Achieving CO2 reduction with H2O on metal photocatalysts and understanding the corresponding mechanisms at the molecular level are challenging. Herein, we report that quantum-sized Au nanoparticles can photocatalytically reduce CO2 to CO with the help of H2O by electron-hole pairs mainly originating from interband transitions. Notably, the Au photocatalyst shows a CO production rate of 4.73 mmol g−1 h−1 (~100% selectivity), ~2.5 times the rate during CO2 reduction with H2 under the same experimental conditions, under low-intensity irradiation at 420 nm. Theoretical and experimental studies reveal that the increased activity is induced by surface Au–O species formed from H2O decomposition, which synchronously optimizes the rate-determining steps in the CO2 reduction and H2O oxidation reactions, lowers the energy barriers for the *CO desorption and *OOH formation, and facilitates CO and O2 production. Our findings provide an in-depth mechanistic understanding for designing active metal photocatalysts for efficient CO2 reduction with H2O.
Coupling CO2 reduction and H2O oxidation over photocatalysts using solar energy is an appealing strategy to alleviate the greenhouse effect and simultaneously produce value-added chemicals or fuels1,2. In recent decades, photocatalytic studies have almost exclusively focused on semiconductors, which provide an opportunity for reducing CO2 with H2O at room temperature3,4. However, the reaction rates of semiconductor photocatalysts are usually proportional to the square root of light intensity, thereby limiting their efficiency in practical applications, in which relatively high light intensity is required5.
Other than semiconductor photocatalysts, some plasmonic (such as Ag, Au, and Cu)6,7,8,9,10,11,12 or nonplasmonic (such as Ni and Ru)13,14 metal nanoparticles (NPs) have been recently reported to show photocatalytic activities for CO2 reduction. Under illumination, a collective oscillation of free electrons (conduction band electrons) is sometimes induced in metal NPs, a phenomenon defined as localized surface plasmon resonance (LSPR). Moreover, valence electrons in metal NPs can be excited to a much higher energy level through interband transitions15. Compared with the photoinduced electron-hole pairs in semiconductors, charge carriers produced in metal NPs usually possess sufficient energy for activating chemical bonds6,16. In addition, metal photocatalysts also have the ability to combine light and thermal energies to activate inert molecules in catalytic reactions15. Notably, a linear increase in reaction rates with increasing reaction temperature and light intensity was reported for metal photocatalysts, which sharply contrasts with semiconductor photocatalysts17. These intriguing properties make metal photocatalysts effective candidates for overcoming the severe challenges to photocatalytic CO2 reduction18.
However, for photocatalytic CO2 conversion on metal NPs, strong reductants (such as CH4, H2, isopropyl alcohol, and triethanolamine) are generally required7,19,20,21. For application purposes, using inexpensive and abundant H2O as an electron donor should be the direction and goal of photocatalytic CO2 reduction22. In this reaction, a series of complex and consecutive multi charge transfer processes followed by simultaneous CO2 reduction and H2O oxidation reactions at two different sites should occur under illumination3. The hot carriers generated from the LSPR effect of metal photocatalysts always have a much shorter lifetime than those from semiconductors23, which leads to difficulty in accomplishing CO2 reduction with H2O. To solve this problem, combining plasmonic metal nanostructures with semiconductors (such as TiO2) has been a widely used method by which the lifetime of hot charge carriers is prolonged24. In this situation, the metal NPs work as photosensitizers via the LSPR effect, and their low interfacial charge transfer efficiency greatly limits the application of these composite photocatalysts25.
In contrast to the LSPR effect, hot electron-hole pairs generated from the interband transitions of metal NPs always possess higher redox potentials and longer lifetimes, which are more conducive to driving the H2O oxidation and CO2 reduction reactions26. However, to the best of our knowledge, photocatalytic CO2 reduction with H2O has been rarely realized on metal-based photocatalysts by interband transitions. Moreover, molecular-level insights into the fundamentals of photocatalytic reactions on metal photocatalysts are still challenging to obtain and lacking, greatly restricting the rational design of high-efficiency metal photocatalysts. Gold (Au) NPs have received increasing attention for the photocatalytic conversion of CO2 into C1 or C2 hydrocarbons by the LSPR effect, which is mainly performed in the presence of H2 or sacrificial agents6,19,20,21,27. Recently, it has been proven that the interband absorption of Au NPs can dominate plasmon absorption with decreasing particle size28. Therefore, quantum-sized Au NPs with unique interband excitations under visible light irradiation may provide unprecedented possibilities for driving the photocatalytic conversion of CO2 with H2O.
In this work, quantum-sized Au NPs with a diameter of ~4 nm, which can be used as efficient photocatalysts for reducing CO2 to CO with H2O by interband transitions, are synthesized by a one-pot reduction approach. Molecular insight into the photocatalytic reaction mechanism reveals that the CO2 reduction and H2O oxidation reactions on Au followed the CO2→*COOH→*CO→CO(g) and H2O→*OH→*O−→*O→*OOH→O2 routes, respectively. Notably, the surface Au–O species formed from decomposing H2O are proven to be able to lower the activation energy of the CO2 reduction reaction (CO2RR) and improve the evolution of CO and O2 on the Au surface.
Characterizations of the Au NPs
The ~4 nm Au NPs were fabricated by reducing HAuCl4·xH2O with NaBH4 at room temperature. The representative transmission electron microscopy (TEM) image in Fig. 1a shows that the as-prepared Au NPs exhibit quasispherical morphology. As shown in the enlarged image (Supplementary Fig. 1), accidental aggregation of the Au NPs is observed, which is understandable considering that the surfactant-free synthesis strategy was used in the present study. The high-resolution TEM (HRTEM) image in Fig. 1b shows that the lattice spacing of the nanoparticle is 0.235 nm, consistent with the spacing of the (111) planes of metallic Au. As shown in Fig. 1c, the Au NPs display a narrow size distribution in the range of 4.0 ± 0.5 nm. X-ray photoelectron spectroscopy (XPS) was employed to study the valence states of the obtained samples (Fig. 1d). The binding energies centered at 83.6 and 87.2 eV prove the metallic character of the Au NPs29.
