Atomically dispersed s-block metal calcium site modified mesoporous g-C₃N₄ for boosting photocatalytic N₂ reduction

Abstract

Alkaline-earth metal elements in the s-block of the periodic table have rarely been studied as active sites for nitrogen (N₂) photofixation. Herein, we report a single-atom calcium (Ca)-modified mesoporous g-C₃N₄ (Ca/m-g-C₃N₄) for promoting the photocatalytic N₂ reduction reaction (pNRR) under ambient conditions. Moreover, the atomically dispersed Ca as active sites of Ca/m-g-C₃N₄ could achieve high adsorption of N₂ molecules, which could be confirmed by nitrogen temperature-programmed desorption test (N₂-TPD). In this regard, Ca single atoms can not only serve as the active centers of N₂ but also optimize the energy band structure of m-g-C₃N₄, facilitating the photocatalytic synthesis of ammonia (NH₃). Therefore, the optimal 0.5 Ca/m-g-C₃N₄ demonstrates a remarkable NH₃ generation amount of 42.23 μg g_cat.⁻¹ h⁻¹, which is 2.1 times that of pure g-C₃N₄ (20.41 μg g_cat.⁻¹ h⁻¹). Furthermore, the present 0.5 Ca/m-g-C₃N₄ demonstrates good stability, and the NH₃ production is still remarkably unvaried after four cycling tests. More interestingly, it was also found that other alkaline earth metal elements, including magnesium (Mg), strontium (Sr), and barium (Ba), can also activate N₂ molecules. Accordingly, the photocatalytic NH₃ synthesis yield of Ba/m-g-C₃N₄ is 29.52 μg g_cat.⁻¹ h⁻¹, slightly more than that of Mg/m-g-C₃N₄ (20.52 μg g_cat.⁻¹ h⁻¹) and Sr/m-g-C₃N₄ (20.88 μg g_cat.⁻¹ h⁻¹). We hope that this work could provide a novel insight for developing other high-performance N₂-photofixation systems based on low-cost s-block alkaline-earth metal materials in the future.

---

Introduction

Photocatalytic synthesis of ammonia (NH₃) offers a promising prospect for the conversion of nitrogen (N₂) to NH₃ through solar energy with low energy consumption and zero carbon emissions. Since the triple bond of N₂ molecule is extremely stable and the binding energy is as high as 941 kJ mol⁻¹, the effective activation of N₂ molecule is a great bottleneck.^1,2 Inspired by the biomimetic mechanism of the Fe–Mo–S cofactor of nitrogenase, many transition metals have been reported to form a strong bond with N₂ (transition metals–N₂, TM–N₂), which can effectively weaken the N≡N bond. Many reports have confirmed that there is a strong electron back-donation into the antibonding orbitals of the TM, between the d orbital electron of the TM and the p orbitals of the N₂ molecule, where the N₂ molecule could accept electrons from the metal sites and thus induce polarization for the adsorption and activation of N₂.^3–5 For example, Skulason et al.^6,7 demonstrated that scandium (Sc), yttrium (Y), titanium (Ti), and zirconium (Zr) show strong coordination bonding of N₂, which is conducive for the capture and activation of N₂ molecules. Unfortunately, the main-group elements do not have the structural characteristics of the d orbitals of the transition metal valence layer; thus, it has long been believed that the activation of N₂ molecules using the main-group elements is not feasible.⁸

Recently, some interesting work has proved that the p-block element boron (B) can also activate N₂ molecules and further enable the complete cyclic process of NH₃ formation.^9,10 For example, the Holger Braunschweig's group introduced a series of remarkable achievements using main-group element compounds in the field of nitrogen fixation. They also showed that the p-block elements can not only activate N₂ molecules in the form of transition metals but also achieve a coupling reaction of nitrogen dimer reduction that is not yet achievable through TM-based catalysts.^3,9 However, the main-group element materials that can adsorb and activate N₂ molecules are limited to some p-block elements, but for s-block elements, the related reports are few or only limited to lithium and its nitrides, such as Li₃N.¹¹ For a long time, other s-block elements including alkali metals and alkaline-earth metals, except Li, were not considered to be effective for adsorbing and activating N₂ molecules. Until 2021, Sjoerd Harder and co-workers used stable monovalent calcium complexes to activate N₂ molecules.¹⁰ They confirmed that the calcium (Ca) sites of the alkaline-earth metal elements can use its p orbitals to participate in the adsorption and activation of N₂ molecules, just like the d orbital electron action of TM. The discovery of Ca active sites has become a new breakthrough in the field of N₂ fixation. Since then, Chen et al.¹² confirmed that the unique p-orbital electronic structure of s-block metals, such as the Ca active centers, could enhance the adsorption of intermediates and accelerate electron transfer during the ORR reaction process. Afterward, Liu and co-workers¹³ reported the efficient electrocatalytic reduction of CO₂ by loading s-block magnesium (Mg) single atoms on g-C₃N₄. Concomitantly, Qiao's group¹⁴ evaluated the NRR activity of the main-group elements through the functional mechanism selectively promoting NRR and inhibiting HER, which remarkably accelerated the development of main-group elements for N₂ fixation reaction. A valuable study for developing s-block element-based NRR system with satisfactory properties, therefore, seems imperative.

