Evaluation of Efficient and Noble-Metal-Free NiTiO₃ Nanofibers Sensitized with Porous gC₃N₄ Sheets for Photocatalytic Applications

Abstract

One-dimensional nickel titanate nanofibers (NiTiO₃ NFs) were synthesized and loaded with acetic acid-treated exfoliated and sintered sheets of graphitic carbon nitride (AAs-gC₃N₄) to fabricate a unique heterogeneous structure. This novel fabrication method for porous AAs-gC₃N₄ sheets using acetic acid-treated exfoliation followed by sintering provided gC₃N₄ with a surface area manifold larger than that of bulk gC₃N₄, with an abundance of catalytically active sites. Hybrid photocatalysts were synthesized through a two-step process. Firstly, NiTiO₃ NFs (360 nm in diameter) were made by electrospinning, and these NiTiO₃ NFs were sensitized with exfoliated gC₃N₄ sheets via a sonication process. Varying the weight ratio of NiTiO₃ fibers to porous AAs-gC₃N₄ established that NiTiO₃ NFs containing 40 wt% of porous AAs-gC₃N₄ exhibited optimal activity, i.e., removal of methylene blue and H₂ evolution. After 60 min exposure to visible light irradiation, 97% of the methylene blue molecules were removed by the hybrid photocatalyst, compared with 82%, 72%, and 76% by pristine AAs-gC₃N₄, NiTiO₃ NFs, and bulk gC₃N₄, respectively. The optimal structure also displayed excellent H₂ evolution performance. The H₂ evolution rate in the optimal sample (152 μmol g⁻¹) was 2.2, 3.2 and 3-fold higher than that in pure AAs-gC₃N₄ (69 μmol g⁻¹), NiTiO₃ NFs (47 μmol g⁻¹) and bulk gC₃N₄ (50 μmol g⁻¹), respectively. This clearly shows that the holey AAs-gC₃N₄ nanosheets interacted synergistically with the NiTiO₃ NFs. This extended the lifetime of photogenerated charge carriers and resulted in superior photocatalytic activity compared with pristine NiTiO₃ NFs and bulk gC₃N₄. The higher Brunauer-Emmett-Teller surface area and the presence of many catalytically active sites also enhanced the photocatalytic performance of the hybrid sample. Moreover, through photoluminescence and photocurrent response analysis, a significant decrease in the recombination losses of the hybrid photocatalysts was also confirmed. Thus, this is a novel strategy to fabricate highly efficient photocatalysts with precisely tunable operating windows and enhanced charge separation.

1. Introduction

Today, with burgeoning water contamination issues and a looming energy crisis, the utilization of renewable energy is of particular interest [1]. The most efficient and ecofriendly approach to renewable energy utilization is the use of photoenergy for photocatalytic removal of pollutants and H₂ evolution via water splitting [2]. Numerous photocatalysts have been developed over the last several decades, but a more efficient photocatalyst is still needed for practical applications; low solar-to-chemical energy conversion efficiency and complex chemical reactions need to be overcome [3].

Recently, the photocatalytic activity of titanium (Ti)-based perovskite oxides, such as SrTiO₃, CoTiO₃, ZnTiO₃, and NiTiO₃, have attracted significant attention [4,5,6,7]. Nickel titanate (NiTiO₃) has the ilmenite structure in which both Ni and Ti have octahedral coordination with alternating cation layers occupied by Ni and Ti [8]. This titanate has bandgap energy in the range of around 2.14 to 2.50 eV depending on the morphology, size, and synthetic route [9]. The material is highly stable in an oxidizing environment and under light irradiation. However, its low bandgap energy reduces its quantum efficiency when used as an individual photocatalyst [10].

Graphitic carbon nitride (gC₃N₄), a novel metal-free polymeric semiconductor, is another extensively used photocatalyst because of its superior photocatalytic performance [11]. However, gC₃N₄ suffers from low sunlight absorption, low quantum efficiency, and high recombination rates. Several techniques have been used to improve the photocatalytic activity of gC₃N₄ by increasing its surface area, doping, and formation of heterojunctions with other semiconductors. Many heterostructures, such as NiTiO₃/gC₃N₄, NiTiO₃/TiO₂, gC₃N₄/Bi₂WO₆, gC₃N₄/BiOCl, gC₃N₄/SnO₂, gC₃N₄/WO₃, Fe₂O₃/NiTiO₃, and gC₃N₄/ZnO [12,13,14,15,16,17], display enhanced photocatalytic activity. Notably, two-dimensional (2D) amorphous nanosheets display superior electrocatalytic performance compared with their crystalline counterparts due to the abundant active sites in amorphous structures [18]. First, the high density of well-distributed pores provides abundant active sites at the pore edges for redox reactions. Second, the abundant pores in the highly porous nanosheets can serve as permeable channels for the diffusion of reactants, thus making more subsurface materials available to be catalytically active. Third, this unique structure can shorten the diffusion lengths of photoexcited electrons from the interior to the edges for improved electron transport. The abundant interconnected pores throughout the nanosheets ensure ready access to active sites and suppress charge carrier recombination.