As revealed by the UV–Vis absorption spectrum in Fig. 1e, the Au NPs exhibit absorption in the wavelength range of 320–800 nm. A small absorption peak from 485 to 650 nm with maximum absorption at 520 nm is ascribed to the LSPR effect due to intraband absorption, which occurs by the excitation of free conduction electrons near the Fermi surface from 6sp-hybridized atomic orbitals of Au (Fig. 1f). In addition to absorption from the LSPR effect, a dominant absorption band with a tail that can reach 800 nm is observed in Fig. 1e, which can be attributed to the 5d–6sp interband transitions of Au NPs (Fig. 1f)26,30. According to theoretical calculations31, the interband absorption of Au NPs is derived from two types of excitations. As shown in Fig. 1f, one type of excitation occurs by the transition of electrons from the top of the 5d band to states just above EF (Fermi level) in the 6sp band with a threshold at 2.4 eV (λ < 517 nm, occurs near the L-point in the Brillouin zone), and the other is from the 5d band to unoccupied states in the 6sp band above EF with a tail at 1.8 eV (λ < 689 nm, occurs near the X-point in the Brillouin zone). Considering that minor discrepancies between the theoretical prediction and experimental observation should be acceptable, the intraband and interband absorption regions of Au are marked with different colors in Fig. 1e. As shown in Fig. 1e, the absorption of Au NPs in the range of 320–800 nm is attributed to the overlap of the LSPR intraband absorption and the interband absorption.
The photocatalytic activity of the as-obtained Au NPs towards CO2 reduction with H2O was evaluated in a steel reactor with a quartz window on top, with temperature-controlled by a heating element. To evaluate the contributions of the LSPR and interband transitions to photocatalysis, monochromatic light-emitting diode (LED) light sources at wavelengths in the visible region (from 365 to 620 nm) were used. The intensities of the LED light sources are summarized in Supplementary Table 1. Figure 2a shows the photocatalytic CO production activities at 200 °C under 420 nm LED light irradiation. The Au NPs show CO evolution up to 4.73 mmol g−1 h−1 accompanied by 1.98 mmol g−1 h−1 O2 evolution (based on a 4 h reaction), which are much higher than most of the semiconductor-based photocatalysts, even with the low-intensity LED light used in our study (Supplementary Table 2).
Figure 2b shows the apparent quantum efficiency (AQE) of the Au NPs for CO production under light with different wavelengths, which decreases with increasing wavelength in the range of 365 to 620 nm, except for a slight increase at 520 nm. The enhanced AQE at 520 nm should be ascribed to the improved light absorption due to the LSPR effect of the Au NPs32. Notably, the AQE is fairly low under irradiation ranging from 450 to 620 nm, which is unlike previous observations in which the action spectra were well-matched with the LSPR spectra of Au NPs32. This observation provides clear evidence that the CO evolution reaction was not primarily driven by the LSPR effect of Au26,30. In contrast, the main parts of the AQEs at 365 nm and 420 nm match well with the absorption induced by the interband transitions (Fig. 2b), proving that the photocatalytic CO2-to-CO conversion on Au NPs should be mainly attributed to the hot electrons from interband transitions26,30. Furthermore, the absorption cross-section of the Au NPs in the range of 320–485 nm increases at shorter wavelengths (Supplementary Fig. 2), which is in agreement with the AQE trend in Fig. 2b. This result further confirms that the photocatalytic CO2RR activity is correlated with the 5d–6sp interband transitions of Au33.
To gain insight into the dynamics of the photogenerated carriers, the time-resolved photoluminescence (TRPL) decay spectrum of Au NPs was obtained (Supplementary Fig. 3). The carrier lifetime of the Au sample was calculated to be approximately 0.2 ns, which is close to that of semiconductor photocatalysts and is long enough to drive photocatalytic reactions34. Interestingly, this lifetime is also much longer than that produced with the LSPR effect (approximately 30 fs)23, which is beneficial for the transfer of photogenerated charge carriers and thereby leads to improved photocatalytic activity on Au NPs. The TRPL result proves the superiority of the charge carriers induced by the interband transition of Au for photocatalysis.
As a control experiment, in the absence of catalyst or CO2, there no products were detected even after 4 h of reaction (Fig. 2c). To confirm the role of light and temperature in the catalytic CO2 reduction reactions, control experiments without light or heat input were also performed. As shown in Fig. 2c, negligible products were observed when CO2 reduction was performed at 200 °C on Au NPs without light, while CO production was only 0.181 mmol g−1 h−1 under continuous illumination with 420 nm LED light without additional heating. The above results reflect that CO evolution over Au is caused by CO2 reduction under contributions from both light and thermal energy.
As we know, thermal splitting of CO2 or H2O is only reportedly achieved on metal oxide-based catalysts by using oxygen vacancies as the reaction intermediate, which usually requires significantly high temperatures above 1000 °C35. Although thermal conversion of CO2 on Au can be achieved at approximately 200 °C, H2 must be used as a reducing agent in this reaction5. Therefore, thermocatalytic reduction of CO2 with H2O can hardly be achieved on Au at 200 °C (Fig. 2c). Furthermore, the photothermal effect on catalytic CO2 reduction on Au NPs can be neglected because 420 nm LED light with low intensity (73 mW cm–1, even lower than the solar intensity) was used in the present study. Under 420 nm LED light irradiation, the temperature of the sample can only increase from ambient temperature (24 °C) to 33 °C. Based on the above considerations, we can speculate that CO2-to-CO conversion on Au NPs should be reduced by the photogenerated electrons.
However, a linear dependence of the photocatalytic CO production rate on the reaction temperature at constant light intensity is shown in Fig. 2d, which indicates that the heat input can play a role in improving the photocatalytic CO2RR activity of Au. This phenomenon is possibly induced by the improved populations of the adsorbate in excited vibrational states at higher temperatures17. In the future, the influence of the reaction temperature on photocatalytic CO2 conversion on Au needs to be systematically studied. Moreover, a nearly linear increase in CO production with increasing light intensity is shown in Fig. 2e, which is a characteristic of reactions on metal photocatalysts17. This result confirms that the CO2RR on Au is in fact a light-driven process. The results in Fig. 2d and Fig. 2e make the Au NPs practical for increasing the reaction rate by increasing the temperature and light intensity, indicating that the Au NPs in our work can efficiently couple thermal and light energy to drive chemical transformations, which is in sharp contrast to semiconductor photocatalysts25.
To reveal the role of H2O in the photocatalytic CO2 reduction process over Au, desorption of absorbed H2O from the surface of Au was carried out by calcining the sample at 450 °C for 1 h in Ar gas, and the complete desorption of H2O was confirmed by FT-IR spectra (Supplementary Fig. 4). After that, photocatalytic CO2 reduction without H2O was conducted at 200 °C under 420 nm LED illumination, and hardly any CO evolution was detected (Fig. 2f, black curve). However, when H2O was added to the reactor, the photocatalytic CO evolution rate on Au reached 3.7 mmol g−1 h−1 (Fig. 2f, red curve), which was similar to the rate in Fig. 2a, confirming that H2O is essential for the CO2RR. Furthermore, 13CO2 and H218O isotopic labeling experiments confirm that CO production was derived from the reduction of CO2 along with the oxidation of H2O into O2 (Fig. 2g).
