Stabilized Bi(III) Sites Direct *NH₂OH Pathway for Efficient Cyclohexanone Oxime Electrosynthesis

ABSTRACT

Coupling electrocatalytic nitrite reduction reaction (NO₂ ⁻RR) with cyclohexanone conversion enables a sustainable route to cyclohexanone oxime (CHO) electrosynthesis, a key feedstock for the nylon‐6 industry. However, this approach is fundamentally constrained by the difficulty of sustaining selective *NH₂OH intermediate formation. Here, we reveal for the first time that the Bi(III) sites can enable the catalytic system to efficiently follow the *NH₂OH pathway, avoiding the undesired *N pathway. Accordingly, a BiPO₄/SiO _x interface was designed in which amorphous SiO _x functions as an electron‐buffer to stabilize Bi(III) active sites. As a result, the BiPO₄/SiO _x catalyst exhibits a high faradaic efficiency (FE_CHO) of 77.0 ± 3.4% and a CHO yield rate of 0.64 ± 0.01 mmol h⁻¹ cm⁻², surpassing all previously reported catalysts in H‐cellMoreover, the BiPO₄/SiO _x catalyst delivers a nearly 100% carbon and nitrogen selectivity to CHO and retains 91.8% of its initial efficiency after extended cycling, substantially outperforming pristine BiPO₄. Combined experimental and theoretical analyses reveal that the stabilized Bi(III) site suppresses the formation of surface K⁺ H₂O, effectively suppressing competing hydrogen evolution and over hydrogenation, thereby enabling efficient CHO electrosynthesis.

The electron‐buffering effect stabilizes Bi³⁺ sites under electroreductive conditions, preserving a favorable interfacial water configuration that sustains the selective *NH₂OH pathway and enables efficient cyclohexanone oxime electrosynthesis.

1. Introduction

Cyclohexanone oxime (CHO) is a critical feedstock for the nylon‐6 production, which has an annual demand of approximately 10 million tons and a market size that amounts to $25 billion [1, 2]. Traditional CHO synthesis relies on a two‐step process, including hydroxylamines (NH₂OH) production via the Raschig process and then the oximation of cyclohexanone (CYC) by NH₂OH [3]. This process requires energy‐intensive feedstocks, expensive noble metal catalysts, and harsh reaction conditions, resulting in massive energy consumption and significant carbon emissions [4]. Renewable energy‐driven electrochemical construction of C─N bond from nitric oxide and carbon sources presents a sustainable route for carbon‐neutral chemical production (Scheme 1a,b) [5, 6, 7]. Mechanistic studies have shown that the *NH₂OH intermediate generated during the electroreduction of nitric oxide spontaneously condenses with CYC to form CHO. However, the complexity of this process arises from several parallel and sequential competing pathways, which is accompanied by the production of numerous by‐products, ultimately resulting in low CHO selectivity [8, 9].

SCHEME 1.

Schematic illustration of CHO production via (a) traditional method and (b) electrosynthesis method. (c) Reaction pathways of NO₂ ⁻RR. I. *NH₂OH pathway for CHO electrosynthesis; II. *N pathway for ammonia (NH₃) electrosynthesis.

As shown in Scheme 1c, the reduction reaction using nitrite (NO₂ ⁻) as the initial reactant can be divided into two parallel reaction pathways, namely, the *NH₂OH pathway and the *N pathway. The *NH₂OH intermediate required for the electrochemical CHO synthesis only exists in the *NH₂OH pathway. Therefore, the selectivity of CHO is closely related to the reaction pathway and the directional utilization efficiency of active hydrogen (*H) species during the reaction. Various metal‐based electrocatalysts have been investigated for the reactions involving *NH₂OH intermediate. The bismuth (Bi)‐based catalyst has attracted considerable interest due to its unique ability to selectively reduce nitric oxide to *NH₂OH intermediate. This high selectivity originates from its inherently low activity toward the hydrogen evolution reaction (HER) and regulation of *H supply, which effectively prevents further reduction of *NH₂OH. Furthermore, as a p‐block metal, Bi exhibits filled p‐orbitals and strong oxophilicity, allowing it to modulate the adsorption strength of key oxygen‐containing intermediates and thereby enhance the catalytic performance [10, 11, 12].

Recent experimental and theoretical studies have shown that the non‐zero valence Bi sites could promote the electrochemical CHO synthesis by strengthening the adsorption of key nitrogenous intermediates and adjusting the supply of *H species, which lower the energy barrier of the potential‐determining step and lead to markedly improved selectivity [13]. This finding indicate an underlying relationship between the valence state of Bi sites and the reaction pathway. Therefore, we conclude that directly modulating the non‐zero valence Bi sites could provide effective control over the reaction pathway. However, the self‐reduction of the high‐valence active sites to zero‐valence metallic phase under the reduction potential is the main problem for Bi‐based materials, resulting in reduced catalytic activity or even complete deactivation of catalysts [14, 15, 16]. Interface engineering might be an effective strategy to maintain the stability and activity of the non‐zero valence Bi sites. Although interface engineering has been widely applied in various catalytic systems to modulate the electronic structure and stabilize the active phase [17, 18, 19, 20], few studies have focused on maintaining the stability of metal sites during the construction of electrochemical C─N bonds through this strategy [21, 22].

