Skip to main content
NCBI home page
Chemical Science logoLink to Chemical Science
. 2026 Sep 18. Online ahead of print. doi: 10.1039/d6sc03539f

Atomically dispersed and electron-enriched Cu–O4 active sites on BiOCl atomic layers for high selectivity methanol photosynthesis

Ting Peng a,b, Yiqing Wang b, Ke Wang c,d, Zhaoke Zheng e, Zhenhui Liu e, Haotian Zhou b, Daming Zhao f, Yiduo Wang b, Qingqing Guan b, Binglan Wu b, Guofu Wang b, Bo Kou g, Shaohua Shen a,b,✉
PMCID: PMC13617759  PMID: 42807867

Abstract

High selectivity photosynthesis of CH3OH via photocatalytic CO2 reduction has been severely challenged by retarded charge carrier dynamics and sluggish multi-electron reaction kinetics. Herein, single Cu atoms (CuSA) were introduced on BiOCl atomic layers (BOCAL) by a facile two-step method of electrostatic self-assembly and pyrolysis for efficient photocatalytic CO2 reduction, with CH3OH selectively produced at an evolution rate of 103.98 µmol g−1 h−1 and a selectivity of 99.14%. This excellent photocatalytic activity should benefit from the atomically dispersed Cu–O4 coordination structure on BOCAL, which could accelerate the separation and migration of photogenerated charge carriers by creating the impurity energy level (IEL) in the forbidden band of BiOCl with a lowered work function. In situ spectral investigations and density functional theory (DFT) calculations unveil that the Cu–O4 coordination structures with electron density increased by photogenerated oxygen vacancies (VO) could act as the active sites to deliver electrons to CO2 molecules for efficient activation, as well as promote CH3OH generation by inhibiting *CO intermediate desorption and reducing the energy barrier of the rate-determining *COH + H → *CHOH step. This work provides a facile approach to the design of electronic structure-tuned single-atom active sites on two-dimensional (2D) nanostructures towards high selectivity CH3OH photosynthesis, and also deepens the understanding of the single-atom and vacancy interaction between active sites and semiconducting supports for efficient CO2 photoreduction.

The electron-rich Cu-O4 sites on BiOCl atomic layers create impurity levels in the bandgap that boost charge separation and migration, inhibit *CO desorption, and lower the rate-determining barrier, enabling highly selective photocatalytic CO2-to-methanol conversion.graphic file with name d6sc03539f-ga.webp

1. Introduction

Photocatalysis, simulating botanic photosynthesis for converting CO2 and H2O into carbon-based fuels or high-value chemicals, has been considered as a sustainable strategy to address the challenges of the energy dilemma and greenhouse effect.1–3 During the past few decades, diverse C1 fuels or chemicals have been extensively produced via photocatalytic CO2 reduction, such as CO,4,5 HCOOH,6,7 HCHO,8 CH3OH,9–11 and CH4.12,13 In particular, CH3OH has fully exerted its industrial value, for its high energy density as a clean fuel and wide applications as a commercial reagent or chemical feedstock.14 However, the CO2-to-CH3OH conversion, undergoing a six-electron reduction process, has been competed by some kinetically favored reactions, which prefer to generate H2 or other C1 products (e.g., CO, HCOOH, and HCHO) via less-electron-involved reduction processes.15 When looking into the reaction pathway for CH3OH production, *CO, formed from *COOH intermediates and regarded as the key intermediate for CO2 reduction, would be sequentially hydrogenated by following the *CO → *CHO/*COH → *CHOH → *CH2OH → CH3OH or *CO → *CHO → *CH2O → *OCH3 → CH3OH process to selectively produce CH3OH.16 Thus, the CH3OH selectivity could be expected to improve by surface active site engineering, with the *CO adsorption strengthened and/or the *CO hydrogenation energy barrier reduced to facilitate the generation of *CHO/*COH intermediates.

In comparison to nanoparticles or clusters with different atomic coordinations at the surface and in the bulk, surface-anchored or coordinated metal single atoms (SAs), serving as the homogeneous and highly dispersed active sites, are believed to benefit *CO hydrogenation towards targetable CO2 reduction to produce CH3OH.17,18 For example, by highly dispersing Re SAs on In2O3 to serve as active sites for photocatalytic CO2 reduction, the significantly improved adsorption of *CO intermediates at the surface could prohibit *CO desorption and then promote *CO hydrogenation to produce CH3OH.19 With Ru SAs highly dispersed on porous carbon nitride, the well-coordinated Ru–C/N atomic structure would build the bridges for accelerated electron transfer from N/C atoms to Ru atoms and act as electron-enriched active sites to effectively activate and selectively convert CO2 to CH3OH.20 Despite various SAs of noble and rare metals (e.g., Au, Ru, Pt, Ir, Rh, etc.) being evidenced to be effective for photocatalytic CO2 reduction for selective CH3OH production,19–25 non-noble metal SAs (e.g., Fe, Ni, Co, Ce, Zr, Cu, etc.)26–32 are more inclined to serve as active sites for high selectivity CO2 reduction from the economic perspective. In particular, by serving as the excellent active sites that accelerate photogenerated charge carrier transfer for CO2 activation and *CHO intermediate generation, Cu SAs featured with d electron configuration could steer the photogenerated charge carrier behaviors and then significantly improve the photocatalytic activity for CO2-to-CH3OH conversion.33 For example, Wang et al. constructed Cu SA active sites on UiO-66–NH2 by a photoinduction method, which was evidenced to promote the formation of *CO and CHO* intermediates, and thus achieved 5.33 µmol h−1 g−1 productivity and 55.81% selectivity towards CH3OH.30 It was also verified that Cu SA active sites anchored on BiOBr nanosheets are energy-favorable for *CO hydrogenation to form *CHO intermediates, exhibiting an excellent selectivity of 90% towards CH3OH production.34

To support SAs as efficient active sites for surface reactions, 2D ultrathin architectures with thickness reduced to atomic layers have been extensively investigated and are believed to favor photocatalysis by accelerating charge carrier transfer from the bulk to the surface. Our previous study demonstrated that, benefiting from the shortened charge carrier transfer distance and the increased built-in electric field intensity, BiOCl atomic layers (BOCAL), exposed with electron-enriched Bi sites, delivered outstanding photocatalytic activity for CO2 overall splitting, with the CO evolution rate reaching 134.8 µmol h−1 g−1 under simulated solar irradiation.35 BOCAL with an electronic structure being further tunable by the introduction of oxygen vacancies (VO) could be anticipated as an excellent platform featured with excellent charge carrier transport ability to support highly dispersed SAs acting as active sites for photocatalytic CO2 reduction.36