To prove that the Au sample has a suitable electronic band structure for the CO2 reduction and H2O oxidation reactions, the valence-band XPS spectra of the sample were examined. As shown in Fig. 2h, the zero point of the energy scale corresponds to the EF position. Two broad peaks at 3 eV and 6 eV result from the splitting of the 5d5/2 and 5d3/2 levels by spin-orbit interaction, while the 5d band edge of Au is approximately 1.34 eV36. Moreover, an electronic state induced by the 6sp electrons of Au tailing above EF reflects the metallic character of the Au NPs28. Based on the result in Fig. 2h, the electronic band structure of the Au NPs and the charge transfer processes involved in the photocatalytic CO2 reduction and H2O oxidation reactions are schematically shown in Fig. 2i. The value of hot electron energy (Ee,hot) is –3.49 eV when the Au NPs are illuminated by 420 nm LED light, which is more positive than the reduction potential of CO/CO2 (–3.92 eV vs. vacuum level)37. The value of the corresponding hot hole energy (Eh,hot) is determined to be –6.44 eV, which is more negative than the oxidation potential of O2/H2O (–5.67 eV vs. vacuum level)33. The details for calculating Ee,hot and Eh,hot are shown in the Method part. These results indicate that the Au NPs exhibit a suitable electronic band structure to achieve CO2 reduction and H2O oxidation simultaneously by interband transitions.
It is well known that selectivity is also a fundamental issue for the CO2RR with H2O due to the competing H2 evolution reaction (HER). In the present study, no H2 production on Au was detected during photocatalytic CO2 reduction with H2O (Supplementary Fig. 5). The results in Fig. 3c-d prove that the *COOH intermediate can be strongly adsorbed on the active sites of the Au NPs (discussed below), which is a possible reason for the HER being suppressed since the active sites are occupied by *COOH. On the other hand, the absorbed *COOH can increase the reaction barrier for the HER on Au NPs38, which can also contribute to the suppression of the HER. In addition, the weak affinity between Au and H might be another factor responsible for suppressing the HER on Au39. For the above three reasons, selective CO2 reduction with the inhibition of the side HER is achieved on the Au photocatalyst in the present study.
Fourier transform infrared spectroscopy studies
To elucidate the catalytic mechanism, in situ Fourier transform infrared (in situ FT-IR) spectroscopy was conducted to detect the surface-adsorbed species on Au during the photocatalytic process. As shown in Fig. 3a, a broad peak centered at 1650 cm−1 corresponding to the surface-adsorbed H2O appears with prolonged reaction time, suggesting that H2O was adsorbed on the Au NPs. As the reaction proceeded, a new peak at 1698 cm−1 that can be assigned to the vibration of carboxylate (CO2–) was observed, which represents the chemical adsorption state of CO2 on Au40. The interaction between the CO2 and H2O (and/or surface oxygen atom) coadsorbed on the Au surface leads to the monodentate carbonate (m-CO32−) peaks at 1510 cm−1 and 1380 cm−1 and bidentate bicarbonate (b-HCO32−) peak at 1458 cm−1 40,41. Moreover, two new peaks at 1545 and 1180 cm−1 that can be ascribed to the *COOH group (the asterisk denotes the surface adsorption site)40 appear with increasing reaction time, demonstrating that *COOH is a key intermediate of the CO2-to-CO conversion process. Moreover, the changes in the peak intensities of the critical intermediates with reaction time are shown in Fig. 3b, which increase with the extension of the reaction time from 0 to 15 min. After that, the peak intensities are slightly reduced with the reaction time being prolonged from 15 to 30 min and are nearly stable after 30 min of reaction. The changes in peak intensities may be understood by considering the cooperative effect between the H2O and CO2 adsorption process and the photocatalytic conversion process of the intermediate species during the reduction of CO2 on Au NPs.
As shown in Fig. 2a, the CO2-to-CO conversion rate is greatly reduced after 3 h of photocatalytic reaction. To study the reasons for this phenomenon, FT-IR spectra over long time ranges were tested (Fig. 3c). As shown in Fig. 3c, the intensities of the peaks belonging to *COOH increase gradually as the reaction time proceeds to 9 h, suggesting that an accumulation of *COOH on the Au surface occurs during the CO2RR process. XPS analysis was also conducted to further verify the gradual accumulation of *COOH species on the Au surface (Fig. 3d). For the Au catalyst after photoreactions, there is an increase in the C–O peak intensity at 286.2 eV and a simultaneous emergence of a peak at 288.5 eV, which can be attributed to the adsorbed intermediate *COOH on the Au surface during the photocatalytic CO2 reduction reaction42. This is basically consistent with the FT-IR result in Fig. 3c. *COOH accumulation might induce the deactivation of the Au photocatalysts observed in Fig. 2a by blocking active sites. To desorb the *COOH species, the Au sample after 12 h of CO2RR was outgassed under vacuum for 12 h, and the elimination of *COOH was verified by the FT-IR (Supplementary Fig. 6) and XPS (Supplementary Fig. 7) spectra. Then, the decarboxylated sample was reused to convert CO2 and H2O under the same reaction conditions as those in Fig. 2a, and the CO production activity recovered immediately. As shown in Fig. 3e, the Au NPs remained stable for at least 10 cycles over the entire 120 h reaction, indicating that the Au catalyst possesses good stability.
It is well known that the CO2RRs on metal photocatalysts were mainly performed in H2 gas in the previous reports5. Therefore, photocatalytic CO2 conversion on Au was also assessed by substituting H2O with H2 under the same conditions. Sample dehydration before the CO2RR was carried out to eliminate the influence of adsorbed H2O. With the help of H2, the CO yield on Au is 1.86 mmol g−1 h−1 (Fig. 3f), which is less than 1/2 of that formed in Fig. 2a. Gas chromatography-mass spectrometry (GC–MS) demonstrates that CO is the only reduced product in the CO2RR with H2 (Supplementary Fig. 8). This result clearly reveals that the unique quantum-sized Au NPs exhibit superiority in photocatalytic CO2 reduction with H2O. Therefore, the present work opens an exciting pathway to efficiently drive CO2 reduction with H2O on metal NPs by interband transitions, which is essential for fundamental research and applications. To understand this counterintuitive phenomenon, the role of H2O in CO2RR is systematically studied based on experiments and theoretical calculations in the following.
To study the reaction pathway for H2O during the CO2RR, 1H solid-state nuclear magnetic resonance (1H ssNMR) spectra of the Au NPs before and after the photocatalytic reaction were obtained. As shown in Fig. 4a, in the spectra of the fresh Au NPs, two different types of H species can be clearly observed. The broad peak with a chemical shift of 6.36 ppm is caused by highly immobilized H2O molecules with restricted mobility near the surface region43. The sharp peak at 3.43 ppm can be attributed to the weakly adsorbed H2O on the outer surface of the Au44,45. Furthermore, surface H2O species on Au can be also confirmed by thermogravimetric analysis (TGA). As shown in Fig. 4b, there are two major stages in weight loss from room temperature to 450 °C in the derivative thermogravimetry (DTG) curve. The first weight loss of ~2.3% occurred from the initial temperature to 258 °C, which is attributed to the sustained release of weakly adsorbed H2O (physisorbed water) from the surface of Au46. The second weight loss of ~1.2% in the temperature range of 258-450 °C can be attributed to the removal of highly immobilized water from the surface of Au47. This is well consistent with the 1H ssNMR results.