In this work, we designed interface engineering strategy to stabilize Bi³⁺ (Bi(III)) sites for efficient CHO electrosynthesis via *NH₂OH pathway. To implement this concept, a catalyst with a tailored interaction between the BiPO₄ and amorphous SiO _x (BiPO₄/SiO _x ) was designed, which utilizes the SiO _x as an electron‐buffer to prevent the reduction of Bi(III) sites under negative potential. This electron‑buffering effect means that SiO _x can accept excess electrons from BiPO₄, thereby stabilizing the Bi(III) oxidation state. Systematic experimental characterizations and theoretical analyses reveal that SiO _x exhibits a remarkable electron‐buffering effect, which can transfer electron from BiPO₄ to SiO _x during the electroreduction process, thus preventing the reduction of Bi(III) sites. After stabilization, the as‐prepared BiPO₄/SiO _x catalyst exhibits an outstanding CHO electrosynthesis performance, including a high faradaic efficiency (FE_CHO) of 77.0 ± 3.4%, a CHO yield rate of 0.64 ± 0.01 mmol h⁻¹ cm⁻² and a nearly 100% carbon and nitrogen selectivity to CHO which is highly advantageous in terms of atomic economy. Moreover, BiPO₄/SiO _x catalyst retains 91.8% of its initial efficiency after 14 h of operation at −1.3 V versus RHE, substantially outperforming pristine BiPO₄. In addition, systematic in‐situ and ex‐situ characterizations elucidate that the Bi(III) site suppresses the formation of surface K⁺ H₂O, the main source of H₂, effectively suppressing competing HER and over hydrogenation, thereby enabling efficient CHO electrosynthesis.

2. Results and Discussion

2.1. Material Synthesis and Characterization

A composite of the BiOCl nanosheets and amorphous SiO _x spheres (marked as BiOCl/SiO _x ) was designed as the precursor. The BiOCl/SiO _x precursor was synthesized using a solvothermal method analogous to that reported by Tan et al. [22]. Scanning electron microscopy (SEM) images exhibit a composite structure consisting of SiO _x nanospheres and BiOCl nanosheets (Figure S1). The X‐ray diffraction (XRD) pattern of BiOCl/SiO _x precursor is only indexed to tetragonal BiOCl (PDF No. 82–0485), confirming amorphous and sub‐stoichiometric characteristic of SiO _x (x < 2) (Figure S2), as previously reported for Stöber method‐derived silica. After electrochemical reconstruction in phosphate buffered solution (PBS) (see the details in the Supporting Information) by a linear sweep voltammetry (LSV) from −0.2 to −2.4 V versus Hg/Hg₂SO₄ until the current remains stable (Figure S3), the precursor has transformed into BiPO₄/SiO _x and is subjected for further characterization.

As shown in Figure S2, the XRD pattern of BiPO₄/SiO _x catalyst displays three distinct peaks at 20.07°, 29.50°, and 31.32°, assigned to the (101), (200), and (102) planes of the hexagonal BiPO₄ (PDF No. 15–0766). SEM and transmission electron microscopy (TEM) images show that the SiO _x still maintains the spherical structure and the BiPO₄ has completely combined with SiO _x (Figure S4a,b). This observation is further corroborated by elemental line scans at the selected location (Figure S4c). At higher magnification, the high‐resolution TEM (HRTEM) images manifest that the BiPO₄/SiO _x displays lattice fringes with the spacing of 0.34, 0.30, and 0.46 nm, corresponding to the (110), (200), (101) and plane of BiPO₄ (Figure 1a and Figure S5).

FIGURE 1.

(a) TEM image and (b) EDX elemental mappings for Bi, Si, P, and O. (c) High‐resolution XPS spectra of Bi 4f for BiPO₄/SiO _x and BiPO₄. (d) Bi L₃‐edge XANES spectra of BiPO₄/SiO _x , BiPO₄, Bi foil, and Bi₂O₃. (e) FT k²‐weighted χ(k)‐function of the EXAFS spectra at Bi L₃‐edge. (f) WT‐EXAFS of k²‐weighted k‐space spectra of Bi for Bi foil, Bi₂O₃, BiPO₄, and BiPO₄/SiO _x . (g) Charge density difference diagram of BiPO₄/SiO _x . (h) Schematic of the structure‐performance relationship in the BiPO₄/SiO _x .

