Boosting Charge Separation in NiS/C₃N₄ Type-II Heterojunction for Efficient Photoelectrocatalytic Water Reduction

Abstract

To tackle the intrinsic limitations of fast charge recombination and sluggish reaction kinetics in carbon nitride (C₃N₄) for photoelectrocatalytic (PEC) water reduction reaction, we constructed a NiS/C₃N₄ heterojunction photoelectrode via a sequential approach combining chemical vapor deposition and hydrothermal treatment. Compared with pristine C₃N₄, the introduction of NiS significantly reduced interfacial charge transfer resistance and effectively suppressed the photogenerated carrier recombination. Among all compositions investigated, the NiS-0.2 photoelectrode demonstrated a maximum photocurrent density of −13.44 mA cm⁻² at −0.8 V vs. RHE, representing a more than 6.7-fold enhancement in comparison to bare C₃N₄ (−2.00 mA cm⁻²). This remarkable improvement is attributed to the construction of an efficient type-II heterojunction between C₃N₄ and NiS. Under the driving force of the internal electric field at the interface, photoinduced electrons migrate from the conduction band of C₃N₄ to NiS, whereas holes move from the valence band of NiS to C₃N₄. This spatial separation mechanism, coupled with the role of NiS as an efficient active site for the water reduction reaction, synergistically enhances the overall PEC performance. This work offers a rational and feasible approach for designing efficient, stable, and cost-effective C₃N₄-based photoelectrodes.

1. Introduction

The growing global energy scarcity, along with the problems of environmental pollution, has made the prompt development of sustainable and clean alternative energy sources imperative [1]. Hydrogen, with its high energy density and totally carbon-neutral emissions, is considered one of the most promising energy carriers for the next generation [2,3]. Photoelectrochemical (PEC) water splitting, which utilizes solar energy to directly dissociate water into hydrogen and oxygen, enables the efficient conversion and storage of solar energy into chemical energy [4,5,6]. This process offers a potential pathway toward a carbon-neutral energy cycle [7,8]. The efficiency and practicality of solar-to-hydrogen conversion critically depend on the performance of photoelectrodes, which regulate three fundamental processes of light absorption, charge separation and transfer, and surface catalytic reactions [9,10]. Therefore, the pursuit of efficient, stable, and cost-effective photoelectrodes has become a central focus of scientific research in this field. Extensive endeavors have been devoted to investigating various semiconductor materials for PEC applications. To date, PEC water reduction systems based on inorganic p-type semiconductors, including metal oxides [11,12,13] and silicon [14,15], have reached advanced development stages. However, photocathodes still suffer from intrinsic limitations, including self-photo-corrosion in aqueous environments and rapid surface electron–hole recombination, which, respectively, result in severe stability issues and efficiency degradation.

Among various candidates, as a metal-free polymeric semiconductor, carbon nitride (C₃N₄) has emerged as a highly promising photocathode material [16,17]. It is endowed with advantages such as an optimal bandgap (~2.7 eV) that enables visible light harvesting, favorable band energy structures that encompass the water redox potentials, superior chemical stability, and the abundance of its constituent elements. Recent research advances have demonstrated that C₃N₄’s performance was enhanced through multiple tactics: elemental doping to optimize electronic structure and charge carrier concentration [18], along with morphology engineering to shorten charge diffusion distances to the surface [19]. Despite these inherent advantages and significant progress, the practical implementation of pristine C₃N₄ photocathodes remains fundamentally restricted by several inherent drawbacks: (i) fast recombination of photogenerated electron–hole pairs arising from poor charge carrier mobility, (ii) insufficient active sites for the kinetically sluggish water reduction, and (iii) poor interfacial contact between photocathode and substrate. To tackle these challenges, constructing heterojunctions through the coupling of C₃N₄ with secondary functional materials has been validated as a highly effective strategy, as well as using chemical vapor deposition (CVD) to overcome the poor film-forming properties of traditional methods such as drop coating or spin coating [20,21]. This approach can facilitate spatial charge separation and enhance the surface reaction kinetics. In particular, coupling with cocatalysts featuring high electrical conductivity and superior catalytic activity is highly desirable.

