γ-Graphyne-modulated g-C₃N₄ nanostructures for boosted photocatalytic ammonia production

Abstract

Photocatalytic nitrogen reduction is a promising route for sustainable ammonia production, yet its efficiency remains limited by poor charge separation and sluggish N₂ activation. Herein, a two-dimensional g-C₃N₄ nanosheet/γ-graphyne (CNNS/GY) heterostructure is designed to modulate interfacial charge dynamics and strengthen nitrogen binding under illumination. The CNNS/GY interface delivers markedly improved charge separation and accelerates electron transfer, enabling an NH₄⁺ production rate of 803.67 µmol g⁻¹ h⁻¹, which is 2.45 times higher than that of pristine CNNS. DFT calculations reveal strong electronic coupling and rapid charge migration across the heterojunction, providing a mechanistic basis for the enhanced activity. This work demonstrates the effectiveness of γ-graphyne in tuning the interfacial electronic structure and highlights CNNS/GY heterostructures as a promising platform for advancing photocatalytic ammonia production.

---

1. Introduction

Ammonia serves as both a critical chemical feedstock and a promising energy carrier, playing an indispensable role in sustaining human society and supporting global economic progress. While the Haber–Bosch process currently dominates industrial ammonia synthesis, its reliance on high temperature and pressure results in substantial energy consumption and CO₂ emissions.¹ Photocatalytic approaches leverage abundant N₂ and H₂O as feedstocks to drive the nitrogen reduction reaction (NRR) under benign conditions, offering a viable substitute for traditional ammonia synthesis technologies.^2,3 Since Schrauzer and Guth first reported the photocatalytic synthesis of ammonia from N₂ over titanium dioxide (TiO₂) under UV irradiation in 1977,⁴ research on photocatalysts for nitrogen fixation has developed rapidly.^5,6 Among them, graphitic carbon nitride (g-C₃N₄) stands out as a highly promising metal-free photocatalytic material owing to its visible-light response, tunable bandgap (≈2.7 eV), thermal and chemical stability, abundant surface defects and functional groups, low cost, and facile synthesis.⁷ The layered architecture is bound by weak van der Waals interactions, enabling bulk g-C₃N₄ to be readily exfoliated into single-layer nanosheets.⁸ Compared with the bulk counterpart, g-C₃N₄ nanosheets (CNNS) exhibit a much larger specific surface area and more exposed reactive sites. Moreover, their two-dimensional architecture provides shortened charge-transport pathways, allowing photogenerated electron–hole pairs to rapidly migrate onto the interface before recombination, thus largely boosting the photoelectrochemical activity.⁹ Although these strategies, including doping, defect engineering, and heterojunction construction, collectively confirm the importance of interfacial charge management, a persistent need for co-catalysts capable of simultaneously enhancing conductivity, stabilizing reactive sites, and facilitating N₂ activation is also highlighted.^10–12

Graphynes (GYs), two-dimensional carbon allotropes featuring both sp- and sp²-hybridized carbon atoms, have recently attracted attention as highly conductive and chemically robust platforms.^13–15 A growing body of studies have demonstrated that the integration of GYs with semiconductors represents a viable strategy to enhance photocatalytic performance. Yu et al.,¹⁶ for the first time, fabricated a TiO₂ nanofiber/graphdiyne (GDY) composite photocatalyst using a simple electrostatic self-assembly strategy for catalytic CO₂ conversion. The concept of GDY serving as a co-catalyst is put forward, serving as an electron reservoir and offering adsorption sites during the photocatalytic process. The integration of γ-graphyne (GY) with TiO₂ nanorod arrays has led to the construction of a 2D–1D heterojunction, achieving a 2.2-fold enhancement of photoconversion efficiency.¹⁷ The ternary structure of TiO₂ nanonet/SrTiO₃/GY has been reported with improved photoresponse, resulting in a 1.7-fold increase in light conversion efficiency.¹⁸ What's more, dual-defected N-doped γ-GY/Ti³⁺–TiO₂ nanocomposites were prepared, resulting in a sevenfold enhancement in NRR photocatalytic efficiency.¹⁹ In systems such as Ag₃PO₄@GY,²⁰ TiO₂/GY,²¹ and TiO₂/MoSe₂/GY,²² γ-graphyne has been theoretically identified as the monoalkyne-bonded carbon allotrope exhibiting the highest structural stability and a conjugated framework. Its characteristic triangular pores promote efficient ion insertion and transport, thereby offering significant potential for development in the design of junction-modulated photocatalysis.^23,24

Prior studies on incorporating GY into semiconductor systems have demonstrated improved charge transport and broadened photoresponse, suggesting its potential as an efficient electron mediator.²⁵ Nevertheless, existing reported systems have primarily focused on photocatalytic reactions such as CO₂ reduction and H₂ evolution, whereas studies related to the photocatalytic NRR remain relatively limited. In the context of the NRR, previous studies have shown that the introduction of GDY can contribute to enhanced photocatalytic performance, often associated with improved charge separation.^26,27 In comparison, the influence of γ-graphyne incorporation on light adsorption, interfacial interaction and its correlation with N₂ reduction behavior remains an area that could benefit from further investigation. Therefore, we design g-C₃N₄ nanosheets/γ-graphyne composites to establish the role of γ-graphyne in modulating interfacial electronic coupling and enhancing the N₂-fixation capability of g-C₃N₄. A uniform composite consisting of g-C₃N₄ nanosheets (CNNS) integrated with GY was obtained through a two-step strategy. The resulting GY-modified CNNS exhibit a superior NH₄⁺ production rate of 803.67 µmol g⁻¹ h⁻¹ under flowing N₂, which is 2.45-fold that of pristine CNNS. Combined characterization and DFT analyses elucidate that the charge separation is strengthened and N₂ activation is promoted through band engineering and interfacial electronic modulation. The resulting system enables a mechanistically informed pathway toward improving the photocatalytic efficiency and durability of g-C₃N₄-based NRR catalysts.