The electrochemical reconstruction process induces the formation of the heterointerface between BiPO₄ and amorphous SiO _x , as clearly visualized in the HRTEM. An illustrative BiPO₄/SiO _x interface was directly observed, comprising a (110) plane of BiPO₄ with a lattice spacing of 0.34 nm in contact with an amorphous SiO _x region (Figure 1a). In Figure 1b, the elemental mapping images reveal that the BiPO₄ nanoparticles are tightly integrated with the SiO _x nanospheres, with all constituent elements homogeneously distributed throughout the composite. The Bi content of BiPO₄/SiO _x and BiPO₄ is determined to be 39.23 and 70.39 wt% by EDX (Figure 1b; Figure S6 and Table S1).

X‐ray photoelectron spectroscopy (XPS) and X‐ray absorption spectroscopy (XAS) were performed to analyze the electronic structure of BiPO₄/SiO _x and BiPO₄ catalysts. For BiPO₄/SiO _x catalyst, the high‐resolution XPS spectrum of Bi 4f exhibits typical Bi³⁺ features with 4f_7/2 and 4f_5/2 peaks centered at 159.91 and 165.20 eV, respectively, corresponding to BiPO₄ species (Figure 1c). By contrast, the binding energy of Bi³⁺ 4f_7/2 for BiPO₄ is ∼0.10 eV lower than that in BiPO₄/SiO _x . This suggests both a higher oxidation state of Bi in the composite and electron transfer from BiPO₄ to SiO _x , confirming a strong electronic interaction. Meanwhile, the P 2p XPS spectra obtained for BiPO₄/SiO _x and BiPO₄ (Figure S7a) can be fitted into two peaks which were attributed to the P 2p_3/2 and P 2p_1/2, respectively. The O 1s XPS spectra of BiPO₄/SiO _x can be deconvoluted into four characteristic peaks at 531.04, 532.67, 533.65, and 535.73 eV, corresponding to the PO₄ ³⁻, Si‐O, hydroxyl‐O, and absorbed‐O, respectively (Figure S7b). For BiPO₄, the O 1s XPS spectra were deconvoluted into three characteristic peaks at 531.12, 532.72, and 535.74 eV, corresponding to the PO₄ ³⁻, hydroxyl‐O, and absorbed‐O. The Si 2p signal of BiPO₄/SiO _x appears at 103.52 eV (Figure S7c).

X‐ray absorption near‐edge structure (XANES) and extended X‐ray absorption fine structure (EXAFS) were analyzed to verify the electronic and coordination structures in detail. As shown in XANES spectra of Bi L₃‐edge (Figure 1d), BiPO₄/SiO _x exhibits a higher oxidation state than the Bi₂O₃ reference and BiPO₄ sample, which was in accordance with the XPS results. Therefore, the above evidence reveals that the tailored SiO _x electron‐buffering zone induces the electron transfer and reordering of local electron distribution, potentially making Bi sites electron‐deficient by modulating the electron supply during the reaction to avoid the reduction of Bi³⁺ sites.

The local coordination structure of Bi was explored by the EXAFS (Figure 1e) and wavelet transform (WT) spectra (Figure 1f). Bi foil and Bi₂O₃ exhibit the Bi─Bi and Bi─O bond coordination at around 2.95 and 1.63 Å, respectively. The EXAFS of BiPO₄/SiO _x and BiPO₄ displayed a main peak at 1.53 Å, corresponding to the Bi─O bond. Moreover, no peaks attributed to the Bi─Bi bond were observed around 2.95 Å, indicating that BiOCl only transformed into BiPO₄ without forming metallic Bi after the electrochemical reconstruction. These coordination structure results can also be clearly observed from the corresponding WT spectra (Figure 1f).

To better understand the electron transfer between BiPO₄ and SiO _x , we conducted a density functional theory (DFT) calculation for BiPO₄/SiO _x catalyst (Figure 1g and Figure S8). Bader charge analysis of BiPO₄/SiO _x reveals that the introduction of SiO _x causes an electron transfer of 1.6 e⁻ from BiPO₄ to SiO _x , thereby resulting in a higher oxidation state of Bi in BiPO₄/SiO _x . Based on the fine spectral characterizations, a unique electron‐buffering effect of SiO _x in BiPO₄/SiO _x is demonstrated, which remains stable Bi³⁺ sites. As illustrated in Figure 1h, the electron‐buffering effect provided by SiO _x in the system prevented the self‐reduction of the BiPO₄ active phase under the reduction potential [23, 24].

2.2. Pathway Selection for Electrochemical CHO Synthesis

Building on our hypothesis that the valence state of Bi sites could control CHO selectivity, we sought to establish the advantage of the Bi(III) sites in the pathway selection for electrochemical CHO synthesis by in‐situ differential electrochemical mass spectrometry (DEMS). The mass‐to‐charge (m/z) ratios of 33, 14, and 15 for the *NH₂OH, *N, and *NH intermediates in the electrochemical CHO synthesis were used as characteristic descriptions for the two reaction pathways. The DEMS was used to detect the intermediate generated on BiPO₄/SiO _x , BiPO₄, and Bi NPs catalysts during the nitrite reduction reaction (NO₂ ⁻RR). As shown in Figure 2a–c, all three catalysts exhibited strong signals for the *NH₂OH intermediate, proving that all three catalysts can conduct the reaction via the *NH₂OH pathway during the reaction. For the BiPO₄/SiO _x catalyst, almost no mass spectral signals corresponding to the *N and *NH intermediates of the *N pathway were detected. In contrast, the Bi NPs catalyst resulted in a strong mass spectral signal, corresponding to *N intermediate. While the mass spectral signal obtained using the BiPO₄ catalyst could observe the presence of both *N and *NH intermediates signals (Figure 2b,c).