As inspired by the above demonstrations, herein, Cu SAs (CuSA) were highly dispersed on BOCAL via a facile electrostatic self-assembly and pyrolysis two-step method, with a Cu–O4 coordination structure created for highly efficient photocatalytic CO2 reduction to selectively produce CH3OH. In comparison to the pristine BOCAL, CuSA–BOCAL exhibits much improved photocatalytic activity for CO2 reduction, with CH3OH evolution rate and selectivity reaching 103.98 µmol g−1 h−1 and 99.14% as high, respectively. Such a high photocatalytic activity is believed to benefit from the Cu–O4 coordination structure, which could introduce the IEL in the forbidden band of BiOCl with a reduced work function, accelerating charge carrier separation and transfer to CuSA active sites for CO2 photoreduction. In situ spectral investigations and theoretical calculations further reveal that the CuSA active sites with electron density increased by photogenerated VO are conducive to deliver electrons to activate the surface-adsorbed CO2 molecules and lower the energy barrier of the rate-determining step (RDS) for *COH hydrogenation to form *CHOH intermediates, contributing to the high selectivity CH3OH photosynthesis.

2. Experimental section

2.1. Synthesis of BiOCl atomic layers (BOCAL)

0.4 g of PVP was added to 25 mL of 0.04 M mannitol solution under stirring for complete dissolution. 0.481 g of Bi(NO3)3·5H2O was dissolved into the resultant solution under vigorous stirring to keep it transparent. By adding dropwise 5 mL of saturated NaCl solution into the transparent solution under vigorous stirring for 30 min, BiOCl nanosheets were synthesized via a hydrothermal process of the resultant solution for 24 h at 160 °C. The obtained BiOCl nanosheets were dried at 60 °C after being washed with ethanol several times. 200 mg of dried BiOCl nanosheets was ultrasonically treated in isopropanol for 4 h at 30 °C. After centrifuging the resultant suspension liquid at 11 000 rpm for 10 min, the obtained supernatant was rotary evaporated at 40 °C. The resultant humid powder was immersed in ultrapure water and freeze-dried for 48 h. The resultant powder was denoted as BOCAL.

2.2. Synthesis of [Cu2(bpy)2(µ-ox)]Cl2

10 mL of 0.16 M oxalic acid solution was added dropwise to 20 mL of 0.08 M CuCl2 solution simultaneously under stirring. After dropwise addition of 10 mL of 0.16 M 2,2′-bipyridine solution to the resultant solution, [Cu2(bpy)2(µ-ox)]Cl2 was synthesized by drying the obtained sediment at 60 °C after being washed repeatedly with ethanol.

2.3. Synthesis of CuSA–BOCAL

With 100 mg of BOCAL and 2.38 mg of [Cu2(bpy)2(µ-ox)]Cl2 added to 10 mL of ultrapure water, the resultant liquid–solid mixture was homogeneously dispersed by ultrasonic treatment for 15 min. The resultant suspension was rotary evaporated at 40 °C, and CuSA–BOCAL was synthesized by calcining the obtained powder at 250 °C for 10 h in a mixed atmosphere of Ar (92%) and H2 (8%) gas.

3. Results and discussion

Herein, BOCAL, characterized by a tetragonal phase structure and nanosheet morphology exposed with (001) facets (Fig. S1), as identified by the lattice spacing of 2.75 Å,35,37 was obtained by ultrasonically exfoliating the hydrothermally synthesized BiOCl nanosheets in isopropanol.38 With Cu SAs anchored on BOCAL via an electrostatic self-assembly and pyrolysis two-step method using [Cu2(byp)2(µ-ox)]Cl2 as the precursor (Fig. 1a and S2), the obtained CuSA–BOCAL exhibits the well-maintained morphology of nanosheets with a tetragonal phase structure (Fig. 1b, c and S3) and abundant VO (Fig. S4 and S5).39,40 Elemental mappings unveil the homogeneous dispersion of Cu on CuSA–BOCAL without any agglomerated Cu species (Fig. 1d–h). With the atomically layered structure verified for CuSA–BOCAL by atomic force microscopy (AFM) to be 2–3 nm in thickness (Fig. S6), an aberration-corrected high-angle annular dark-field scanning transmission electron microscopy (AC HAADF-STEM) image reveals that Bi atoms (bright dots) exposed on the (001) facet are arranged with the orthometric (110) facet lattice fringe,35,41 while Cu SAs (dim dots tagged by red dashed circles) are highly dispersed on CuSA–BOCAL (Fig. 1i).42

Fig. 1. (a) Schematic process for the synthesis of CuSA–BOCAL, (b and c) TEM images, (d–h) elemental mapping, and (i) AC HAADF-STEM image of CuSA–BOCAL.

Fig. 1

Open in a new tab

With Cu SAs identified to be highly dispersed on BOCAL, the electronic and coordination structures of Cu SAs were investigated by X-ray absorption spectroscopy (XAS) with X-ray absorption near-edge structure spectra (XANES) and extended X-ray absorption fine structure (EXAFS) spectra recorded at the Cu K-edge. As shown in Fig. 2a, CuSA–BOCAL reveals the XANES absorption edge closer to CuO rather than Cu2O as references, suggesting an oxidation state of approximately +2 for Cu SAs in CuSA–BOCAL. The oxidation state of Cu SAs in CuSA–BOCAL could be further estimated to be +1.95 (Fig. 2b), depending on the linear relationship between the absorption edge energy position and the oxidation states for CuSA–BOCAL with Cu foil, Cu2O and CuO as references (Fig. S7).43 Then, the coordination structure of Cu SAs in CuSA–BOCAL was investigated by Cu R-space Fourier transformed k3-weighted EXAFS (FT-EXAFS) spectra. The distinct peaks located at ∼2.24, 2.53 and 2.76 Å could be assigned to the Cu–Cu bond for Cu foil and Cu–O–Cu bond for CuO and Cu2O (Fig. 2c), respectively, which are absent in CuSA–BOCAL, indicating that the Cu species would not exist as agglomerated metallic or oxide clusters and nanoparticles in CuSA–BOCAL. Instead, the only one distinct peak observed for CuSA–BOCAL at ∼1.52 Å is approaching the Cu–O peak in CuO, indicative of highly dispersed Cu SAs in CuSA–BOCAL. The FT-EXAFS fitting result further confirms that Cu SAs are coordinated with four neighbouring O atoms, with a Cu–O4 coordination structure created in CuSA–BOCAL (Fig. 2d and Table S1).

Fig. 2. (a) Cu K-edge XANES spectra of CuSA–BOCAL and references, (b) Cu oxidation states of CuSA–BOCAL, (c) FT-EXAFS spectra of CuSA–BOCAL and references, (d) FT-EXAFS fitting curve of CuSA–BOCAL.