2. Experimental

2.1. Materials

All the used chemicals and materials for the synthesis, analysis, and measurements were of analytical grade and the details are provided in the SI. The schematic diagram for the preparation process of CNNS/GY is shown in Fig. S1.

2.2. Synthesis of g-C₃N₄ nanosheets (CNNS)

g-C₃N₄ nanosheets were synthesized via thermal oxidative exfoliation. First, 3 g of melamine were placed in a 50 mL alumina crucible, covered and transferred to a muffle furnace. The reaction system was then heated to 550 °C at a controlled rate of 5 °C min⁻¹ and maintained at this temperature for 4 h. Afterward, the sample was allowed to cool naturally inside the furnace. The resulting product, which appeared as yellowish g-C₃N₄, was collected. Subsequently, 500 mg of the obtained product was placed in an open crucible and heated to 500 °C at the same ramping rate under static air, followed by a 2 h dwell. After the reaction, the sample was removed and rapidly cooled in air, yielding white g-C₃N₄ nanosheets (CNNS).

2.3. Synthesis of γ-graphyne (GY)

GY was synthesized via a mechanochemical route following procedures reported in previous studies.²⁸ In a 250 mL stainless-steel milling vessel, 2 mL of benzene, 10.0 g of CaC₂, 35 mL of absolute ethanol, and 375 g of stainless-steel balls were added. Planetary ball milling (TENCAN QXQM-2) was carried out at 600 rpm for 8 h and then continued at 450 rpm for an additional 8 h. The obtained milled mixture was subsequently annealed at 260 °C for 3 h under a nitrogen flow. The resulting powder was subsequently subjected to two sequential purification steps using excess dilute nitric acid (0.1 mol L⁻¹) and glacial acetic acid (2 mol L⁻¹), respectively. Afterward, the purified GY was collected via vacuum filtration and dried at 60 °C to give the final high-purity product.

2.4. Preparation of γ-graphyne-modified g-C₃N₄ nanosheets (CNNS/GY)

Two-dimensional CNNS/GY composites were synthesized via an impregnation method. Specifically, 100 mg of CNNS and a specified amount of GY powder were each dispersed in 3 mL of anhydrous ethanol and subjected to ultrasonication for 30 min to achieve homogeneous dispersion. The obtained mixture was subsequently transferred to a water bath and heated at 80 °C with continuous magnetic stirring until the solvent completely evaporated. The obtained solid was collected as CNNS/GY powder. Based on the amount of GY added (1 mg, 5 mg, and 10 mg), the corresponding samples were designated as CNNS/GY-1, CNNS/GY-5, and CNNS/GY-10, respectively.

2.5. Preparation of γ-graphyne-modified g-C₃N₄ nanosheet electrodes

CNNS/GY composite electrodes were fabricated to evaluate their electrochemical performance. 20 mg of CNNS/GY powder was suspended in a mixture of 500 µL ultrapure water, 1 mL anhydrous ethanol, and 15 µL Nafion solution (5 wt%). The suspension was subjected to ultrasonication for 30 min to achieve homogeneous dispersion. Subsequently, an appropriate amount of the suspension was deposited onto a 10 × 15 mm FTO glass substrate (effective physical area: 1 cm²) and dried under ambient conditions.

2.6. General characterization studies

X-ray diffraction (XRD) patterns were obtained using a Bruker D8 Advance diffractometer equipped with Cu Kα radiation. Fourier transform infrared (FTIR) spectra were collected on a Nicolet iS20 spectrometer (Thermo Scientific, USA). Scanning electron microscopy (SEM) was carried out using a ZEISS GeminiSEM 300 microscope. Transmission electron microscopy (TEM) and selected area electron diffraction (SAED) were conducted using a FEI Talos F200S microscope. X-ray photoelectron spectroscopy (XPS) was performed on a Thermo Scientific K-Alpha spectrometer. Ultraviolet-visible diffuse reflectance spectroscopy (UV-vis DRS) was performed using a Shimadzu UV-2600 spectrophotometer. Photoluminescence (PL) spectra were recorded using a Hitachi F-7000 fluorescence spectrophotometer.

2.7. Photoelectrochemical tests

Photoelectrochemical tests were conducted in a three-electrode system, with a platinum (Pt) electrode serving as the counter electrode, a saturated calomel electrode (SCE) as the reference electrode, and a 0.5 mol L⁻¹ Na₂SO₄ solution (pH = 7) as the electrolyte. Linear sweep voltammetry (LSV) and electrochemical impedance spectroscopy (EIS) measurements were performed on an electrochemical workstation (PGSTAT302N, Metrohm, Switzerland). The LSV scans were conducted at a sweep rate of 50 mV s⁻¹, and the EIS measurements were carried out over a frequency range of 0.1 to 1 × 10⁵ Hz. Transient photocurrent measurements were conducted on another electrochemical workstation (CHI 760E, CH Instruments, China) under an applied bias of 0.3 V (vs. SCE).