Transition metal sulfides, particularly nickel sulfides [22,23,24], have emerged as high-performance cocatalysts in PEC water reduction, addressing critical limitations of traditional semiconductor photoelectrodes such as sluggish HER kinetics and severe recombination of photogenerated charge carriers. Recent research has focused on nanoengineering strategies to optimize their efficacy; for example, loading nickel sulfide nanoparticles or nanosheets onto photoelectrode surfaces enables efficient extraction and transfer of photogenerated electrons, reduces HER overpotential, and enhances photocurrent and hydrogen production rates [25]. Constructing intimate heterojunctions optimizes interfacial charge separation/transport and synergistically broadens light absorption [26]. Crystal facet exposure and microstructure optimization increase active sites, electrical conductivity, and electrolyte diffusion, boosting intrinsic catalytic activity [27]. Overall, nickel sulfide research has advanced from performance validation to in-depth optimization of charge transport and stability via microstructure design, interface modulation, and multi-component synergy, providing a critical material basis for efficient, stable, and low-cost PEC systems.

Herein, we conceived and fabricated a novel heterostructure photocathode by in situ deposition of NiS nanoparticles onto a C₃N₄ thin film. We hypothesize that the tight interfacial contact between C₃N₄ and NiS will form an efficient heterojunction, promoting the efficient separation and rapid migration of photoinduced charge carriers. The metallic NiS particles will function as ultra-active water reduction sites, drastically reducing the reaction overpotential. Consequently, this synergistic effect is expected to simultaneously mitigate the key limitations of C₃N₄: charge recombination and slow surface kinetics. Through a series of structural, optical, electrochemical, and PEC characterizations, we systematically investigate the mechanism of performance enhancement. The optimized NiS/C₃N₄ heterojunction photoanode demonstrates a significantly enhanced photocurrent density and excellent stability compared to pristine C₃N₄. This study provides a rational design strategy for developing high-performance, low-cost C₃N₄-based photoanodes for PEC water reduction.

2. Experimental

2.1. Materials

Melamine (C₃H₆N₆, ≥99%), nickel (II) nitrate hexahydrate (Ni(NO₃)₂·6H₂O, ≥98%), and thiourea (CH₄N₂S, ≥99%) were acquired from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). Before use, the fluorine-doped tin oxide (FTO, Nippon Sheet Glass, Osaka, Japan, 1 × 2.5 cm²) substrates were cleaned in sequence with ethanol, acetone, and ethyl acetate (Sinopharm Chemical Reagent Co., Ltd., Shanghai, China). All reagents were employed directly as received without additional purification.

2.2. Fabrication of C₃N₄ Photoanode

The FTO-supported C₃N₄ photoanode was fabricated via chemical vapor deposition. To be concise, an FTO substrate and 1 g of melamine were positioned at the back and front of a crucible boat, respectively. After covering the boat with tin foil, it was then moved into a tube furnace and annealed at 550 °C for 1 h, 2 h, and 4 h, respectively, under nitrogen atmosphere with a heating rate of 2 °C/min.

2.3. Preparation of NiS/C₃N₄ Photoanode

A simple hydrothermal process was employed to synthesize the NiS/C₃N₄ photoanode. Briefly, nitrate hexahydrate (0.2 mmol) and thiourea (0.2 mmol) were dissolved and mixed in ethylene glycol (30 mL). The resulting mixed solution was then moved into a Teflon-lined stainless-steel autoclave holding two pieces of C₃N₄ placed back-to-back, followed by reaction at 160 °C for 16 h. The as-obtained samples were rinsed thoroughly with distilled water, subsequently dried under a nitrogen flow, and abbreviated as NiS-0.2. To examine the effect of NiS loading on the PEC performance of C₃N₄, additional samples (i.e., NiS-0.05, NiS-0.1, and NiS-0.25) were prepared using the same hydrothermal protocol, with only the amounts of thiourea and Ni(NO₃)₂·6H₂O adjusted.