Fig. 2

Open in a new tab

The photocatalytic performances for CO2 reduction over CuSA–BOCAL as well as BOCAL and Cu nanoparticle (Cu NP)-modified BOCAL (CuNP–BOCAL, Fig. S8–S10) as references were comparatively evaluated in a gas–liquid–solid photoreaction system under the illumination of a 300 W Xe lamp, fed with ultrapure water and CO2 gas. As shown in Fig. 3a, BOCAL exhibits 100% selectivity towards CO gas, with an evolution rate of 11.21 µmol g−1 h−1. In comparison, the CO evolution rate is reduced to 9.74 µmol g−1 h−1 for CuNP–BOCAL, with a much decreased selectivity of 27.87%, while CH3OH is generated as the major product at an evolution rate of 25.28 µmol g−1 h−1, with selectivity reaching 72.13% as high. These comparative results suggest that Cu NPs loaded on BOCAL could serve as active sites to effectively drive the CO2 reduction reaction for CH3OH production. More interestingly, with highly dispersed Cu SAs anchored on BOCAL, CuSA–BOCAL exhibits significantly improved photocatalytic activity and selectivity for CH3OH production. The optimized content of Cu SAs (Fig. S11 and Table S2) endows CuSA–BOCAL with a quite high CH3OH evolution rate of 103.98 µmol g−1 h−1 at a selectivity of 99.14% and a low CO evolution rate suppressed to be only 0.81 µmol g−1 h−1 (Fig. 3a, S12 and S13). These results indicate that Cu SAs anchored on BOCAL as active sites could prohibit the *CO desorption from the surface to generate CO, and instead promote the *CO intermediates to be hydrogenated and further converted to CH3OH. Notably, the apparent quantum efficiency (AQE) of CuSA–BOCAL for photocatalytic CO2 reduction reaches 2.60% at 365 nm (Fig. S14). In addition, no H2 gas was detected over all samples, indicating the excellent competitiveness against the water overall splitting reaction. Encouragingly, the obtained CuSA–BOCAL photocatalyst has surpassed all the BilOmXn (X: Cl, Br, I)-based photocatalysts for CO2 reduction (Table S3), and especially for CH3OH selective production. During the five cycles of CO2 photoreduction reactions (Fig. 3b), the CH3OH evolution rate remains almost unchanged, with CH3OH selectivity kept at 98.5–99.5%, indicating the excellent stability of CuSA–BOCAL. Moreover, the excellent stability of CuSA–BOCAL for CO2 photoreduction stems from the well-preserved crystal phase and the intact single-atom Cu configuration (Fig. S15–S18). In addition, O2 as the oxidation product could also be detected (Fig. S19), with O2 and CH3OH produced at a stoichiometric molar ratio of 1.5 : 1 (Fig. S20–S24). To verify that the products originated from CO2 photoreduction, control experiments were carried out over CuSA–BOCAL under various reaction conditions. With CO2 gas replaced by N2, in the dark or without a photocatalyst, no product could be detected, inferring that CH3OH dominating in products should be produced by CO2 photoreduction induced by light irradiation (Fig. 3c). As further cleared by 13CO2 isotope labeling experiments, 13C labeled CO (13CO at m/z = 29) and CH3OH (13CH3OH at m/z = 33) detected in the gas and liquid phases, respectively, confirm that both CH3OH and CO are undoubtedly originated from photocatalytic CO2 reduction (Fig. 3d, e, S25 and S26).

Fig. 3. (a) Product evolution rates of BOCAL, CuNP–BOCAL and CuSA–BOCAL, (b) cycling tests for photocatalytic CO2 reduction over CuSA–BOCAL, and (c) product evolution rates under various reaction conditions. Mass spectra of (d) CO and (e) CH3OH evolved during photocatalytic 13CO2 (99%) reduction.

Fig. 3

Open in a new tab

With the very similar structural, morphological and optical properties (Fig. S3–S5 and S7–S9) excluded from contributing to the different photocatalytic activities, the photogenerated charge transfer ability was explored by single-particle photoluminescence (PL) spectroscopy to unveil the motivation for the dramatically improved photocatalytic performance for CO2 reduction over CuSA–BOCAL. By recording the time-resolved transient photoluminescence (TRPL) map on a single nanosheet of BOCAL, which illustrates the charge carrier lifetime (0–5 ns) at each active site, the homogeneous PL emission distribution (Fig. 4a) hints at the similar charge carrier migration ability across BOCAL.44 This result could be corroborated by PL spectra recorded at three selected sites from the center to the edge on the BOCAL nanosheet, with similar intensities noted for the distinct PL emission at ca. 575 nm (Fig. 4b). Such obvious charge carrier recombination across the whole BOCAL nanosheet would result in the short charge carrier lifetime (Fig. 4c), explaining the poor photocatalytic activity of BOCAL without introduction of foreign active sites. For CuNP–BOCAL, one could observe very different PL emissions at different regions on the nanosheet, which indicates that the introduced Cu NPs could steer the charge carrier transfer process by acting as nanoregional electron capture centers on BOCAL (Fig. 4d),45 as evidenced by the heterogeneous PL emission distribution recorded at different sites (Fig. 4e).46,47 However, the negligible prolongation of the charge carrier lifetime monitored for CuNP–BOCAL as compared to BOCAL suggests that the introduced Cu NPs would not significantly promote the charge carrier migration in the entire BOCAL nanosheet (Fig. 4f). With Cu SAs anchored on BOCAL, the homogeneous PL emission distribution could be again noted for the CuSA–BOCAL nanosheet (Fig. 4g), with much weakened PL emission intensity, suggesting the inhibited combination of photogenerated charge carriers (Fig. 4h), which should account for the prolonged charge carrier lifetime (Fig. 4i) and thus benefit the photocatalytic reaction. Steady-state PL spectra (Fig. S27) and TRPL spectra (Fig. S28) recorded on these powdery photocatalysts display the attenuated PL emission quenching and the prolonged average carrier lifetime, again confirming the promoted separation and transfer of photogenerated electrons and holes upon the introduction of Cu SAs as active sites. Such improved charge transfer properties for CuSA–BOCAL could be further evidenced by the electrochemical impedance spectra fitted with Nyquist plots using an equivalent circuit, with much reduced charge transfer resistances determined for CuSA–BOCAL relative to BOCAL and CuNP–BOCAL (Fig. S29 and Table S4), which could be also supported by the significantly increased transient photocurrent density for CuSA–BOCAL (Fig. S30).

Fig. 4. Single-particle TRPL mapping of (a) BOCAL, (d) CuNP–BOCAL and (g) CuSA–BOCAL. Single-site PL spectra of (b) BOCAL, (e) CuNP–BOCAL and (h) CuSA–BOCAL. Single-site TRPL spectra of (c) BOCAL, (f) CuNP–BOCAL and (i) CuSA–BOCAL.