Herein, the Ca single-atom-modified mesoporous g-C₃N₄ nanosheets (denoted as Ca/m-g-C₃N₄) was synthesized through calcination and continuous stirring approach. Undoubtedly, the single mesoporous g-C₃N₄ (m-g-C₃N₄) generally restricts the surface active sites, which limits its N₂ photofixation activity. In this regard, the modification strategy using the s-block Ca single-atom endowed Ca active sites to achieve the effective enhancement of the pNRR performance of m-g-C₃N₄. The Ca single atoms were decorated on the surface of m-g-C₃N₄ as active sites to realize high adsorption and activation of the N₂ molecules. Furthermore, the N₂-TPD data further proved that the Ca single-atom sites can significantly promote the N₂ chemisorption of g-C₃N₄ and improve its pNRR activity. Simultaneously, from the energy band structure, the VB potential of Ca/m-g-C₃N₄ clearly has a greater potential to produce hydrogen protons, promoting the process of electron-coupled hydrogenation during pNRR. Using the advantage of the Ca single atoms as the active centers together with the sufficient proton source derived from the valence band oxidation process, N₂ could be effectively reduced to NH₃ over Ca/m-g-C₃N₄ under ambient conditions.

Experimental section

Preparation of mesoporous g-C₃N₄ nanosheets

Mesoporous g-C₃N₄ nanosheets (m-g-C₃N₄) were prepared according to previous reports with minor modifications.¹⁵ In general, 1.0 g melamine was mixed with 5.0 g cyanuric acid and then ground for 30 min with the addition of 2.0 mL ethanol. The mixture was then placed in a tube furnace and subjected to a two-stage heating process under N₂ atmosphere, starting with calcination at 400 °C for 10 min at a heating rate of 5.0 °C min⁻¹, followed by calcination at the same rate to 550 °C for 4.0 h. Finally, the light-yellow product was ground into a powder to produce mesoporous g-C₃N₄ nanosheets (denoted as m-g-C₃N₄).

Preparation of Ca single-atom-modified m-g-C₃N₄ nanosheets

The Ca single-atom-modified mesoporous g-C₃N₄ nanosheets was synthesized via calcination and continuous stirring method (for detailed steps, see Fig. S1†). In brief, 100 mg m-g-C₃N₄ was dispersed in 60 mL water. Subsequently, different masses of calcium chloride dihydrate (the added amounts of Ca were controlled as 30 mg, 50 mg, 100 mg, and 200 mg) were added to the above dispersion. The mixture was stirred for 24 h and then collected by centrifugation and washing several times. Next, the collected materials were dried at 60 °C in vacuum for 4 h. Finally, the powder was calcined at 300 °C for 1 h in an air atmosphere with a heating rate of 2.0 °C min⁻¹. g-C₃N₄ loaded with different mass ratios of Ca was denoted as 0.3 Ca/m-g-C₃N₄, 0.5 Ca/m-g-C₃N₄, 1.0 Ca/m-g-C₃N₄, and 2.0 Ca/m-g-C₃N₄, respectively.

Preparation of Mg, Sr, or Ba single-atom-modified g-C₃N₄ nanosheets

Mg, Sr, and Ba-modified m-g-C₃N₄ (denoted as Mg/m-g-C₃N₄, Sr/m-g-C₃N₄, and Ba/m-g-C₃N₄, respectively) were prepared with a similar method as the synthesis of Ca/m-g-C₃N₄, except that CaCl₂·2H₂O was replaced with MgCl₂·6H₂O, SrCl₂·6H₂O, and BaCl₂·2H₂O.

Apparatus

The crystal phase of the as-prepared catalysts was evaluated using a PANalytical X'Pert powder X-ray diffractometer (XRD, Netherlands). Scanning electron microscopy (SEM, JEOL JSM-7001F) and transmission electron microscopy (TEM, JEOL 2100F) were used to analyze the porous morphology, and the accompanying energy dispersive X-ray spectroscopy (EDS) was used to analyze the elemental composition of the samples. High-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM, Titan Cubed Themis G230) was used to analyze the surface single-metal atoms. Further, the surface element properties were investigated using a Thermo Fisher Scientific K-alpha X-ray photoelectron spectrometer (XPS, Thermo ESCALAB250Xi), and all binding energies were referenced to the C 1s peak at 284.8 eV. The N₂ temperature-programmed desorption assessments were performed (Micromeritics Auto Chem II) by implementing a TCD as the detector. Transient and steady photoluminescence for the as-fabricated samples were achieved under 800 nm excitation on a FLS1000 fluorescence lifetime spectrophotometer (Edinburgh Instruments, UK). The specific surface area (BET) was recorded using an ASAP2460 instrument (Micromeritics).

