Development and Mechanistic Studies of Ternary Nanocomposites for Hydrogen Production from Water Splitting to Yield Sustainable/Green Energy and Environmental Remediation.

Photocatalysts lead vitally to water purifications and decarbonise environment each by wastewater treatment and hydrogen (H2) production as a renewable energy source from water-photolysis. This work deals with the photocatalytic degradation of ciprofloxacin (CIP) and H2 production by novel silver-nanoparticle (AgNPs) based ternary-nanocomposites of thiolated reduce-graphene oxide graphitic carbon nitride (AgNPs-S-rGO2%@g-C3N4) material. Herein, the optimised balanced ratio of thiolated reduce-graphene oxide in prepared ternary-nanocomposites played matchlessly to enhance activity by increasing the charge carriers' movements via slowing down charge-recombination ratios. Reduced graphene oxide (rGO), >2 wt.% or <2 wt.%, rendered H2 production by light-shielding effect. As a result, CIP degradation was enhanced to 95.90% by AgNPs-S-rGO2%@g-C3N4 under the optimised pH(6) and catalyst dosage(25 mg/L) irradiating beneath visible-light (450 nm, 150 watts) for 70 min. The chemical and morphological analysis of AgNPs-S-rGO2%@g-C3N4 surface also supported the possible role of thiolation for this enhancement, assisted by surface plasmon resonance of AgNPs having size < 10 nm. Therefore, AgNPs-S-rGO2%@g-C3N4 has 3772.5 μmolg-1 h-1 H2 production, which is 6.43-fold higher than g-C3N4 having cyclic stability of 96% even after four consecutive cycles. The proposed mechanism for AgNPs-S-rGO2%@g-C3N4 revealed that the photo-excited electrons in the conduction-band of g-C3N4 react with the adhered water moieties to generate H2.

1. Introduction

Hydrogen (H2) production through the photolysis of water is an alternative renewable green energy source with the prime advantages of harvesting energy and eco-friendly nature. The splitting of water under solar radiation or a light source of suitable wavelength is pertinent to the advent of new photocatalysts [1,2]. For this, various nanostructured semiconducting materials have been explored, such as titanium dioxide (TiO2), (g-C3N4), cadmium sulphide (CdS), zinc oxide (ZnO) and cerium(IV) [3]. However, the efficiency of single semiconductor photocatalysts is low due to the absorption of light in a small range of spectrum and the poor charge separation abilities leading to a high rate of electron-hole pair recombination. These problems have been addressed by engineering the appropriate bandgap and improving charge-recombination ratios or electron affinities for a radiation source using a combination of semiconducting materials with conducting oxides of metals or non-metals [4,5]. This combination develops the heterojunctions at recombinants’ interfaces and results in band-bending to charge separations by forcing their photogenerated charge carriers to move in reverse directions [6,7]. Therefore, by using the combination of binary or tertiary nanocomposite-based photocatalysts, the maximum H2 production can be obtained from water under certain experimental conditions by overcoming the listed issues [8].

Ciprofloxacin (CIP) is a broad-spectrum antibiotic with low-biodegradability being discharged at high volumes ranging from μg/L to mg/L into wastewaters of hospitals and pharmaceutical industries [9]. Its rise into water bodies poses a significant threat to the environment by increasing drug resistance in bacteria. Its elimination from the environment thus requires an urgent and efficient means. Accordingly, different techniques such as physical adsorption onto various substrates and biological, electrochemical, and photocatalytic degradations have been reported [10]. Among these, the photocatalytic approach has been most distinguished in eliminating the CIP pollutants due to high effectivity and ecological sustainability. The charge-carriers such as electron and hole (e−and h+) pairs produced by these photocatalysts, all under the influence of light-irradiation, are the source to generate the reactive-oxygen spices such as hydroxyl and superoxide (·OH and O₂⁻) radicals, responsible for the degradation of CIP molecules. It has been observed that the degradation of CIP molecules has been enhanced using heterojunctions or composites of semiconducting materials with metal nanoparticles [11]. The doping of sulphur (S) in g-C3N4 and rGO helped us modify their electrical conductivity and bandgap required for photodegradation of CIP molecules. This doping of S into these semiconductors further supports forming the smaller sized AgNPs. It results from changes in their pristine optical absorptions by activating these photocatalysts under higher visible ranges.

Among the previously mentioned photocatalysts, g-C3N4 has attained the scientific community’s attention for photocatalytic H2 production. The popularity of g-C3N4 is attributed to its non-toxicity, ease of synthesis, and tuneable nature of its bandgap [12]. Furthermore, its physio-chemical stability is due to covalent bonding between nitrogen and carbon, allowing it to function equally well in acidic and basic media. Additionally, the negative band potential of g-C3N4 (around 1.3 eV) facilitates photocatalytic H2 production and water oxidation reaction under visible light [13]. However, the competence of pure g-C3N4 for photocatalytic H2 production is far less due to rapid charge-recombination ratios and the existence of a wide bandgap (of around 2.7 eV). Therefore, doping of g-C3N4 with other materials such as cadmium sulphide, molybdenum disulphide, gold, graphene and low dimension carbon-based materials supports the suppression of rapid charge recombinations and narrows the bandgap, which is highly required for large-scale photocatalytic H2 production [14,15]. Among these materials, reduced graphene oxide (rGO) seems a promising co-catalyst due to its conjugated structure, high specific surface area, and the tuneable nature of electronic bands [16].