When the catalyst was subjected to photocatalytic CO2RR for 3 h, the peak at 6.36 ppm disappeared (Fig. 4a), suggesting that the immobilized H2O molecules were decomposed under illumination. Simultaneously, a dominant resonance centered at 3.96 ppm emerged in the red curve, which is ascribed to the overlapping signals between OH and free H2O, indicating the formation of an AuOH-H2O species on the Au surface by hydrogen bonds between AuOH groups and free H2O44. In addition, the peak centered at 1.47 ppm in the red curve can be related to the isolated *OH species (without hydrogen bonds) formed on the Au surface44. These results are very similar to the observations of the production of SiOH–H2O and isolated SiOH species on mesoporous silica NPs44.
Based on the 1H ssNMR spectra, it can be concluded that molecular H2O was dissociated into *OH species under photocatalytic reaction conditions and that partial *OH species were not bound to H2O. This is consistent with the in situ FT-IR characterizations in Fig. 4c, in which a stretching band (at 3740 cm−1) attributed to isolated *OH species on the Au surface was observed48. And a characteristic band at 1260 cm−1 that can be regarded as the vibration of the *OOH group appeared after the CO2RR (Fig. 3a), which is regarded as an important intermediate for O2 production in the H2O oxidation process49.
To study the species formed on Au during the photocatalytic process, quasi in situ XPS measurements were carried out. As shown in Fig. 4d, prior to irradiation, the high-resolution XPS spectrum of Au 4 f exhibits two peaks at 83.3 eV and 86.9 eV, which represent the Au 4f7/2 and Au 4f5/2 states of metallic Au (Au0)50, respectively. Upon irradiation at 200 °C, the peaks corresponding to the oxide state of Au (Au+) emerged at 84.8 eV and 88.4 eV50, which demonstrates that the Au–O species formed by H2O decomposition on Au during the reaction process. As shown in Fig. 4d, the concentration of the Au–O species gradually increased as the reaction time was prolonged from 1 h to 3 h, which coincides with the production of Au–O species in the photocatalytic CO2RR on Au.
To further confirm the formation of Au oxide species, electrochemical linear scan voltammetry (LSV) was performed on the Au samples before and after the photocatalytic reaction. As shown in Fig. 4e, compared to the fresh Au sample, a peak at approximately 0.8 V vs. the reversible hydrogen electrode (RHE) can be clearly observed for the Au sample after 3 h of photocatalytic reaction. As proven by the cyclic voltammetry (CV) curve of Au NPs (Supplementary Fig. 9), the onset potential during the back scan is not sufficient to oxidize Au. Therefore, the peak at 0.8 V can be attributed to the reduction of the generated Au oxide species during the photocatalytic process51. In addition, Raman spectroscopy was also used to characterize the Au–O species formed on the Au surface. As shown in Fig. 4f, a characteristic peak at 480 cm−1 appears during the photocatalytic CO2RR process, which can be attributed to the stretching vibrations of Au–O species52. Both the electrochemistry and Raman characterizations confirm that Au–O species formed on the surface of Au NPs during the photocatalytic CO2RR, which is consistent with the XPS result in Fig. 4d.
Furthermore, the high-resolution O 1 s XPS spectra are conducted to distinguish the oxygen species on Au surface. Supplementary Fig. 10 shows the changes of XPS spectra before and after reaction under 420 nm LED light irradiation. Two peaks at 531.7 eV and 533.1 eV in both samples are observed, which are attributed to the hydroxyl groups and surface adsorbed H2O molecules, respectively53. While the O 1 s spectrum of the Au sample after 3 h reaction shows two additional shoulder peaks at 530.0 eV and 535.0 eV, the former can be attributed to the Au–O species54, which is consistent with the XPS results of Au 4 f. The latter peak at 535.0 eV can be attributed to chemisorbed oxygen species from COOH or/and H2O55. As discussed above through 1H ssNMR, FT-IR, XPS, electrochemistry, and Raman analyses, we can confirm that Au–O species were formed on the Au surface in the photocatalytic CO2RR process by decomposing H2O. Thus, we speculate that the formed Au–O species might play a crucial role in CO2RR on Au. Meanwhile, the above results prove that the *COOH, *OH, and *OOH species are the main intermediates in photocatalytic CO2 reduction with H2O.
Based on the above experimental results, we employed density functional theory (DFT) calculations based on an electrochemical method to examine the possible CO2RR mechanisms occurring on the surface of Au. The experimental results above have proven that the photocatalytic CO2 reduction and H2O oxidation processes over Au NPs undergo pathways (Eqs. (1–9)) that are the same as the electrochemical reduction pathways of CO2 in H2O56. Furthermore, the Gibbs free energy diagram estimated by the electrochemical simulation method has been widely used in investigating the mechanism of photocatalytic CO2RR considering that the thermodynamic properties of the chemical reactions should not be changed by light stimulation9,41,57. Based on the above considerations, the electrochemical simulation method might be suitable to investigate the photocatalytic CO2RR mechanism on Au in terms of thermodynamics.
We first simulated the energetic barriers for the H2O dissociation reaction on Au (111) surface without photoinduced charges. As shown in Fig.5a, an extremely high energy barrier (1.35 eV) with an endothermic process (1.09 eV) is observed, indicating that it is difficult to produce H+ on the Au surface by direct H2O decomposition without the participation of photoinduced charges. Considering that CO2 reduction via the proton-coupled electron transfer process is favourable21, it was reasonable that the CO2RR on the Au surface was negligible in H2O without the presence of light since no available H+ and e− pairs were generated. However, according to the previous theoretical calculations, the energy barrier for H2O decomposition would be significantly reduced if the crossing of excited states is considered58,59. Therefore, both the presence of H2O and light illumination are critical for the CO2RR on Au, which is consistent with our experimental observations in Fig. 2c. Furthermore, according to the above 1H ssNMR, FT-IR, and XPS analyses, light irradiation is helpful for dissociating H2O into *OH and H*, then goes through the *O and *OOH intermediates and eventually forms O2. In this process, O bound to the Au surface inevitably formed under light irradiation. Therefore, the improved CO2RR activity on Au with H2O in Fig. 3f most likely resulted from the effect of surface Au–O species. As a result, a model of the Au surface with preadsorbed O was constructed to simulate the CO2RR with H2O.