Results and discussion

Characterization

The structures and morphologies were investigated by the means of TEM and HAADF-STEM. As shown in Fig. 1a, the m-g-C₃N₄ sample exhibits a bubbly-fold nanosheet structure with rich pores. Notably, no obvious single atoms were observed in Fig. 1b. Furthermore, Fig. 1c exhibits that the whole structure of m-g-C₃N₄ did not change after Ca single-atom modification, indicating that the structure of the catalyst is relatively stable. As presented by the HAADF-STEM image for 0.5 Ca/m-g-C₃N₄, many bright spots (i.e., Ca single atoms) could be seen in Fig. 1d, suggesting that monatomic Ca is successfully decorated on the surface of m-g-C₃N₄ nanosheets. Besides, the corresponding EDS mapping characterization was used to analyze the distribution of Ca, N, and C elements in 0.5 Ca/g-C₃N₄, and it was clearly observed that the Ca single atoms were uniformly distributed (Fig. 1e and f). Thus, the special porous sheet structure of the present 0.5 Ca/g-C₃N₄ sample can facilitate the rapid transfer of reactants and products during photocatalytic reactions.^16–18

Fig. 1 TEM (a) and HAADF-STEM (b) images of m-g-C₃N₄. TEM (c), HAADF-STEM (d), and corresponding EDS-mapping images (e and f) of 0.5 Ca/m-g-C₃N₄.

The XRD patterns of the as-prepared samples are displayed in Fig. 2a. In Fig. 2a, two characteristic diffraction peaks at ∼13.1° and ∼27.4° appeared in all m-g-C₃N₄-based samples, which correspond to the (100) crystal surface of the tri-s-triazine structural unit of g-C₃N₄ and the (002) crystal surface with aromatic stacking, respectively (Fig. 2a).¹⁹ It is worth noting that after doping Ca single atoms, the diffraction peak became sharper and stronger, indicating that Ca single atoms could improve the crystallinity of m-g-C₃N₄. Crucially, the enhancement of crystallinity accelerates the movement of photogenerated electrons on the nanosheets, further improving the photocatalytic activity of g-C₃N₄.²⁰ No diffraction peak of calcium oxide (CaO) was observed in all the samples, indicating that Ca does not exist as an oxide. Moreover, with the increase in the Ca content, the typical (002) diffraction peak of m-g-C₃N₄ was significantly shifted to a lower value, which is caused by the successful modification of Ca single atoms into the lattice of m-g-C₃N₄. We used FTIR spectroscopy to further analyze the surface groups of these catalysts, as shown in Fig. 2b. The prominent absorption at about 808 cm⁻¹ is due to out-of-plane bending vibration of tri-s-triazine. The strong bands at 1100–1700 cm⁻¹ were assigned to the typical stretching vibration of the heptazine ring. The broad weak band between 3000 and 3500 cm⁻¹ could be ascribed the presence of NH and/or NH₂ groups.¹⁹ Compared to the single m-g-C₃N₄, the NH₂ peak of Ca/m-g-C₃N₄ was slightly enhanced, indicating that there are more C vacancies on Ca/m-g-C₃N₄.²¹ The above data confirm the existence of a bonding force between calcium monoatomic Ca and m-g-C₃N₄, showing that the single-atom structure of 0.5 Ca/m-g-C₃N₄ has good stability.

Fig. 2 The XRD patterns (a) and FTIR spectra (b) of the as-prepared m-g-C₃N₄, 0.3 Ca/m-g-C₃N₄, 0.5 Ca/m-g-C₃N₄, 1.0 Ca/m-g-C₃N₄, and 2.0 Ca/m-g-C₃N₄ samples.