2.4. Characterization

A Bruker D8 Advance diffractometer equipped with CuKα radiation (λ = 1.5406 Å) was employed for XRD measurements (Billerica, MA, USA). The instrument was operated at a voltage of 40 kV and a current of 40 mA, with scanning conducted over a 2θ range of 10° to 80° at a rate of 10° min⁻¹. Field emission scanning electron microscope (FESEM) images and energy-dispersive X-ray spectroscopy (EDS) analysis were acquired using a Zeiss SU 8010 microscope (Oberkochen, Germany). X-ray photoelectron spectroscopy (XPS) was employed to investigate the chemical states of elements in NiS-0.2 utilizing a Thermo VG Scientific Escalab 250 spectrometer with a monochromatic AlKα source (Waltham, MA, USA). An Edinburgh FI/FSTCSPC 920 spectrophotometer (Tokyo, Japan) was used to measure photoluminescence (PL) spectra under ambient conditions, setting the excitation wavelength at 355 nm. For Fourier transform infrared (FTIR) characterization, a Nicolet 5700 FT-IR Spectrometer (Bavaria, Germany) was employed to collect spectra over 650–4000 cm⁻¹ via the attenuated total reflection (ATR) method.

2.5. Photoelectrochemical (PEC) Measurements

A conventional three-electrode configuration was adopted to assess the PEC water reduction performance, with 0.5 M Na₂SO₄ used as the electrolyte. Herein, as-prepared samples functioned as the working electrode, while Pt wire served as the counter electrode, and the Ag/AgCl electrode functioned as the reference electrode, respectively. A HAMAMATSU LC8 L9588 Xe lamp (Beijing, China, 200 W) was employed for PEC measurements, maintaining the incident light intensity on the photoelectrode surface at 100 mW/cm². Linear sweep voltammetry (LSV) experiments were performed with potentials ranging from 0 to −0.8 V vs. RHE at 10 mV/s. Photo-assisted electrochemical impedance spectroscopy (EIS) was characterized under illumination at −0.4 V vs. RHE, with frequency varied from 10⁵ to 0.1 Hz. Moreover, chopped-light chronoamperometric (I-t) curves and photostability tests were conducted at −0.4 V vs. RHE.

3. Results and Discussion

The crystalline phase composition of the obtained samples was analyzed via XRD measurements. As depicted in Figure 1a, only diffraction peaks attributed to FTO and C₃N₄ are observed, with strong peaks at 37.6° and 43.9° identified as the FTO ’s characteristic diffraction peaks. To better observe the diffraction peak of C₃N₄, the magnified XRD patterns show a distinct diffraction peak attributed to the (002) crystal plane of C₃N₄ which can be clearly observed at 27.0° (Figure S1a) [28], and no obvious change in the diffraction peak is detected after NiS deposition (Figure S1b). This is primarily attributed to the poor crystallinity of NiS in NiS/C₃N₄ photoelectrodes [29]. The unaltered diffraction peaks of C₃N₄ in these composites indicate that the introduced NiS did not impair the nanostructure of C₃N₄. Furthermore, FTIR spectra of C₃N₄, NiS-0.05, NiS-0.1, NiS-0.2, and NiS-0.25 was performed to analyze the functional group information. As observed from Figure 1b, the absorption peak around 3300 cm⁻¹ is associated with N-H bonds, while the absorption peaks in the range of 1600–1200 cm⁻¹ correspond to CN heterocycles [30]. Additionally, the absorption peak in the 800 cm⁻¹ region is attributed to the s-triazine ring structure in C₃N₄ [31]. The peaks of samples with varying NiS contents exhibit only slight differences in intensity and shape compared to those of pure C₃N₄, indicating that loading NiS has minimal effect on the s-triazine ring structure of C₃N₄. The above results demonstrate that NiS/C₃N₄ has a structure similar to that of bare C₃N₄, and thus the NiS/C₃N₄ synthesized via the hydrothermal method possesses a stable chemical structure.