Morphological tuning of the material, especially one having a one-dimensional (1D) structure, can significantly improve photocatalytic activity by increasing the specific surface area and improving optical emission, electron transport, and chemical stability [19,20]. A 1D nickel titanate nanofiber (NiTiO₃ NF) based photocatalyst coupled with a 2D porous laminar structure would form an efficient photocatalyst addressing the intrinsic drawbacks of NiTiO₃ and gC₃N₄ for photocatalytic applications. To the best of our knowledge, no similar research has been reported concerning 1D/2D NiTiO₃/AAs-gC₃N₄-based photocatalysts for water-treatment applications [21].

Based on the above considerations, we report a cost-effective route to a unique heterostructure with enhanced photocatalytic performance formed by combining 1D NiTiO₃ NFs with holey 2D AAs-gC₃N₄ nanosheets. The fabrication process involved electrospinning NiTiO₃ NFs and their heterostructure with porous AAs-gC₃N₄ nanosheets obtained from one-step acid etching followed by thermal pyrolysis. A series of NiTiO₃/AAs-gC₃N₄ heterostructures, varying weight fractions were successfully fabricated and used for the removal of methylene blue (MB, a model pollutant) and H₂ evolution. The optimal combination of NiTiO₃/AAs-gC₃N₄ (60/40) rapidly removed MB and generated the highest H₂ yield. The improved performance was attributed to enhanced optical absorbance and efficient separation and movement of photogenerated charge carriers. This work is expected to increase interest in developing highly efficient and novel heterogeneous photocatalysts.

2. Results and Discussion

2.1. Field-Emission Scanning Electron Microscopy (FE-SEM)

The textural structure and morphology of pure NiTiO₃ NFs, gC₃N₄, AAs-gC₃N₄ (holey gC₃N₄), and hybrid structures were analyzed by FE-SEM (Figure 1). The observed rough-surfaced NiTiO₃ NFs were several micrometers long and of relatively uniform diameter (Figure 1a). Bulk gC₃N₄ consisted of disorderly stacked and irregularly sized clusters (Figure 1b), while AAs-gC₃N₄ had a loose, rough, and etched surface with abundant pores in its structure (Figure 1c). These abundant pores in AAs-gC₃N₄ serve as reaction centers by providing a large number of active sites at the pore edges and also effectively improve photogenerated carrier separation by shortening their transfer lengths. The hybrid structures were composed of NiTiO₃ NFs and AAs-gC₃N₄ distributed throughout the samples. High-magnification SEM images also revealed a coating of AAs-gC₃N₄ over the NiTiO₃ NFs (Figure 1d,e). SEM-energy-dispersive X-ray spectroscopy (Figure S1) confirmed the presence of essential elements. The observed uniform distribution of N over the NiTiO₃ NFs was indicative of the coating of AAs-gC₃N₄ sheets. In addition to this coating, a separate bulky structure of AAs-gC₃N₄ was also observed. This structure may improve light absorbance and then charge separation characteristics to enhance the photocatalytic performance of the photocatalysts. It was also confirmed that AAs-gC₃N₄-rich areas concealing uncoated NiTiO₃ NFs were formed with increasing AAs-gC₃N₄ loading (gnT4060) (Figure 1f).

2.2. Transmission Electron Microscopy (TEM) Analysis

The morphologies and crystal structures of pure NiTiO₃ NFs, gC₃N₄, AAs-gC₃N₄, and hybrid structures were further substantiated using TEM (Figure 2). The NiTiO₃ NFs had an average diameter of 360 nm. In the bulk gC₃N₄, aggregated irregular thick black particles were observed, contrary to the thin transparent porous graphene-like sheets in AAs-gC₃N₄ (Figure 2a–d).

The presence of most pores and thin layers in acid-etched gC₃N₄ was confirmed in AAs-gC₃N₄ (Figure 2d). The average pore size was approximately 14.5 nm. The presence of NiTiO₃ NFs with AAs-gC₃N₄ sheets in the hybrid structures was confirmed which reveals the formation of the hybrid photocatalyst. Moreover, through elemental mapping in scanning transmission electron microscopy, the NiTiO₃ NFs were found anchored to the surface of the AAs-gC₃N₄ sheets (Figure S5).

High magnification TEM images (Figure 2f,h) of the hybrid structures also revealed that the crystalline structure of NiTiO₃ NFs was maintained in the hybrid samples. A thin layer of ca. 10 nm and 13 nm of AAs-gC₃N₄ was found coated on the surface of gnT4060 and gnT6040 respectively. This uniform coverage of AAs-gC₃N₄ behaved as a junction between the fibers, as already revealed by SEM. The estimated interplanar distance in NiTiO₃ NFs was approximately 0.25 nm (Figure 2g), which is similar to the values already reported in the literature [22,23]. Selected area electron diffraction (Figure 2h inset) confirmed the polycrystalline nature of the hybrid structure, which was consistent with the presence of various planes of NiTiO₃ NFs. This close interaction between NiTiO₃ NFs and holey AAs-gC₃N₄ could enhance charge separation and transfer and thereby improve efficient photocatalytic activity.