The high-resolution spectra of XPS for C 1s, O 1s, N 1s, and Ca 2p were recorded particularly for studying the chemical state and surface composition of m-g-C₃N₄ and 0.5 Ca/m-g-C₃N₄ catalysts. In the high-resolution spectrum of C 1s (Fig. 3a), the peak at 288.3 eV originates from the N=C–N structure, and the peak at 284.8 eV is due to graphitic carbon species (C_sp2–C_sp2 bonds). Compared to m-g-C₃N₄, the graphitic carbon peak intensity of 0.5 Ca/C₃N₄ was significantly lower, which helps to optimize the charge distribution of the aromatic π-conjugated system to promote charge separation and reduce the centers of charge recombination in the conjugated system.¹⁹ This indicates that the modified Ca atoms can effectively promote the separation and migration of photogenerated carriers. Notably, relative to m-g-C₃N₄, the peak assigned to the C–NH_x species in 0.5 Ca/m-g-C₃N₄ showed a slight blue shift, indicating a decrease in the charge density at the edges of tri-s-triazine, which is caused by the exposed N defects in tri-s-triazine.¹⁷ Relative to m-g-C₃N₄, a characteristic peak at ∼533.5 eV of 0.5 Ca/m-g-C₃N₄ was attributed to the water (H₂O) adsorbed on the surface in the high-resolution O 1s XPS spectrum (Fig. 3b). This data indicates that the 0.5 Ca/m-g-C₃N₄ sample may possess better hydrophilicity and more easily adsorb H₂O molecules, which will be conducive to the consumption of photogenerated holes. In the N 1s spectra, the existence of the C–N=C structure is demonstrated by the binding energy at ∼398.8 eV (Fig. 3c). Also, the peak at 399.9 eV is the signal of N connecting three C atoms (N–(C)₃), whereas the contribution at ∼401.2 eV may be caused by the amino group.¹⁸ Notably, the spectrum of N 1 s shows that the ratio of peak areas between C=N–C and N–C₃ decreases from 5.53 to 4.45 after the Ca atoms are loaded on m-g-C₃N₄, indicating that the Ca sites may be coordinated with C=N–C in 0.5 Ca/m-g-C₃N₄.¹⁴ Simultaneously, the typical 2p orbital signal of Ca element was detected (Fig. 3d), indicating that the Ca single atoms are present in a bivalent state, firmly confirming that Ca single atoms of the 0.5 Ca/m-g-C₃N₄ material exist in coordination.^11,22 Meanwhile, the peak of N–(C)₃ displayed a slight red shift, which was caused by Ca–N coordination. Presumably, the partial valence electron of the Ca atoms shifts to the tri-s-triazine unit structure in the g-C₃N₄ framework, resulting in an increase in the electron density of N in N–(C)₃.²³ Thus, the XPS characterization results prove that Ca single atoms mainly coordinated to the cavity edge N formed by the tri-s-triazine ring.²⁴

Fig. 3 High-resolution XPS spectra of C 1s (a), O 1s (b), N 1s (c), and Ca 2p (d) of the as-synthesized photocatalysts.

The specific surface area (S_BET) was investigated by N₂ adsorption–desorption isotherm analysis (Fig. S2†). The S_BET values of m-g-C₃N₄ and 0.5 Ca/m-g-C₃N₄ are 71.1 m² g⁻¹ and 40.9 m² g⁻¹, respectively. The two samples belong to type IV isotherms, indicating that our samples are mesoporous materials and have comparable average pore sizes. Although the introduction of Ca single atoms decreased the specific surface area and pore volume, the activity of pNRR was significantly enhanced, demonstrating that the S_BET is not the main factor. In addition, to further understand the pNRR activity of 0.5 Ca/m-g-C₃N₄, the essence of photocatalytic active sites was revealed by the N₂-TPD test (Fig. 4). The N₂ desorption peak at ∼150 °C corresponds to the physisorption peak of N₂ molecules and ∼420 °C corresponds to the chemisorption peak of N₂ molecules. As shown in Fig. 4, the sample of 0.5 Ca/m-g-C₃N₄ demonstrates an obvious chemisorption of N₂ molecules, corresponding to the desorption peak at ∼420 °C, while m-g-C₃N₄ lacks this effective chemisorption of N₂. The N₂-TPD data confirms that the 0.5 Ca/m-g-C₃N₄ catalyst could provide a large amount of chemical adsorption sites for capturing N₂ from water. Since N₂ adsorption is the first step for NH₃ synthesis, the remarkable N₂ absorption on 0.5 Ca/m-g-C₃N₄ will be beneficial for the whole pNRR procedure.

Fig. 4 N₂-TPD spectra of the as-prepared various photocatalysts.

Photocatalytic activities

The photocatalytic N₂ fixation performance of the as-prepared samples was determined by employing water as the solvent and proton source under full spectrum by the spectrophotometric measurement of the generated NH₃ with indophenol indicator (Fig. S3†).²⁵ The 0.5 Ca/m-g-C₃N₄ sample showed an excellent NH₃ production rate of 42.23 μg g_cat.⁻¹ h⁻¹ under full-spectrum light irradiation (Fig. 5a and b), which is about 2.1 times higher than that of m-g-C₃N₄ (20.41 μg g_cat.⁻¹ h⁻¹). Meanwhile, under 5 h light irradiation, the synthetic NH₃ yield in per unit time exhibits a “volcanic” relationship with Ca content, as shown in Fig. 5b, the optimal 0.5 Ca/m-g-C₃N₄ exhibiting excellent pNRR performance. Thus, the optimized catalyst of the 0.5 Ca/m-g-C₃N₄ sample was conducted for other related characterizations and experiments. Accordingly, no N₂H₄ was detected during the process (Fig. S4† and 5c), demonstrating an excellent selectivity for NH₃ production.^26–29 After four cycles, the activity of Ca/m-g-C₃N₄ for reducing N₂ to NH₃ decreased slightly, indicating remarkable catalytic stability (Fig. 5d). Owing to the successful modification of the surface single-atom Ca sites, the 0.5 Ca/m-g-C₃N₄ catalyst demonstrates excellent pNRR activity.

