Excitonic Au₄Ru₂(PPh₃)₂(SC₂H₄Ph)₈ cluster for light-driven dinitrogen fixation

Abstract

The surface plasmon resonance of metal nanoparticles has been widely used to improve photochemical transformations by plasmon-induced charge transfer. However, it remains elusive for the molecular-like metal clusters with non-metallic or excitonic behavior to enable light harvesting including electron/hole pair production and separation. Here we report a paradigm for solar energy conversion on an atomically precise Au₄Ru₂ cluster supported on TiO₂ with oxygen vacancies, in which the electron–hole pairs can be directly generated from the excited Au₄Ru₂ cluster and the TiO₂ support, and the photogenerated electrons can transfer to the Ru atoms. Importantly, the Ru atoms in the Au₄Ru₂ cluster are capable of injecting the electrons into adsorbed N₂ to activate N₂ molecules. The cooperative effect in the supported Au₄Ru₂ catalyst efficiently boosts the photocatalytic activity for N₂ fixation in comparison with homogold (Au_n) clusters.

---

Introduction

Atomically precise metal clusters with exact formulas, molecular purity, and total structures have gathered momentum in recent years, owing to their unique physical and chemical properties.^1–6 The metal clusters in the quantum size regime possess discrete electron energy levels and show non-metallic or excitonic behaviours,⁷ which are totally different from the larger metallic-state nanoparticles exhibiting a distinct surface plasmon resonance.^8,9 Significant advances in chemical synthesis of the clusters provide an exciting opportunity to unveil previously unknown or inaccessible insights into the applications in optics, catalysis, and biochemistry.^10,11 Especially the cluster-based heterogeneous catalysts have exhibited new catalytic properties in many chemical reactions compared to the plasmonic metal nanoparticles.^12,13 Furthermore, the precise relation of the properties with atomic-level structures not only reveals the origin of metal catalysis, but also promotes the exploration of important chemical processes with these clusters as well-defined, highly efficient catalysts.^14,15

One important chemical process is the reduction of dinitrogen to ammonia, which is an essential chemical and energy carrier.¹⁶ However, high temperature and pressure are necessary to drive the reaction of N₂ with H₂ to NH₃, due to the strong nonpolar triple bond in N₂ and its high activation barrier.¹⁷ Many efforts have been made to develop less energy-consuming alternatives that can overcome the kinetic limitation of NH₃ production.^18–21 Inspired by nitrogenase enzymes that can fix nitrogen under ambient conditions, nanostructured metal catalysts are springing up to enable N₂ fixation with the help of photosynthesis.^22–25 Despite the important advances in homogeneous complex systems, construction of heterogeneous metal sites for N₂ fixation is currently still challenging. Considering that atomically precise metal clusters can bridge the gap between homogeneous and heterogeneous catalysts, we speculate whether metal clusters with excitonic behaviour can convert N₂ into ammonia under mild conditions, that is, whether these clusters are capable of generating hot electrons driven by solar light, ensuring the charge separation, and then donating electrons into N₂. If this scenario is feasible, it can not only unravel the mystery of non-plasmon-induced solar energy conversion but also offer fundamental insights into exact, heterogeneous metal sites to govern the N₂ transformation at the unprecedented level of atomic precision.

Since no example of N₂ conversion on atomically precise metal clusters has been documented, a series of ligand-protected Au_n (n = gold atom number) clusters with different atomic structures were first screened to catalyse the N₂ conversion under light irradiation. As shown in Fig. 1, these Au_n clusters failed to give a convincing activity for photocatalytic reaction of N₂ fixation, mainly because the N₂ molecule cannot coordinate to clusters and be activated on the gold sites of the Au_n clusters, according to our density functional theory (DFT) calculations (inset of Fig. 1). We next turned our attention to the bimetal clusters. Since Ru is recognized as a suitable candidate for N₂ fixation,²⁴ we sought to explore the wet chemical synthesis of atomically precise Au–Ru clusters.