Notably, the proportion of rGO as co-catalyst in the composite catalyst is crucial for photocatalytic H2 production and energy application [17]. The higher amount of rGO produces negative shielding effects [18], which slow down the charge carriers’ movement from the valence band (VB) to the conduction band (CB), ultimately affecting the H2 production. Consequently, the efficiency of rGO@g-C3N4 can be further enhanced after adjusting the amount of rGO in g-C3N4 to a specific ratio. Additionally, doping with elements with low electronegativities, such as boron (B) and sulphur (S), can also boost the catalytic efficiency by raising the Fermi level [19]. Thus, the thiolation of composite materials, rGO and g-C3N4, can enhance photocatalytic H2 production in two ways, by dropping the charge-recombination ratios and increasing the light absorption adeptness. The thiolation help to conquer the negative shielding effect of thiolated rGO (HS-rGO), increasing its absorption efficiency and the increased carbon vacancies in thiolated g-C3N4 (HS-g-C3N4) act as active territories to reduce the chance of charge recombinations [20]. In summary, thiolation increases the charge separation capabilities of both HS-rGO and HS-g-C3N4. Interestingly, photocatalytic H2 production can be further improved by coupling thiolated composites with some noble metal nanoparticles (NPs), such as silver. The silver nanoparticles (AgNPs) are highly conductive and support photocatalysis for H2 production through surface plasmon resonance (SPR). This SPR property also assists the movement of charge carriers, enhancing photocatalytic H2 production [21,22].

A literature review showed that various groups had synthesised rGO and its composites with g-C3N4 at a fixed amount of rGO for photocatalytic hydrogen production. In this context, Jaiswal et al. prepared the rGO-based composites of g-C3N4 using a constant 5 wt.% [23]. Likewise, Ibrahim and his colleagues investigated the TiO2/rGO/g-C3N4 system for H2 production with a fixed 1 wt.% of rGO [24]. In the same way, Pan et al. [25] explored the effect of 0.3% rGO in their rGO@g-C3N4 composites for hydrogen production. However, the impact of thiolation has not been analysed, and the concentration of thiolated rGO (HS-rGO) in ternary-nanocomposites for H2 production can be further optimised.

In view of the previous reports [12,19,20,23,24,25], rGO-based composites of g-C3N4 were prepared at different concentrations of HS-rGO, from 1 to 4 wt.%. The thiolated composite (HS-rGO@g-C3N4) with an optimised concentration of HS-rGO of 2 wt.% was selected based on photoluminescence (PL) studies which indicated lesser charge recombinations. HS-rGO2%@g-C3N4 was found to have improved structural and chemical properties and is highly desirable for photocatalytic H2 production. Next, the HS-rGO2%@g-C3N4 was decorated with AgNPs to produce AgNPs-S-rGO2%@g-C3N4. The charge transfer, structural, surface chemical, optical conduction, and photocatalytic H2 production properties of this final nanocomposite, AgNPs-S-rGO2%@g-C3N4, were explored under the same experimental conditions compared to HS-rGO2%@g-C3N4 and rGO2%@g-C3N4.

2. Experimental Details

2.1. Synthesis and Thiolation of rGO

Sheets of graphene oxide (GO) were synthesised using a modified Hummer method as per our previous work [26]. After this, GO sheets were heated to 200 °C for 2 h to obtain rGO. This rGO was thiolated following an earlier report with minor modifications [27]: 50 mg of rGO was dispersed in 50 mL of ultrapure deionised-water and ultrasonicated for 1.5 h, followed by the addition of 3 mL of HBr and further sonication for 2 h. After this, 2.5 g of thiocarbamide was added at 80 °C and stirred continuously for 24 h. The reaction was stopped, and the reaction media was allowed to cool to room temperature. Afterwards, 25 mL of 4 M aqueous solution of KOH was added and stirred for a further 1 h. The black suspension of thiolated GO (HS-rGO) was filtrated through a PTFE membrane filter (0.2 µm) and washed many times with dimethyl ether, ethanol, and water. The resulting HS-rGO nanosheets were redispersed into ultrapure water by sonication and vacuum dried before further use.

2.2. Synthesis of Graphitic Carbon Nitride Nanosheets (g-C3N4)

To prepare g-C3N4, a pale-yellow powder, 2.5 g of melamine was heated in the muffle-furnace to 525 °C by providing 10 °C min−1 heating rate as per previous reports by Zhang et al. [28].

2.3. Synthesis of Thiolated Graphitic Carbon Nitride Nanosheets (HS-g-C3N4)

HS-g-C3N4 was prepared by following the Wang’s protocol via the single-step thermal-polymerisation of thiocarbamide [13]. For this, 30 g of thiocarbamide was kept in a porcelain crucible with no lid. This open crucible was placed into another large crucible bolted with a lid and placed to heat in a muffle-furnace to 550 °C for 5 h by providing a temperate rate of 2 °C min−1. This process resulted in a pale-yellow coloured powder, HS-g-C3N4, which accumulated in the larger crucible.