To explore the effect of Au–O species on the CO2RR and oxygen evolution reaction (OER) activities, we calculated the free energy diagrams for CO and O2 evolution on Au with (O-preadsorbed Au) and without (clean Au) surface Au–O species. As shown in Fig. 5b, the electronic configuration regulated by the Au–O species can modify the adsorption of intermediates formed during the CO2RR. Furthermore, the required free energy of the rate-determining step (the first protonation process of CO2 to form *COOH) on O-preadsorbed Au (0.59 eV) is lower than that (the last step of *CO desorption) on clean Au (0.68 eV), implying the higher activity of the O-preadsorbed Au surface in the CO2RR. Figure 5b also shows that the affinity between CO and Au is weakened by the presence of the Au–O species. As shown in Fig. 5b, the reaction energy for *CO desorption is reduced from 0.68 eV to 0.33 eV in the presence of surface Au–O species, indicating that the adsorbed CO can be more easily eliminated on Au if H2O is used for CO2 reduction. This is also helpful for enhancing the CO2-to-CO conversion activity on Au. As shown in Fig. 5c, the thermodynamic barrier for OER (determined by the rate-determining step of *O + H2O→*OOH + H+ + e−) is greatly reduced on Au in the presence of surface Au–O species (1.59 eV) compared with a clean Au surface (2.64 eV), demonstrating that the preadsorbed O can act as an effective structure to boost the OER activity by weakening the adsorption of the formed intermediates in the H2O oxidation process.
To understand the mechanism by which the Au–O species enhance the above CO2RR and OER performances, we further examined the electronic structure of the Au centers in the Au–O species using Bader charge analysis. As shown in Fig. 5d, apparent charge transfer occurs on the O-preadsorbed Au surface by transferring the electrons from the d orbital of Au to O. As a result, the d electrons present on the Au atoms involved in adsorbing the intermediates (*COOH/*CO/*OH/O*/*OOH species) of the CO2RR and OER processes are reduced, verifying the superiority of Au–O for the enhancement of CO and O2 production. The relationship of catalytic activity, adsorption energy and d electrons obtained here is consistent with our previous studies60. Overall, based on the DFT calculation results, we can confirm that the generated Au–O species play a critical role in synchronously improving the CO2RR and OER activities on Au, which is consistent with experimental results in Fig. 3f.
According to experimental results and DFT calculations, a possible reaction mechanism for the photocatalytic reduction of CO2 with H2O is proposed as follows. Under illumination, hot electron-hole pairs are generated on the Au NPs through interband transitions (Fig. 5e). The photogenerated holes can efficiently dissociate H2O into O2 based on the following equations:
Unlike the H+ generated from direct H2O decomposition, the H+ released from Au–OH (Au–OH→Au–O− + H+) can possibly participate in photocatalytic CO2 reduction by the proton-coupled electron transfer process. In this process, negatively charged Au–O− species are left on the Au surface61. The formation of Au–O− species was confirmed by zeta potential measurements. Compared with the zeta potential of the fresh Au sample (–2.71 mV), the Au sample after 9 h of photocatalytic reaction possesses a much more negative potential (–14.3 mV), consistent with the formation of Au–O− species on the surface of Au in the photocatalytic process61. Moreover, the adsorbed CO2 on the Au NP surface reacts with photogenerated e− and H+ to yield CO based on the following equations:
Au NPs were routinely prepared by a one-step reduction method. In detail, 3 mL ice-cooled NaBH4 (98%, Aladdin) solution (0.1 M) was rapidly added into a 100 mL aqueous solution containing 0.25 mM HAuCl4·xH2O (99%, Adamas) under vigorous magnetic stirring. In this process, the color of the solution changed from yellow to wine red, which matched the formation of Au NPs. After stirring for 30 min, the obtained solution was aged for 24 h to completely hydrolyze the unreacted NaBH4. Then, the aged solution containing Au NPs was dialyzed in deionized water for a week to remove unnecessary ions, in which the water was exchanged for fresh water each day.
XPS and valence band XPS characterization was conducted with a ThermoFisher ESCALAB 250Xi instrument equipped with an excitation source of monochromatic aluminium, and the C1s peak at 284.6 eV can be regarded as the reference for the calibration binding energy. TEM and HRTEM images were acquired on an FEI Tecnai G2 F20 microscope working at 200 kV. The absorbance spectra of the samples were collected on a Shimadzu UV-2600 UV−Vis spectrophotometer, in which the dialyzed Au colloidal solution was dropped onto a sample plate with BaSO4 as the background, and the diffuse reflectance spectra were measured after drying. The absorption cross-section of 4 nm Au NPs was calculated using the method in a previous report33. Raman microscopy was performed using the emission of a 633 nm laser. The TRPL spectrum was obtained on an Edinburgh FS-5 spectrometer with an excitation wavelength of 340 nm. FT-IR was carried out on a Bruker VERTEX 70 V. TGA was measured on a TGA 5500 instrument with a heating rate of 10 °C min−1 under nitrogen flow. 1H ssNMR spectra were obtained on an Agilent 600 DD2 spectrometer at room temperature. The resonance frequency and the spinning rate used in this test were 600.19 MHz and 15 kHz, respectively. Zeta potential measurements were performed on a NAMO-ZS Zetasizer (Malvern, UK). Isotopic distributions of the gas products were analysed on a gas chromatograph-mass spectrometer (GCMS-QP2010 plus, Shimadzu).
Electrochemical measurements were performed in 0.1 M H2SO4 electrolyte at room temperature with an electrochemical workstation (CHI 760E, CH Instruments, Inc.) with a standard three-electrode system. The prepared sample, Pt wire, and a saturated Ag/AgCl electrode were used as working, counter, and reference electrodes, respectively. The working electrode was prepared by attaching Au samples onto a 1 \(\times\) 1 cm fluorine-doped tin oxide glass substrate through dilute Nafion solution (1%). CV and LSV curves were acquired from 0.6 V to 0 V vs. Ag/AgCl at a scan rate of 20 mV/s. The potentials vs. Ag/AgCl were converted to the RHE scale by the following equation:
In situ FT-IR measurements
In situ FT-IR measurements were carried out on a Bruker Vertex 70 v spectrometer equipped with a liquid nitrogen-cooled HgCdTe detector. All spectra were obtained with a resolution of 2 cm−1 and an accumulation time of 32 s. A photo of the instrument used for in situ FT-IR measurements is shown in Supplementary Fig. 11. Prior to the measurement, a vacuum pump was used to evacuate all the gases from the reaction chamber for 1 h. Then, CO2 and H2O vapour, as reaction gases, were introduced into the reaction chamber through the gas valve. The temperature of the reaction cell was raised to 200 °C with a heating rate of 5 °C min−1. After that, the background spectrum in the presence of the sample was collected. Next, the FT-IR spectra were recorded as a function of time, in which 420 nm LED light was used to illuminate the sample by a top quartz window.