2.8. Photocatalytic experiments

A schematic diagram of the reactor is shown in Fig. S2. A 300 W xenon arc lamp (CEL-HXF300-T3) was employed as the light source, with a fixed distance of 15 cm between the lamp and the reactor. No optical filter was used, and the reaction was carried out under UV-vis irradiation. A total of 40 mg of catalyst powder was dispersed in 40 mL of methanol–water mixture containing 20 vol% methanol and subjected to ultrasonication for 15 min to achieve homogeneous dispersion. High-purity nitrogen gas (0.4 MPa) was steadily bubbled through the solution under magnetic stirring, with a circulating cooling water jacket. Following 30 min of dark equilibration, the reaction was carried out under light irradiation for 2 h, with aliquots collected every 30 min. After each photocatalytic run, the catalyst was collected by centrifugation (8000 rpm, 5 min). The obtained precipitate was redispersed in deionized water, followed by centrifugation. This washing process was repeated three times to remove residual reactants. The recovered catalyst was then dried at 60 °C overnight and reused for subsequent photocatalytic cycles.

2.9. Detection of ammonium (NH₄⁺) concentration

The NH₄⁺ concentration was measured by UV-vis spectrophotometry using Nessler's reagent at 420 nm. Since the NH₄⁺ concentration exceeded the linear range of the calibration curve, the solution was diluted accordingly before analysis. For quantification, 1 mL of the reaction mixture was filtered through a 0.22 µm membrane and diluted to 10 mL with deionized water and then sequentially mixed with 20 µL of potassium sodium tartrate, followed by 20 µL of Nessler's reagent. The solution colors at different ammonia concentrations are presented in Fig. S3, and the absorbance values of the ammonia standard solutions are listed in Table S1. After standing at 25 °C for ∼10 min, its absorbance was recorded at 420 nm. The ammonia concentration was determined via UV-vis spectrophotometry, using the calibration curve established from standard NH₄Cl solutions with concentrations of 0, 0.1, 0.2, 0.4, 0.8, 1.2, 1.6, and 2.0 mg L⁻¹ (Fig. S4, y = 0.17468x + 0.00144, R² = 0.99802).

The average NH₄⁺ yield was calculated using eqn (1):

In the above equation, denotes the NH₄⁺ yield of the photocatalyst; represents the concentration of NH₄⁺ determined from the calibration curve; V is the volume of the reaction system; t is the reaction time; m_cat. is the mass of the photocatalyst used and is the relative molecular mass of the ammonium ion.

2.10. Density functional theory (DFT) calculations

All DFT calculations were performed using the CASTEP module of Materials Studio,²⁹ within the framework of the Perdew–Burke–Ernzerhof (PBE) functional under the generalized gradient approximation (GGA).³⁰ To avoid interactions between periodic images, a vacuum layer of 15 Å was introduced along the non-periodic direction. For all calculations, a plane-wave cutoff energy of 500 eV and a self-consistent field (SCF) convergence threshold of 1 × 10⁻⁶ eV were adopted. A Monkhorst–Pack k-point mesh of 2 × 2 × 1 was used for both geometry optimizations and subsequent property calculations. During structural relaxation, all atoms were allowed to move freely without any positional constraints. The convergence criteria were set as follows: energy variation per atom below 1.0 × 10⁻⁵ eV, maximum force less than 0.03 eV Å⁻¹, maximum stress below 0.05 GPa, and maximum atomic displacement under 0.001 Å.

3. Results and discussion

3.1. Characterization analysis

The XRD patterns of CNNS and CNNS/GY are illustrated in Fig. 1(a), both of which exhibit characteristic diffraction features of g-C₃N₄ (JCPDS No. 87-1526). The weak peak at approximately 12.8° is attributed to the (100) plane (d = 0.69 nm), corresponding to the in-plane periodic arrangement of triazine units (C₆N₇).³¹ The intense peak at 27.9° is attributed to the (002) plane (d = 0.32 nm), which originates from the interlayer π–π stacking of aromatic layers. The characteristic diffraction peaks of g-C₃N₄ are retained in the composite, suggesting successful hybridization without structural disruption. No obvious peak shift is observed (Fig. 1(b)). The reduced intensity of the (100) and (002) peaks may be attributed to the increased GY content, which lowers the thermodynamic stability of these planes and weakens the interlayer π–π stacking. A relatively weak peak observed at approximately 44.3° is assigned to the (300) plane of GY (d = 0.21 nm).³²

Fig. 1 (a) XRD patterns of CNNS and CNNS/GY; (b) enlarged (002) peaks of XRD patterns; (c) FTIR spectra of CNNS, CNNS/GY-1, CNNS/GY-5, and CNNS/GY-10.

The FTIR spectra shown in Fig. 1(c) are used to investigate the functional groups present in CNNS and CNNS/GY. The characteristic absorption bands of g-C₃N₄ are mainly located in the following regions. The broad absorption between 3000 and 3500 cm⁻¹ is ascribed to the stretching vibrations of terminal –NH₂ groups and edge –NH bonds. Multiple peaks in the 1200–1650 cm⁻¹ region, including those at approximately 1235, 1318 and 1560 cm⁻¹, are assigned to C–N and C=N stretching vibrations as well as conjugated ring modes associated with the tri-s-triazine framework.³³ The sharp peak near 810 cm⁻¹ is associated with the out-of-plane bending vibration of the tri-s-triazine ring.³⁴ Additionally, a weak band near 2200 cm⁻¹ corresponds to the presence of C≡C bonds, demonstrating the successful incorporation of GY.³⁵