FIGURE 2.

Mass signals of (a) *NH₂OH, (b) *N, and (c) *NH determined from in‐situ DEMS with the corresponding currents measured for BiPO₄/SiO _x , BiPO₄, and Bi nanoparticles (Bi NPs) catalyst. d) Schematic illustration of reaction pathway for electrochemical CHO synthesis on BiPO₄/SiO _x , BiPO₄, and Bi NPs catalyst surface.

Based on the above DEMS results, it can be concluded that the Bi NPs catalyst simultaneously undergoes the *N and *NH₂OH pathways in NO₂ ⁻RR, while the BiPO₄/SiO _x catalyst has high selectivity for the *NH₂OH pathway. However, the evidence for two pathways was also present in the BiPO₄ catalytic system. This may be due to the uncontrolled self‐reduction of the BiPO₄ catalyst during the reaction, causing part of the catalyst to transform into metallic Bi, and thus the catalytic system tends toward the two pathways (Figure 2d). Therefore, maintaining the pre‐designed structure of the BiPO₄ active site at the reduction potential is particularly important for improving the selectivity of the electrochemical CHO synthesis.

2.3. Electrocatalytic Performance for CHO Electrosynthesis

The electrochemical CHO synthesis activity and selectivity of the catalysts were investigated in a three‐electrode H‐cell system. Figure S9 provides that LSV curves of BiPO₄/SiO _x and BiPO₄ recorded in 0.5 M PBS + 0.1 M KNO₂ with and without 1.0 mmol CYC. To evaluate the intrinsic activity of the catalysts, we determined the electrochemical active surface area (ECSA) from the double layer capacitance (C _dl). The C _dl of BiPO₄/SiO _x (0.181 mF cm⁻²) is much lower than that of BiPO₄ (0.535 mF cm⁻²). This decrease is attributed to the poor conductivity of amorphous SiO _x (Figures S10 and S11). After normalizing the LSV curves by the ECSA, Figure 3a shows the higher ECSA‐normalized current density on the LSV curves for BiPO₄/SiO _x than for BiPO₄, indicating the preferred intrinsic activity of the BiPO₄/SiO _x catalyst for electrochemical CHO synthesis.

FIGURE 3.

(a) ECSA‐normalized LSV curves of BiPO₄/SiO_x and BiPO₄ recorded in 0.5 M PBS + 0.1 M KNO₂ with and without 1.0 mmol CYC. (b) CHO yield rates and FEs of BiPO₄/SiO _x for CHO electrosynthesis at given potentials. (c) Comparison of the electrochemical CHO synthesis performance under the same conditions (Nitrogen source: NO₂ ⁻, Reactor: H‐cell) of BiPO₄/SiO _x with the recently reported catalysts. (d) Comparison of mass activity of BiPO₄/SiO _x and BiPO₄ based on Bi content at given potentials. (e) Comparison of CHO FE of BiPO₄/SiO _x , BiPO₄, and SiO _x at given potentials. (f) Potential‐dependent NO₂ ⁻ to CHO (upper panel) and CYC to CHO (lower panel) selectivity of BiPO₄/SiO _x catalyst in CHO electrosynthesis. (g) Cycling stability of BiPO₄/SiO _x catalyst at −1.3 V versus RHE for CHO electrosynthesis. (h) Plant‐gate levelized cost per kilogram of CHO from TEA under potentiostatic electrolysis.

Chronoamperometry was used to systematically study the distribution of products from the above‐mentioned cathode reaction (Figure S12). The products were analyzed according to the reported methods: NH₃ was determined using the indophenol blue method (Figure S13), while CHO was determined and quantified by gas chromatography (GC), respectively (Figure S14). As shown in Figure 3b and Table S2, the main product of the cathode reaction over BiPO₄/SiO _x was CHO, with a maximum FE of 77.0 ± 3.4% at −1.3 V versus reversible hydrogen electrode (vs. RHE) and a CHO yield rate of 0.64 ± 0.01 mmol h⁻¹ cm⁻² at −1.3 V versus RHE, which is at the forefront of the previously reported catalysts (Figure 3c and Table S3) [25, 26, 27, 28, 29, 30, 31, 32, 33]. Control experiments confirmed that no detectable CHO crossed the proton exchange membrane into the anolyte (Figure S15), thus validating the accuracy of product quantification based on catholyte analysis.

