Electro-Oxidation of Metal Oxide-Fabricated Graphitic Carbon Nitride for Hydrogen Production via Water Splitting

Abstract

Hydrogen is a great sourcez of energy due to having zero emission of carbon-based contents. It is found primarily in water, which is abundant and renewable. For electrochemical splitting of water molecules, it is necessary to use catalytic materials that minimize energy consumption. As a famous carbon material, graphitic carbon nitride, with its excellent physicochemical properties and diversified functionalities, presents great potential in electrocatalytic sensing. In the present work, graphitic carbon nitride-fabricated metal tungstate nanocomposites are synthesized by the hydrothermal method to study their applications in catalysis, electrochemical sensing, and water splitting for hydrogen production. Nanocomposites using different metals, such as cobalt, manganese, strontium, tin, and nickel, were used as a precursor are synthesized via the hydrothermal process. The synthesized materials (g-C₃N₄/NiWO₄, g-C₃N₄/MnWO₄, g-C₃N₄/CoWO₄, g-C₃N₄/SnWO₄, g-C₃N₄/SrWO₄) were characterized using different techniques, such as FTIR and XRD. The presence of a functional groups between the metal and tungstate groups was confirmed by the FTIR spectra. All the nanocomposites show a tungstate peak at 600 cm⁻¹, while the vibrational absorption bands for metals appear in the range of 400–600 cm⁻¹. X-ray diffraction (XRD) shows that the characteristic peaks matched with the JCPDS in the literature, which confirmed the successful formation of all nanocomposites. The electrochemical active surface area is calculated by taking cyclic voltammograms of the potassium–ferrocyanide redox couple. Among the entire series of metal tungstate, the g-C₃N₄/NiWO₄ has a large surface area owing to the high conductive properties towards water oxidation. In order to study the electrocatalytic activity of the as-synthesized materials, electrochemical water splitting is performed by cyclic voltammetry in alkaline medium. All the synthesized materials proved to be efficient catalysts with enhanced conductive properties towards water oxidation. Among the entire series, g-C₃N₄-NiWO₄ is a very efficient electrocatalyst owing to its higher active surface area and conductive activity. The order of electrocatalytic sensing of the different composites is: g-C₃N₄-NiWO₄ > g-C₃N₄-SrWO₄ > g-C₃N₄-CoWO₄ > g-C₃N₄-SnWO₄ > g-C₃N₄-MnWO₄. Studies on electrochemically synthesized electrocatalysts revealed their catalytic activity, indicating their potential as electrode materials for direct hydrogen evolution for power generation.

1. Introduction

The rapid growth of energy use demands more sustainable and renewable energy production. The primary renewable energy resources such as solar and wind power are ecofriendly, but they possess seasonal intermittence and variabilities due to regional differences. To avoid such problems, such unstable energy sources should be replaced with the stable hydrogen energy [1]. The production of hydrogen is becoming increasingly popular due to its environment-friendly qualities and potential as a source of energy that is clean, non-polluting, and recyclable. In coping with the future energy crisis, hydrogen can play an important role [2]. In view of hydrogen generation, water electrolysis is a common and efficient method. Generally, water splitting is based on the two half processes: the cathodic oxygen evolution reaction (OER) and the anodic hydrogen evolution reaction (HER). Although water electrocatalysis produces very pure and ecofriendly hydrogen, due to the sluggish kinetics of OER and HER and large overpotential value, the practice of energy generation via water splitting is limited [3]. To solve this problem, the use of electrocatalysts should be implemented in order to minimize the overpotential value for cathodic hydrogen evolution (HER) and anodic oxygen evolution (OER) [4]. The electrocatalysts make the water splitting an energy-efficient process [5]. Recently, research on electrochemical water splitting using various electrocatalysts has been conducted to achieve this objective [4]. The noble metals such as platinum-, iridium-, and ruthenium-based nanomaterials proved to be efficient regarding HER and ORE processes [6,7]. The existence of such catalysts makes them indispensable for many technologically significant chemical processes, although their spontaneous aggregation and growth limit the lifetime and efficiency. Additionally, the ultrahigh price of noble metals has severely limited their future use. Due to the rapid development of modern industry, a critical need of today is highly active, stable, low-cost, and recyclable nanomaterial alternatives to these noble metals, designed for water splitting [8]. Recently, the conductive polymers (CPs) have been shown as a potential alternative to Pt-based materials due to their high electrical properties, large surface area, and greater physical and chemical stability [9,10,11]. Among the 2D CPs, graphitic carbon nitride (g-C₃N₄) is a graphene-like material having excellent structural features, such as high porosity, large surface area, and greater content of nitrogen. g-C₃N₄ has wide applications in electrochemical sensing, optoelectronics, electrochemical oxidations, and energy storage and conversion devices [12,13]. g-C₃N₄ can be used as a good supportive material for manufacturing 3D nanomaterials for different electrochemical applications. Doping of different nanomaterials with g-C₃N₄ generates highly active electrode materials by connection within hybrids [14]. In the past few years, scientists have enhanced the properties of CPs by doping with transition metals/metal oxides [15]. Recently, a series of low-cost metal tungstates, such as Bi₂WO₆, CuWO₄, NiWO₄, and CoWO₄, have been extensively studied as efficient dopants, with g-C₃N₄ showing various applications because of its outstanding optical, magnetic, and catalytic properties [16]. Hence, from these studies, it seems that by doping g-C₃N₄ with CoWO₄ [17] and NiWO₄ [18], its photocatalytic activity is enhanced. Similarly, strontium, tin, manganese, zirconium, and cobalt are the most promising transition metals and are reported as dopants with g-C₃N₄ for numerous applications [15,19,20]. These metals have a large bandgap, rendering them with weak ion transport kinetics as semiconductors or even insulators [19], and electrode film pulverization as a result of the pronounced expansion and contraction of volume during the charging/discharging processes [20].