Fig. 1 The photocatalytic activity of the ligand-protected Au_n clusters supported on TiO₂-Ov for N₂ reduction. The inset shows the atomic structures of the Au_n clusters, which can only physisorb N₂ molecule with distance between N₂ and Au sites larger than 3.2 Å. The C, N, S, Au, P and Cl are shown in gray, blue, yellow, orange, green and violet colors, respectively. H atoms are omitted for clarity.

In this work, we successfully synthesized a new Au₄Ru₂ cluster protected by thiolate and triphenylphosphine ligands (namely, Au₄Ru₂(PPh₃)₂(SC₂H₄Ph)₈) and solved its crystal structure. Excitingly, the Au₄Ru₂ cluster supported on TiO₂ with oxygen vacancies (hereafter denoted as TiO₂-Ov) exhibited a drastic increase in the photocatalytic activity for N₂ reduction compared to the supported Au_n catalysts. Furthermore, we explicitly demonstrated the cooperative mechanism within the Au₄Ru₂/TiO₂-Ov catalyst for achieving light-driven N₂ fixation.

Results and discussion

The crystal structure of the Au₄Ru₂(PPh₃)₂(SC₂H₄Ph)₈ cluster is shown in Fig. 2A. This cluster resembles a distorted hexahedron, in which four Au atoms are located at the midpoints of four side edges, two Ru atoms reside on the centres of the top and the bottom planes, and eight S atoms are fixed at the vertexes. The two apex Ru atoms are coordinated by two PPh₃ with the average Ru–P bond length of 2.204 Å. Four S atoms binding to a Ru atom are within the same plane as indicated by the average S–Ru–S angle of 90°. The average S–Au–S angle is 172.9°, where S–Au bond distances are 2.310 and 2.318 Å, respectively. The Au–Au distances fall in a very narrow range of 3.045–3.144 Å, which are shorter than the sum of van der Waals radii of two Au atoms (3.32 Å), suggesting the presence of d¹⁰–d¹⁰ metallophilic contact within the Au₄Ru₂ cluster.²⁶ Electrospray ionization mass spectrometry (ESI-MS) further confirmed the formula of the cluster, where the m/z 2745 peak was assigned to [Au₄Ru₂(PPh₃)₂(SC₂H₄Ph)₈ + Cs]⁺ adduct supported by the agreement between experimental and simulated isotopic patterns (Fig. 2B). Thermogravimetric analysis further confirmed that the Au₄Ru₂ cluster was highly pure (Fig. S1†).

Fig. 2 (A) Atomic structure (H atoms are omitted for clarity) and (B) ESI-MS profile of Au₄Ru₂(PPh₃)₂(SC₂H₄Ph)₈. (C) UV-vis spectrum of Au₄Ru₂(PPh₃)₂(SC₂H₄Ph)₈. (D) UV-vis spectrum plotted on the photon energy scale. (E) Kohn–Sham orbitals (bottom panel) and density of states (top panel) of Au₄Ru₂(PPh₃)₂(SC₂H₄Ph)₈ from DFT calculations and the corresponding charge density distributions of HOMO and LUMO (insets).

UV-vis absorption spectrum of the Au₄Ru₂ cluster shows two prominent peaks at 349 and 640 nm and one weak peak at 453 nm (Fig. 2C), corresponding to excitation energies of 3.55, 1.94 and 2.74 eV (Fig. 2D), respectively. Accordingly, the optical gap is determined to be 1.33 eV based on the photon-energy scale spectrum, which is basically consistent with the computed gap of 1.25 eV between the highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) from our DFT calculations (Fig. 2E). As revealed by the computed electronic density of states (DOS) in Fig. 2E, the emergence of discrete electronic states and a moderate HOMO–LUMO gap for the Au₄Ru₂ cluster indicates single electron excitations, that is an exciton.⁷ Importantly, the projected DOS shows that the low-lying unoccupied states (LUMO and LUMO+1) are mostly localized on the Ru atoms, suggesting that the excited carriers in the Au₄Ru₂ cluster will be on the Ru atoms, which may act as reaction centers and utilize the excess electrons for N₂ activation.