For a meaningful comparison, the BiPO₄/SiO _x and BiPO₄ catalysts were also evaluated the mass activity of CHO electrosynthesis which was normalized by the mass of Bi (Figure 3d). Apparently, the mass activity and FE of CHO over BiPO₄/SiO _x exceed those on BiPO₄ under all tested potentials, indicating that the introduction of SiO _x can significantly enhance the CHO electrosynthesis performance of BiPO₄ (Figure 3d,e, Figure S16, Tables S4 and S5). Specifically, the BiPO₄/SiO _x exhibits the highest mass yield rate of 2.96 ± 0.07 mmol h⁻¹ mg_Bi ⁻¹ for CHO (Figure 3d). It is worth noting that as mentioned above BiPO₄ has a larger ECSA, which indicates more total active sites and leads to a higher geometric current density. However, its mass activity and FE are lower than those of BiPO₄/SiO _x . This confirms that BiPO₄/SiO _x sacrifices some active site density but gains higher site quality. The Figure S17 show that SiO _x exhibits much lower CHO yield rate and FE than BiPO₄/SiO _x at all tested potentials, confirming that the direct catalytic contribution of SiO _x is limited. Therefore, the enhanced performance might be mainly attributed to the interfacial electron‐buffering effect of SiO _x for stabilizing Bi(III) sites. Moreover, BiPO₄/SiO _x catalyst achieved an N‐selectivity of over 90% at all tested potentials and the highest C‐selectivity of 98.6% at −1.3 V versus RHE, both targeted CHO as the product (Figure 3f).

For the electrochemical CHO synthesis reaction, NH₃ and H₂ are main by‐products attributed to the over‐hydrogenation of the intermediates and competitive HER, respectively. The BiPO₄/SiO _x catalyst demonstrates remarkable suppression of both undesirable by‐products. For NH₃, it maintains a very low NH₃ FE of below 15% across the tested potentials (from −1.0 to −1.6 V vs. RHE). In contrast, BiPO₄ catalyst exhibits significantly higher NH₃ FE, confirming that BiPO₄/SiO _x effectively suppresses the over‐hydrogenation pathway to NH₃ (Figure S18a,b and Table S6). Equally notable is its potent suppression of the competitive HER, which leads to an extremely low H₂ FE of 3.7% and 5.7% at −1.3 and −1.5 V versus RHE respectively, overcoming the critical problem of competitive HER (Figure S18c). As a contrast, BiPO₄ shows a higher H₂ FE of 12.6% and 26.5% at −1.3 and −1.6 V versus RHE. Correspondingly, BiPO₄/SiO _x exhibited a lower partial current density of HER at each reaction potential (Figure S19).

Favorable stability is equally important for electrochemical CHO synthesis. The cycling stability of the catalysts was evaluated by chronoamperometry at −1.3 V versus RHE. Benefiting from the electron‐buffering effect of SiO _x in BiPO₄/SiO _x catalyst, the BiPO₄/SiO _x catalyst retained 91.8% of its initial FE after the 14 times cycling stability test (Figure 3g). In contrast, the pure BiPO₄ catalyst, which is absent of the protection of the electron‐buffering effect, exhibited markedly reduced stability, only retaining 74.9% of its initial FE after the 6 times cycling stability test. The results highlight the versatility of SiO _x incorporation in enhancing performance and stability of electrochemical CHO synthesis. Thus, the superior performance of BiPO₄/SiO _x stems not only from its higher site quality but also from the enhanced stability provided by the electron‑buffering effect of SiO _x . Therefore, BiPO₄/SiO _x shows a clear advantage for CHO electrosynthesis.

Based on the above performance advantages, we further evaluated the economic feasibility of this electrosynthesis strategy. A preliminary techno‐economic analysis (TEA) on plant‐gate levelized cost per kilogram of CHO was performed (Note S1) [34, 35]. The electrosynthesis strategy is much more profitable at the optimum condition (the price of electricity generated from renewable sources is 3 cents kWh⁻¹) (Figure 3h), showing the industrial application potential of this electrosynthesis strategy. In this work, BiPO₄/SiO _x catalyst with a FE_CHO of 77.0% under operational conditions further supports the scalability and practical viability of the process, bridging the gap between laboratory‐scale innovation and potential industrial deployment.

2.4. Stabilized Structure during CHO Electrosynthesis

To investigate the structural evolutions of catalysts during the CHO electrosynthesis process, potential‐dependent in‐situ electrochemical Raman spectroscopy from −1.0 to −1.6 V versus RHE was conducted (Figure 4a,b). Both BiPO₄/SiO _x and BiPO₄ present three typical Raman peaks, which are assigned to Bi─O stretching vibration (100 and 170 cm⁻¹), the ν ₂ bending vibration of the PO₄ unit (464 cm⁻¹) at open circuit potential (OCP) [36, 37]. The Bi─O stretching vibration and ν ₂ band located at 170 and 464 cm⁻¹ corresponding to BiPO₄ in BiPO₄/SiO _x are maintained over increasing potential from −1.0 to −1.6 V versus RHE (Figure 4a), indicating that the Bi³⁺ sites remain structural‐stable and reconstruction‐resistant over a broad potential window during the CHO electrosynthesis process, which is attributed to the electron‐buffering effect of SiO _x .