The morphology of the samples is analyzed using SEM images. Fig. 2(a–c) present the SEM images of CNNS, which display a typical two-dimensional layered structure with smooth surfaces and well-defined edges. Fig. 2(d–f) present the SEM images of CNNS/GY, where GY is distributed on the CNNS layers. The flake-like morphology of CNNS remains largely unchanged after the incorporation of GY. Fig. 2(g–i) present the EDX elemental mapping images of CNNS/GY, revealing a uniform elemental distribution across the sample surface. The C and N elemental images exhibit a uniform and largely overlapping spatial distribution, while the relatively stronger C signal corresponds to the GY-rich regions. The spatially resolved elemental mapping reveals substantial C–N co-localization within the composite. The absence of detectable elemental segregation or impurities further confirms the formation of the CNNS/GY heterostructure. Moreover, Fig. 3(a and b) present the TEM images of CNNS, which exhibit a typical two-dimensional nanosheet morphology. The TEM images of the CNNS/GY composite (Fig. 3c and d) reveal the intimate coexistence of CNNS and GY, indicating successful heterojunction formation.

Fig. 2 SEM images of (a)–(c) CNNS and (d)–(f) CNNS/GY-5; (g)–(i) EDX mapping images of CNNS/GY-5.

Fig. 3 TEM images of (a and b) CNNS (inset: the SAED image of CNNS) and (c and d) CNNS/GY-5 (inset: the SAED image of CNNS/GY-5).

XPS was used to analyze the elemental composition and chemical states of CNNS and CNNS/GY (Fig. 4). The survey spectrum of CNNS/GY (Fig. 4(a)) shows C, N and O as the main elements; their atomic percentages are approximately 57.98% (C), 34.54% (N) and 7.48% (O), indicating a carbon–nitride framework with some oxygen-containing groups and no detectable additional impurity elements. Deconvolution of the high-resolution N 1s spectrum for CNNS/GY (Fig. 4(b)) reveals three distinct characteristic peaks. The peak at 398.7 eV originates from sp²-hybridized nitrogen in the triazine rings (C=N–C) of CNNS, whereas the peak at 399.6 eV corresponds to interlayer-bridging tertiary amine nitrogen (N–(C)₃). The high binding energy peak at 401.1 eV is attributed to edge amino functional groups (C–N–H).³⁶ The weak peak at 404.7 eV results from the charge effect induced by π bond excitation, consistent with that observed in pure CNNS.³⁷Fig. 4(c and d) display the C 1s high-resolution spectra of CNNS and CNNS/GY, respectively, both displaying the same peaks. CNNS exhibits a typical C 1s characteristic peak at 288.2 eV, attributed to the triazine unit N–C=N (sp²) coordination. The peak at 284.8 eV corresponds to adventitious carbon. Compared to pure CNNS, the CNNS/GY composite exhibits four distinct peaks in the C 1s spectrum at 284.5, 285.1, 286.1, and 287.6 eV, which are assigned to C=C (sp²), C≡C (sp), C–O, and C=O bonds, respectively. In addition, two peaks at 288.2 eV and 288.7 eV are observed in CNNS/GY, assigned to CNNS, corresponding to sp²-hybridized N–C=N bonds in the triazine units.²⁵ These spectral features collectively substantiate the incorporation of GY into the CNNS framework.

Fig. 4 (a) Full XPS spectra, (b) XPS N 1s core-level spectra, (c) XPS C 1s core-level spectra, and (d) XPS C 1s core-level spectra of CNNS/GY-5; (e) UV-vis DRS and (f) photoluminescence spectra of CNNS, CNNS/GY-1, CNNS/GY-5, and CNNS/GY-10.

UV-vis DRS and PL spectra were used to analyze the optical properties of the CNNS and CNNS/GY samples. As shown in Fig. 4(e and f), the light absorption characteristics of the CNNS/GY composites change significantly with increasing GY content. In Fig. 4(e), compared to pure CNNS (with an absorption edge around 457 nm), the absorption edges of CNNS/GY-1 (484 nm), CNNS/GY-5 (492 nm), and CNNS/GY-10 (502 nm) exhibit a progressive red shift. Based on the relationship E (eV) = hc/λ (nm), the corresponding bandgap energy decreases with increasing GY content, leading to a broadened visible-light response range. The redshift phenomenon results from the electronic coupling between the hybridized carbon network of GY and the π-conjugated system of CNNS, which facilitates an optimal balance between light absorption and charge transport, thereby providing a crucial foundation for enhancing its photocatalytic performance. Fig. 4(f) presents the PL spectra of the samples, recorded at an excitation wavelength of 320 nm. The PL intensity correlates directly with the recombination rate of photogenerated carriers. A lower PL strength indicates suppressed carrier recombination, suggesting increased availability of electrons and holes for photocatalytic reactions, thus enhancing photocatalytic activity.³⁸ After varying the degrees of CNNS/GY combination, the PL intensities are significantly reduced, suggesting that the highly conductive graphyne network in GY effectively facilitates the extraction and transport of photogenerated electrons, thereby suppressing the recombination of electron–hole pairs.