Fig. 4

Open in a new tab

Density functional theory (DFT) calculations were then conducted to theoretically investigate the Cu SA-induced electronic structure evolution and its intrinsic contribution to the improved charge carrier transfer ability. With structure models optimized for BOCAL, CuNP–BOCAL and CuSA–BOCAL (Fig. S31), the band structure and projected density of states (PDOS) calculations exhibit that the valence band (VB) of BOCAL is composed of O 2p and Cl 3p orbitals, while the conduction band (CB) is dominated by the Bi 6p orbital (Fig. S32a). Upon the introduction of Cu NPs onto BOCAL, the Cl 3p orbitals and Cu 3d–O 2p hybrid orbitals could mainly contribute to the VB of CuNP–BOCAL (Fig. S32b). Moreover, CuSA–BOCAL exhibits the well-maintained band structure with BOCAL, except for the O 2p and Cu 3d orbitals overlapping at 0.95–1.22 eV, revealing the impurity energy level (IEL) created in the forbidden band of BOCAL by the introduced Cu SAs, which is supposed to capture electrons to accelerate the separation of photoinduced charge carriers (Fig. S32c).4 With the up-shift of the Fermi level, the calculated electrostatic potentials observe the gradual decrease in the work function (Φ) by following BOCAL (8.16 eV) > CuNP–BOCAL (7.71 eV) > CuSA–BOCAL (7.29 eV) (Fig. S33), depending on the introduction of Cu NPs and Cu SAs, which signifies that the highest density of charge carriers in CuSA–BOCAL would contribute to the rapid migration of photogenerated charge carriers.48 The differential charge density map depicts the accumulation of electrons around Cu NPs and holes around O atoms at the interface of Cu NPs and BOCAL for CuNP–BOCAL (Fig. S34a). Such a short-distance electron–hole distribution might cause electron and hole recombination and make it difficult for electrons to migrate to surface active sites for effective CO2 activation. In comparison, CuSA–BOCAL exhibits a spatial charge distribution, with electrons mainly enriched on Cu SAs and holes accumulated on O atoms (Fig. S34b), which effectively inhibits the recombination of electrons and holes and thus contributes to the efficient CO2 activation on the Cu–O4 coordination structure.48 All the DFT calculations demonstrate that the Cu–O4 coordination structure could enhance the migration of photogenerated charge carriers and act as the active sites for photocatalytic CO2 reduction.

Given the much higher selectivity towards CH3OH over CuSA–BOCAL than CuNP–BOCAL, in situ electron paramagnetic resonance (EPR) spectroscopy was carried out to unveil the mechanistic fundamentals of electronic structure evolution during photocatalytic CO2 reduction. For CuNP–BOCAL and CuSA–BOCAL, a single Lorentzian line with a g-value of 2.005 assigned to the VO signal could be observed at the mixed atmosphere of CO2 and vapor in the dark, which increased significantly after irradiation for 30 minutes, suggesting the generation of additional VO in both CuNP–BOCAL and CuSA–BOCAL (Fig. 5a), with O atoms escaping from the surface when subject to light energy. Then, the electron-enriched Cu–O4 active sites are created on BOCAL, as evidenced by the calculated charge transfer value (Δq) of 0.04e (Fig. 5b). To reveal the dynamic electron structure evolution induced by photogenerated VO, in situ X-ray photoelectron spectroscopy (XPS) was carried out to recognize the active sites and monitor the electron transfer path. As revealed by O 1s XPS spectra, fresh CuNP–BOCAL exhibits three peaks at 529.75 eV, 531.25 eV and 532.60 eV (Fig. 5c),40 respectively, attributed to lattice oxygen, defect oxygen and surface-adsorbed H2O. The unchanged binding energy (BE) and the increased peak area ratio observed for the defect oxygen signals recorded on CuNP–BOCAL in the mixed atmosphere of CO2 and vapor under irradiation imply the photogeneration of VO with electronic inertness to CO2 and H2O molecules during photocatalytic CO2 reduction. Notably, the 0.10 eV negative shift in BE observed for the lattice oxygen signal suggests that the lattice oxygen would accept photogenerated electrons, resulting in insufficient photogenerated electron availability at the Cu NPs for CO2 activation. The more positive BE for all these O 1s signals in CuSA–BOCAL than CuNP–BOCAL (Fig. 5d) implies that electrons are more favorable to be distributed at Cu–O4 sites in CuSA–BOCAL than at Cu NPs in CuNP–BOCAL to maintain electrical neutrality. During photoreaction, the increased peak area ratio and the 0.10 eV positive shift in BE for the defect oxygen signal, together with the 0.05 eV positive shift in BE for the lattice oxygen signal, indicate the photogeneration of VO in CuSA–BOCAL upon irradiation, which could lead to the migration of photogenerated electrons from lattice oxygen to Cu–O4 sites. The Cu 2p XPS spectrum exhibits two metallic Cu peaks of Cu 2p3/2 and Cu 2p1/2 located at 932.45 eV and 952.25 eV for CuNP–BOCAL (Fig. 5e),49,50 which are negatively shifted by 0.15 eV during photoreaction, suggesting that the electrons could be supplied to activate CO2 molecules on Cu NPs. In comparison, the Cu signals are located at the higher BE, with a larger negative shift (0.25 eV) happened to CuSA–BOCAL during photoreaction (Fig. 5f), indicative of the enriched electrons at Cu SAs for effective CO2 activation. The Bi 4f XPS spectrum displays two peaks of Bi 4f7/2 and Bi 4f5/2 at 159.00 eV and 164.30 eV for Bi3+ in fresh CuNP–BOCAL (Fig. 5g),42 with only a 0.05 eV positive shift in BE observed during photoreaction, which means few electrons donated by Bi atoms, corresponding to the sluggish H2O activation at Bi sites. In comparison, CuSA–BOCAL displays more positive BE, with a larger positive shift of 0.10 eV during photoreaction (Fig. 5h), manifesting the distinct hole accumulation on Bi sites for the water oxidation reaction. Consequently, the Cu–O4 sites are identified as the dominant active centers, suggested by the capacity of gathering photogenerated electrons. Thus, the more active electronic interaction of Cu and Bi sites with CO2/H2O for CuSA–BOCAL than CuNP–BOCAL could be responsible for the excellent photocatalytic activity for CO2 reduction. After photoreaction, all the O, Cu and Bi signals are almost recovered to the original states (Fig. 5c–h), especially for CuSA–BOCAL, which suggests the excellent dynamic stability for photocatalytic CO2 reduction. In situ EPR spectra combined with in situ XPS spectra unveil the different electronic effects of VO in CuNP–BOCAL and CuSA–BOCAL. In CuNP–BOCAL, photogenerated VO exhibits electronic inertness. Lattice oxygen preferentially captures photogenerated electrons, limiting electron availability for CO2 activation at Cu sites. In CuSA–BOCAL, photogenerated VO promotes electron migration from lattice oxygen to the Cu–O4 active sites, resulting in electron enrichment at Cu SAs for efficient CO2 activation, while simultaneously inducing hole accumulation at Bi sites to facilitate water oxidation. This dramatic electronic interaction underpins the superior photocatalytic activity of CuSA–BOCAL for CO2 reduction.