2.3. X-ray Diffraction Analysis

The phase structures of the pristine (NiTiO₃ NFs, gC₃N₄, and AAs-gC₃N₄) and hybrid (NiTiO₃/AAs-gC₃N₄) photocatalysts were determined by X-ray diffraction analysis (Figure 3a). For the pristine NiTiO₃ NFs sample, the diffraction peaks at 24.1°, 33.1°, 35.7°, 49.5°, 49.5°, and 69.2° were indexed to (012), (104), (110), (024), and (208) crystal planes of ilmenite NiTiO₃ NFs, respectively, according to Joint Committee on Powder Diffraction Standards (JCPDS) 01-083-0198. The other two diffraction peaks observed in the NiTiO₃ NFs structure at 25.29° and 27.39° were attributed to the rutile (101) and anatase (110) planes TiO₂ (JCPDS card numbers 01.071-1168 and 01-076-0317, respectively). The relatively high sintering temperature (600 °C) in the air may have favored the formation of the TiO₂ phases [24]. Pristine gC₃N₄ and AAs-gC₃N₄ showed two characteristic diffraction peaks at 27.8° and 14.1°, which were associated with the (002) for conjugated aromatic segments (CN) and (100) for the structure of tri-s-triazine units in the gC₃N₄ structure, respectively [25]. The prominent peak (002) in AAs-gC₃N₄ is broader and shifted slightly towards the high-angle side as compared with that of pure gC₃N₄; this can be due to exfoliation and higher amorphous regions. The diffraction patterns of the various NiTiO₃/AAs-gC₃N₄ hybrid structures exhibited peaks matching the NiTiO₃ NFs and AAs-gC₃N₄ phases, which confirmed the hybrid structure formation. Additionally, the intensities of the NiTiO₃ NFs and AAs-gC₃N₄ peaks in the hybrid structures resembled those of the pristine materials, although increased loading of AAs-gC₃N₄ decreased the intensities of the NiTiO₃ NFs peaks. A similar trend has been reported for other gC₃N₄-based hybrids [10,26].

2.4. Fourier Transform Infrared (FT-IR) Spectroscopy Analysis

To investigate functional groups and their chemical states, FT-IR spectroscopy was used for NiTiO₃ NFs, gC₃N₄, AAs-gC₃N₄, and the hybrid samples (Figure 3b). Peaks observed at 455, 561, and 687 cm⁻¹ were attributed to vibrations of the Ni−O, Ti−O, and Ni−O−Ti functional groups of NiTiO₃ NFs, respectively [27]. These peaks appeared both in the pristine and hybrid materials. The spectrum of gC₃N₄ and AAs-gC₃N₄ showed several peaks in the region of 1700−1200 cm⁻¹ corresponding to the breathing mode of C−N heterocycles. The characteristic peak that appeared at 810 cm⁻¹ was attributed to the hydroxyl (−OH) and the amine (−NH) groups from water molecules and the partial hydrogenation of N atoms in gC₃N₄ and AAs-gC₃N₄. The main peaks in the heterostructures corresponded to the NiTiO₃ NFs and AAs-gC₃N₄ nanostructures, which provided evidence of their coexistence in a hybrid structure.

2.5. Surface Area and Pore Size Distribution

The effect of loading exfoliated and porous sheets of AAs-gC₃N₄ on the surface of 1D NiTiO₃ NFs was assessed by measuring the N₂ adsorption-desorption isotherms and the Barrett-Joyner-Halenda (BJH) pore size distribution of the pure and hybrid samples (Figure 4). The samples were degassed overnight (12 h) at 100 °C in vacuum condition. All samples displayed H3 hysteresis and the type-IV isotherm (Figure 4a), which indicated the presence of mesopores. The calculated Brunauer-Emmett-Teller (BET) surface areas of NiTiO₃ NFs, gC₃N₄, AAs-gC₃N₄ and gNT6040 at 0.5–1.0 P/P₀ were 11.048, 11.681, 126.84 and 46.0890 m² g⁻¹, respectively. The enhanced surface area of the gnT6040 sample was approximately four-fold higher than that of the NiTiO₃ NFs and bulk gC₃N₄. This was attributed to the introduction of porous sheets of AAs-gC₃N₄ on the NiTiO₃ NFs; sheets of AAs-gC₃N₄ had a higher specific surface area compared with bulk gC₃N₄ (Table S1). Using the BJH method (Figure 4b), the pore volume of the hybrid structure (0.2279 cm³ g⁻¹) was six-times higher than that of pristine NiTiO₃ NFs (0.0357 cm³ g⁻¹). This analysis suggested that the porous hybrid structure, with its greater number of accessible active sites, was responsible for the improved photocatalytic performance. It should also be pointed out that in comparison with AAs-gC₃N₄ (126.84 m² g⁻¹), the hybrid structure gnT6040 (46.08 m² g⁻¹) showed a lower specific surface area, but higher photoactivity. Lower surface area could be due to a higher amount of NiTiO₃ NFs (60 wt%) in the hybrid structure as compared to AAs-gC₃N₄ (40 wt%). Also, this observation is an indirect indication that the dynamics process of photoinduced charge transfer occurring through the interface between AAs-gC₃N₄ and NiTiO₃ NFs may be crucial to the photocatalytic activity. Successful charge transfer occurring in the heterostructure interface is also confirmed by PEC and PL analysis.