Motivated by these results, we have designed this research work to explore the catalytic activity of transition metal tungstate-fabricated graphitic carbon nitride for electrochemical applications. In this work, the nanocomposites (g-C₃N₄-NiWO₄, g-C₃N₄-MnWO₄, g-C₃N₄-SnWO₄, g-C₃N₄-CoWO₄, g-C₃N₄-SrWO₄) were synthesized via the hydrothermal method and characterized by XRD, FTIR, and electrochemical cyclic voltammetry. The study of the electrocatalytic splitting of water for hydrogen production was carried out in alkaline medium via cyclic voltammetry. All the synthesized nanocomposites represent excellent catalytic activity towards water oxidation.

2. Materials and Methods

2.1. Materials

Melamine (purity: 99.95%), cyanuric acid (purity: >99%), sodium tungstate (Sigma Aldrich, St. Louis, MO, USA), sodium hydroxide (Sigma Aldrich), potassium hexacyanoferrate (III) (K₄[Fe(CN)₆]), alumina powder, nafion, and different metal precursors were used. All chemicals were purchased from Sigma Aldrich (St. Louis, MO, USA).

2.2. Preparation of g-C₃N₄

Following the protocol of Vilian et al., C₃N₄ was prepared [21]. The melamine and cyanuric acid were dissolved in water (75 mL) and ultrasonically maintained for 4 h. An oven (Memmert, Germany) was used at 200 °C for 12 h to heat the solution after sonication in a 100 mL stainless Teflon lining. Centrifugation, followed by cleaning with ethanol and water, was performed to eradicate any remaining contaminations from the precipitate, then centrifuging was carried out at 5000 RPM for 30 min (Figure 1). After drying at 100 °C for 24 h, the precipitate was removed. Consequently, g-C₃N₄ (Vilian et al., 2020) was synthesized [21].

2.3. Preparation of g-C₃N₄/Metal Tungstate Composite

The graphitic carbon nitride g-C₃N₄/MetalWO₄ nanocomposite was synthesized by utilizing a simple hydrothermal procedure without any surfactants. Initially, a solution of sodium tungstate and graphitic carbon nitride was made. For this purpose, 0.01 g of g- C₃N₄ and 15 mM of Na₂WO₄·2H₂O were placed in 37 mL of deionized water. This solution was then kept under sonication for 20 min. The temperature was set at room temperature. After sonication, the mixture was placed aside. Then, 15 mM of salt solution in deionized water was prepared and added into the above mixture to form a homogeneous solution, and 1 mL of aqueous NaOH solution (0.1 M) was gradually added to the homogeneous solution. The mixture solution was then dispensed into an autoclave and heated at 180 °C for 12 h. After autoclaving, the resulting mixture was centrifuged to obtain the maximum analyte. To extract the excess unreacted starting materials, the obtained precipitate was collected and thoroughly rinsed three times with ethanol and water and desiccated at 60 °C. Finally, the resulting precipitation was strengthened in a vacuum oven at 300 °C for five hours to attain the g-C₃N₄/MetalWO₄ nanocomposites. The obtained g-C₃N₄/MetalWO₄ nanocomposite was circulated in ethanol to perform electrochemical measurements in Figure 2.