Fig. 5. (a) In situ EPR spectra of CuNP–BOCAL and CuSA–BOCAL. (b) Bader charge of CuSA–BOCAL before and after irradiation. In situ O 1s, Cu 2p and Bi 4f XPS spectra of (c, e and g) CuNP–BOCAL and (d, f and h) CuSA–BOCAL.

Fig. 5

Open in a new tab

To deepen the fundamental understanding of high selectivity for the CO2-to-CH3OH conversion over CuSA–BOCAL, in situ diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) was carried out to monitor the reaction intermediates, especially the key species for CH3OH generation, and further to unveil the reaction pathways. With the mixed gas of H2O vapour and CO2 introduced into the reaction, the *CO2− species could be detected on BOCAL (1622, 1681 cm−1), CuNP–BOCAL (1683 cm−1) and CuSA–BOCAL (1681 cm−1) (Fig. 6a–c), which are gradually intensified with prolonging irradiation, indicating the CO2 chemisorption at the surface.51 Moreover, the intensity of the *CO2− signal is significantly stronger for CuSA–BOCAL than BOCAL and CuNP–BOCAL, suggesting that introduced Cu SAs on BOCAL could act as the active sites to effectively absorb and activate CO2 molecules. The H2O signal observed for BOCAL (1661 cm−1) is more distinct than CuNP–BOCAL (1654 cm−1), while it disappears over CuSA–BOCAL.52 In addition, four signals related to *OH species could be detected at 3598–3730 cm−1 on the three samples (Fig. S35–S37), which are generated from the dissociation of adsorbed H2O (*H2O → H+ + *OH).5,35 It is clear that the *OH signals are much stronger for CuSA–BOCAL than BOCAL and CuNP–BOCAL, indicating that the dissociation of H2O molecules could be promoted by the introduced Cu SAs. The above observations on superior accumulation of *CO2− species and enhanced H2O dissociation over CuSA–BOCAL would favor the subsequent processes for CO2 reduction. The peaks at 1551, 1342 and 1284 cm−1 over CuSA–BOCAL could be attributed to the vibration signal of C Created by potrace 1.16, written by Peter Selinger 2001-2019 O stretching, C–O stretching and O–H deformation for *COOH species,53,54 which are the key intermediate for CO2-to-CO conversion via the H+ + *CO2− → *COOH step.5,52 In comparison, the intensities of *COOH peaks are obviously decreased for BOCAL and CuNP–BOCAL, which in turn supports that Cu SAs acting as active sites could promote *COOH generation. The negligible signals at 2058 and 2077 cm−1 over BOCAL (Fig. S38), attributed to the *CO species generated from the hydrogenation and dehydration of *COOH (H+ + *COOH → *CO + H2O),35,55 support the more favorable *CO desorption and CO production over BOCAL, instead of CH3OH. In comparison to BOCAL, a slightly stronger *CO signal could be observed over CuNP–BOCAL (2077 cm−1), and even CuSA–BOCAL exhibits two more distinct signals of *CO species at 2058 and 2077 cm−1.35,55,56 This increased accumulation of *CO species over CuSA–BOCAL suggests that the strengthened adsorption of *CO species would favor the subsequent *CO-to-CH3OH conversion. With *CO-to-CH3OH conversion considered as a hydrogenation process, which could generate *COH and *CH3OH species (H+ + *CO → *COH, H+ + *COH → *CHOH, H+ + *CHOH → *CH2OH, *CH2OH → *CH3OH), the *COH species could be detected at 1078 cm−1 for CuSA–BOCAL,57 which is absent over CuNP–BOCAL. Moreover, one could note two signals at 1035 and 1122 cm−1 for CuSA–BOCAL, associated with the *CH3O species, which are more distinct than those for CuNP–BOCAL (1032 and 1137 cm−1).58 These observations imply that the *CO hydrogenation process occurs more favorably on the Cu–O4 coordination structure than Cu NPs. The absence of *COH and *CH3O species indicates the incompetence for *CO hydrogenation to produce CH3OH over BOCAL. Moreover, several signals detected at 2881, 2935 and 2972 cm−1 over CuSA–BOCAL, assigned to the *CH2OH and *CH3OH species, are more distinct than CuNP–BOCAL,55,59 which certainly accounts for the higher photocatalytic activity towards CH3OH production over CuSA–BOCAL than CuNP–BOCAL.

Fig. 6. In situ DRIFTS spectra of (a) BOCAL, (b) CuNP–BOCAL and (c) CuSA–BOCAL. (d) Energetic reaction pathways of photocatalytic CO2 reduction over BOCAL, CuNP–BOCAL, and CuSA–BOCAL. (e) Proposed mechanism for photocatalytic CO2 reduction over CuSA–BOCAL.