Quasi in situ XPS measurements
Quasi in situ XPS spectra were acquired with a customized SPECS UHV system equipped with a SPECS PHOIBOS 150 analyser, a SPECS XR50 X-ray source, and a SPECS 1D DLD detector. The basic pressure of the analysis chamber was less than 3 × 10−10 mbar. The Au catalyst used for the XPS measurement was pressed onto stainless steel mesh. The CO2 photoreduction process was conducted in a load-lock chamber. After loading the catalyst, the reaction chamber was first treated with 5 mbar CO2 and H2O vapour and then heated to 200 °C for the photocatalytic reaction. In this process, 420 nm LED light was used to illuminate the sample. The distance between the light source and the sample was approximately 13.5 cm. After the photocatalytic reactions proceeded for different reaction times (1 h, 2 h, or 3 h), the load-lock chamber was pumped down to high vacuum (<10–9 mbar). Then, the sample was transferred to the analysis chamber through the sample transfer axis for XPS measurements without exposing the sample to air.
Photocatalytic activity tests
Photocatalytic CO2 conversion with H2O was carried out in a custom-built stainless-steel reaction chamber (450 mL) with a quartz window on top for light irradiation. The temperature of the reaction system was controlled by a heating element. Before the photocatalytic reaction test, 3 mL of the prepared Au colloidal solution (~100 mg L−1) was transferred into a quartz container and then dried in a vacuum freeze dryer overnight. The obtained Au photocatalyst was evenly distributed in the stainless-steel reaction chamber. After that, the gas-tight reactor was purged with Ar (99.999%) gas for 30 min to ensure air removal. Subsequently, 50 mL CO2 (99.999%) and 0.2 mL H2O were injected into the reactor. Monochromatic LED light sources at 365 nm, 420 nm, 450 nm, 520 nm, 590 nm and 620 nm (Perfectlight, Beijing) were used to drive photocatalytic CO2 conversion. The power of the light was measured with a PLS-LED100B photoradiometer (Perfect Light, Beijing, China). The gas products were analysed by a gas chromatograph (Agilent 7890B) equipped with two thermal conductivity detectors using He and Ar as the carrier gases.
The AQEs at different wavelengths were calculated by the following equation. The measured intensities of LED light at 365 nm, 420 nm, 450 nm, 520 nm, 590 nm, and 620 nm are shown in Supplementary Table 1. The reaction area was 9.0 cm2.
E e,hot and E h,hot calculations
The Ee,hot and Eh,hot generated under light excitation by the interband transitions of Au NPs can be calculated from the following two equations62:
where EF is the Fermi level of Au sample (–5.1 eV vs. vacuum level)33, and hv represents photon energy, which can be given by hv = 1240/λ. Meanwhile, the relationship between the standard hydrogen electrode (ESHE) and the vacuum energy level (Evac) is33:
Theoretical calculations were carried out using DFT with Vienna ab-initio simulation package63. The projector augmented wave method was adopted to describe the electron-ion interactions64. The exchange-correlation potentials were expressed by the Perdew–Burke–Ernzerhof functional with generalized gradient approximation65. The wave functions at each k-point were expanded with a plane wave basis set and a 400 eV cutoff energy was set. A 1 × 1 × 1 Monkhorst-Pack grid was selected to conduct the integration of the Brillouin zone66. The convergence threshold was set to 1.0 × 10−4 eV for electronic relaxation and 0.05 eV/Å for force. The Van der Waals (vdW) with vdW-DF approximation and spin polarization was involved in our current study67. The transition states were obtained by the climbing image nudged elastic band method68.
The Au(111) surface was constructed by cutting the bulk Au along (111) direction and 3 × 3 supercell with three layers was selected as the computational model. During the optimization, the atoms in the last layer were fixed to maintain the bulk structure and the other atoms were allowed to fully relax. A vacuum layer of 15 Å was used along the c direction normal to the surface to avoid periodic interactions.
The free energy diagrams of CO2RR and OER were estimated (15)
according to the method presented by Nørskov et al.69:
Where ΔE is the energy difference of reactants and products obtained by DFT calculations; ΔZPE and ΔS are the energy difference of zero-point energy and entropy, which are evaluated from the vibrational frequencies; T is the temperature, here 298.15 K is considered; The free energy of H2O in bulk water was calculated in the gas phase under a pressure of 0.035 bar (the equilibrium vapor pressure of H2O at 298 K). The free energy of O2 was obtained from the free energy change of the reaction O2 + 2H2→2H2O under the standard condition and the value of −4.92 eV was chosen. The free energy of (H++e−), according to a computational hydrogen electrode model proposed by Nørskov et al, was estimated as the energy of 1/2 H2 under standard conditions. The entropies and vibrational frequencies of the molecules in the gas phase were extracted from the NIST Chemistry WebBook (URL webbook.nist.gov, National Institute of Standards and Technology Chemistry WebBook).
Source data that support the findings of this study are provided in this paper, which can also be available from the corresponding author upon reasonable request. Source data are provided in this paper.
Gong, E. et al. Solar fuels: research and development strategies to accelerate photocatalytic CO2 conversion into hydrocarbon fuels. Energy Environ. Sci. 15, 880–937 (2022).
Lee, B.-H. et al. Electronic interaction between transition metal single-atoms and anatase TiO2 boosts CO2 photoreduction with H2O. Energy Environ. Sci. 15, 601–609 (2022).
Wang, Y. et al. Visible-light driven overall conversion of CO2 and H2O to CH4 and O2 on 3D-SiC@2D-MoS2 heterostructure. J. Am. Chem. Soc. 140, 14595–14598 (2018).
Iwase, A. et al. Water splitting and CO2 reduction under visible light irradiation using Z-Scheme systems consisting of metal sulfides, CoOx-Loaded BiVO4, and a reduced graphene oxide electron mediator. J. Am. Chem. Soc. 138, 10260–10264 (2016).
Zhang, X. et al. Product selectivity in plasmonic photocatalysis for carbon dioxide hydrogenation. Nat. Commun. 8, 14542 (2017).
Meng, X. et al. Nanometals for solar-to-chemical energy conversion: from semiconductor-based photocatalysis to plasmon-mediated photocatalysis and photo-thermocatalysis. Adv. Mater. 28, 6781–6803 (2016).
Yu, S., Wilson, A. J., Heo, J. & Jain, P. K. Plasmonic control of multi-electron transfer and C–C coupling in visible-light-driven CO2 reduction on Au nanoparticles. Nano Lett. 18, 2189–2194 (2018).
Li, S. et al. Recent advances in plasmonic nanostructures for enhanced photocatalysis and electrocatalysis. Adv. Mater. 33, 2000086 (2021).
Kumari, G., Zhang, X., Devasia, D., Heo, J. & Jain, P. K. Watching visible light-driven CO2 reduction on a plasmonic nanoparticle catalyst. ACS Nano 12, 8330–8340 (2018).