2.6. X-ray Photoelectron Spectroscopy (XPS) Analysis

XPS was used to investigate the chemical states and elemental composition of the samples. The survey spectra (Figure S3) show that the main elements on the surface of NiTiO₃ NFs are Ni, Ti, O, and some amount of external carbon contaminant and the main elements of AAs-gC₃N₄ are C, N, and adsorbed O. The as-prepared gnT6040 was composed of NiTiO₃ NFs and AAs-gC₃N₄. XPS spectra of gnT6040 were further analyzed (Figure 5) to investigate the chemical states of its elements and the interaction between NiTiO₃ NFs and AAs-gC₃N₄. The Ni 2p signal (Figure 5a) appeared as a doublet at 855.45 and 873.38 eV, which were attributed to 2p3/2 and 2p1/2 states, respectively. Two broad peaks at 861.38 and 880.98 eV were assigned as Ni²⁺ satellites. In the Ti 2p spectra (Figure 5b), peaks at 458.28 (2p3/2) and 464.37 eV (2p1/2) were assigned to the Ti⁴⁺ oxidation state and those at 457.44 and 463.80 eV to the Ti³⁺ state. The appearance of the Ti–N peak at 458.93 eV indicated a strong interaction between Ti from NiTiO₃ NFs and N from the AAs-gC₃N₄ skeleton [28,29]. The formation of this Ti–N linkage decreased the recombination rate of photogenerated charges in the hybrid structure due to higher charge separation efficiency.

The O 1s peak (Figure 5c) was deconvoluted into three peaks, with the major one at 529.47 eV assigned to O²⁻ in the NiTiO₃ NFs structure. The peak at 532.68 eV resulted from the ˙OH radicals adsorbed on the surface of NiTiO₃ NFs, while the peak appearing at 530.57 eV was attributed to lattice O²⁻ groups. The C 1s peak was deconvoluted into four peaks (Figure 5d) having binding energies at 284.29, 285.40, 287.91, and 288.79 eV, which were assigned to C–C, C–NH₂, C–(NH)_x, and N–C=N bonds, respectively. The N 1s signal was also deconvoluted into three peaks. The peaks at 398.48 and 399.72 eV were derived from sp²-hybridized N atoms (C=N–C) and tertiary N atoms (N–C₃), respectively. The peak at 400.68 eV was assigned to the amino group (C–N–H) (Figure 5e). Another small peak at 396.75 eV corresponded to the Ti–N peak linkage between NiTiO₃ NFs and AAs-gC₃N₄ [27,28]. Thus, the overall XPS analysis indicated the successful formation of the NiTiO₃/AAs-gC₃N₄ composite with strong interaction between NiTiO₃ NFs and AAs-gC₃N₄. The XPS results were consistent with the SEM, TEM, X-ray diffraction, and FT-IR analyses confirming the formation of the heterostructure between NiTiO₃ NFs and AAs-gC₃N₄ nanosheets.

2.7. Optical Absorbance

The light-harvesting capacity and bandgap energy of the pristine and hybrid photocatalysts were measured by analyzing their ultraviolet-visible (UV-vis) absorbance spectra (Figure 6a,b). The absorption band edges of pure NiTiO₃ NFs and gC₃N₄ were observed at approximately 430 and 400 nm, respectively. The composite structures extended the absorbance to the visible region from 450 to 900 nm due to the introduction of porous and thin sheets of AAs-gC₃N₄ between the NiTiO₃ NFs. The observed progressive color change of the different photocatalysts resulted from multiple reflections of incident light and was attributed to the porous framework (Figure 6a inset). The slight blue shift noted in the intrinsic absorption edge of the hybrid structure compared with bulk gC₃N₄ suggested the quantum confinement effect [30]. Bandgap energies calculated from the Kubelka-Munk function increased slightly from 2.38 eV for bulk gC₃N₄ to 2.41 eV for the hybrid structure. These changes in the band edge absorption indicated the ability of the photocatalysts to absorb light in the visible range. It is well-known that enhanced surface area, a porous framework, and increased bandgap are beneficial to light absorption, charge separation, and mass transfer and lead to improved photocatalytic properties. Furthermore, the increasing bandgaps from bulk gC₃N₄ to AAs-gC₃N₄ were also confirmed by a hypsochromic shift of the photoluminescence spectra (Figure 6c).