Fig. 5 (a) Quantitation measurement of the generated NH₃ by full spectra illumination with an indophenol indicator over different systems in N₂. (b) Comparison of the photocatalytic NH₃ production rate within 5.0 h of the corresponding catalysts. (c) UV-vis absorption spectra of the reaction solution by different catalysts. No N₂H₄ was detected at 458 nm. (d) Photocatalytic cycling tests for 0.5 Ca/m-g-C₃N₄.

Next, several rigorous contrast experiments were conducted for investigating the NRR activity of 0.5 Ca/m-g-C₃N₄ different conditions (Fig. S5†). When the NRR process was carried out under Ar atmosphere (without N₂), under dark or without catalyst, and only trace NH₃ could be detected, confirming that the produced NH₃ over 0.5 Ca/m-g-C₃N₄ indeed originates from the photocatalytic reaction. As the quencher for the photo-induced hole, CH₃OH is critical for sufficient charge separation and efficient NH₃ production. Adding CH₃OH (20 vol%) as the solvent, NH₃ formation demonstrates a significant enhancement.

Moreover, to further confirm whether photocatalytic N₂ fixation on the 0.5 Ca/m-g-C₃N₄ was authentic, photocatalytic N₂ fixation under ¹⁵N isotope-labeled N₂ was conducted (with purity not less than 99%). The produced NH₄⁺ could react with phenol and hypochlorite for the formation of ¹⁵N-labeled indophenol,^30,31 which could be assayed accurately by liquid chromatography-mass spectrometry (LC-MS). Fig. S6a† displays a potent mass spectroscopy signal of ¹⁴N-labeled indophenol anion at about 198 m/z in LC-MS research when using ¹⁴N₂ as the feeding gas. Notably, ¹⁵N-labeled indophenol negative anion displays a remarkable enhanced mass spectrum signal at about 199 m/z in LC-MS analysis (Fig. S6b†). The signal gives a higher intensity relative to the 14 N : 15 N natural abundance ratio after 30 min illumination. These data identify that the generate NH₄⁺ ions detected in this work originated from N₂ photofixation.

Possible mechanisms

With the intention of exploring the kinetic behaviors of the carriers in-depth, steady-state photoluminescence (PL) measurements and time-resolved photoluminescence decay measurements were carried out. In comparison to the potent emission peak of m-g-C₃N₄, the 0.5 Ca/m-g-C₃N₄ composites demonstrated a rather weak PL peak, illustrating that the modification of Ca single atoms can effectively reduce the recombination velocity of electrons and holes. (Fig. 6a). Relative to the time-resolved photoluminescence decay measurements of g-C₃N₄ (Fig. 6b), 0.5 Ca/m-g-C₃N₄ displayed longer average decay times of 9.6 ns, indicating the better condition for carrier separation in 0.5 Ca/m-g-C₃N₄. A series of electrochemical characterizations were carried out to evaluate the photocatalytic performance of the as-synthesized samples.^32,33 As shown in Fig. 6c, the electrochemical impedance spectroscopy (EIS) of 0.5 Ca/m-g-C₃N₄ exhibits a smaller semicircle than that of the pure m-g-C₃N₄, indicating that 0.5 Ca/m-g-C₃N₄ presents a lower charge transference resistance and a greater charge transfer rate. In contrast to m-g-C₃N₄, the transient photocurrent response of 0.5 Ca/m-g-C₃N₄ under full-spectrum light irradiation was significantly stable over 11 on–off cycles, indicating a stable photocatalytic process (Fig. 6d). As shown in Fig. 6d, the catalysts demonstrate a rapid light current response to periodic light on–off. Among these, 0.5 Ca/m-g-C₃N₄ displays a remarkably higher photocurrent intensity relative to m-g-C₃N₄, providing that 0.5 Ca/m-g-C₃N₄ takes possession of more beneficial carrier separation. This provides firm evidence that the modification of Ca single-atom sites can enhance the activity of g-C₃N₄. Hence, the modification of Ca single-atom sites presents greatly promoted carrier separation efficiency and effective N₂ adsorption sites, which are beneficial for pNRR.

Fig. 6 (a) The steady-state photoluminescence spectra (PL) and (b) time-resolved transient photoluminescence decay spectra of m-g-C₃N₄ and 0.5 Ca/m-g-C₃N₄, respectively. (c) Electrochemical impedance spectroscopy response for m-g-C₃N₄ and 0.5 Ca/mg-C₃N₄ samples in 0.1 M Na₂SO₄ aqueous solution. (d) The transient photocurrent spectra of the as-prepared samples.