Fig. 6

Open in a new tab

DFT calculations were further performed to investigate the energetic pathways of photocatalytic CO2 reduction over BOCAL, CuNP–BOCAL and CuSA–BOCAL to unveil the significant contribution of the Cu–O4 coordination structure to the high selectivity towards CH3OH production. With the intermediates well detected by in situ DRIFTS experiments, the reaction pathway of photocatalytic CO2 reduction for CO generation over BOCAL is reasonably determined to be *CO2− → *COOH → *CO → CO, while the reaction pathways for photocatalytic CO2 reduction to produce CH3OH over CuNP–BOCAL and CuSA–BOCAL are theoretically proposed as *CO2− → *COOH → *CO → *COH → *CHOH → *CH2OH → *CH3OH → CH3OH. With the active sites for CO2 reduction well identified by in situ XPS spectra, the CO2 molecules are considered to be preferentially adsorbed at Cu SAs and Cu NPs. Moreover, with the geometrical structures of intermediates at the active sites constructed and optimized for BOCAL, CuNP–BOCAL and CuSA–BOCAL (Fig. S39), the Gibbs free energy of each elementary step for photocatalytic CO2 reduction was then calculated (Fig. 6d) and the catalytic cycle for CuSA–BOCAL was depicted according to aforementioned reaction pathways (Fig. S40). Specifically, the free energy difference (ΔG) is calculated to be 0.58 eV for the hydrogenation of adsorbed CO2 molecules to form *COOH over BOCAL, which is distinctly reduced to 0.17 eV for CuSA–BOCAL and further to 0.01 eV for CuNP–BOCAL. It could be thus deduced that the Cu species introduced onto BOCAL should serve as the active sites to effectively activate CO2 molecules, with *COOH intermediates generated and then converted to *CO intermediates by the following hydrogenation and dehydration processes. This ΔG of *COOH formation (0.17 eV) over CuSA–BOCAL is much lower than that of hydrogen evolution (2.08 eV), accounting for the higher selectivity towards CO2 reduction (Fig. S41). In comparison, CuNP–BOCAL exhibits a higher ΔG (0.96 eV) for *CO desorption than both CuSA–BOCAL and BOCAL, suggesting that *CO desorption to form gaseous CO is more unfavourable on CuNP–BOCAL (Fig. S42). Moreover, the *CO hydrogenation step (CO + H+ → *COH) is exergonic for both CuSA–BOCAL (0.79 eV) and CuNP–BOCAL (1.17 eV), indicating that the reaction preferentially follows the *CO hydrogenation pathway to form *COH rather than *CO desorption to release CO gas. Notably, the ΔG (0.65 eV) required by CuSA–BOCAL for the formation of *CHOH intermediates, regarded as the RDS for photocatalytic CO2 reduction towards CH3OH generation, is much lower than CuNP–BOCAL (1.70 eV). In the subsequent steps, the generation of various intermediates (*CH2OH, and *CH3OH) via multiple hydrogenation processes for CH3OH production requires smaller ΔG for CuSA–BOCAL than CuNP–BOCAL, indicating that the Cu–O4 coordination units as active sites would benefit photocatalytic CO2 reduction to CH3OH. All these careful comparisons on free energy in the reaction pathways for CO2 reduction illustrate that BOCAL is conducive to the desorption of *CO for CO production, while the introduced Cu species, especially Cu SAs, could provide abundant active sites to activate CO2 molecules to generate *CO intermediates, which would then get involved in the following hydrogenation processes for favored CH3OH production by lowering the ΔG of the RDS (*COH + H+ → *CHOH).

Based on the aforementioned experimental investigations and theoretical calculations, the mechanism of photocatalytic CO2 reduction with improved selectivity towards CH3OH production over CuSA–BOCAL could be reasonably proposed (Fig. 6e and S43). Upon irradiation, electrons are photoexcited from the VB to the CB for CO2 reduction, with holes left on the VB in BOCAL for water oxidation. With Cu SAs introduced onto BOCAL, the charge carrier transfer processes could be significantly accelerated by creating IEL in the forbidden band and lowering the work function of BiOCl. Moreover, the electron-enriched Cu–O4 coordination structure, with electrons density increased by photogenerated VO, would act as the active sites to effectively transfer electrons to activate the adsorbed CO2 molecules, which could strengthen the *CO adsorption and lower the energy barrier of the RDS to facilitate the generation of *CHOH intermediates, contributing to the excellent selectivity towards CH3OH production.

4. Conclusions

In this work, highly dispersed single Cu atoms (CuSA) on BiOCl atomic layers (BOCAL) featuring a Cu–O4 coordination structure were successfully synthesized through a facile electrostatic self-assembly and pyrolysis two-step method. The obtained CuSA–BOCAL photocatalyst exhibits an excellent performance for photocatalytic CO2 reduction with a much improved selectivity of 99.14% towards CH3OH at a high evolution rate of 103.98 µmol g−1 h−1. This superior photocatalytic performance of CuSA–BOCAL should benefit from the Cu–O4 coordination structure, which could accelerate charge carrier migration by the introduced impurity energy level and the lowered work function. Moreover, in situ spectral characterization and theoretical calculations reveal that the Cu–O4 coordination structure with electron density increased by photogenerated VO, could act as the active sites to transfer electrons for effective CO2 activation and reduce the energy barrier of the RDS (*CO + H → *COH), distinctly improving selectivity towards CH3OH production. This work provides an available strategy to the design of single-atom active sites with tuned electronic structures for high selectivity CH3OH photosynthesis and inspires the understanding of electronic interaction between active sites and semiconducting supports for efficient CO2 photoreduction.

Author contributions

T. Peng completed the experimental investigations and the original drafts of the write-ups. Y. Wang, H. Zhou and Y. Wang helped with theoretical calculations. K. Wang conducted the synchrotron radiation measurements. Z. Zheng and Z. Liu performed single-particle photoluminescence spectroscopy. H. Zhou, D. Zhao, Q. Guan, Y. Wang, B. Wu and G. Wang helped analyze the data. B. Kou helped with the writing and editing. S. Shen funded the project and supervised the project, and completed the reviewing and editing. All the authors discussed the results and commented on the manuscript.

Conflicts of interest

There are no conflicts to declare.

Supplementary Material

SC-OLF-D6SC03539F-s001

Acknowledgments

The authors acknowledge the financial support from the National Key R&D Program of China (2025YFF0522700), the National Natural Science Foundation of China (52225606 and 52488201) and the Leadership Talent Program (Science and Education) of SIP.

Data availability

The data supporting this manuscript are available from the corresponding author upon reasonable request.

Supplementary information (SI) is available. See DOI: https://doi.org/10.1039/d6sc03539f.