Lai, H. et al. Plasmon-induced carrier separation boosts high-selective photocatalytic CO2 reduction on dagger-axe-like Cu@Co core–shell bimetal. Chem. Eng. J. 417, 129295 (2021).
Cortés, E. & Camargo, P. H. Plasmonic Catalysis: From Fundamentals to Applications. (Wiley, 2021).
Devasia, D., Wilson, A. J., Heo, J., Mohan, V. & Jain, P. K. A rich catalog of C–C bonded species formed in CO2 reduction on a plasmonic photocatalyst. Nat. Commun. 12, 2612 (2021).
Anouar, A. et al. Remarkable activity of 002 facet of ruthenium nanoparticles grown on graphene films on the photocatalytic CO2 methanation. Adv. Sustain. Syst. 6, 2100487 (2022).
Lin, X. et al. Magnetic hollow spheres assembled from graphene-encapsulated nickel nanoparticles for efficient photocatalytic CO2 reduction. ACS Appl. Energy Mater. 2, 7670–7678 (2019).
Sarina, S. et al. Viable photocatalysts under solar-spectrum irradiation: nonplasmonic metal nanoparticles. Angew. Chem. Int. Ed. 53, 2935–2940 (2014).
Aslam, U., Rao, V. G., Chavez, S. & Linic, S. Catalytic conversion of solar to chemical energy on plasmonic metal nanostructures. Nat. Catal. 1, 656–665 (2018).
Christopher, P., Xin, H., Marimuthu, A. & Linic, S. Singular characteristics and unique chemical bond activation mechanisms of photocatalytic reactions on plasmonic nanostructures. Nat. Mater. 11, 1044–1050 (2012).
Kazuma, E. & Kim, Y. Mechanistic studies of plasmon chemistry on metal catalysts. Angew. Chem., Int. Ed. 58, 4800–4808 (2019).
Cui, X. et al. Turning Au nanoclusters catalytically active for visible-light-driven CO2 reduction through bridging ligands. J. Am. Chem. Soc. 140, 16514–16520 (2018).
Song, H. et al. Light-enhanced carbon dioxide activation and conversion by effective plasmonic coupling effect of Pt and Au nanoparticles. ACS Appl. Mater. Interfaces 10, 408–416 (2018).
Li, Y. et al. Boosting thermo-photocatalytic CO2 conversion activity by using photosynthesis-inspired electron-proton-transfer mediators. Nat. Commun. 12, 123 (2021).
Li, K., Peng, B. & Peng, T. Recent advances in heterogeneous photocatalytic CO2 conversion to solar fuels. ACS Catal. 6, 7485–7527 (2016).
Brown, A. M., Sundararaman, R., Narang, P., Goddard, W. A. & Atwater, H. A. Nonradiative plasmon decay and hot carrier dynamics: effects of phonons, surfaces, and geometry. ACS Nano 10, 957–966 (2016).
Wang, K. et al. Operando DRIFTS-MS investigation on plasmon-thermal coupling mechanism of CO2 hydrogenation on Au/TiO2: the enhanced generation of oxygen vacancies. Appl. Catal., B 296, 120341 (2021).
Li, Y. et al. Plasmonic hot electrons from oxygen vacancies for infrared light-driven catalytic CO2 reduction on Bi2O3−x. Angew. Chem., Int. Ed. 60, 910–916 (2021).
Zhao, J. et al. A comparison of photocatalytic activities of gold nanoparticles following plasmonic and interband excitation and a strategy for harnessing interband hot carriers for solution phase photocatalysis. ACS Cent. Sci. 3, 482–488 (2017).
Yu, S. & Jain, P. K. Plasmonic photosynthesis of C1–C3 hydrocarbons from carbon dioxide assisted by an ionic liquid. Nat. Commun. 10, 2022 (2019).
Balamurugan, B. & Maruyama, T. Evidence of an enhanced interband absorption in Au nanoparticles: size-dependent electronic structure and optical properties. Appl. Phys. Lett. 87, 143105 (2005).
Yan, J., Shan, L., Gu, X., Zhang, X. & Cai, J. Steric effects of mesoporous silica supported bimetallic Au-Pt catalysts on the selective aerobic oxidation of aromatic alcohols. Catalysts 10, 1192 (2020).
Wang, W., Fang, J. & Huang, X. Different behaviors between interband and intraband transitions generated hot carriers on g-C3N4/Au for photocatalytic H2 production. Appl. Surf. Sci. 513, 145830 (2020).
Beversluis, M. R., Bouhelier, A. & Novotny, L. Continuum generation from single gold nanostructures through near-field mediated intraband transitions. Phys. Rev. B 68, 115433 (2003).
Chen, L.-W. et al. Metal–organic framework membranes encapsulating gold nanoparticles for direct plasmonic photocatalytic nitrogen fixation. J. Am. Chem. Soc. 143, 5727–5736 (2021).
Kim, Y., Smith, J. G. & Jain, P. K. Harvesting multiple electron–hole pairs generated through plasmonic excitation of Au nanoparticles. Nat. Chem. 10, 763–769 (2018).
Miao, Z., Wang, Q., Zhang, Y., Meng, L. & Wang, X. In situ construction of S-scheme AgBr/BiOBr heterojunction with surface oxygen vacancy for boosting photocatalytic CO2 reduction with H2O. Appl. Catal., B 301, 120802 (2022).
Boretti, A. Technology readiness level of solar thermochemical splitting cycles. ACS Energy Lett. 6, 1170–1174 (2021).
Ohta, T. et al. Size and structure dependence of electronic states in thiolate-protected gold nanoclusters of Au25(SR)18, Au38(SR)24, and Au144(SR)60. J. Phys. Chem. C. 117, 3674–3679 (2013).
Wu, J., Huang, Y., Ye, W. & Li, Y. CO2 reduction: from the electrochemical to photochemical approach. Adv. Sci. 4, 1700194 (2017).
Lee, J. H. et al. Understanding the role of functional groups in polymeric binder for electrochemical carbon dioxide reduction on gold nanoparticles. Adv. Funct. Mater. 28, 1804762 (2018).
Hinnemann, B. et al. Biomimetic hydrogen evolution: MoS2 nanoparticles as catalyst for hydrogen evolution. J. Am. Chem. Soc. 127, 5308–5309 (2005).
Wu, J. et al. Efficient visible-light-driven CO2 reduction mediated by defect-engineered BiOBr atomic layers. Angew. Chem. Int. Ed. 57, 8719–8723 (2018).
Li, J. et al. Interfacial engineering of Bi19Br3S27 nanowires promotes metallic photocatalytic CO2 reduction activity under near-infrared light irradiation. J. Am. Chem. Soc. 143, 6551–6559 (2021).