Intriguingly, the pNRR performance of other alkaline-earth metal elements (e.g., Mg, Sr, and Ba) was also probed. The NH₃ concentration over Mg/m-g-C₃N₄ and Sr/m-g-C₃N₄ samples increased with increasing irradiation time, indicating that the modified Mg and Sr can also improve the pNRR activity of m-g-C₃N₄ to a certain extent (Fig. S7a†). In contrast, the loading of Ca or Ba atoms can significantly promote the NRR performance of m-g-C₃N₄, which may be attributed to the more effective adsorption and activation ability of Ca and Ba atoms to N₂ molecules. At the same time, no N₂H₄ was detected during the pNRR process (Fig. S7b†). We use UV-vis diffuse reflectance spectra (DRS) to estimate the light absorption performance and calculate the band gap of the samples. Compared with m-g-C₃N₄, the 0.5 Ca/m-g-C₃N₄ sample possesses a wider light absorption, demonstrating that its solar energy utilization is further improved (Fig. 6a). Moreover, the ability of 0.5 Ca/m-g-C₃N₄ to absorb light is also improved in the UV region, which could imply enhancements in the photocatalytic activity. The bandgap of 0.5 Ca/m-g-C₃N₄ is 2.62 eV, which could be obtained from the Tauc diagram transformed from UV-vis DRS (Fig. S8a†). It increases a little by 0.09 eV compared with m-g-C₃N₄ (2.53 eV). The flat-band potential of the materials can be measured by the Mott-Schottky test to further investigate the valence band (VB) and conduction band (CB) position (Fig. S8c and d†). Conversion between the measured potential (vs. Ag/AgCl) and NHE is achieved through the following equation.³⁴

E_NHE = E_Ag/AgCl + 0.1976(25 °C)

Thus, the flat band potentials of m-g-C₃N₄ and 0.5 Ca/m-g-C₃N₄ are ca. −1.25 V and −1.22 V (vs., NHE, pH = 7.0), respectively. In general, for n-type semiconductors that exhibit a positive slope in the Mott-Schottky plots, the CB potential is slightly higher than the flat band potential by 0.1 to 0.3 eV, and the difference is taken as 0.1 eV in this paper.¹⁸

Therefore, the CB potential of m-g-C₃N₄ and 0.5 Ca/m-g-C₃N₄ is −1.35 V (vs. NHE) and −1.32 V (vs. NHE), respectively. Based on E_g = E_VB − E_CB, the valence band (VB) potential can be estimated as 1.18 V (vs. NHE, m-g-C₃N₄) and 1.30 V (vs. NHE, 0.5 Ca/m-g-C₃N₄), respectively. Apparently, the VB potential deviation of both was 0.12 eV, which is consistent with the results of the XPS valence band spectrum (Fig. S8b†).¹⁹ Notably, although the photogenerated electrons of m-g-C₃N₄ have a stronger N₂ reduction potential than 0.5 Ca/m-g-C₃N₄, the opposite results are obtained by yield experiments. This is owing to the fact that the valence band of 0.5 Ca/m-g-C₃N₄ is +1.30 V, which the photogenerated holes could oxidize H₂O to produce more H protons (H₂O − 4e⁻ − O₂ + H⁺, +1.23 V vs. NHE), whereas the VB potential of m-g-C₃N₄ (+1.18 V vs. NHE) was not. It is inferred that the chelated Ca single atoms not only serve as the active sites for N₂ molecules but also optimize the bandgap energy structure of m-g-C₃N₄. The VB potential is positively shifted to produce more H⁺, thus significantly facilitating the coupled proton-electron process of pNRR compared to g-C₃N₄.

From the above analysis, a continuous hydrogenation pathway of N₂ molecules over the 0.5 Ca/m-g-C₃N₄ catalyst was proposed (Fig. 7). Because no N₂H₄ was detected during the pNRR process (Fig. 5c), it can be assumed that the N₂ molecules reduction pathway is mainly the associative distal pathway. During the pNRR process, the adsorption of N₂ molecules is accompanied by chemical reactions. In the associative distal pathway,^35–38 the N₂ molecules adsorb on the Ca single-atom sites in an end-on configuration, which causes the N≡N bond to be weakened. Simultaneously, the N atom farthest from the catalytic surface in the adsorbed N₂ molecule prefers to conjugate H⁺ and couple the photogenerated electrons, facilitating the process of hydrogenation of the N₂ molecules to form an NH₃ molecule (*N≡N, *N=NH, *N–NH₂, *N–NH₃, NH₃). Then, one NH₃ is released and the other N atom of the adsorbed state proceeds to the hydrogenation reduction process. Finally, another NH₃ molecule was formed by the same hydrogenation process.

Fig. 7 The energy band structure and the proposed pNRR path of the continuous hydrogenation process on the Ca active sites of 0.5 Ca/m-g-C₃N₄.