2.4. Electrochemical Studies

The electrochemical water oxidation of all synthesized nanocomposites was studied by using the Gamry Potentiostat interface 1000 (Gamry, Warminster, PA, USA) with a 3-electrode cell (100 mL) assembly in alkaline medium. The working electrode was glassy carbon (GC), the counter electrode was platinum wire (Pt), and the reference electrode was silver/silver chloride (Ag/AgCl). Before modification, the surface of glassy carbon was cleaned with alumina slurry by polishing the electrode in a figure eight motion for 5 min. Then, it was rinsed with deionized water and sonicated for 1 min in deionized water to remove remnants of alumina powder. It was dried at room temperature. The catalyst was then deposited on the GC surface by the drop casting method, where 5% nafion was used as a binder.

3. Characterization Studies

3.1. Fourier-Transform Infrared (FTIR) Spectroscopy

The chemical and structural characteristics of nanocomposites, as well as their atomic and molecular vibrations, were investigated using FTIR (Nicolet 5PC, Nicolet Analytical Instrument (Protea, Cambridgeshire, UK) in a 500 to 4000 cm⁻¹ range. The FTIR spectrum of the cobalt tungstate-fabricated nanocomposite is depicted in Figure 3, where the two prominent peaks appeared at 1627.5 and 3350 cm⁻¹, representing N-H bond and hydroxyl bond stretching vibration, respectively [22]. The conjugated g-C₃N₄-based heterocycles are represented by several absorption peaks in the 1600–1200 cm⁻¹ region [21], which appeared in nearly all spectra of g-C₃N₄/MetalWO₄. The band at 611.8 cm⁻¹ corresponds to Co-O-W symmetric vibration [23].

Figure 4 presents the FTIR spectrum of g-C₃N₄/MnWO₄. There should be obvious peaks at 875, 827, 710, 605, and 512 cm⁻¹ for pristine MnWO₄. The peaks at 870, 704, 621, and 514 cm⁻¹ correspond to the W–O (symmetric), W–O (asymmetric), Mn–O, and Mn–O–Mn stretching vibrations, respectively, which confirms the formation of manganese tungstate [21]. In Figure 4, two strong peaks appeared at 794 and 670 cm⁻¹, demonstrating W–O (symmetric) and Mn–O–Mn stretching vibrations, respectively. Hence, Figure 4 confirms the successful preparation of g-C₃N₄/MnWO₄.

The FTIR spectrum of graphitic carbon nitride fabricated on nickel tungstate is shown in Figure 5. Strong peaks appeared at 1621 and 3322 cm⁻¹, showing conjugation between graphitic carbon nitride and metal oxide, as in previous spectra. Peaks in the series below 500 cm⁻¹ are due to the vibrations of the NiO₆ polyhedron [24], which is not observed in this figure as the spectrum range is from 500 to 1500 cm⁻¹.

The FTIR of the tin tungstate–graphitic carbon nitride sample in Figure 6 depicts a band at 1000 cm⁻¹ which is due to the W=O bond. The band at 650 cm⁻¹ corresponds to Sn-O vibrational modes, which slightly appeared in Figure 6.

In Figure 7, the strong peak at 768 cm⁻¹ is due to the W-O bond in the tetrahedral. The peak at 3350 cm⁻¹ is due to OH stretching vibrations. Absorption peaks at 471 cm⁻¹ are due to the bond between strontium and oxygen [25], which did not appear here due to the selected frequency range, which is 500–1500 cm⁻¹.

3.2. X-ray Diffraction (XRD)

X-ray diffraction (PANalytical X’PERT High Score’s diffractometer, Malvern, UK) patterns of nanocomposites prepared by the hydrothermal method are shown in Figure 8. All the diffraction peaks in Figure 8 are sharp, indicating the good crystallinity of the material. Figure 8a represents the peaks of 2θ for g-C₃N₄/SnWO₄ at 28.3° and 53.1° with crystallographic plans of (121) and (161), with JCPDS card No. 29-1354 [26], which confirms the formation of orthorhombic α-SnWO₄. The spectrum in Figure 8b shows the XRD of g-C₃N₄/MnWO₄, which shows intensity peaks for MnWO₄ at 30.1°, 30.57°, 36.25°, and 67.1°, with crystal lattice (−111), (111), (120), and (140), respectively, with JCPDS card No. 72-0478 [27], along with the low-intensity peak of g-C₃N₄ at 26.7° (002). This spectrum in Figure 8b depicts the formation of g-C₃N₄/MnWO₄. The XRD spectrum in Figure 8c of g-C₃N₄/CoWO₄ shows the characteristic peak of g-C₃N₄ at 13.1°, indexed as (100) with JCPDS card No. 50-1512 [28], and this is due to the in-plane structural packing motif. The peaks for CoWO₄ appeared at 30.8°, 36.4°, and 54.3°, indexed as (111), (200), and (202), respectively, with JCPDS card No. 72-0479 [29]. The peaks of g-C₃N₄/SrWO₄ presented in Figure 8d are 33.6° and 55.8°, with corresponding miller indices (204) and (312) and JCPDS card No. 08-0490, which agree well with the reported data of SrWO₄ for Scheelite phase [30]. Figure 8e represents the XRD spectrum of g-C₃N₄/NiWO₄, having peaks at 31.1°, 36.2°, 55.4°, and 65.2° for NiWO₄ with JCPDS card No. 15-0755 [31]; thus, this confirmed the synthesis of g-C₃N₄/NiWO₄. The intensity of the peaks in many spectra decreased and the peak width increased, which indicated the interaction between metal tungstate and C₃N₄ nanoparticles, and thus confirmed the successful formation of nanocomposites. No characteristic peaks for other impurity phases were observed in all five XRD patterns, showing that the selected synthetic method is a feasible route to prepare pure phases of catalysts.