Notes and references

  1. Li X. Sun Y. Xu J. Shao Y. Wu J. Xu X. Pan Y. Ju H. Zhu J. Xie Y. Nat. Energy. 2019;4:690–699. [Google Scholar]
  2. Yu J. Huang L. Tang Q. Yu S.-B. Qi Q.-Y. Zhang J. Ma D. Lei Y. Su J. Song Y. Eloi J.-C. Harniman R. L. Borucu U. Zhang L. Zhu M. Tian F. Du L. Phillips D. L. Manners I. Ye R. Tian J. Nat. Catal. 2023;6:464–475. [Google Scholar]
  3. Wang Z. Cai R. Zhang Y. Huang X. Bi Y. Angew. Chem., Int. Ed. 2025;64:e202502834. doi: 10.1002/anie.202502834. [DOI] [PubMed] [Google Scholar]
  4. Peng T. Wang Y. Wang K. Zhang K. Wang Y. Xu Y. Guan Q. Wang G. Zhang W. Wu B. Shen S. Green Energy Environ. 2025;10:1948–1955. [Google Scholar]
  5. Li Y. Xue Y. Gao X. Wang L. Liu X. Wang Z. Shen S. Adv. Funct. Mater. 2023;34:2312634. [Google Scholar]
  6. Kang X. He Z. Wang F. Liu Y. Guo L. Adv. Funct. Mater. 2024;35:2419802. [Google Scholar]
  7. Zhang H. Liu S. Zhou X. Zheng Z. Wang P. Cheng H. Wang Z. Dai Y. Liu Y. Huang B. ACS Catal. 2025;15:3000–3007. [Google Scholar]
  8. Zhao S. Liang H. Q. Hu X. M. Li S. Daasbjerg K. Angew. Chem., Int. Ed. 2022;61:e202204008. doi: 10.1002/anie.202204008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Wang R. Zhang M. Liu J. Wu X. Zhang S. Zhong Q. Yao J. Nat. Commun. 2025;16:7844. doi: 10.1038/s41467-025-63179-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Chatterjee R. Reza S. Samui S. Thapa R. Bhaumik A. Angew. Chem., Int. Ed. 2025;64:e202511602. doi: 10.1002/anie.202511602. [DOI] [PubMed] [Google Scholar]
  11. Li H. Zhang X. MacFarlane D. R. Adv. Energy Mater. 2014;5:1401077. [Google Scholar]
  12. Jiang P. Liu W.-R. Wang K. Ding Y. Wang C.-H. Ng Y. H. Chen L.-H. Su B.-L. CCS Chem. 2025;7:2149–2160. [Google Scholar]
  13. Zheng K. Liu S. Zhu J. Dai Z. Liu C. Li B. Zheng Y. Chen X. Zhai L. Wu Y. Liu W. Fan M. Hu J. Pan Y. Zhu J. Sun F. Sun Y. Xie Y. Angew. Chem., Int. Ed. 2025;64:e202508259. doi: 10.1002/anie.202508259. [DOI] [PubMed] [Google Scholar]
  14. Ding J. Tang Q. Fu Y. Zhang Y. Hu J. Li T. Zhong Q. Fan M. Kung H. H. J. Am. Chem. Soc. 2022;144:9576–9585. doi: 10.1021/jacs.1c13301. [DOI] [PubMed] [Google Scholar]
  15. Wang R. Zhang M. Zhang S. Zheng J. Zeng Y. Yang Y. Ding J. Wu X. Zhong Q. ACS Nano. 2023;17:24363–24373. doi: 10.1021/acsnano.3c10352. [DOI] [PubMed] [Google Scholar]
  16. Wu Y. Cao S. Hou J. Li Z. Zhang B. Zhai P. Zhang Y. Sun L. Adv. Energy Mater. 2020;10:2000588. [Google Scholar]
  17. Wang S. Wang L. Wang D. Li Y. Energy Environ. Sci. 2023;16:2759–2803. [Google Scholar]
  18. Xue Z.-H. Luan D. Zhang H. Lou X. W. Joule. 2022;6:92–133. [Google Scholar]
  19. Shen C. Meng X. Y. Zou R. Sun K. Wu Q. Pan Y. X. Liu C. J. Angew. Chem., Int. Ed. 2024;63:e202402369. doi: 10.1002/anie.202402369. [DOI] [PubMed] [Google Scholar]
  20. Liu Z. Chen Z. Li M. Li J. Zhuang W. Yang X. Wu S. Zhang J. ACS Catal. 2023;13:6630–6640. [Google Scholar]
  21. Huang L. Lu R. Zhang W. Fan Y. Du Y. Ni K. Zhu Y. Zhu M. Angew. Chem., Int. Ed. 2024;63:e202412964. doi: 10.1002/anie.202412964. [DOI] [PubMed] [Google Scholar]
  22. Vennapoosa C. S. Varangane S. Abraham B. M. Bhasin V. Bhattacharyya S. Wang X. Pal U. Chatterjee D. J. Phys. Chem. Lett. 2023;14:11400–11411. doi: 10.1021/acs.jpclett.3c02347. [DOI] [PubMed] [Google Scholar]
  23. Sharma P. Kumar S. Tomanec O. Petr M. Chen J. Z. Miller J. T. Varma R. S. Gawande M. B. Zboril R. Small. 2021;17:e2006478. doi: 10.1002/smll.202006478. [DOI] [PubMed] [Google Scholar]
  24. Pieta I. S. Kadam R. G. Pieta P. Mrdenovic D. Nowakowski R. Bakandritsos A. Tomanec O. Petr M. Otyepka M. Kostecki R. Khan M. A. M. Zboril R. Gawande M. B. Adv. Mater. Interfaces. 2021;8:2001822. [Google Scholar]
  25. Kale H. B. Kute A. D. Gaikwad R. P. Fornasiero P. Zbořil R. Gawande M. B. Coord. Chem. Rev. 2024;502:215602. [Google Scholar]
  26. Huang J. R. Shi W. X. Xu S. Y. Luo H. Zhang J. Lu T. B. Zhang Z. M. Adv. Mater. 2023;36:2306906. [Google Scholar]
  27. Millet M.-M. Algara-Siller G. Wrabetz S. Mazheika A. Girgsdies F. Teschner D. Seitz F. Tarasov A. Levchenko S. V. Schlögl R. Frei E. J. Am. Chem. Soc. 2019;141:2451–2461. doi: 10.1021/jacs.8b11729. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Ma M. Huang Z. Doronkin D. E. Fa W. Rao Z. Zou Y. Wang R. Zhong Y. Cao Y. Zhang R. Zhou Y. Appl. Catal., B. 2022;300:120695. [Google Scholar]
  29. Wang W. Song S. Wang P. He M. Fang Z. Yuan X. Li H. Li C. Wang X. Wei Y. Song W. Xu H. Li Z. ACS Catal. 2023;13:4597–4610. [Google Scholar]
  30. Wang G. He C.-T. Huang R. Mao J. Wang D. Li Y. J. Am. Chem. Soc. 2020;142:19339–19345. doi: 10.1021/jacs.0c09599. [DOI] [PubMed] [Google Scholar]
  31. Liu L. Li M. Chen F. Huang H. Small Struct. 2022;4:2200188. [Google Scholar]
  32. Bakandritsos A. Kadam R. G. Kumar P. Zoppellaro G. Medved M. Tuček J. Montini T. Tomanec O. Andrýsková P. Drahoš B. Varma R. S. Otyepka M. Gawande M. B. Fornasiero P. Zbořil R. Adv. Mater. 2019;31:1900323. doi: 10.1002/adma.201900323. [DOI] [PubMed] [Google Scholar]