Conclusion

In summary, the photocatalytic activity of m-g-C₃N₄ driven by sunlight for N₂ fixation was significantly boosted after the modification of Ca single atoms. The reinforced activity is due to the high-efficiency separation of photogenerated charges and the improvement of surface Ca sites; 0.5 Ca/m-g-C₃N₄ demonstrates an excellent pNRR activity. Accordingly, it is confirmed that the Ca single atom can not only act as an active site to realize high adsorption and activation of N₂ molecules but also improve the band gap structure of m-g-C₃N₄ to induce more hydrogen protons and promote the hydrogenation process during the pNRR procedure. Therefore, 0.5 Ca/m-g-C₃N₄ displays an excellent NH₃ generation amount of 42.23 μg g_cat.⁻¹ h⁻¹, which is 2.1 times than that of m-g-C₃N₄ (20.41 μg g_cat.⁻¹ h⁻¹). Besides, we also found that other alkaline-earth metals (e.g., Mg, Sr, and Ba.) can also promote the pNRR perfomance of m-g-C₃N₄. It is expected that our study will provide insights into the future development of the photocatalytic synthesis of ammonia by s-block metal materials.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was sponsored by the National Natural Science Foundation of China (22172040, 21974031), the Department of Science and Techniques of Guangdong Province (2021A1515010180, 2019B010933001), Guangzhou Municipal Science and Technology Bureau (202102010449), and the Department of Guangdong Provincial Public Security (GZQC20-PZ11-FD084). We also thank Shiyanjia Lab platform (https://www.shiyanjia.com/) for providing the XPS and N₂-TPD experimental testing and analysis services.