FIGURE 4.

Time‐dependent in‐situ Raman spectra of (a) BiPO₄/SiO _x and (b) BiPO₄ from −1.0 to −1.6 V versus RHE. (c) XRD patterns of BiPO₄/SiO _x and BiPO₄ after cycling stability test. High‐resolution Bi 4f XPS spectra of (d) BiPO₄/SiO _x and (e) BiPO₄ after cycling stability test. (f) PO₄ ³⁻ content in catalysts after cycling stability test. TEM images of (g) BiPO₄/SiO _x and (h) BiPO₄ after cycling stability test.

However, the intensity of the peak located at 170 cm⁻¹ significantly vanished in pristine BiPO₄ over the same potential range (Figure 4b), implying that the pristine BiPO₄ is unable to resist the electrochemical reconstruction to generate metallic Bi under the cathodic potential. Notably, the Bi─O stretching vibration located at 100 cm⁻¹ in BiPO₄ exhibited a red shift of approximately 5 cm⁻¹ as the reduction potential becomes more negative compared to the spectrum of BiPO₄ at OCP. This is attributed to the reduction of the Bi³⁺ sites to metallic Bi under the reduction potential with electrons filling the anti‐bonding orbitals of Bi─O, resulting in an increase in the Bi─O bond length, thereby causing a decrease in the vibration frequency [12]. In contrast, the Bi─O stretching vibration in BiPO₄/SiO _x exhibits no shift under the reduction potential ranging from −1.0 to −1.6 V versus RHE. This absence of peak shift indicates that the structure of the Bi─O bond remains stable under the reduction potential.

After the cycling stability test, the XRD patterns in Figure 4c show that the crystalline structure of BiPO₄ is retained in BiPO₄/SiO _x catalyst during the CHO electrosynthesis process. In contrast, the pristine BiPO₄ nanoparticles inevitably suffer from spontaneous undesirable self‐reduction under cathodic potential. As shown in Figure 4c, the XRD patterns of BiPO₄ show a new diffraction peak around 27.0° after the cycling stability test, corresponding to the (012) plane of the metallic Bi. Without the protection of electron‐buffering effect, BiPO₄ readily undergoes electrochemical reduction to metallic Bi. This self‐reduction phenomenon could explain the decline in stability of pristine BiPO₄. The above results directly reflect the enhanced structural stability of BiPO₄/SiO _x owing to the electron‐buffering effect of SiO_x which mitigates electron transfer toward Bi sites, thereby preventing the reduction of the Bi(III) sites and preserving the original Bi─O bond strength.

To further probe the evolution of electronic structure of catalysts during CHO electrosynthesis process, XPS measurement was performed. After the cycling stability test, the high‐resolution Bi 4f XPS spectrum of BiPO₄/SiO _x exhibits two typical 4f_7/2 and 4f_5/2 peaks located at 159.95 and 165.27 eV respectively (Figure 4d), which is consistent with the binding energy of pristine BiPO₄/SiO _x . Compared with BiPO₄/SiO _x , the high‐resolution Bi 4f XPS spectra of BiPO₄ are fitted into four peaks at 159.84/165.11 and 159.37/164.62 eV attributed to Bi³⁺ and Bi⁰ (Figure 4e), indicating that partial BiPO₄ has transformed into metallic Bi during the reaction. Furthermore, in order to further demonstrate the reconstruction degree of the BiPO₄ component, we used O 1s XPS spectra to quantify the content of PO₄ ³⁻ before and after the reaction (Figure 4f and Figure S20a,b). The content of PO₄ ³⁻ in BiPO₄/SiO _x remains constant. In contrast, the content of PO₄ ³⁻ in BiPO₄ decreases from 58% to 40.5%, which also proves that some BiPO₄ transforms into the metallic Bi during the reaction. As shown in Figure S20, the high‐resolution XPS spectra of Si 2p and P 2p in BiPO₄/SiO _x show no obvious change during the cycling stability test.

The structural evolution of catalysts was explored by the SEM and TEM. BiPO₄/SiO _x still maintains the original morphology with uniform size after the stability test (Figure S21). The TEM images of post‐test BiPO₄/SiO _x (Figure 4g and Figure S22) show that BiPO₄ and SiO _x still retain in a closely integrated composite structure, and display lattice fringes with the spacing of 0.34 and 0.30 nm corresponding to the (110) and (200) planes of BiPO₄. By contrast, post‐test BiPO₄ displays lattice fringes with the spacings of 0.32 nm corresponding to the (012) of metallic Bi (Figure 4h), which is consistent with XRD, in‐situ Raman, and XPS results. Combining the above characterization, it can be concluded that the BiPO₄/SiO _x interface exhibited high electrochemical stability under cathodic reduction potentials, which enabled the BiPO₄ species to maintain excellent catalytic performance over the stability test.