3.2. Photoelectrochemical performances

The transient photocurrent response displayed in Fig. 5(a) shows that pristine CNNS deliver a photocurrent density of approximately 0.45 µA cm⁻². Introducing GY markedly enhances the photocurrent, with the 5% GY sample reaching a maximum of about 2.5 µA cm⁻², corresponding to a ∼5.6-fold improvement over CNNS. This enhancement arises from the efficient electron-transport channels provided by the sp–sp² hybridized network of GY and the promoted spatial separation of charge carriers enabled by the Type-II band alignment at the heterointerface. Further increasing the GY content to 10% results in a reduced photocurrent (∼1.5 µA cm⁻²), indicating that excessive GY loading adversely affects charge collection. EIS is employed to investigate the charge transport dynamics at the semiconductor/electrolyte interface. As shown in Fig. 5(b), this figure presents an enlarged view of the Nyquist plot in the high-frequency region for the CNNS and CNNS/GY samples under dark conditions. Among these, the CNNS/GY-5 composite exhibits the smallest arc radius, reflecting the lowest charge-transfer resistance (R_ct) among all samples tested. Samples containing a higher GY content display larger radii, underscoring the importance of compositional balance for efficient interfacial charge transport. The oxygen evolution reaction (OER) performance of the samples is analyzed using LSV curves (Fig. 5(c)). As the GY content increases, the LSV curve shifts toward lower overpotential accompanied by a higher current density, indicating improved charge transfer kinetics. These results demonstrate that the incorporation of GY markedly enhances the water-splitting activity of CNNS and facilitates the transport of photogenerated holes, reaching optimal performance at 5% GY content. This effect is mainly ascribed to the nanoporous conductive network of GY, which accelerates hole transport, and the electron-rich regions formed at the heterointerface, which lower the OER energy barrier. The Tafel slopes are shown in Fig. 5(d). The CNNS/GY-5 composite demonstrates the smallest Tafel slope (89 mV dec⁻¹), substantially lower than that of pristine CNNS (133 mV dec⁻¹), indicating more favorable reaction kinetics.

Fig. 5 (a) Transient photocurrent curves under UV-visible light irradiation, (b) EIS under dark, (c) LSV curve under dark, and (d) the corresponding Tafel curve of CNNS, CNNS/GY-1, CNNS/GY-5, and CNNS/GY-10.

It can be concluded that a moderate amount of GY notably improves charge separation efficiency and interfacial carrier transport. The observed performance trends are consistent with the formation of efficient electron-transport pathways, thereby enhancing charge transfer kinetics, reducing overpotential and lowering the activation energy barrier, which would collectively contribute to the enhanced photocatalytic activity.

3.3. Photocatalytic properties

Fig. 6(a) shows the photocatalytic nitrogen fixation activity of CNNS/GY. As the light exposure duration extends, the ammonia concentration gradually increases in both the CNNS and CNNS/GY photocatalytic nitrogen reduction systems. The NH₄⁺ concentration increased nearly linearly with irradiation time, indicating a stable photocatalytic ammonia generation rate under the applied conditions. Fig. S5 shows the measured NH₄⁺ concentration catalyzed by CN and CNNS, which primarily demonstrates the superiority of exfoliated CNNS on photocatalytic nitrogen fixation. With the assistance of GY, the CNNS/GY-5 composite exhibits the optimal nitrogen fixation performance. As shown in Fig. 6(b), the NH₄⁺ yield of pristine CNNS is the lowest, at 327.89 µmol g_cat⁻¹ h⁻¹. In contrast, the NH₄⁺ yields of the CNNS/GY composites are 478.81 µmol g_cat⁻¹ h⁻¹ (CNNS/GY-1), 803.67 µmol g_cat⁻¹ h⁻¹ (CNNS/GY-5), and 629.25 µmol g_cat⁻¹ h⁻¹ (CNNS/GY-10). A series of control experiments were conducted to identify the optimal reaction system. As shown in Fig. 6(c), the NH₄⁺ yield without a sacrificial agent is 127.45 µmol g_cat⁻¹ h⁻¹, significantly lower than that observed in the methanol solution system. This is because methanol, acting as a hole scavenger, irreversibly traps holes on the illuminated semiconductor surface, suppressing electron–hole recombination. Subsequently, argon gas is injected in place of nitrogen gas. During this process, the NH₄⁺ yield is negligible, confirming that NH₄⁺ primarily originates from the N₂ gas supplied to the system. Fig. 6(d) illustrates the stability of the CNNS/GY-5 composite after three cycles of photocatalytic testing. The NH₄⁺ yield of CNNS/GY-5 remains at 82.5% of the initial value. Catalyst loss during the sequential washing process before each photocatalytic cycle is likely responsible for the observed decrease in photocatalytic performance. To further evaluate the structural stability of the catalyst after repeated use, the XRD patterns of CNNS/GY composites after three cycles were examined (Fig. S6). No obvious changes were observed compared with those of the fresh samples, indicating that the crystal structure is well maintained during the photocatalytic process. These results collectively demonstrate that the CNNS/GY photocatalyst exhibits good stability and recyclability. Overall, incorporating a suitable quantity of GY markedly improves the photocatalytic nitrogen-fixation activity of CNNS, which is in accordance with the photoelectrochemical results. The performance reaches its maximum at a GY loading of 5%, whereas a further increase to 10% leads to a noticeable decline in the overall N₂-fixation rate. This deterioration is likely associated with GY overloading, which introduces excessive conductive domains that facilitate charge recombination, undermining the transport of photogenerated holes to the catalyst surface, thereby suppressing the reaction efficiency.

Fig. 6 (a) Typical time course of NH₄⁺ production of CNNS, CNNS/GY-1, CNNS/GY-5, and CNNS/GY-10; (b) comparison of the NH₄⁺ synthesis rate; (c) NH₄⁺ synthesis rate of CNNS/GY-5 under different conditions; (d) three cycles of NH₄⁺ production rate over CNNS/GY-5 in photocatalytic nitrogen fixation reaction.