Lu, L. et al. Surface chemistry imposes selective reduction of CO2 to CO over Ta3N5/LaTiO2N photocatalyst. J. Mater. Chem. A 6, 14838–14846 (2018).
Zhu, L. et al. Characterization of the mobility and reactivity of water molecules on TiO2 nanoparticles by 1H solid-state nuclear magnetic resonance. ACS Appl. Mater. Interfaces 5, 10352–10356 (2013).
Trébosc, J., Wiench, J. W., Huh, S., Lin, V. S. Y. & Pruski, M. Solid-state NMR study of MCM-41-type mesoporous silica nanoparticles. J. Am. Chem. Soc. 127, 3057–3068 (2005).
Nosaka, A. Y., Fujiwara, T., Yagi, H., Akutsu, H. & Nosaka, Y. Photoinduced changes of surface and adsorbed water in TiO2 photocatalytic systems as studied by solid state 1H-NMR spectroscopy. Chem. Lett. 31, 420–421 (2002).
Elkhalifah, A. E. I., Murugesan, T. & Bustam, M. A. Characterization of different cationic forms of montmorillonite by FTIR, XRD and TGA techniques. In Proceedings of the 2011 IEEE National Postgraduate Conference (NPC), Kuala Lumpur, Malaysia, 1–6 (2011).
Saavedra, J., Doan, H. A., Pursell, C. J., Grabow, L. C. & Chandler, B. D. The critical role of water at the gold-titania interface in catalytic CO oxidation. Science 345, 1599–1602 (2014).
Panda, R. N., Hsieh, M. F., Chung, R. J. & Chin, T. S. FTIR, XRD, SEM and solid state NMR investigations of carbonate-containing hydroxyapatite nano-particles synthesized by hydroxide-gel technique. J. Phys. Chem. Solids 64, 193–199 (2003).
Zhang, R. et al. Insight into the effective aerobic oxidative cross-esterification of alcohols over Au/porous boron nitride catalyst. ACS Appl. Mater. Interfaces 11, 46678–46687 (2019).
Tian, B., Zhang, J., Tong, T. & Chen, F. Preparation of Au/TiO2 catalysts from Au(I)–thiosulfate complex and study of their photocatalytic activity for the degradation of methyl orange. Appl. Catal., B 79, 394–401 (2008).
Xu, C. et al. Low temperature CO oxidation over unsupported nanoporous gold. J. Am. Chem. Soc. 129, 42–43 (2007).
Pfisterer, J. H. K. et al. Role of OH intermediates during the Au oxide electro-reduction at low pH elucidated by electrochemical surface-enhanced raman spectroscopy and implicit solvent density functional theory. ACS Catal. 10, 12716–12726 (2020).
Al-Ajlony, A.-M. B., Kanjilal, A., Harilal, S. S. & Hassanein, A. Carbon contamination and oxidation of Au surfaces under extreme ultraviolet radiation: an X-ray photoelectron spectroscopy study. J. Vac. Sci. Technol., B 30, 041603 (2012).
Yeo, B. S., Klaus, S. L., Ross, P. N., Mathies, R. A. & Bell, A. T. Identification of hydroperoxy species as reaction intermediates in the electrochemical evolution of oxygen on gold. ChemPhysChem 11, 1854–1857 (2010).
Xu, B., Duan, H., Chu, M., Cao, G. & Yang, Y. Facile synthesis of nitrogen-doped porous carbon for supercapacitors. J. Mater. Chem. A 1, 4565–4570 (2013).
Mi, Y. et al. Cobalt−iron oxide nanosheets for high-efficiency solar-driven CO2−H2O coupling electrocatalytic reactions. Adv. Funct. Mater. 30, 2003438 (2020).
Li, Y. et al. Facile Top-down strategy for direct metal atomization and coordination achieving a high turnover number in CO2 photoreduction. J. Am. Chem. Soc. 142, 19259–19267 (2020).
Martirez, J. M. P. & Carter, E. A. Excited-state N2 dissociation pathway on Fe-functionalized Au. J. Am. Chem. Soc. 139, 4390–4398 (2017).
Zhang, X.-G. et al. Light driven mechanism of carbon dioxide reduction reaction to carbon monoxide on gold nanoparticles: a theoretical prediction. J. Phys. Chem. Lett. 12, 1125–1130 (2021).
Xiao, M. et al. Engineering energy level of metal center: Ru single-atom site for efficient and durable oxygen reduction catalysis. J. Am. Chem. Soc. 141, 19800–19806 (2019).
Sylvestre, J.-P., Kabashin, A. V., Sacher, E., Meunier, M. & Luong, J. H. T. Stabilization and size control of gold nanoparticles during laser ablation in aqueous cyclodextrins. J. Am. Chem. Soc. 126, 7176–7177 (2004).
Govorov, A. O., Zhang, H. & Gun’ko, Y. K. Theory of photoinjection of hot plasmonic carriers from metal nanostructures into semiconductors and surface molecules. J. Phys. Chem. C. 117, 16616–16631 (2013).
Kresse, G. & Furthmüller, J. Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set. Phys. Rev. B 54, 11169–11186 (1996).
Kresse, G. & Joubert, D. From ultrasoft pseudopotentials to the projector augmented-wave method. Phys. Rev. B 59, 1758–1775 (1999).
Perdew, J. P., Burke, K. & Ernzerhof, M. Generalized gradient approximation made simple. Phys. Rev. Lett. 77, 3865–3868 (1996).
Monkhorst, H. J. & Pack, J. D. Special points for Brillouin-zone integrations. Phys. Rev. B 13, 5188–5192 (1976).
Dion, M., Rydberg, H., Schröder, E., Langreth, D. C. & Lundqvist, B. I. Van der Waals density functional for general geometries. Phys. Rev. Lett. 92, 246401 (2004).
Henkelman, G., Uberuaga, B. P. & Jónsson, H. A climbing image nudged elastic band method for finding saddle points and minimum energy paths. J. Chem. Phys. 113, 9901–9904 (2000).
Nørskov, J. K. et al. Origin of the overpotential for oxygen reduction at a fuel-cell cathode. J. Phys. Chem. B 108, 17886–17892 (2004).
Y. Li acknowledges the support from the National Natural Science Foundation of China (Grant No. 21972082). Y. Wang thanks the support from the National Natural Science Foundation of China (Grant Nos. 21831003, 21673220), and the Jilin Province Science and Technology Development Program (20200201001JC).
The authors declare no competing interests.
Peer review information
Nature Communications thanks Yi Zhao and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Peer reviewer reports are available.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
About this article
Cite this article
Shangguan, W., Liu, Q., Wang, Y. et al. Molecular-level insight into photocatalytic CO2 reduction with H2O over Au nanoparticles by interband transitions. Nat Commun 13, 3894 (2022). https://doi.org/10.1038/s41467-022-31474-2