Typical morphologies of the as-fabricated C₃N₄ and NiS-0.2 were investigated via SEM. As displayed in Figure 2a, the pure C₃N₄ film deposited on the FTO substrate exhibits a compact nanosheet-like architecture and a smooth surface, which confirms that C₃N₄ was synthesized successfully. As for NiS-0.2 (Figure 2b), it is evident that the nanosheet structure of C₃N₄ remains intact following the deposition of NiS nanoparticles. Notably, C₃N₄’s surface becomes significantly rougher, owing to the uniform decoration of NiS nanoparticles on its surface. This retained structural integrity of C₃N₄ is crucial for maintaining its inherent properties, while the introduction of NiS-induced surface roughness could favor the accessibility of catalytic active sites. To accurately quantify the film thickness, additional SEM cross-sectional images were acquired (Figure 2c,d). Quantitative analysis reveals that C₃N₄ has a thickness of around 100 nm, whereas the NiS layer is roughly 250 nm. This thickness difference indicates the successful deposition of a continuous NiS layer on the C₃N₄ surface. Furthermore, energy-dispersive X-ray spectroscopy (EDS) analysis was performed to verify the elemental composition of NiS-0.2. The EDS results confirm the presence of C, N, Ni and S elements in NiS-0.2 (Figure S2).

To characterize the elemental compositions and oxidation states of the as-prepared NiS-0.2 photocatalysts, X-ray photoelectron spectroscopy (XPS) was employed. XPS spectra confirm that NiS-0.2 mainly consists of C and N elements, with relatively low contents of Ni and S elements, thus directly verifying the successful deposition of NiS on C₃N₄’s surface (Figure 3). To obtain a more in-depth understanding of the C-N chemical bonding in the composite, high-resolution C 1s and N 1s XPS spectra were analyzed. As illustrated in Figure 3a, deconvolution of the C 1s spectrum yields two peaks at 284.8 eV and 287.9 eV, with the former attributed to C-C/C=C groups and the latter to N-C=N groups, respectively [32]. Meanwhile, for the N 1s spectrum (Figure 3b), three resolved peaks are detected at 398.3 eV, 399.7 eV, and 400.9 eV, corresponding to sp²-hybridized nitrogen (C=N-C), tertiary nitrogen (N(C)₃), and amino groups, respectively [33]. For the Ni 2p spectrum (Figure 3c), four deconvoluted peaks are attributed to Ni²⁺ species: the peaks at 857.1 eV and 874.6 eV correspond to Ni 2p_3/2 and Ni 2p_1/2, respectively, while the remaining two are assigned to Ni 2p satellite peaks [34]. The characteristic peak at 169.9 eV is featured in the S 2p spectrum (Figure 3d), which attributes to the surface sulfur-based Ni-O-S species that originate from the sample undergoing oxidation when exposed to air [35].

To explore the effect of calcination time on C₃N₄ properties, we carried out linear sweep voltammetry (LSV) measurements on samples prepared with calcination durations of 1 h, 2 h, and 4 h (Figure S3). When the potential is −0.8 V vs. RHE, the corresponding current density of CN-1h, CN-2h, and CN-4h are 1.4, 5.5, and 3.2 mA cm⁻², respectively. Moreover, the onset overpotential of CN-2h is significantly lower than those of CN-1h and CN-4h, indicating that CN-2h initiates current generation earlier than the other two samples [36]. CN-2h exhibits relatively superior electrocatalytic activity under light irradiation and is more susceptible to electrochemical reactions. Notably, CN-2h shows favorable performance under both light and dark conditions, suggesting that a calcination time of 2 h is more conducive to enhancing its PEC performance. Furthermore, light-on/light-off cycling experiments (Figure S4a) confirmed that CN-2h exhibits a significantly higher photocurrent response than the other samples, accompanied by excellent stability. Electrochemical impedance spectroscopy (EIS, Figure S4b) measurements revealed a smaller arc radius for CN-2h, confirming it has lower charge transfer resistance compared to the other samples. Additionally, Mott–Schottky (M-S, Figure S5) analysis yielded a more negative flat band potential for CN-2h, indicating that it possesses a more negative conduction band position.