The size of the prepared materials was calculated by using the Debye–Scherrer equation [32]:

$${\mathrm{D}}_{\mathrm{avrg}}=\frac{\mathrm{k}\mathsf{\lambda }}{\mathsf{\beta } \cos \mathsf{\theta }}$$

(1)

where D_avrg is equal to the average crystallite size, λ is the X-ray wavelength, θ is the Bragg angle of the desired peak, and β is the line broadening evaluated from the peak width at half height. The average crystalline size of all synthesized particles is presented in

Table 1. Calculated values of D_avrg.

Samples	Davrg (XRD) (nm)
g-C3N4/SnWO4	9.0
g-C3N4/MnWO4	16.3
g-C3N4/CoWO4	52.6
g-C3N4/SrWO4	84.4
g-C3N4/NiWO4	30.3

. In XRD patterns, the peak width is inversely proportional to the crystallite size. Normally, nanoparticles of small-sized peaks have a small particle size, while those of sharp peaks have large particle sizes. The small particle size is responsible for more catalytic activity [33].

3.3. Estimation of Active Surface Area of Modified Electrode

To determine the active surface area of electrodes, which is altered by catalysts having diverse loadings of nanocomposites, cyclic voltammograms were recorded at a scan rate of 100 mV s⁻¹ in a 0.1 M KCl solution, with 5 mM of potassium ferrocyanide as a model redox mediator (Figure 9). The peak currents show that the highest value corresponds to the NiWO₄-modified electrode. This means that the sample NiWO₄-presents better conductive properties towards the redox probe as related to the rest of the catalysts.

The active surface area was estimated using the Randles–Sevcik equation for reversible reactions [34]:

$${\mathrm{Ip}}_{a }=2.69\times {10}^5{\mathrm{n}}^{3/2}A{\mathrm{D}}^{1/2}{\mathsf{\upsilon }}^{1/2}C$$

(2)

where D is the diffusion coefficient (0.76 × 10⁻⁵ cm²s⁻¹), n = 1, and C is the concentration of analytes (5 mM).

From

Table 2. Active surface area calculated for modified electrodes.

Samples	Active Surface Area (cm−2)
g-C3N4/SnWO4	0.061
g-C3N4/MnWO4	0.050
g-C3N4/CoWO4	0.068
g-C3N4/SrWO4	0.081
g-C3N4/NiWO4	0.088

, it can be seen that g-C₃N₄ with nickel tungstate presented a higher surface area as compared to the other metal tungstates. This means that the nickel tungstate-fabricated graphitic carbon nitride is an efficient electrocatalyst towards an electrochemical redox reaction.

4. Electrochemical Water Oxidation

CV (Gamry Interface 1000) was used to perform electrochemical studies, such as water oxidation. Three electrodes were utilized for the experiment: modified GC as a working electrode, Pt-wire as a counter electrode, and silver-silver chloride as a reference electrode. Water oxidation was carried out in a 1 M KOH solution in a potential window ranging from 0.5 to 1.8 V at different scan rates (10 to 100 mV s⁻¹). The cyclic voltammetry behaviors of nanocomposites were used to investigate the redox kinetics of the electrodes (Figure 10a–e). The anodic peak current with a growing scan rate from 10 to 100 mV s⁻¹ was represented, which shows that the current increased with the increasing sweep rate, from 1.25 to 4.46 mA for g-C₃N₄-CoWO₄/GC, from 0.17 to 0.45 mA for g-C₃N₄/MnWO₄/GC, from 0.2 to 1.7 mA for g-C₃N₄-SnWO₄/GC, from 2.5 to 6.5 mA for g- C₃N₄-NiWO₄/GC, and from 2.0 to 5.5 mA for g-C₃N₄-SrWO₄/GC, and the peak potential shifted towards higher (anodic) potentials. The onset potential for all five nanocomposites is presented in

Table 3. Kinetics parameters for water oxidation at g-C₃N₄/MetalWO₄/GC.

