z.-F. Huang, J. Li, L. Gan, L. Pan, C. Shi, X. Zhang, G. Yang and J.-J. Zou, Mater. Chem. Front., 2022, 6, 2190–2200.
33. C. Chen, X. Zhu, X. Wen, Y. Zhou, L. Zhou, H. Li, L. Tao, Q. Li, S. Du, T. Liu, D. Yan, C. Xie, Y. Zou, Y. Wang, R. Chen, J. Huo, Y. Li, J. Cheng, H. Su, X. Zhao, W. Cheng, Q. Liu, H. Lin, J. Luo, J. Chen, M. Dong, K. Cheng, C. Li and S. Wang, Nat. Chem., 2020, 12, 717–724.
34. W. Gao, N. Guan, J. Chen, X. Guan, R. Jin, H. Zeng, Z. Liu and F. Zhang, Appl. Catal., B, 2003, 46, 341–351.
35. J. Li, R. Chen, J. Wang, Y. Zhou, G. Yang and F. Dong, Nat. Commun., 2022, 13, 1098.
36. Z. Wei, X. Niu, H. Yin, S. Yu and J. Li, Appl. Catal., A, 2022, 636, 118596.
37. Q. Ren, Y. He, H. Wang, Y. Sun and F. Dong, ACS Catal., 2022, 12, 14015–14025.
38. N. Zhang, X. Li, H. Ye, S. Chen, H. Ju, D. Liu, Y. Lin, W. Ye, C. Wang, Q. Xu, J. Zhu, L. Song, J. Jiang and Y. Xiong, J. Am. Chem. Soc., 2016, 138, 8928–8935.
39. S. Chen, H. Wang, Z. Kang, S. Jin, X. Zhang, X. Zheng, Z. Qi, J. Zhu, B. Pan and Y. Xie, Nat. Commun., 2019, 10, 788.
40. K. Ye, K. Li, Y. Lu, Z. Guo, N. Ni, H. Liu, Y. Huang, H. Ji and P. Wang, TrAC, Trends Anal. Chem., 2019, 116, 102–108.
41. X. Ma, Y. Dai, L. Yu and B. Huang, Sci. Rep., 2014, 4, 3986.
42. K. Jiang, P. Wang, S. Guo, X. Zhang, X. Shen, G. Lu, D. Su and X. Huang, Angew. Chem., Int. Ed., 2016, 55, 9030–9035.
43. Z. Guo, X. Kang, X. Zheng, J. Huang and S. Chen, J. Catal., 2019, 374, 101–109.
44. J. Huang, T. Ding, K. Ma, J. Cai, Z. Sun, Y. Tian, Z. Jiang, J. Zhang, L. Zheng and X. Li, ChemCatChem, 2018, 10, 3862–3871.
45. Z. Jin, C. Liu, K. Qi and X. Cui, Sci. Rep., 2017, 7, 39695.
46. X. Yuan, S. Chen, D. Cheng, L. Li, W. Zhu, D. Zhong, Z. J. Zhao, J. Li, T. Wang and J. Gong, Angew. Chem., Int. Ed., 2021, 60, 15344–15347.
47. J. Sá, J. Bernardi and J. A. Anderson, Catal. Letters, 2007, 114, 91–95.
48. C. Zhang, L. Wang, U. J. Etim, Y. Song, O. M. Gazit and Z. Zhong, J. Catal., 2022, 413, 284–296.
49. J. Sun, H. Yang, W. Gao, T. Cao and G. Zhao, Angew. Chem., Int. Ed., 2022, 61, e202211373.
50. B. Bharti, S. Kumar, H. N. Lee and R. Kumar, Sci. Rep., 2016, 6, 32355.
51. W. P. Utomo, H. Wu and Y. H. Ng, Small, 2022, 18, 2200996.
52. M. Nunes, D. M. Fernandes, M. V. Morales, I. Rodríguez-Ramos, A. Guerrero-Ruiz and C. Freire, Catal. Today, 2020, 357, 279–290.
53. V. Augustyn, P. Simon and B. Dunn, Energy Environ. Sci., 2014, 7, 1597.
54. R. Ge, Y. Wang, Z. Li, M. Xu, S. Xu, H. Zhou, K. Ji, F. Chen, J. Zhou and H. Duan, Angew Chem. Int. Ed. Engl., 2022, 61, e202200211.
55. D. A. Aikens, J. Chem. Educ., 1983, 60, A25.
56. Y. Lu, T. Liu, C. L. Dong, Y. C. Huang, Y. Li, J. Chen, Y. Zou and S. Wang, Adv. Mater., 2021, 33, 2007056.
57. P. Li, Z. Zhou, Q. Wang, M. Guo, S. Chen, J. Low, R. Long, W. Liu, P. Ding, Y. Wu and Y. Xiong, J. Am. Chem. Soc., 2020, 142, 12430–12439.
58. F. Lai, N. Chen, X. Ye, G. He, W. Zong, K. B. Holt, B. Pan, I. P. Parkin, T. Liu and R. Chen, Adv. Funct. Mater., 2020, 30, 1907376.
59. Y. Yao, S. Zhu, H. Wang, H. Li and M. Shao, Angew. Chem., Int. Ed., 2020, 59, 10479–10483.
60. M. Y. Mihaylov, V. R. Zdravkova, E. Z. Ivanova, H. A. Aleksandrov, P. S. Petkov, G. N. Vayssilov and K. I. Hadjiivanov, J. Catal., 2021, 394, 245–258.
61. A. Li, S. Kong, C. Guo, H. Ooka, K. Adachi, D. Hashizume, Q. Jiang, H. Han, J. Xiao and R. Nakamura, Nat. Catal., 2022, 5, 109–118.
62. S. Steinberg and R. Dronskowski, Crystals, 2018, 8, 225.
63. J. K. Nørskov, F. Abild-Pedersen, F. Studt and T. Bligaard, Proc. Natl. Acad. Sci., 2011, 108, 937–943.
64. H. Xu, J. Chen, Z. Zhang, C.-T. Hung, J. Yang and W. Li, Adv. Mater., 2023, 35, 2207522.
65. J.-X. Liu, D. Richards, N. Singh and B. R. Goldsmith, ACS Catal., 2019, 9, 7052–7064.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d3ta04155g

---