3.4. DFT calculation

The g-C₃N₄ surface and the CNNS surface were subjected to geometric optimization. The optimized CNNS and GY structures were then used to construct the CNNS/GY heterojunction, to which a 15 Å vacuum layer was added, followed by geometric optimization of the entire system. Fig. S7 shows the theoretical model of the electronic structure of the CNNS/GY heterojunction. In the differential charge density map of the CNNS/GY heterostructure, a contour value of 0.001 e Bohr⁻³ was adopted. The yellow and blue regions denote electron depletion and accumulation, respectively. As shown in Fig. 7(a and b), CNNS behaves as an electron-rich region whereas GY appears as an electron-deficient region, indicating that electrons transfer from the underlying CNNS layer to the overlying GY layer. This interlayer charge transfer promotes effective separation of photogenerated electron–hole pairs, thereby improving the photocatalytic activity of the heterojunction.³⁹ In the differential charge density maps of the CNNS/GY heterostructure after nitrogen adsorption (Fig. 7c and d), the region associated with the adsorbed nitrogen molecules appears blue, whereas the surrounding area is predominantly yellow. This distribution indicates that electrons are transferred from the CNNS/GY heterostructure to the nitrogen molecules, confirming the successful adsorption of nitrogen onto the CNNS/GY heterojunction.

Fig. 7 Differential charge density distributions of CNNS/GY: (a) side view and (b) oblique view; CNNS/GY with adsorbed N₂: (c) side view and (d) oblique view (isosurface value: 0.002e Å⁻³).

The calculated density of states (DOS) plot (Fig. 8) reveals that CNNS/GY exhibits a slightly higher electronic density near the Fermi level compared with pristine CNNS, suggesting an enhanced ability to supply electrons during catalytic processes. After N₂ adsorption, the total DOS of the heterostructure undergoes noticeable perturbations, with distinct changes appearing around −3 and −5 eV, corresponding to p-orbital electrons (Fig. S8).⁴⁰ These features indicate that the electronic states of CNNS/GY participate in the interaction with the adsorbed N₂ molecule, reflecting the formation of adsorption-induced electronic coupling between the heterojunction and N₂.

Fig. 8 Density of states (DOS) of CNNS, GY, and the CNNS/GY heterostructure (top panel), and DOS comparison of the CNNS/GY heterostructure before and after N₂ adsorption.

The electron density difference isosurface slices of CNNS and CNNS/GY are displayed in Fig. 9(a and b). In these plots, the blue and red regions represent electron accumulation and depletion, respectively. Compared with the N₂–CNNS system (Fig. 9(a)), the N₂–CNNS/GY heterostructure (Fig. 9(b)) exhibits markedly intensified charge redistribution around both the adsorbed N₂ molecule and the underlying CNNS layer. To further evaluate the N₂ activation capability, the adsorption energies on different surfaces were calculated (Fig. 9(c)). The adsorption energy of N₂ on CNNS is −0.89 eV, which further decreases to −1.26 eV after constructing the CNNS/GY heterostructure, indicating a strengthened interaction between the catalyst surface and N₂ molecules. This result is consistent with the charge redistribution observed above. This enhanced redistribution reflects a strengthened electronic interaction at the CNNS/GY interface, which facilitates more effective charge separation within the composite. Upon N₂ adsorption, charge accumulation around the adsorption site enables effective orbital interactions between the N₂ molecule and the π-conjugated framework of GY, thereby inducing polarization and facilitating the activation of the N≡N bond. The introduction of GY therefore modifies the local electronic environment of CNNS and establishes an interface capable of promoting charge transfer toward adsorbed N₂ molecules, providing a more favorable electronic configuration for subsequent photocatalytic reduction steps. Based on previous theoretical studies, the NRR process is generally described by an associative mechanism, and the distal pathway is adopted in this work for mechanistic analysis.^41–43 The proposed reaction pathway is illustrated in Fig. 9(d).

Fig. 9 (a) CNNS and (b) CNNS/GY heterostructure: electron density difference isosurface slices for N₂ adsorption (isosurface value: 0.002e Å⁻³); (c) N₂ adsorption energies on CNNS, GY, and CNNS/GY; (d) proposed schematic illustration of the reaction pathway for NH₃ formation on the CNNS/GY surface during photocatalytic N₂ reduction.

3.5. Mechanisms

The enhanced photocatalytic nitrogen-fixation performance of the CNNS/GY heterostructure arises from a synergistic interplay between its interfacial electronic structure, carrier dynamics, and N₂ activation capability. Integrating the experimental findings with DFT calculations reveals a coherent mechanistic picture as follows. As shown in Fig. 10, previous studies have determined the bandgap of CNNS to be 2.81 eV and that of GY to be 2.7 eV with the corresponding conduction band (CB) and valence band (VB) positions also illustrated.^9,44 This band structure analysis reveals a typical type-II alignment, with the CB of CNNS positioned more negatively than that of GY and the VB of GY more positive than that of CNNS. This configuration ensures that photogenerated electrons preferentially accumulate in GY while holes migrate toward CNNS. This alignment also meets the requirement of N₂/NH₃ redox potential (0.55 V vs. RHE).⁴⁵ The resulting spatial separation correlates with the high photocurrent density of the CNNS/GY-5 sample (Fig. 5a) and the significantly reduced Tafel slope (Fig. 5d). Owing to the unique structure of GY, the CNNS/GY interface provides a continuous sp–sp² hybridized conduction network, enabling more efficient interfacial electron transport. On the other hand, differential charge-density analysis shows evident electron transfer from CNNS to GY at equilibrium, establishing a built-in electric field across the intimate interface. This further proves that γ-graphyne can modulate the electronic environment of polymeric carbon nitride, thereby broadening the applicability of graphyne-based electronic regulation. This intrinsic interfacial field accounts for the reduced charge-transfer resistance and suppressed radiative recombination observed experimentally. What's more, pronounced charge redistribution around the N₂ molecules on the CNNS/GY interface has been revealed by DFT calculations, displaying enhanced electron donation into the antibonding orbitals of N₂ which favors N≡N bond weakening. Therefore, the electron-rich framework and enlarged pore structure of γ-graphyne create a more favorable adsorption microenvironment, consistent with the significantly higher NH₄⁺ yield of the CNNS/GY composite. Simultaneously, the improved OER kinetics, reflected by the lower overpotential and smaller Tafel slope, confirm that photogenerated holes are efficiently transported to the CNNS surface, while electrons accumulated in GY drive the multi-electron reduction of activated N₂. This cooperative enhancement of both oxidation and reduction half-reactions provides a balanced pathway for the dual-electron/hole requirements of photocatalytic ammonia synthesis.