  33. Wang J. Heil T. Zhu B. Tung C.-W. Yu J. Chen H. M. Antonietti M. Cao S. ACS Nano. 2020;14:8584–8593. doi: 10.1021/acsnano.0c02940. [DOI] [PubMed] [Google Scholar]
  34. Wang K. Cheng M. Xia F. Cao N. Zhang F. Ni W. Yue X. Yan K. He Y. Shi Y. Dai W. Xie P. Small. 2023;19:2207581. doi: 10.1002/smll.202207581. [DOI] [PubMed] [Google Scholar]
  35. Peng T. Wang Y. Dong C.-L. Nga T. T. T. Wu B. Wang Y. Guan Q. Zhang W. Shen S. Nano-Micro Lett. 2025;17:223. doi: 10.1007/s40820-025-01723-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Li J. Li H. Zhan G. Zhang L. Acc. Chem. Res. 2016;50:112–121. doi: 10.1021/acs.accounts.6b00523. [DOI] [PubMed] [Google Scholar]
  37. Shi Y. Shou H. Li H. Zhan G. Liu X. Yang Z. Mao C. Cheng J. Zhang X. Jiang Y. Zhao S. Wang J. Liu X. Song L. Sun H. Zhang L. Angew. Chem., Int. Ed. 2023;62:e202302286. doi: 10.1002/anie.202302286. [DOI] [PubMed] [Google Scholar]
  38. Zhang L. Niu C.-G. Zhao X.-F. Liang C. Guo H. Zeng G.-M. ACS Appl. Mater. Interfaces. 2018;10:39723–39734. doi: 10.1021/acsami.8b14227. [DOI] [PubMed] [Google Scholar]
  39. Gong S. Zhu G. Wang R. Rao F. Shi X. Gao J. Huang Y. He C. Hojamberdiev M. Appl. Catal., B. 2021;297:120413. [Google Scholar]
  40. Li W. Mao Y. Liu Z. Zhang J. Luo J. Zhang L. Qiao Z. A. Adv. Mater. 2023;35:2300396. doi: 10.1002/adma.202300396. [DOI] [PubMed] [Google Scholar]
  41. Sun J. Wu S. Yang S.-Z. Li Q. Xiong J. Yang Z. Gu L. Zhang X. Sun L. Appl. Catal., B. 2018;221:467–472. [Google Scholar]
  42. Zhao L. Hou H. Wang S. Wang L. Yang Y. Bowen C. R. Wang J. Liao Z. Yang D. Yan R. Yang W. Adv. Funct. Mater. 2024;35:2416346. [Google Scholar]
  43. Lin Z. Wang Y. Peng Z. Huang Y. C. Meng F. Chen J. L. Dong C. L. Zhang Q. Wang R. Zhao D. Chen J. Gu L. Shen S. Adv. Energy Mater. 2022;12:2200716. [Google Scholar]
  44. Li B. Lv M. Zhang Y. Gong X. Lou Z. Wang Z. Liu Y. Wang P. Cheng H. Dai Y. Huang B. Zheng Z. ACS Nano. 2024;18:25522–25534. doi: 10.1021/acsnano.4c05351. [DOI] [PubMed] [Google Scholar]
  45. Lv M. Zhang X. Li B. Huang B. Zheng Z. ACS Nano. 2024;18:30247–30268. doi: 10.1021/acsnano.4c10702. [DOI] [PubMed] [Google Scholar]
  46. Yu Y. He Y. Yan P. Wang S. Dong F. Proc. Natl. Acad. Sci. U. S. A. 2023;120:e2307320120. doi: 10.1073/pnas.2307320120. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Ou H. Qian Y. Yuan L. Li H. Zhang L. Chen S. Zhou M. Yang G. Wang D. Wang Y. Adv. Mater. 2023;35:2305077. doi: 10.1002/adma.202305077. [DOI] [PubMed] [Google Scholar]
  48. Zhao D. Wang Y. Dong C.-L. Meng F. Huang Y.-C. Zhang Q. Gu L. Liu L. Shen S. Nano-Micro Lett. 2022;14:223. doi: 10.1007/s40820-022-00962-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Wang G. Huang R. Zhang J. Mao J. Wang D. Li Y. Adv. Mater. 2021;33:2105904. doi: 10.1002/adma.202105904. [DOI] [PubMed] [Google Scholar]
  50. Xie W. Li K. Liu X. H. Zhang X. Huang H. Adv. Mater. 2022;35:2208132. doi: 10.1002/adma.202208132. [DOI] [PubMed] [Google Scholar]
  51. Shi X. Dong X. a. He Y. Yan P. Dong F. Sci. Bull. 2022;67:1137–1144. doi: 10.1016/j.scib.2022.01.013. [DOI] [PubMed] [Google Scholar]
  52. Shangguan W. Liu Q. Wang Y. Sun N. Liu Y. Zhao R. Li Y. Wang C. Zhao J. Nat. Commun. 2022;13:3894. doi: 10.1038/s41467-022-31474-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Liu P. Huang Z. Gao X. Hong X. Zhu J. Wang G. Wu Y. Zeng J. Zheng X. Adv. Mater. 2022;34:2200057. doi: 10.1002/adma.202200057. [DOI] [PubMed] [Google Scholar]
  54. Ran L. Li Z. Ran B. Cao J. Zhao Y. Shao T. Song Y. Leung M. K. H. Sun L. Hou J. J. Am. Chem. Soc. 2022;144:17097–17109. doi: 10.1021/jacs.2c06920. [DOI] [PubMed] [Google Scholar]
  55. Zou W. Cheng Y. Ye Y. X. Wei X. Tong Q. Dong L. Ouyang G. Angew. Chem., Int. Ed. 2023;62:e202313392. doi: 10.1002/anie.202313392. [DOI] [PubMed] [Google Scholar]
  56. Zhang G. Wang T. Zhang M. Li L. Cheng D. Zhen S. Wang Y. Qin J. Zhao Z.-J. Gong J. Nat. Commun. 2022;13:7768. doi: 10.1038/s41467-022-35450-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Kou M. Wang Y. Xu Y. Ye L. Huang Y. Jia B. Li H. Ren J. Deng Y. Chen J. Zhou Y. Lei K. Wang L. Liu W. Huang H. Ma T. Angew. Chem., Int. Ed. 2022;61:e202200413. doi: 10.1002/anie.202200413. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Barman S. Singh A. Rahimi F. A. Maji T. K. J. Am. Chem. Soc. 2021;143:16284–16292. doi: 10.1021/jacs.1c07916. [DOI] [PubMed] [Google Scholar]
  59. Dai W. Wang P. Long J. Xu Y. Zhang M. Yang L. Zou J. Luo X. Luo S. ACS Catal. 2023;13:2513–2522. [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

SC-OLF-D6SC03539F-s001
SC-OLF-D6SC03539F-s001.pdf (5.3MB, pdf)

Data Availability Statement

The data supporting this manuscript are available from the corresponding author upon reasonable request.

Supplementary information (SI) is available. See DOI: https://doi.org/10.1039/d6sc03539f.

Articles from Chemical Science are provided here courtesy of Royal Society of Chemistry

RESOURCES