References

1. Y. Xin, S. M. Wang, H. B. Yuan, T. T. Hou, W. K. Zhu, Y. X. Liu, Y. Yao, W. H. Zhang, S. Q. Liang and L. B. Wang, Chem, 2021, 7, 2118–2136.
2. J. H. Yang, H. Y. Bai, Y. Z. Guo, H. Zhang, R. B. Jiang, B. C. Yang, J. F. Wang and J. C. Yu, Angew. Chem., Int. Ed., 2021, 60, 927–936.
3. M. A. Legare, G. Belanger-Chabot, R. D. Dewhurst, E. Welz, I. Krummenacher, B. Engels and H. Braunschweig, Science, 2018, 359, 896–899.
4. H. Tanaka, Y. Nishibayashi and K. Yoshizawa, Acc. Chem. Res., 2016, 49, 987–995.
5. W. S. Zhang, D. F. Han, M. J. Dai, Y. Y. Fan, G. L. Pan, W. Q. Liang, Q. T. Zheng, D. D. Qin, D. X. Han, Y. He and L. Niu, ACS Sustainable Chem. Eng., 2021, 9, 15331–15343.
6. Y. Abghoui, A. L. Garden, V. F. Hlynsson, S. Bjorgvinsdottir, H. Olafsdottir and E. Skulason, Phys. Chem. Chem. Phys., 2015, 17, 4909–4918.
7. E. Skulason, T. Bligaard, S. Gudmundsdottir, F. Studt, J. Rossmeisl, F. Abild-Pedersen, T. Vegge, H. Jonsson and J. K. Norskov, Phys. Chem. Chem. Phys., 2012, 14, 1235–1245.
8. M. A. Legare, G. Belanger-Chabot, M. Rang, R. D. Dewhurst, I. Krummenacher, R. Bertermann and H. Braunschweig, Nat. Chem., 2020, 12, 1076–1080.
9. M. A. Legare, M. Rang, G. Belanger-Chabot, J. I. Schweizer, I. Krummenacher, R. Bertermann, M. Arrowsmith, M. C. Holthausen and H. Braunschweig, Science, 2019, 363, 1329–1332.
10. B. Rosch, T. X. Gentner, J. Langer, C. Farber, J. Eyselein, L. Zhao, C. Ding, G. Frenking and S. Harder, Science, 2021, 371, 1125–1128.
11. J. G. Kay, Science, 1964, 144, 703.
12. Z. Y. Lin, H. Huang, L. Cheng, W. Hu, P. P. Xu, Y. Yang, J. M. Li, F. Y. Gao, K. Yang, S. Liu, P. Jiang, W. S. Yan, S. Chen, C. L. Wang, H. G. Tong, M. X. Huang, W. Zheng, H. Wang and Q. W. Chen, Adv. Mater., 2021, 33, 7103–7113.
13. Q. Y. Wang, K. Liu, J. W. Fu, C. Cai, H. J. W. Li, Y. Long, S. Y. Chen, B. Liu, H. M. Li, W. Z. Li, X. Q. Qiu, N. Zhang, J. H. Hu, H. Pan and M. Liu, Angew. Chem., Int. Ed., 2021, 60, 25241–25245.
14. L. Q. Li, C. Tang, H. Y. Jin, K. Davey and S. Z. Qiao, Chem, 2021, 7, 3232–3255.
15. H. Y. Niu, W. J. Zhao, H. Z. Lv, Y. L. Yang and Y. Q. Cai, Chem. Eng. J., 2021, 411, 400–411.
16. X. S. Wang, C. Zhou, R. Shi, Q. Q. Liu, G. I. N. Waterhouse, L. Z. Wu, C. H. Tung and T. Zhang, Nano Res., 2019, 12, 2385–2389.
17. S. Y. Gao, X. Y. Wang, C. J. Song, S. J. Zhou, F. Yang and Y. Kong, Appl. Catal., B, 2021, 295, 3873–3883.
18. W. J. Luo, X. J. Chen, Z. Wei, D. Liu, W. Q. Yao and Y. F. Zhu, Appl. Catal., B, 2019, 255, 117761.
19. X. J. Chen, R. Shi, Q. Chen, Z. J. Zhang, W. J. Jiang, Y. F. Zhu and T. R. Zhang, Nano Energy, 2019, 59, 644–650.
20. Y. Zhang, L. L. Wu, X. Y. Zhao, Y. N. Zhao, H. Q. Tan, X. Zhao, Y. Y. Ma, Z. Zhao, S. Y. Song, Y. H. Wang and Y. G. Li, Adv. Energy Mater., 2018, 8, 1–9.
21. S. H. Cao, B. Fan, Y. C. Feng, H. Chen, F. Jiang and X. Wang, Chem. Eng. J., 2018, 353, 147–156.
22. H. J. Chen, Z. Chen, G. X. Zhao, Z. B. Zhang, C. Xu, Y. H. Liu, J. Chen, L. Zhuang, T. Haya and X. K. Wang, J. Hazard. Mater., 2018, 347, 67–77.
23. S. Z. Hu, X. Chen, Q. Li, F. Y. Li, Z. P. Fan, H. Wang, Y. J. Wang, B. H. Zheng and G. Wu, Appl. Catal., B, 2017, 201, 58–69.
24. Y. T. Xiao, G. H. Tian, W. Li, Y. Xie, B. J. Jiang, C. G. Tian, D. Y. Zhao and H. G. Fu, J. Am. Chem. Soc., 2019, 141, 2508–2515.
25. S. Z. Andersen, V. Colic, S. Yang, J. A. Schwalbe, A. C. Nielander, J. M. McEnaney, K. Enemark-Rasmussen, J. G. Baker, A. R. Singh, B. A. Rohr, M. J. Statt, S. J. Blair, S. Mezzavilla, J. Kibsgaard, P. C. K. Vesborg, M. Cargnello, F. S. Bent, T. F. Jaramillo, I. E. L. Stephens, J. K. Norskov and I. Chorkendorff, Nature, 2019, 570, 504–508.
26. Z. F. Zhao, H. J. Ren, D. Yang, Y. Han, J. F. Shi, K. An, Y. Chen, Y. H. Shi, W. J. Wang, J. D. Tan, X. Xin, Y. Zhang and Z. Y. Jiang, ACS Catal., 2021, 11, 9986–9995.
27. G. W. Watt and J. D. Chrisp, Anal. Chem., 1952, 24, 2006–2008.
28. J. X. Zhao, X. J. Liu, X. Ren, X. Sun, D. X. Tian, Q. Wei and D. Wu, Appl. Catal., B, 2021, 284, 1–8.
29. W. B. Qiu, X. Y. Xie, J. D. Qiu, W. H. Fang, R. P. Liang, X. Ren, Q. X. Ji, G. W. Cui, A. M. Asiri, G. L. Cui, B. Tang and X. P. Sun, Nat. Commun., 2018, 9, 3485–3493.
30. H. Hirakawa, M. Hashimoto, Y. Shiraishi and T. Hirai, J. Am. Chem. Soc., 2017, 139, 10929–10936.
31. G. H. Dong, W. K. Ho and C. Y. Wang, J. Mater. Chem. A, 2015, 3, 23435–23441.
32. K. Chen, X. Zhao, X. J. Zhang, W. S. Zhang, Z. F. Wu, H. Y. Wang, D. X. Han and L. Niu, Catal. Sci. Technol., 2021, 11, 2713–2717.
33. X. J. Zhang, D. F. Han, M. J. Dai, K. Chen, Z. Y. Han, Y. Y. Fan, Y. He, D. X. Han and L. Niu, Catal. Sci. Technol., 2021, 11, 6248–6256.
34. X. Yang, X. J. Lian, S. J. Liu, J. Tian, C. P. Jiang and G. Wang, Int. J. Electrochem. Sci., 2013, 8, 3721.
35. S. Zhang, Y. X. Zhao, R. Shi, G. I. N. Waterhouse and T. R. Zhang, EnergyChem, 2019, 1, 1–40.
36. T. T. Hou, Y. Xiao, P. X. Cui, Y. N. Huang, X. P. Tan, X. S. Zheng, Y. Zou, C. X. Liu, W. K. Zhu, S. Q. Liang and L. B. Wang, Adv. Energy Mater., 2019, 9, 2319–2327.
37. C. X. Guo, J. R. Ran, A. Vasileff and S. Z. Qiao, Energy Environ. Sci., 2018, 11, 45–56.
38. Y. N. Bo, H. Y. Wang, Y. X. Lin, T. Yang, R. Ye, Y. Li, C. Y. Hu, P. Y. Du, Y. G. Hu, Z. Liu, R. Long, C. Gao, B. J. Ye, L. Song, X. J. Wu and Y. J. Xiong, Angew. Chem., 2021, 60, 16085–16092.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d2cy01507b

‡ These authors contributed equally to this work.

---