With the newly synthesized cluster, we explored the proposed light-driven N₂ fixation using Au₄Ru₂ as a heterogeneous catalyst. As shown in Fig. 3A, the Au₄Ru₂/TiO₂-Ov catalyst gave rise to an ammonia production rate of 44.5 μmol g⁻¹ h⁻¹ under full spectrum illumination, which exhibited over 3-fold increase in photocatalytic activity compared to the Au_n/TiO₂-Ov and pure TiO₂-Ov catalysts. As much, the Au₄Ru₂/TiO₂-Ov catalyst resulted in a 4-time higher activity than TiO₂-Ov in visible light-driven N₂ reduction (Fig. 3A), suggesting a strong synergistic effect between Au₄Ru₂ and TiO₂-Ov. Time-dependent photocatalytic ammonia production over the Au₄Ru₂ catalysts revealed that, not only the ammonia concentration increased linearly with the irradiation time in the visible light region (Fig. 3B), but also the Au₄Ru₂ loaded on the TiO₂ support without abundant oxygen vacancies gave a much lower activity driven by either UV-vis or visible light (Fig. 3A and B). In fact, both TiO₂-Ov and TiO₂ substrates were in anatase phase (Fig. S2A†). The difference in the two TiO₂ samples was that the former contained oxygen vacancies, but the latter not, which was confirmed by electron paramagnetic resonance (EPR). TiO₂-Ov showed a characteristic EPR signal at approximately g = 1.998, suggesting the presence of oxygen vacancies,²⁷ whereas no EPR signal was observed on the other TiO₂ sample (Fig. S2B†). It can be conjectured that the abundant oxygen vacancies in TiO₂ facilitate the photochemical reaction of N₂ reduction.¹⁷

Fig. 3 (A) The photocatalytic performances of the Au₄Ru₂/TiO₂-Ov, Au₄Ru₂/TiO₂, TiO₂-Ov and TiO₂ catalysts for N₂ reduction. (B) Time-dependent photocatalytic ammonia production over the Au₄Ru₂/TiO₂-Ov, Au₄Ru₂/TiO₂, TiO₂-Ov and TiO₂ catalysts. (C) Reaction energy diagrams of water dissociation on the anatase TiO₂(101) surface without defect (TiO₂) and with an oxygen vacancy per supercell on surface (TiO₂-Ov₁) and subsurface (TiO₂-Ov₂), respectively. ΔE is relative to the energy of each substrate plus a free water molecule. Insets are the structures of water adsorption (top) and dissociation into *H and *OH species adsorbed on TiO₂-Ov₁ (bottom), and the transition state (middle). The H, O and Ti are shown in cyan, red and yellow. (D) UV-DRS and quantum efficiency of Au₄Ru₂/TiO₂-Ov for NH₄⁺ evolution by photocatalytic N₂ reduction under monochromatic light of different wavelengths. (E) In situ FTIR spectra of the photocatalytic N₂ reduction over Au₄Ru₂/TiO₂-Ov during full spectrum illumination in the presence of N₂ and water. (F) Performances of the Au₄Ru₂ clusters protected by different ligands for photocatalytic N₂ reduction.

DFT calculations demonstrated that TiO₂ with oxygen vacancies can efficiently promote the photolysis of water to produce hydrogen as the proton source of ammonia (Fig. S3†). The anatase TiO₂(101) surfaces with an oxygen vacancy on the surface (TiO₂-Ov₁) and subsurface (TiO₂-Ov₂), have low kinetic barriers (ΔE_a) of 0.09 and 0.02 eV for water dissociation, respectively, compared to 0.23 eV for the perfect TiO₂(101) surface (Fig. 3C). Moreover, they all provide moderate binding strength with H atoms (binding energy ΔE_H = 0.14–0.43 eV relative to the energy of H₂ molecule), which is beneficial for protons transfer from TiO₂ to Au₄Ru₂. Photocatalytic N₂ fixation on the Au₄Ru₂/TiO₂-Ov catalyst in CH₃CN solvent did not produce ammonia, again corroborating the origin of protons in ammonia from water splitting. For comparison, the Au₄Ru₂ cluster supported on SiO₂ gave a low ammonia production rate of 2.4 μmol g⁻¹ h⁻¹, implying the key role of TiO₂-Ov in water splitting.