2.5. Activity Origin

To further reveal the influence of phase stabilization on the reaction mechanism, in‐situ Fourier transform infrared spectroscopy (FTIR) and DEMS measurements were performed to identify the intermediates of CHO electrosynthesis and the configuration of the interfacial H₂O. As shown in Figure 5a and Figure S23, potential‐dependent in‐situ FTIR of BiPO₄/SiO _x and BiPO₄ catalysts show the vibration bands at approximately 1268 and 1166 cm⁻¹ corresponding to the characteristic vibration adsorption of *NH₂ and *NH₂OH, respectively. With the increasing potential, the intensity of *NH₂OH vibration gradually increases, indicating that *NH₂OH accumulates over the catalysts.

FIGURE 5.

(a) Potential‐dependent in‐situ FTIR spectra of NO₂ ⁻RR conducted on BiPO₄/SiO _x . Potential‐dependent in‐situ FTIR spectra of the O─H stretching mode of interfacial water on (b) BiPO₄/SiO _x and (c) BiPO₄ catalyst surface. Comparison of the (d) 2H‐H₂O and (e) K⁺‐H₂O content on BiPO₄/SiO _x , BiPO₄, and Bi NPs catalyst surface. (f) Tafel slopes of BiPO₄/SiO _x and BiPO₄ catalyst in 0.5 M PBS. (g) Mass signals of H₂ determined from in‐situ DEMS with the corresponding currents measured for BiPO₄/SiO _x and BiPO₄ catalyst. (h) Schematic illustration of the interfacial water effect on BiPO₄/SiO _x and BiPO₄ catalyst surface.

Considering that the cation in the water‐cation interaction with the transition state in H₂O dissociates to form *H for H₂ evolution and hydrogenation reactions, the microscopic effect of catalysts on interfacial water dissociation should be further explored [38]. The configuration of the interfacial H₂O was studied by fitting the peak of the O─H bond stretching band (3000–3800 cm⁻¹) in the in‐situ FTIR [39]. The O─H peak could be deconvoluted into three distinct components, including 4‐coordinated hydrogen‐bonded water (4H‐H₂O, yellow shadow), 2‐coordinated hydrogen‐bonded water (2H‐H₂O, blue shadow), and K⁺ ion hydrated H₂O (K⁺‐H₂O, red shadow) (Figure 5b,c; Figures S24 and S25). K⁺‐H₂O in the solution is regarded as the main source of H₂, which leads to excessive *H accumulation and further over‐hydrogenation of the *NH₂OH intermediate to NH₃ during the reaction. Figure 5d,e shows the relative proportion of 2H‐H₂O and K⁺‐H₂O at all applied potentials during the reaction. The BiPO₄/SiO _x catalyst showed an increasing proportion of 2H‐H₂O and a decreasing proportion of K⁺‐H₂O compared with the BiPO₄ catalyst. The above results indicate that BiPO₄/SiO _x catalyst can reduce the proportion of K⁺‐H₂O, thereby controlling the amount of *H and preventing the formation of H₂ and NH₃ by‐products [40, 41]. Conversely, the BiPO₄ catalyst with a higher K⁺‐H₂O content exhibits stronger HER and over‐hydrogenation during the reaction due to the accumulation of *H, which is consistent with experimental results.

In order to explore the correlation between the self‐reduction of BiPO₄ and the interfacial H₂O configuration, the interfacial H₂O configuration of the metallic Bi (the final self‐reduction phase) was analyzed. The Bi NPs catalyst exhibited the highest proportion of K⁺‐H₂O and the lowest proportion of 2H‐H₂O. This result demonstrates that water is more favored to dissociate on the surface of metallic Bi and generate *H during the reaction, thereby leading to the formation of by‐products such as H₂ and NH₃. It can thus be inferred that the uncontrolled self‐reduction of BiPO₄ into metallic Bi induces a transformation of the interfacial water configuration, thus causing a decline in the selectivity toward CHO. In contrast, owing to the presence of electron‐buffering SiO _x zone, BiPO₄/SiO _x catalyst maintains the stable BiPO₄ phase, thereby avoiding the transformation of the interfacial water configuration and efficiently synthesizing CHO.

Moreover, BiPO₄/SiO _x has a more negative HER overpotential and a lower current density compared with BiPO₄ (Figure S26). The corresponding Tafel slope of the BiPO₄/SiO _x (328.4 mV dec⁻¹) is also greater than that of the BiPO₄ (241.0 mV dec⁻¹) and Bi NPs reference (256.6 mV dec⁻¹), indicating slow kinetics for HER (Figure 5f) [38]. The H₂ evolution during LSV from −1.0 to −1.6 V versus RHE was monitored by DEMS. As shown in Figure 5g, the BiPO₄ exhibited a lower onset potential, and the H₂ signal could be detected as early as 350 s, indicating more pronounced HER activity compared to BiPO₄/SiO _x [42]. Together, these features of BiPO₄/SiO _x create an efficient interfacial environment that suppresses over‐hydrogenation and minimizes HER (Figure 5h).