Samples	Eonset (V)	Ipa (mA)	α	Do/10−6 cm2s−1
g-C3N4/SnWO4	1.08 (NHE =1.23)	1.5	0.2	40.72
g-C3N4/MnWO4	0.999 (NHE = 1.19)	0.45	0.3	14.09
g-C3N4/CoWO4	0.92 (NHE = 1.12)	4.5	0.2	130.00
g-C3N4/SrWO4	0.86 (NHE = 1.06)	5.3	0.2	132.06
g-C3N4/NiWO4	0.80 (NHE = 1.01)	6.5	0.1	133.3

. The reported onset potential for water oxidation on CoWO₄ was ~1.54 V (NHE) [35], while CoWO₄-fabricated g-C₃N₄ was 1.12 V (NHE). An ideal potential value for the water oxidation reaction is 1.23 V (NHE) [36,37,38]. From Table 3, it can be seen that all synthesized electrocatalysts presented lower onset potential values (NHE) for OER, which revealed that all the nanocomposites performed well in cyclic voltammetry, inherently possessing superior oxygen evolution activity. The peak current of different synthesized nanocomposites is shown in Figure 10f and Table 3, which was in the order: g-C₃N₄/NiWO₄ > g-C₃N₄/SrWO₄ > g-C₃N₄/CoWO₄ > g-C₃N₄/SnWO₄ > g-C₃N₄/MnWO₄. Hence, these synthesized nanocomposites are favorable for being excellent electrochemical catalysts towards water oxidation.

The linear relationship between peak current (Ip_a) and scan rates (υ^1/2) is presented in Figure 11 and illustrates that the water oxidation on the surface of metal oxide catalysts is diffusion-controlled.

Determination of Diffusion Coefficient (D^o)

The diffusion coefficient was determined by using the Randles–Sevcik formula for irreversible reactions [32]:

Ip_a = 2.69 × 10⁵ n ^3/2 AC × (1 − a) n D^1/2 ν^1/2

(3)

where A is the surface area of the electrode (0.07 cm²), C is the concentration of KOH (1 M), N is the number of electrons for OER (2), and α is the transfer coefficient (0.2–0.7). From the slope of the plot of Ip_a vs. ν^1/2, we obtained the diffusion coefficients tabulated in Table 3. Table 3 depicts that g-C₃N₄/NiWO₄ has a higher value of the diffusion coefficient as compared to the other catalysts, showing that g-C₃N₄/NiWO₄ nanoparticles are an excellent candidate for rapid redox reactions towards water oxidation.

5. Conclusions

The present study presented a simple fabrication strategy for the fabrication of g-C₃N₄ on metal tungstates as a potential electrode material for the electrochemical splitting of water samples. Metal tungstates (nickel tungstate, copper tungstate, stannous tungstate, strontium tungstate, and manganese tungstate) were distributed on the graphitic carbon nitride surface. The composites were successfully formed, as revealed by XRD and FTIR analysis. In FTIR spectroscopy for functional group detection, all the composites showed a tungstate peak at 600 cm⁻¹. X-ray diffraction (XRD) was performed at 2 theta, showing characteristic peaks according to their JCPDS card numbers. The electrochemical properties of the electrodes were confirmed via cyclic voltammetry. The g-C₃N₄-NiWO₄ composite electrode displayed a unique electrochemical sensing behavior compared to the other composite electrodes, exhibiting a high peak current and low peak potential. The order of electrocatalytic sensing of different composites towards water oxidation was: g-C₃N₄-NiWO₄ > g-C₃N₄ SrWO₄ > g-C₃N₄-CoWO₄ > g-C₃N₄-SnWO₄ > g-C₃N₄-MnWO₄. Based on the results from the fabricated electrochemical sensor, the g-C₃N₄/NiWO₄ composite displayed the highest active surface area at 0.088 cm⁻². In this case, the peak current was higher and the sensitivity was also high. Aside from this, g-C₃N₄/NiWO₄ can also be employed in the manufacture of electrodes used in clinical, pharmaceutical, and medical applications. As a result, the proposed modified electrode can be employed in the long term for electrocatalytic sensing, with good selectivity and reproducibility. The findings indicated that the investigated approach could be useful in regular analytical applications.