Furthermore, the action spectrum for NH₃ formation on the Au₄Ru₂/TiO₂-Ov catalyst was determined under monochromatic light irradiation at wavelengths of 334, 420, 520, 600, and 700 nm. The trend of apparent quantum efficiencies (AQEs) well matched that of the optical absorption spectrum of the Au₄Ru/TiO₂-Ov (Fig. 3D). This proved that the photocatalytic N₂ fixation originated from the light absorption by the Au₄Ru₂ cluster. In addition, the catalytic activity decreased slightly with multiple cycles (Fig. S4†), mainly due to the partial detachment of Au₄Ru₂ from TiO₂-Ov (∼8 wt% metal loss detected by inductively coupled plasma-atomic emission spectroscopy (ICP-AES) analysis). The diffuse reflectance optical spectra (DRS) of the Au₄Ru₂/TiO₂-Ov sample did not significantly change after the reaction (Fig. S5A and B†) and transmission electron microscopy (TEM) studies showed that the spent catalyst had no obvious aggregation (Fig. S5C and D†), suggesting that the Au₄Ru₂ cluster was robust throughout the reaction.

To directly visualize the N₂ conversion on the Au₄Ru₂/TiO₂-Ov catalyst, in situ infrared Fourier transform (DRIFT) spectroscopy was utilized to monitor the time-dependent change of the functional nitrogenous intermediates on the surface of Au₄Ru₂/TiO₂-Ov. No signal change was observed in the DRIFT spectra within 30 min of incident light exposure in the absence of water (Fig. S6†), suggesting that the H atoms in ammonia indeed came from water. After water was introduced into the reaction cell, several absorption peaks appeared gradually with the irradiation time (Fig. 3E). The broad band at 3590 cm⁻¹ is assigned to the ν(N–H) stretching vibration, and the two absorption bands at 1705 and 1559 cm⁻¹ are attributed to the σ(N–H) bending vibration.^25,28 Besides, the bands at 1405 and 2912 cm⁻¹ assigned to the NH₄⁺ deforming vibration became stronger gradually with the irradiation time.^29,30 The result validated that the Au₄Ru₂/TiO₂-Ov catalyst can convert N₂ into ammonia under the light irradiation.

Considering that the Au₄Ru₂ cluster contained thiolate and PPh₃ ligands, it is natural to ask whether the ligands can affect the catalytic conversion of N₂. To address this, the comparison experiments were conducted, where a series of Au₄Ru₂ clusters protected by different ligands were prepared (Fig. S7†). As shown in Fig. 3F, the Au₄Ru₂ clusters with different ligands showed no drastic difference in the photocatalytic performance for N₂ reduction, manifesting that the catalytic reaction was mainly determined by metal sites, rather than the ligands. Furthermore, when all the ligands in Au₄Ru₂ were removed via the thermal treatment, the Au₄Ru₂ clusters crashed into large nanoparticles and hence lost the activity for N₂ fixation (Fig. S8†).

We now discuss the mechanism that Au₄Ru₂/TiO₂-Ov can achieve an extraordinary activity for photocatalytic N₂ reduction, but Au_n/TiO₂-Ov and TiO₂-Ov cannot. Mott–Schottky (M–S) plots were first collected to provide the flat band potentials of Au₄Ru₂ and TiO₂-Ov. The obtained tangent positive slopes indicated that both Au₄Ru₂ and TiO₂-Ov were likely n-type semiconductors (Fig. 4A). The flat band potentials of Au₄Ru₂ and TiO₂-Ov versus the saturated Ag/AgCl were −0.18 and −0.38 V, respectively. Based on the UV-DRS and M–S measurements, the band alignments of Au₄Ru₂ and TiO₂-Ov were shown in Fig. 4B. From the thermodynamic point of view, photogenerated electron carriers can transfer from TiO₂-Ov to Au₄Ru₂, while the Au₄Ru₂ cluster was more capable of light-driven N₂ reduction than the TiO₂-Ov substrate.