2.8. Photoluminescence (PL) Analysis

Charge carrier recombination is one of the most serious issues in photocatalysis. We analyzed the room-temperature photoluminescence (PL) emission spectra to determine the effective charge separation and transport in the pristine and hybrid structures (Figure 6c). PL was measured at an excitation wavelength of 325 nm. The strongest emission peak at 490 nm for gC₃N₄ indicated a high rate of carrier recombination which was further reduced in AAs-gC₃N₄. When AAs-gC₃N₄ was added to NiTiO₃ NFs, remarkable peak quenching was observed. This remarkable reduction in the intensity of the PL peak for the hybrid structure relative to that of bulk gC₃N₄ indicated a much lower photogenerated charge carrier recombination rate. Pure NiTiO₃ NFs showed much weaker PL emission (Figure S4) due to the lesser number of photogenerated charge carriers. Thus, PL analysis revealed that the hybrid structures exhibited fewer recombination losses compared with the pristine photocatalysts. Quantum confinement was also corroborated by the PL peak shifting from 490 to 460 nm.

2.9. Photoelectrochemical Performance

An efficient photocatalyst absorbs incident photons efficiently, converts them into photogenerated charge carriers (electrons and holes), and keeps them separated to successfully participate in surface reactions. This ability of the photocatalyst can be evaluated by measuring the transient photoresponse. In our study, we analyzed the intensity of photocurrent generated from pristine NiTiO₃ NFs, gC₃N₄, AAs-gC₃N₄, and the optimum hybrid structure (gnT6040) during several one-minute on/off cycles of light irradiation (Figure 6d). The intensity of photocurrent showed by the hybrid structure was much higher than the separate NiTiO₃ NFs, gC₃N₄, and AAs-gC₃N₄ photocatalysts. This enhanced photocurrent in the hybrid structure indicated that the presence of abundantly available charge carriers for rapid photocatalytic activity. This behavior also discovered a strong interaction between NiTiO₃ NFs and AAs-gC₃N₄ sheets and low interfacial transfer and separation of photogenerated charges via heterojunction formation. These results also reinforced the minimal recombination losses revealed by the PL analysis.

2.10. Removal of MB Dye

The water treatment capabilities of the pristine and the hybrid photocatalysts were evaluated using percentage removal of MB as a model pollutant under visible light irradiation (Figure 7). In this experiment, the removal of MB dye molecules was analyzed via an adsorption process. In the experiment, MB removal was measured by maintaining the MB suspension in the dark for one hour with constant stirring. This hour-long stirring period ensured adsorption equilibrium before light irradiation. A controlled MB removal experiment was carried out without adding any catalysts to assess the stability (photolysis) of MB. As observed, change in the MB concentration was negligible initially when there was no catalyst added in the solution. This confirms that MB was stable under visible light for up to 1 h. However, adding catalyst significantly decreased the MB concentration. The MB removal performance of pristine NiTiO₃ NFs (72%), gC₃N₄, (76%), and AAs-gC₃N₄ (82%) was much poorer (Figure 7a) due to their poor optical properties and higher charge carrier recombination rates. When the NiTiO₃ NFs were combined with AAs-gC₃N₄ sheets, the MB removal rate increased substantially. The MB removal percentage of 97% was higher for the optimal photocatalyst gnT6040. These results showed that the gnT6040 hybrid was the best photocatalyst for water treatment. Moreover, the obtained MB removal performance was substantially better than that of reported NiTiO₃/gC₃N₄-based hybrid photocatalysts (Table S2). This improvement in photocatalytic performance observed with the hybrid structure was attributed to heterojunctions between the NiTiO₃ NFs and the porous AAs-gC₃N₄ sheets. These heterojunctions hampered charge carrier recombinations while providing abundant catalytically active sites.

The reaction kinetics involved during the removal of MB was determined by measuring the first-order reaction equation of ln (C₀/C_n) as a function of kt, where k shows the slope of this curve corresponding to the apparent kinetic rate constant (Figure 7b). The apparent rate constant of gnT6040 was 58 × 10⁻³ min⁻¹, which is 1.8 times higher than that of pristine AAs-gC₃N₄ (28 ×10⁻³ min⁻¹), 2.5-fold higher than that of pristine NiTiO₃ NFs (23 × 10⁻³ min⁻¹), and 2.4-fold higher than that of bulk gC₃N₄ (24 × 10⁻³ min⁻¹). The higher performance of the hybrid structure may have resulted from higher charge carrier separation at the interfacial region and a broader range of light absorbance in the visible region. At lower AAs-gC₃N₄ content (gnT6040), a good dispersion of AAs-gC₃N₄ nanosheets on NiTiO₃ NFs could be formed. However, excess AAs-gC₃N₄ loading (gnT4060) causes a reduction of light for the covered NiTiO₃ NFs and decreases its excitation process for interfacial charge transfer. Therefore, a balance would be built between AAs-gC₃N₄ loading amount and NiTiO₃ NFs contact area, which can maximize the photocatalytic performance for the heterostructure. Through the sequential remediation of heterojunction components (NiTiO₃ NFs and AAs-gC₃N₄), we concluded an optimal loading of 40 wt% AAs-gC₃N₄ onto NiTiO₃ NFs, which exhibited the highest photocatalytic activity in our work.