Fig. 10 The proposed photocatalytic mechanism for the CNNS/GY heterojunction.

4. Conclusion

In summary, 2D/2D highly conjugated nanocomposites of CNNS/GY were successfully prepared. The composite with an optimal GY content exhibits a superior NH₄⁺ production rate of 803.67 µmol g_cat⁻¹ h⁻¹ with great stability. This work demonstrates that integrating γ-graphyne with exfoliated g-C₃N₄ creates an electronically coupled 2D heterostructure capable of fundamentally reshaping charge-transfer pathways and nitrogen-activation behavior. γ-Graphyne acts as an interfacial electronic modulator rather than a conductive component, inducing built-in electric fields, directional carrier migration, and strengthened N₂ adsorption, thereby enabling far more efficient photocatalytic nitrogen fixation. These mechanistic insights establish a transferable design principle for tuning the interfacial electronic structure in carbon-based photocatalysts and highlight γ-graphyne as a promising platform for solar-driven ammonia synthesis.

Author contributions

All authors have contributed sufficiently to the work. Dongmeng Wang: investigation, formal analysis, and writing – original draft; Xiaoqing Ma: supervision, writing – review and editing, and funding acquisition; Qiaodan Li: conceptualization and methodology; Xinyao Tian: data curation; Jilei Si: writing – formal analysis; Jiahao Wu: visualization; Liang Zhao: writing – review and editing; Zheyu Lian: writing – review and editing; Mauricio A. Melo: funding acquisition.

Conflicts of interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

All data generated or analyzed during this study are included in this published article and its supplementary information (SI). The SI includes additional experimental details, characterization data, and supplementary figures supporting the conclusions of this work. See DOI: https://doi.org/10.1039/d5nj04837k.

Acknowledgements

This work was financially supported by the Shanghai Sci-tech Co-research Program (No. 25HB2709300), the National Natural Science Foundation of China (NSFC) (No. 52002241), and the Class III Peak Discipline of Shanghai—Materials Science and Engineering (High-Energy Beam Intelligent Processing and Green Manufacturing). And we are also indebted to the support given by the Rio de Janeiro Research Foundation (FAPERJ, Process E-26/210.812/2021 and E-26/211.016/2026).