Fig. 4 (A) Mott–Schottky plots of Au₄Ru₂ and TiO₂-Ov (potential V vs. Ag/AgCl at pH 6.4). (B) Schematic diagram of band alignment (potential V vs. Ag/AgCl at pH 6.4). VB means valence band, and CB denotes conduction band. (C) Time-resolved PL spectra of the Au₄Ru₂/TiO₂-Ov and TiO₂-Ov under Ar and N₂ atmospheres, respectively. (D) Photocurrent responses of Au₄Ru₂/TiO₂-Ov, Au₂₄/TiO₂-Ov, Au₂₅/TiO₂-Ov and TiO₂-Ov under Ar and N₂ atmospheres, respectively. (E) Electrochemical impedance spectra of Au₄Ru_2/TiO₂-Ov, Au₂₄/TiO₂-Ov, Au₂₅/TiO₂-Ov and TiO₂-Ov in the presence of 300 W xenon lamp. (F) Proposed mechanism for the photocatalytic N₂ reduction on the Au₄Ru₂/TiO₂-Ov catalyst.

The charge carrier kinetics of the Au₄Ru₂ cluster with N₂, including separation, transfer and recombination, was investigated by room temperature steady-state and time-resolved photoluminescence (PL) spectroscopy. As shown in Fig. S9,† when the Ar atmosphere was changed to the N₂ atmosphere, the steady-state PL spectrum of Au₄Ru₂/TiO₂-Ov was significantly quenched, which was related to the non-radiative transfer of the photoexcited electrons from Au₄Ru₂ to the π* antibonding orbitals of N₂ adsorbed on the cluster.³⁰ The time-resolved PL spectroscopy studies (Fig. 4C) showed that the average decay time (τ = 14.40 ns) of Au₄Ru₂/TiO₂-Ov was longer than that of TiO₂-Ov (τ = 5.26 ns). The prolonged lifetime of the photogenerated electrons illustrated that Au₄Ru₂ supported on TiO₂-Ov can reduce the charge recombination, thereby possess highly effective separation of electron–hole pairs.³¹

We argued that the high efficiency of N₂ fixation on the Au₄Ru₂/TiO₂-Ov catalyst was acquired by a synergic effect between Au₄Ru₂ and TiO₂-Ov. To confirm this, transient photocurrent responses were conducted on the Au₄Ru₂/TiO₂-Ov, TiO₂-Ov, and two reference systems (Au₂₄/TiO₂-Ov and Au₂₅/TiO₂-Ov) under the Ar and N₂ atmospheres with light (Fig. 4D), respectively. Compared to TiO₂-Ov, the photocurrents of Au₂₄/TiO₂-Ov and Au₂₅/TiO₂-Ov were enhanced under the Ar atmosphere, suggesting that the Au_n clusters also can serve as trapping sites for the photogenerated electrons.³⁰ However, the transient photocurrent responses of TiO₂-Ov, Au₂₄/TiO₂ and Au₂₅/TiO₂-Ov samples under N₂ atmosphere were similar to those under Ar atmosphere, indicating that the interfacial electron transfer in the three samples was not interfered by surrounding N₂. It partially accounted for the catalytic performance of the Au_n/TiO₂-Ov catalysts shown in Fig. 1 that the Au_n/TiO₂-Ov catalysts did not significantly enhance the photoactivity of N₂ reduction in comparison with the TiO₂-Ov. Notably, the photocurrent of the Au₄Ru₂/TiO₂-Ov sample under the N₂ atmosphere was only a quarter of that under the Ar atmosphere (Fig. 4D), in which the quenching of the other three-quarters of photocurrent was possibly due to the electrons consumed by the adsorbed N₂ molecules.