In comparison, we also prepared and analyzed the MB removal efficiency of the p-gnT6040 sample prepared through physically mixing NiTiO₃ NFs and AAs-gC₃N₄ (Figure S6). Results showed that only 75% MB was removed by p-gnT6040 prepared through physical mixing in comparison with the 97% MB removal efficiency of gnT6040. This analysis shows an interface was formed for gnT6040 while a physically mixed sample (p-gnT6040) does not.

2.11. Photocatalytic hydrogen evolution

To further explore their photocatalytic efficiency, the as-prepared samples were also evaluated in the water-splitting application under light irradiation. Figure 8 shows the detailed H₂ evolution performance of the photocatalysts over several cycles. The H₂ evolution experiment was performed in an aqueous triethanolamine (TEOA) solution (10 vol%) as the sacrificial agent. Hydrogen evolution was lower for bulk gC₃N₄ (50 μmol g⁻¹), AAs-gC₃N₄ (60 μmol g⁻¹), and NiTiO₃ NFs (47 μmol g⁻¹) because of their moderate photocatalytic properties compared with the hybrid photocatalyst gnT6040 (152 µmol g⁻¹) (Figure 8a). When we calculated the efficiency of the photocatalysts, the highest H₂ yield for gnT6040 (21.7 μmol g⁻¹ h⁻¹) was 2.5-, 3.2- and 3-fold higher than that of AAs-gC₃N₄ (8.5 μmol g⁻¹ h⁻¹), NiTiO₃ NFs (6.7 μmol g⁻¹ h⁻¹), and gC₃N₄ (7.1 μmol g⁻¹ h⁻¹), respectively (Figure 8b). This trend is the same as that observed for the MB removal test. The higher H₂ evolution performances recorded for the hybrid structures were ascribed to heterojunctions between the NiTiO₃ NFs and the porous AAs-gC₃N₄ sheets. These heterojunctions successfully hampered charge carrier recombination and hence helped enhance photocatalytic activity.

For any efficient photocatalyst to be used in commercial applications, its stability and reusability are crucial. To evaluate the stability of the optimal gnT6040 photocatalyst, H₂ evolution experiments were carried out three times keeping the same conditions (Figure 8c). After each cycle, the vessel was purged with N₂ gas for 4 h and Ar. for 30 min to ensure an anaerobic environment. The experiment was then run again, and the evolved H₂ was collected and measured. The absence of a significant change in the performance of the photocatalyst in terms of H₂ evolution indicated good stability, which could be due to the incorporation of photo corrosion-reducing Ni [15]. Based on these results, gnT6040 showed efficient and stable photocatalytic performance and could be used for long-term applications.

3. Possible Photocatalytic Mechanism

A possible mechanism for the photocatalytic activity including MB removal and H₂ evolution for the NiTiO₃/AAs-gC₃N₄ photocatalyst is proposed based on our findings (Figure 9). In any photocatalytic activity, charge generation and effective utilization are very important and play a vital role in photocatalytic reactions. We observed that heterostructure formation between NiTiO₃ NFs and AAs-gC₃N₄ amorphous sheets increased the number of catalytically active sites, thereby increasing optical absorbance and charge separation and substantially improving photocatalytic performance. The outstanding charge transfer in the hybrid structure was ascribed to the 1D NiTiO₃ NFs anchored on 2D amorphous nanosheets of AAs-gC₃N₄. When the light was illuminated on the heterostructure, electron-hole pairs were generated in the AAs-gC₃N₄ sheets. The energy levels of the NiTiO₃ NFs and AAs-gC₃N₄ were appropriate for charge transfer and separation at the interface. In our heterostructural photocatalytic system, AAs-gC₃N₄ acted as a light sensitizer owing to its narrow bandgap. As a result of light absorption, electrons from the valence band (VB) of amorphous AAs-gC₃N₄ excited into the conduction band (CB), leaving holes in VB. Due to suitable energy levels, these photoexcited electrons transferred from the CB of AAs-gC₃N₄ sheets to the CB of NiTiO₃ NFs via interfacial interactions. There, these CB electrons participate in surface reactions reacting with O₂ adsorbed on the catalytic surface to generate reactive oxygen species (ROS) i.e., ˙O₂⁻ and ˙OH. These ROS are then responsible for productive surface reactions including removal of MB dye or enabling reduction for H₂ evolution reactions (Figure 9; I, II). Meantime, holes available in the VB of AAs-gC₃N₄ reacted with water (H₂O) molecules to make ˙OH radicals that could oxidize pollutants. This favorable charge transfer mechanism at the interface of NiTiO₃ NFs and AAs-gC₃N₄ facilitated the efficient separation of photogenerated charge carriers (electrons and holes).