3. Conclusion

This work addresses the stability and selectivity challenges in electrochemical CHO synthesis by designing a BiPO₄/SiO _x catalyst with tailored interfacial interaction. The tailored SiO _x electron‐buffering zone induces the electron transfer from BiPO₄ to SiO _x , which modulates the electron supply to avoid the reduction of Bi(III) sites. The stabilized Bi³⁺ sites not only ensure the selective formation of the *NH₂OH intermediate but also maintain an interfacial water configuration that optimizes *H supply to suppress competing reactions. The as‐prepared BiPO₄/SiO _x catalyst exhibits an outstanding CHO electrosynthesis performance, achieving a high FE_CHO of 77.0 ± 3.4%, a CHO yield rate of 0.64 ± 0.01 mmol h⁻¹ cm⁻², and a nearly 100% carbon and nitrogen selectivity in CHO. Moreover, BiPO₄/SiO _x catalyst retains 91.8% of its initial efficiency after 14 h of operation at −1.3 V versus RHE, substantially outperforming pristine BiPO₄. The catalyst design strategy of introducing the electron‐buffering in this work can be extended to other structure sensitive catalytic systems, which provides insight into designing the catalysts with superior stabilities.

4. Experimental Section/Methods

4.1. Catalyst Preparation

4.1.1. Preparation of SiO_x Spheres

The SiO _x spheres were synthesized via the modified Stöber method [43]. First, 50 mL of deionized water, 100 mL of ethanol, and 15 mL of ammonia solution were added sequentially to a beaker and stirred until homogeneous. Then, 30 mL of tetraethyl orthosilicate was slowly added dropwise while stirring, and the reaction was continued for 5 h. Finally, the SiO _x spheres were collected by vacuum filtration, washed with distilled water and ethanol, and dried at 60°C.

4.1.2. Preparation of BiOCl/SiO_x

BiOCl/SiO _x was obtained by a hydrothermal method. The as‐prepared 100 mg of SiO _x spheres were dispersed in a mixture solution of 10 mL ethylene glycol and 20 mL ethanol and ultrasonic treated for 30 min. Then, 0.3 mmol of Bi(NO₃)₃ 5H₂O and 10 mmol of NH₄Cl were added into the above solution with magnetic stirring for 10 min. After that, 1.0 mL of ammonia solution was dropped into the above solution and stirred for 20 min. The mixture was transferred into a 100 mL Teflon‐lined stainless steel autoclave and kept at 140°C for 20 h. After cooling to room temperature, the BiOCl/SiO _x was taken out by vacuum filtration, washed by ethanol, and dried at 60°C. BiOCl was prepared following the same procedures without SiO _x .

4.1.3. Preparation of BiPO₄/SiO_x

The BiOCl/SiO _x powder (10 mg) was mixed with 4 mL of isopropyl alcohol and 20  µL of 5 wt% Nafion solution, followed by ultrasonication for 30 min to form an ink. The above ink was sprayed onto a carbon paper (effective geometric area: 2 cm²). Then, the carbon paper coated the BiOCl/SiO _x served as the working electrode and conducted linear sweep voltammogram (LSV) curves (from 0.5 to −1.5 V vs. RHE) for 10 times in 0.5 M PBS (3.33 g of K₂HPO₄, 65.45 g of KH₂PO₄, and 8.51 g of KNO₂ are dissolved and made up to a total volume of 1.0 L) to obtaining the BiPO₄/SiO _x catalyst for subsequent tests. BiPO₄ was prepared following the same procedures by BiOCl.

4.2. Electrocatalytic Electrochemical CHO Synthesis Measurements

Around 65.45 g of KH₂PO₄ and 3.33 g of K₂HPO₄ were dissolved in water and then made up to 1.0 L to obtain a 0.5 M PBS. All electrochemical CHO synthesis measurements were conducted at room temperature using an electrochemical workstation (CHI 760E) in an H‐type cell separated by a Nafion 211 membrane. A Pt foil and the Hg/Hg₂SO₄ electrode serve as the counter and reference electrodes, respectively. The applied potentials were converted to the RHE scale using the following formula: E (vs. RHE) = E (vs. Hg/Hg₂SO₄) + 0.656 V + 0.0591 V × pH. Potentiostatic tests were performed at different potentials ranging from −1.0 to −1.6 V versus RHE in 40 mL of a freshly prepared 0.5 M PBS + 0.1 M KNO₂ + 1.0 mmol CHO solution for 1 h to evaluate the electrochemical CHO synthesis activity. LSV measurements were performed at a scan rate of 10 mV s⁻¹ from 0.5 to −1.5 V versus RHE.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File: adma73997‐sup‐0001‐SuppMat.docx.

Data Availability Statement

The data that supports the findings of this study are available in the supplementary material of this article.