References

1. T. Kandemir, M. E. Schuster, A. Senyshyn, M. Behrens and R. Schlogl, Angew. Chem., Int. Ed., 2013, 52, 12723–12726.
2. R. Guan, X. Cheng, Y. Chen, Z. Wu, Z. Zhao, Q. Shang, Y. Sun and Z. Sun, Nano Res., 2023, 16, 10770–10778.
3. T. Hou, H. Peng, Y. Xin, S. Wang, W. Zhu, L. Chen, Y. Yao, W. Zhang, S. Liang and L. Wang, ACS Catal., 2020, 10, 5502–5510.
4. G. N. Schrauzer and T. D. Guth, J. Am. Chem. Soc., 1977, 99, 7189–7193.
5. S. Liu, Y. Wang, S. Wang, M. You, S. Hong, T.-S. Wu, Y.-L. Soo, Z. Zhao, G. Jiang, Q. Jieshan, B. Wang and Z. Sun, ACS Sustainable Chem. Eng., 2019, 7, 6813–6820.
6. Z. Zhao, S. Hong, C. Yan, C. Choi, Y. Jung, Y. Liu, S. Liu, X. Li, J. Qiu and Z. Sun, Chem. Commun., 2019, 55, 7171–7174.
7. X. Wang, K. Maeda, A. Thomas, K. Takanabe, G. Xin, J. M. Carlsson, K. Domen and M. Antonietti, Nat. Mater., 2009, 8, 76–80.
8. S. Yang, Y. Gong, J. Zhang, L. Zhan, L. Ma, Z. Fang, R. Vajtai, X. Wang and P. M. Ajayan, Adv. Mater., 2013, 25, 2452–2456.
9. X. Liao and X. Ma, J. Funct. Mater., 2023, 54, 8025–8034.
10. Alauddin, Z. U. Rehman, M. Bilal, J. A. Jrar, Z. Asghar, X. Yang, M. A. Khan, S. Ali, P. M. Ismail and J. Hou, Int. J. Hydrogen Energy, 2025, 127, 801–812.
11. Z. Asghar, Z. U. Rehman, F. K. Butt, K. Zheng, Y. Liu, H. Lin, Y. Zhang, J. Hou and X. Wang, Renewable Energy, 2025, 248, 123134.
12. C. Li, S. Yu, H. Che, X. Zhang, J. Han, Y. Mao, Y. Wang, C. Liu and H. Dong, ACS Sustainable Chem. Eng., 2018, 6, 16437–16447.
13. R. H. Baughman, H. Eckhardt and M. Kertesz, J. Chem. Phys., 1987, 87, 6687–6699.
14. A. Hirsch, Nat. Mater., 2010, 9, 868–871.
15. S. Zhang, J. Yun, L. Zeng, L. Yao, Z. Bi, C. Mai, P. Kang, J. Yan and Z. Zhang, New J. Chem., 2023, 47, 7084–7092.
16. F. Xu, K. Meng, B. Zhu, H. Liu, J. Xu and J. Yu, Adv. Funct. Mater., 2019, 29, 1904256.
17. Z. Wang, X. Ma, Q. Li, X. Tian, J. Si and M. A. Melo, Vacuum, 2024, 220, 112872.
18. J. Si, X. Ma, Q. Li, Z. Wang, J. Wu, D. Wang and M. A. Melo, J. Alloys Compd., 2025, 1037, 182331.
19. X. Tian, X. Ma, Q. Li, J. Si, D. Wang and J. Wu, J. Alloys Compd., 2025, 1022, 180072.
20. Y. Lin, H. Liu, C. Yang, X. Wu, C. Du, L. Jiang and Y. Zhong, Appl. Catal., B, 2020, 264, 118479.
21. J. Si, X. Ma, Q. Li, J. Wu and D. Wang, Int. J. Hydrogen Energy, 2025, 185, 151974.
22. L. Wu, Q. Li, C. Yang, Y. Chen, Z. Dai, B. Yao, X. Zhang and X. Cui, J. Mater. Sci.: Mater. Electron., 2020, 31, 8796–8804.
23. Z. Hao, Y. Chen, L. Wang and X. Cui, Chin. J. Inorg. Chem., 2022, 38, 2321–2341.
24. C. Chen, Y. Chen, J. Sun, Z. Chen, C. Zhou, M. Zhang, C. Sang and C. Zhang, New J. Chem., 2023, 47, 12570–12579.
25. H. Si, Q. Deng, C. Yin, J. Zhou, S. Zhang, Y. Zhang, Z. Liu, J. Zhang, J. Zhang and J. Kong, J. Alloys Compd., 2020, 833, 155054.
26. Y. Fang, Y. Xue, L. Hui, H. Yu and Y. Li, Angew. Chem., Int. Ed., 2021, 60, 3170–3174.
27. Y. Liu, Y. Xue, L. Hui, H. Yu, Y. Fang, F. He and Y. Li, Nano Energy, 2021, 89, 106333.
28. Q. Li, C. Yang, L. Wu, H. Wang and X. Cui, J. Mater. Chem. A, 2019, 7, 5981–5990.
29. M. D. Segall, J. D. L. Philip, M. J. Probert, C. J. Pickard, P. J. Hasnip, S. J. Clark and M. C. Payne, J. Phys.: Condens. Matter, 2002, 14, 2717.
30. J. P. Perdew, K. Burke and M. Ernzerhof, Phys. Rev. Lett., 1996, 77, 3865–3868.
31. H. Wang, Y. Sun, Y. Wu, W. Tu, S. Wu, X. Yuan, G. Zeng, Z. J. Xu, S. Li and J. W. Chew, Appl. Catal., B, 2019, 245, 290–301.
32. W. Ding, M. Sun, B. Gao, W. Liu, Z. Ding and S. Anandan, Dalton Trans., 2020, 49, 10958–10969.
33. Y. Wang, S. Zhao, Y. Zhang, J. Fang, Y. Zhou, S. Yuan, C. Zhang and W. Chen, Appl. Surf. Sci., 2018, 440, 258–265.
34. H. Ou, L. Lin, Y. Zheng, P. Yang, Y. Fang and X. Wang, Adv. Mater., 2017, 29, 1700008.
35. J. Zhao and J. Wang, J. Phys. Chem. C, 2017, 121, 21430–21438.
36. Q. Gu, Z. Gao, H. Zhao, Z. Lou, Y. Liao and C. Xue, RSC Adv., 2015, 5, 49317–49325.
37. P. Xia, B. Zhu, J. Yu, S. Cao and M. Jaroniec, J. Mater. Chem. A, 2017, 5, 3230–3238.
38. L. Yang, X. Liu, Z. Liu, C. Wang, G. Liu, Q. Li and X. Feng, Ceram. Int., 2018, 44, 20613–20619.
39. Z. Wang and F. Labat, J. Mater. Chem. A, 2025, 13, 3115–3131.
40. X. Hui, L. Li, Q. Xia, S. Hong, L. Hao, A. W. Robertson and Z. Sun, Chem. Eng. J., 2022, 438, 135485.
41. S. C. Jesudass, S. Surendran, J. Y. Kim, T.-Y. An, G. Janani, T.-H. Kim, J. K. Kim and U. Sim, Electrochem. Energy Rev., 2023, 6, 27.
42. R. G. de Araujo and J. Perez, ACS Electrochem., 2025, 1, 294–302.
43. H. Xian, H. Guo, Z. Chen, G. Yu, A. A. Alshehri, K. A. Alzahrani, F. Hao, R. Song and T. Li, ACS Appl. Mater. Interfaces, 2020, 12, 2445–2451.
44. T. Xie, X. Ma, Q. Li, X. Tian and J. Si, J. Solid State Chem., 2024, 337, 124793.
45. X. Chen, N. Li, Z. Kong, W.-J. Ong and X. Zhao, Mater. Horiz., 2018, 5, 9–27.

---