4. Materials and Methods

4.1. Materials

All the reagents utilized in our work were of analytical grade and were used as received. Titanium (IV) butoxide, acetylacetone (Reagent plus^® grade, ≥99% purity), and nickel acetate tetrahydrate (98%) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Polyvinylpyrrolidone (PVP)-K90 from Wako Pure Chemical Industries (Osaka, Japan) was used in the synthesis of the NiTiO₃ NFs. Melamine C₆H₆N₆ (1,3,5- triazine-2,4,6- triamine) was used to prepare gC₃N₄ powder. Methylene blue (Daejung Chemical & Metals, Siheung, Korea) was selected as the model pollutant. Acetic acid (AA) used in the exfoliation of melamine was also purchased from Daejung Chemical & Metals, Siheung, Korea. Sodium sulfate (Na₂SO₄) and TEOA were purchased from Samchun Chemicals (Pyeongtaek-si, Korea) to evaluate the photoresponse properties of transient photocurrent and H₂ evolution, respectively. Conductive fluorine-doped tin oxide (FTO) glass (15 mΩ) was used as the substrate to fabricate the films.

4.2. Synthesis of NiTiO₃ NFs

NiTiO₃ NFs were prepared through electrospinning. The solution for electrospinning was prepared by mixing 2 g titanium butoxide (TBOT) with 5 mL ethanol under constant stirring at room temperature. Then, nickel acetate tetrahydrate (1.5 wt%) was added and taken, and dissolved completely with continuous stirring. Polyvinylpyrrolidone (0.7 g) powder was taken and dissolved (8 h) in the above solution to get a viscous gel, and finally, 0.6 mL acetylacetone was added to increase the viscosity of the solution. A uniform light-green colored gel was formed after 2 h of vigorous stirring. This solution gel was transferred into a 12-mL nonpyrogenic plastic syringe connected to a 25-gauge (diameter: 0.26 mm) stainless-steel needle. Then, the syringe was mounted vertically with a pump that was connected to a high-voltage power supply. The flow rate of the solution through the syringe was maintained at 10 µL min⁻¹ (Figure S2). Spinning was performed using an electrospinning system (NanoNC, Geumcheon-gu, Seoul, Korea) operating at 20 kV. The distance from the flat collector to the needle was fixed at 110 mm. The NiTiO₃ NF mats were collected every 3 h and dried in a vacuum oven (60 °C for 12 h) to evaporate the residual solvent. The dried fiber mats were finally sintered at 600 °C for 4 h in the air (ramp rate 10 °C/min).

4.3. Preparation of gC₃N₄

gC₃N₄ was synthesized by thermal polycondensation of melamine. Melamine powder (20 g) was placed in a covered alumina crucible and heated at 550 °C for 4 h at 3 °C min⁻¹ in a muffle furnace. The yellow product was collected and crushed into a powder.

4.4. Preparation of AAs-gC₃N₄

Porous nanosheets of gC₃N₄ were synthesized by exfoliating bulk gC₃N₄ via acetic acid (AA) treatment followed by sintering. Sintering of gC₃N₄ pretreated in an acid environment tends to provide porous gC₃N₄ via pyrolysis of protonated melamine [31]. Under the optimized experimental conditions, 2 g bulk gC₃N₄ was added to 50 mL deionized water and ultrasonicated for 40 min. Then, 1 mL AA was added, and the solution was stirred for 3 h. This solution was washed several times with deionized water and ethanol until it reached pH 7. The washed, AA-treated gC₃N₄ nanosheets were dried overnight at 60 °C in a vacuum oven. Then, 0.2 g dried powder was sintered at 550 °C with 3 °C min⁻¹ for 2 h (Figure S2). We denoted these AA-treated exfoliated and etched nanosheets as AAs-gC₃N₄. They were used without further treatment.

4.5. Preparation of NiTiO₃/AAs-gC₃N₄ Hybrids

Hybrid structures were synthesized according to our previous study [26]. In this experiment, a simple two-step process, as shown in Scheme 1 was used to fabricate hybrid photocatalysts. Firstly, a specific amount of NiTiO₃ NFs was dispersed in 50 mL ethanol under stirring at room temperature and then the required amount of AAs-gC₃N₄ powder was added to the NiTiO₃ NFs solution with constant stirring for 1 h. After proper mixing, the mixture solution was ultrasonicated (30-s pulse, 55% amplitude) for 40 min. The hybrid powder was collected by centrifuging the solution at 5000 rpm for 5 min and then dried the obtained powder in a vacuum oven. In the second step, these dry hybrid powders were sintered at 300 °C for 2 h (10 °C/min) in an N₂ atmosphere to ensure strong interaction between NiTiO₃ NFs and AAs-gC₃N₄ [32]. The final samples (dark grey) were carefully collected soon after the box furnace had reached room temperature. The as-synthesized heterostructures were identified as NiTiO₃/AAs-gC₃N₄ 60/40 (gnT6040) and NiTiO₃/AAs-gC₃N₄ 40/60 (gnT4060).