Moreover, electrochemical impedance spectroscopy (EIS) was measured to investigate the interfacial charge-transfer properties of the above four samples under illumination. As shown in Fig. 4E, the semicircular diameters of Au₂₄/TiO₂-Ov and Au₂₅/TiO₂-Ov measured under light irradiation were slightly smaller than that of TiO₂-Ov, indicating that the Au_n clusters had an inherent ability of electron transport, but this ability was not extraordinary. Notably, the impedance of Au₄Ru₂/TiO₂-Ov was the smallest among the four samples, providing a solid evidence that there existed a fast transfer of the interfacial charges between Au₄Ru₂ and TiO₂-Ov.³² The charge transfer resistance on the Au₄Ru₂/TiO₂-Ov sample without illumination was also investigated (Fig. S10†). It was found that the impedance of Au₄Ru₂/TiO₂-Ov in the absence of light was much higher than that in the presence of light. Therefore, these observations supplied a clue that the Au_n clusters were able to generate the electrons under the light irradiation, but lacked the ability to activate N₂, and thus the Ru atoms in Au₄Ru₂ should be crucial for N₂ binding and activation. To further elucidate the critical contribution of the Ru atoms in the Au₄Ru₂ cluster to the activation of the inert N≡N triple bond, the two Ru atoms of Au₄Ru₂ were replaced by the two Pd atoms (Fig. S11A†), that is, Au₄Pd₂. No increase in the photocatalytic reduction of N₂ on the Au₄Pd₂/TiO₂-Ov was observed when compared to the Au_n/TiO₂-Ov (Fig. S11B†). The result definitely confirmed that the Ru atoms in the Au₄Ru₂ cluster indeed can provide unique reaction sites for the N≡N cleavage by strong coordination.

To gain atomistic insight into the photochemical N₂ reduction on the Au₄Ru₂/TiO₂-Ov catalyst, we performed DFT calculations to determine the active sites and reaction pathways. Our calculations show that the Au atoms do not have any activity for N₂ fixation, but N₂ can be adsorbed onto the Ru atom in the side-on or end-on configuration, with adsorption energies of −0.07 eV and −1.33 eV and Ru–N bond length of 2.14 Å and 1.84 Å, respectively (Fig. 5A–D). The N–N bond is elongated to 1.14–1.16 Å compared to 1.13 Å for the gaseous N₂ molecule, which manifests that N₂ is activated on the Ru site of the Au₄Ru₂ cluster. Moreover, we examined the structure of Au₄Ru₂(SCH₃)₈(P(CH₃)₃)₂ cluster supported on the anatase TiO₂(101) surface (Fig. S12†), which exhibits a weak interfacial interaction with a distance of 2.57 Å between the cluster and substrate and a binding energy of only −0.28 eV per Au(Ru) atom. Therefore, the presence of substrate should not affect the adsorption properties of the Au₄Ru₂ cluster with N₂. Hereafter, we considered N₂-to-NH₃ conversion on the Au₄Ru₂ cluster without support of the substrate.

Fig. 5 The atomic structures of N₂ molecule adsorption via (A and B) end-on and (C and D) side-on patterns on the Au₄Ru₂(SC₂H₄Ph)₈PPh₃ and Au₄Ru₂(SCH₃)₈P(CH₃)₃. The H, C, N, S, Au and Ru are shown in light blue, gray, blue, yellow, orange and blackish green colors, respectively. The adsorption energy of a N₂ molecule is shown for each system, revealing that simplifying the ligands of –PPh₃ and –SC₂H₄Ph by –P(CH₃)₃ and –SCH₃, respectively, has a minor effect on the binding properties of the cluster with N₂. (E) The colors numbers indicate the kinetic barriers of N₂ fixation on the Au₄Ru₂ cluster, and the barriers of rate-limiting step are bolded. The atomic structures of reaction intermediates are displayed as insets. The H, C, N, S, Au and Ru are shown in light blue, gray, blue, yellow, orange and green colors, respectively.

Ammonia synthesis on the Au₄Ru₂ cluster can proceed through three pathways, i.e. distal, alternating and enzymatic mechanisms (Fig. 5E).^24,33 For the former two paths, the N₂ molecule strongly binds with the underlying Ru atom in the end-on configuration. Protonation of the chemisorbed *N₂ species to form a *NNH intermediate is endothermic and has a kinetic barrier of 1.36 eV, which is the rate-limiting step of N₂-to-NH₃ conversion. The following reaction steps are exothermic involving barriers below 0.97 eV or even barrierless. For the enzymatic mechanism, the adsorption strength of *N₂ species on Ru is relatively weak. The reaction proceeds almost down-hill in energy. Reaction from *N₂ to form a *HN–NH species is favorable with a kinetic barrier of only 0.48 eV. Protonation of *H₂N–NH leads to the breaking of N–N bond and generation of two *H₂N species, which requires the largest barrier of 1.68 eV during the whole reaction. Finally, desorption of *NH₃ has to overcome a moderate energy barrier of 0.89 eV. In brief, our DFT calculations suggest that the synergistic effect of the Au₄Ru₂/TiO₂-Ov catalyst stems from the cooperation between cluster and substrate during the catalytic reaction (Fig. 4F): the Ru atom in the Au₄Ru₂ cluster serves as the active site for N₂ fixation and ammonia synthesis through the distal or alternating pathways; the anatase TiO₂(101) substrate plays important roles in water splitting to generate hydrogen protons that transfer to the cluster for the N₂-to-NH₃ reaction.