After optimizing the calcination conditions for C₃N₄, the cocatalyst NiS was loaded onto C₃N₄ via a hydrothermal method to enhance its photogenerated charge separation efficiency. Meanwhile, we explored the impact of NiS content on C₃N₄’s PEC performance, and a series of samples with varying NiS contents (denoted as NiS-0.05, NiS-0.1, NiS-0.2, and NiS-0.25) were prepared. From Figure 4, it is clear that the photocurrent of NiS-loaded C₃N₄ photocatalysts is notably superior to that of pure C₃N₄, with NiS-0.2 exhibiting a particularly remarkable enhancement. At a voltage of −0.8 V vs. RHE, the current densities of C₃N₄, NiS-0.05, NiS-0.1, NiS-0.2, and NiS-0.25 are −2.00 mA/cm², −8.54 mA/cm², −12.53 mA/cm², −13.44 mA/cm², and −8.96 mA/cm², respectively. In comparison, the absolute value of the current density for NiS-0.2 is much larger than that for pure C₃N₄, indicating that NiS-0.2 possesses the best performance in PEC water reduction [37].

In the field of PEC reaction, the performance of a photoelectrode is first and foremost determined by its capacity to absorb light energy. As a key carrier for light energy conversion, the stronger a photoanode’s capacity to absorb light, the more likely it is to efficiently capture light energy and convert it into usable electronic energy, thereby significantly enhancing its own PEC performance. As is evident from Figure 5 and Figure S6, pristine C₃N₄ and NiS have an absorption edge of roughly 447 nm and 712 nm, with a corresponding band gap of 2.77 and 1.74 eV [38]. After loading with NiS, the absorption edges of the samples do not shift significantly, remaining around 450 nm. This result indicates that the optical properties of the samples do not undergo significant change after loading the cocatalyst, suggesting that the loading of the cocatalyst has little impact on the optical properties of C₃N₄.

To study the influence of NiS loading content on the charge transport properties of photogenerated charges at the NiS/C₃N₄ interface and electrolyte/NiS interface, photoelectrochemical impedance measurements were performed [39]. It is apparent from Figure 6a that the impedance of NiS-loaded samples is significantly lower than that of pure C₃N₄. In particular, the semicircle radius of NiS-0.2 in the low-frequency region is much smaller than that of pure C₃N₄ and other samples. This observation indicates that NiS-0.2 exhibits the smallest charge transfer resistance at the solid–liquid interface [40], which corresponds to the highest charge transfer efficiency. Meanwhile, stability is a crucial factor for high-performance photoanodes. Thus, switch photocurrent measurements were employed to evaluate this property. As presented in Figure 6b, under periodic light on–off cycles, all samples exhibited consistent “response-recovery” behavior in their photocurrent: upon light irradiation, the photocurrent rapidly reached a stable plateau, and it immediately returned to the baseline once the light was turned off, with no noticeable attenuation over repeated cycles. This phenomenon clearly demonstrates the excellent photostability of the NiS/C₃N₄ composite photoanodes. Quantitative analysis of the photocurrent data revealed that the photocurrent densities of pure C₃N₄, NiS-0.05, NiS-0.1, NiS-0.2, and NiS-0.25 were −0.38, −0.49, −0.73, −1.58, and −1.47 mA/cm², respectively. Notably, the NiS-0.2 sample exhibited the significantly highest absolute photocurrent density, which confirms the superior conductivity of NiS that optimally facilitates the separation efficiency and migration rate of photogenerated charge carriers [41].