4.6. Photoelectrochemical Performance

The transient photocurrent density was measured in the dark and under UV-vis light irradiation using a three-electrode quartz cell and a potentiostat (VersaSTAT 4; Princeton Applied Research, Oak Ridge, TN, USA) in 0.5 M sodium sulfate electrolyte. The working electrode was prepared by dispersing a 0.2 g powder sample in 1.5 mL ethanol and then spin-coating onto the FTO substrate. The coated FTO substrates were dried in an oven at 60 °C for 2 h and sintering at 150 °C for 1 h. A prepared photoresponsive film, Ag/AgCl, and Pt wire were used as the working, reference, and counter electrodes, respectively. Various one-minute long off/on cycles were noted over several minutes of irradiation with UV-vis light from a 300 W Xe-arc lamp (66984; Newport Corporation, Yongin-Si, Korea).

4.7. MB Removal Test

To measure the photocatalytic performance of the pristine and hybrid catalysts, the MB removal test was evaluated. Visible light was irradiated on an aqueous solution of MB model pollutant placed in a batch-type indigenous-made reactor. In this MB removal experiment, the required amount of the photocatalyst (0.1 g) was dispersed in a 250 mL MB solution of 3.2 mg L⁻¹ concentration. This solution was kept in dark with constant stirring and three sample aliquots were taken and recorded as −60, −30, and 0 min. This stirring in dark and at room temperature helps the solution achieve adsorption-desorption equilibrium. After equilibrium was attained, the suspension was irradiated with visible light (100 W halogen lamp, 380–800 nm) under constant stirring (120 rpm). During light irradiation, 6-mL sample aliquots of the solution were collected after every 10-min interval for analysis; their absorbance corresponded to the concentration of MB in the solution. The extent of pollutant removal was determined by the change in optical absorbance at 663 nm.

4.8. Hydrogen Evolution

The hydrogen evolution performance of the pristine and hybrid photocatalysts was evaluated by using a quartz reactor. In a particular experiment, 10 mg photocatalyst was added to a 10 vol% aqueous solution of TEOA. After complete dispersion of the sample powder, the suspension was shifted into a quartz reactor (in-house made), which was then made air-tight using rubber septum. Nitrogen and argon gases were purged into the reactor to create a completely anaerobic environment. Subsequently, a Xe lamp placed 6cm away from the reactor was used to irradiate the suspension. The suspension was irradiated up to 7 h and then the produced gas (260 µL) was collected after every 1 h intervals using a syringe (500 µL, Agilent, Santa Clara, CA, USA). Thus, gas produced is quantified by using gas chromatography (model 6500GC; YL Instruments, Anyang, Korea). This procedure is repeated for every sample.

4.9. Characterization

The surface morphologies of the samples were analyzed using an FE-SEM operating at 15 kV (S4800; Hitachi, Tokyo, Japan). Additional details were provided by a scanning TEM operating at 200 kV (JEM-2100F; JEOL, Tokyo, Japan). Phase analysis of the pure and hybrid samples was made by X-ray diffraction (D/Max-2500/PC (Cu Kα source, λ = 1.5418 Å); Rigaku, Tokyo, Japan). Chemical binding energies were determined by XPS (K-Alpha⁺ (Al Kα source); Thermo Fisher Scientific, Waltham, MA, USA). The chemical interactions at the surface of the photocatalyst were characterized by FT-IR (iS10; Thermo Fisher Scientific) over the range of 400–4000 cm⁻¹. Ultraviolet-visible spectroscopy (V650; Jasco, Hachioji-shi, Tokyo, Japan) was used to measure the optical absorbance, which is related to the bandgap energy. The separation of charge carriers and their recombination were analyzed by measuring the PL emission spectrum using a spectrofluorometer (SC-100; Dongwoo, Pyeongtaek-si, Gyeonggi-do, Korea) using an excitation wavelength of 325 nm. The BET method was used to measure the specific area of the photocatalysts according to their N₂ adsorption-desorption isotherms using a Quantachrome system (Autosorb iQ-C; Quantachrome Instruments, Boynton Beach, FL, USA). The total pore size distribution was determined by the BJH method.

5. Conclusions

In this work, noble-metal-free, visible-light-active photocatalysts based on heterojunctions of NiTiO₃ NFs and holey AAs-gC₃N₄ with different compositions were synthesized. The NiTiO₃ NFs sensitized with the porous nanosheets exhibited enhanced photocatalytic performance. This enhancement by the 1D/2D heterostructure was ascribed to swift charge transport, higher light absorbance, and effective separation of the charge carriers from the barrier potential produced at the interface. Moreover, a dramatic reduction of recombination losses by the hybrid sample was confirmed by room-temperature PL and photocurrent response measurements. The superior performance of the AAs-gC₃N₄-sensitized NiTiO₃ NFs photocatalyst was also due in part to its large BET surface area and abundance of catalytically active sites. Overall, the results and facile synthesis method offer an exciting avenue for the design of cost-effective catalysts for use in photocatalytic applications.