Conclusions

In conclusion, we have synthesized an excitonic Au₄Ru₂ cluster, which enables light harvesting including electron/hole pair production and separation. The experimental studies combined with theoretical calculations demonstrate that the cooperative effect between Au₄Ru₂ cluster and TiO₂ substrate with oxygen vacancies leads to an extraordinary activity for light-driven N₂ reduction. The electron–hole pairs can be generated from the excited Au₄Ru₂ cluster; the heterojunction between Au₄Ru₂ cluster and the TiO₂-Ov substrate also facilitates photocarriers separation; the photoelectrons transfer to the Ru atoms of the cluster; meanwhile, TiO₂-Ov induces water splitting to produce hydrogen protons for N₂ fixation and conversion on the Ru atoms. Certainly, this work provides deep insights into non-plasmon-induced charge transfer from atomically precise metal clusters and develops a feasible strategy to enable highly efficient solar energy utilization via pursuing heterogeneous catalysts with atomic precision.

Conflicts of interest

No conflicts of interest.

Acknowledgements

We acknowledge financial supports from National Natural Science Foundation of China (21773109, 91845104).

References

1. M. Azubel, J. Koivisto, S. Malola, D. Bushnell, G. L. Hura, A. L. Koh, H. Tsunoyama, T. Tsukuda, M. Pettersson, H. Häkkinen and R. D. Kornberg, Science, 2014, 345, 909–912.
2. R. Jin, C. Zeng, M. Zhou and Y. Chen, Chem. Rev., 2016, 116, 10346–10413.
3. P. Liu, R. Qin, G. Fu and N. Zheng, J. Am. Chem. Soc., 2017, 139, 2122–2131.
4. R. R. Nasaruddin, T. Chen, N. Yan and J. Xie, Coord. Chem. Rev., 2018, 368, 60–79.
5. S. Yuan, C. Xu, J. Li and Q. Wang, Angew. Chem., Int. Ed., 2019, 58, 5967–5970.
6. A. Desireddy, B. E. Conn, J. Guo, B. Yoon, R. N. Barnett, B. M. Monahan, K. Kirschbaum, W. P. Griffith, R. L. Whetten, U. Landman and T. P. Bigioni, Nature, 2013, 501, 399–402.
7. M. Zhou, C. Zeng, Y. Chen, S. Zhao, M. Y. Sfeir, M. Zhu and R. Jin, Nat. Commun., 2016, 7, 13240.
8. T. Higaki, M. Zhou, K. J. Lambright, K. Kirschbaum, M. Y. Sfeir and R. Jin, J. Am. Chem. Soc., 2018, 140, 5691–5695.
9. U. Aslam, V. G. Rao, S. Chavez and S. Linic, Nat. Catal., 2018, 1, 656–665.
10. R. Huang, Y. Wei, X. Dong, X. Wu, C. Du, S. Zang and T. C. W. Mak, Nat. Chem., 2017, 9, 689–697.
11. S. Chen, H. Ma, J. W. Padelford, W. Qinchen, W. Yu, S. Wang, M. Zhu and G. Wang, J. Am. Chem. Soc., 2019, 141, 9603–9609.
12. S. Yamazoe, K. Koyasu and T. Tsukuda, Acc. Chem. Res., 2014, 47, 816–824.
13. E. C. Tyo and S. Vajda, Nat. Nanotechnol., 2015, 10, 577–588.