As documented in the literature, when semiconductor materials absorb photons to generate photogenerated electrons and holes, radiative recombination of these charges gives rise to photoluminescence (PL) signals, thereby yielding a positive correlation between PL intensity and the degree of photogenerated charge recombination. Bare C₃N₄ exhibits a strong PL signal, indicative of severe photogenerated charge recombination [42], as displayed in Figure 7. This, in turn, diminishes the number of charges available for separation, transmission, and photocurrent generation, consequently resulting in a low photocurrent response. In contrast, compared to bare C₃N₄, loading of NiS yields notable quenching of the PL signal, demonstrating that NiS can effectively suppress photogenerated charge recombination and enable more charges to undergo high-efficiency separation and migration [43], and this is beneficial for boosting the photocurrent property. This phenomenon is typically ascribed to the function of NiS as a cocatalyst that promotes charge transfer and reduces internal recombination within C₃N₄. Notably, NiS-0.2 exhibits the lowest PL signal, suggesting the weakest photogenerated charge recombination.

Furthermore, to clarify the reaction mechanism of PEC water reduction over the NiS/C₃N₄ heterojunction, we characterized and analyzed the semiconductor band structure in detail. For the semiconductors, their valence band (VB) positions were obtained through extrapolating the linear segment of VB-XPS spectra to the energy axis. As presented in Figure 8a,b, the valence band maxima (VBM) of C₃N₄ and NiS were calculated to be +1.54 eV and +1.64 eV, respectively, according to the equation [44]

$$E_{VB,NHE}=E_{VB, XPS}+\varphi -E_{VAC}$$

where Φ is an instrument-specific energy level (4.6 eV), and E_VAC is the vacuum energy level (4.5 eV). Integrating results from UV-Vis absorption spectroscopy, the conduction band (CB) position of C₃N₄ and NiS was ascertained to be approximately −1.23 eV and −0.1 eV, respectively, using the band gap relationship (E_g = E_VB − E_CB).

Furthermore, via Mott–Schottky curve analysis, a positive slope indicates that NiS is an n-type semiconductor by extrapolating the linear portion of the curve (Figure S7) [45]. Due to precise energy level alignment with C₃N₄, NiS acts as a “charge transfer bridge” that enables photogenerated electrons in the C₃N₄ CB transfer to the NiS CB, whereas photogenerated holes in the NiS VB migrate to the C₃N₄ VB (Figure 8c). The introduction of NiS fundamentally reconstructs the charge transfer pathway of the heterojunction, which not only effectively inhibits bulk recombination of photogenerated electrons and holes within C₃N₄, but also achieves the directional separation of photogenerated charges, thus remarkedly enhancing photogenerated charge separation efficiency. Moreover, the deposited NiS serves as an active site, capturing photogenerated electrons transferred to its own CB and lowering the activation energy barrier of the water reduction half-reaction, thus facilitating the participation of many photogenerated electrons in the reaction. In summary, via the synergy of energy level alignment, photogenerated carrier separation, and surface catalysis, NiS markedly enhances the PEC water reduction performance of C₃N₄-based heterojunctions.

4. Conclusions

This study successfully designed and fabricated a novel NiS/C₃N₄ heterojunction photoelectrode for efficient PEC water splitting. The C₃N₄ film prepared via chemical vapor deposition demonstrated optimal intrinsic PEC activity and charge transport properties. Deposited NiS nanoparticles not only functioned as efficient HER active sites to significantly reduce reaction overpotential but also formed a tightly bonded heterojunction interface with C₃N₄. Consequently, NiS loading substantially enhanced the photocurrent response and stability of C₃N₄, with the NiS-0.2 composite photoelectrode exhibiting the superior performance, achieving a photocurrent density more than 6.7 times as high as that of bare C₃N₄. The type-II heterojunction formed between C₃N₄ and NiS facilitated the spatial separation and directional transfer of photogenerated electrons and holes under illumination, effectively suppressing carrier recombination and thereby significantly improving charge separation and transfer efficiency. Overall, this present work affords valuable theoretical enlightenment and experimental support for the advancement of high-efficiency C₃N₄-based photoelectrode materials through the synergistic strategy of heterojunction construction.