14. X. Cai, G. Saranya, K. Shen, M. Chen, R. Si, W. Ding and Y. Zhu, Angew. Chem., Int. Ed., 2019, 58, 9964–9968.
15. K. Kwak, W. Choi, Q. Tang, M. Kim, Y. Lee, D. Jiang and D. Lee, Nat. Commun., 2017, 8, 14723.
16. J. G. Chen, R. M. Crooks, L. C. Seefeldt, K. L. Bren, R. M. Bullock, M. Y. Darensbourg, P. L. Holland, B. Hoffman, M. J. Janik, A. K. Jones, M. G. Kanatzidis, P. King, K. M. Lancaster, S. V. Lymar, P. Pfromm, W. F. Schneider and R. R. Schrock, Science, 2018, 360, eaar6611.
17. H. Hirakawa, M. Hashimoto, Y. Shiraishi and T. Hirai, ACS Catal., 2017, 7, 3713–3720.
18. J. M. McEnaney, A. R. Singh, J. A. Schwalbe, J. Kibsgaard, J. C. Lin, M. Cargnello, T. F. Jaramillo and J. K. Nørskov, Energy Environ. Sci., 2017, 10, 1621–1630.
19. M. A. Légaré, G. Bélanger-Chabot, R. D. Dewhurst, E. Welz, I. Krummenacher, B. Engels and H. Braunschweig, Science, 2018, 359, 896–900.
20. J. S. Anderson, J. Rittle and J. C. Peters, Nature, 2013, 501, 84–87.
21. Y. Gong, J. Wu, M. Kitano, J. Wang, T. Ye, J. Li, Y. Kobayashi, K. Kishida, H. Abe, Y. Niwa, H. Yang, T. Tada and H. Hosono, Nat. Catal., 2018, 1, 178–185.
22. A. J. Medford and M. C. Hatzell, ACS Catal., 2017, 7, 2624–2643.
23. K. A. Brown, D. F. Harris, M. B. Wilker, A. Rasmussen, N. Khadka, H. Hamby, S. Keable, G. Dukovic, J. W. Peters, L. C. Seefeldt and P. W. King, Science, 2016, 352, 448–450.
24. S. Wang, F. Ichihara, H. Pang, H. Chen and J. Ye, Adv. Funct. Mater., 2018, 28, 1803309.
25. J. Yang, Y. Guo, R. Jiang, F. Qin, H. Zhang, W. Lu, J. Wang and J. C. Yu, J. Am. Chem. Soc., 2018, 140, 8497–8508.
26. L. Xu, J. Wang, X. Zhu, X. Zeng and Z. Chen, Adv. Funct. Mater., 2015, 25, 3033–3042.
27. Y. Zhao, Y. Zhao, R. Shi, B. Wang, G. I. N. Waterhouse, L. Wu, C. Tung and T. Zhang, Adv. Mater., 2019, 1806482.
28. F. Zuo, L. Wang, T. Wu, Z. Zhang, D. Borchardt and P. Feng, J. Am. Chem. Soc., 2010, 132, 11856–11857.
29. C. Hu, X. Chen, J. Jin, Y. Han, S. Chen, H. Ju, J. Cai, Y. Qiu, C. Gao, C. Wang, Z. Qi, R. Long, L. Song, Z. Liu and Y. Xiong, J. Am. Chem. Soc., 2019, 141, 7807–7814.
30. H. Li, J. Shang, Z. Ai and L. Z. Zhang, J. Am. Chem. Soc., 2015, 137, 6393–6399.
31. C. Li, T. Wang, Z. J. Zhao, W. Yang, J. F. Li, A. Li, Z. Yang, G. A. Ozin and J. Gong, Angew. Chem., Int. Ed., 2018, 57, 5278–5282.
32. J. Xia, J. Di, H. Li, H. Xu, H. Li and S. Guo, Appl. Catal., B, 2016, 181, 260–269.
33. J. Liu, X. Ma, Y. Li, Y. Wang, H. Xiao and J. Li, Nat. Commun., 2018, 9, 1610.

---

Footnotes

† Electronic supplementary information (ESI) available. CCDC 1972938. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c9sc06424a

‡ These authors contributed equally.

---