2.4. Fabrication of rGO@g-C3N4 and HS-rGO@g-C3N4

To fabricate rGO@g-C3N4 with different concentrations of rGO, 4.7 g of g-C3N4 was added separately into 15 mL solution of rGO prepared using 1, 2, 3 and 4 wt.% rGO according to the measured mass of g-C3N4. These four mixtures were then vigorously sonicated under slow stirring for 1 h to coat rGO onto the g-C3N4. Next, the rGO-coated g-C3N4 (rGO@g-C3N4) was filtered from the solution and vacuum dried. The rGO@g-C3N4 with 2 wt.% of rGO (rGO2%@g-C3N4) was found best, as illustrated by the PL study. Further, the same composite was prepared using 2 wt.% of HS-rGO to HS-g-C3N4 (HS- rGO2%@g-C3N4). Finally, the HS- rGO2%@g-C3N4 was decorated with AgNPs to make the ternary composites.

2.5. Synthesis of AgNPs onto the Surface of HS-rGO2%@g-C3N4

AgNPs were prepared over HS-rGO2%@g-C3N4 nanosheets via the reduction of silver ions. In this context, 45 mg of HS-rGO2%@g-C3N4 nanosheets were mixed into 50 mL (2 mM) of AgNO3 solution in a round bottom flask fixed in an ice bath under continuous stirring. Under these conditions, this mixture was sonicated for 30 min to obtain a uniform distribution. Then 20 mL (25 mM) of NaBH4 solution was added dropwise to reduce the silver ions in the prepared dispersion to prepare the AgNPs and attach them to the thiol positions on the nanosheets to yield AgNP-decorated HS-rGO2%@g-C3N4 (AgNPs-S-rGO2%@g-C3N4). Finally, the prepared AgNPs-S-rGO2%@g-C3N4 nanosheets were washed several times with deionised water and methanol by centrifuging them at 8000 rpm, and vacuum dried before further use. The synthetic scheme of AgNPs-S-rGO2%@g-C3N4 (Figure 1).

Figure 1

Scheme for the synthesis of AgNPs-S-rGO2%@g-C3N4.

2.6. Characterisations

Structural properties of g-C3N4, HS-rGO@g-C3N4 and AgNPs-S-rGO@g-C3N4 were studied between diffraction angles from 5 to 80° using X-ray diffraction (XRD), Rigaku Ultima-IV Tokyo Japan. The optical properties were explored using photoluminescence spectroscopy (PL-Kimon 1.K) and UV-Visible diffuse reflectance spectroscopy (HACH LANGE DR 6000). Surface and functional analysis were performed by X-ray photoelectron spectroscopy (PHI-VersaProbII, PHI, Chanhassen, MN, USA) under a high vacuum (10−7 Pa). Morphology and chemical nature of these samples were explored using a Nova NanoSEM 450 (FEI, Hillsboro, OR, USA) equipped with an Everhart-Thornley detector (ETD), a scanning transmission electron detector (STEM), and an energy dispersive X-ray detector (EDX), operated between 15 to 25 kV at a working distance of 5 mm. The evolution of H2 was scrutinised using gas chromatography (Agilent Technologies 7890B-GC System, Santa Clara, CA, USA).

2.7. Hydrogen Production

The photocatalytic study was conducted in a Pyrex beaker that contained 15 mg of AgNPs-S-rGO2%@g-C3N4. The same sacrificial reagent (15% triethanolamine) was used in all experiments. However, before starting the photocatalytic generation of H2, surface-adsorbed oxygen moieties were removed by degassing this reaction assembly with sustained nitrogen gas flow for 1 h. These experiments have been performed under visible light (150 Watt, Xe lamp). The evolution of H2 was scrutinised at regular intervals of 30 min, using gas chromatography (Agilent Technologies, 7890B-GC System). Furthermore, AgNPs-S-rGO2%@g-C3N4 was tested for reusability under the same experimental conditions by four times repeating the experiment.

2.8. Photocatalytic Degradation of Ciprofloxacin Antibiotics

A Ciprofloxacin antibiotic (CIP) was chosen as a model pollutant for the photocatalytic ability of AgNPs-S-rGO2%@g-C3N4. Therefore, 25 mg/L concentration of CIP was prepared. The photocatalytic degradation of CIP was investigated under visible light (450 nm, 150 watts, distance from light source to CIP solution 40 cm, and light intensity 74.64 w/m2) at room temperature while maintaining pH 7 during the study. However, the solution was kept in the dark for 20 min before irradiating to achieve the absorption equilibrium. Afterwards, the CIP solution was started irradiating, and 3 mL was taken with the regular interval of 10 min to analyse the degradation of CIP. The following equation was used to calculate the degradation parameters of CIP [29].

In the above equation, and represents the initial and final concentration of CIP, respectively. Further, the following equation was used to calculate the kinetic rate constant (k) during photocatalytic degradation of CIP [30].

3. Results and Discussion

3.1. Structural Analysis

The diffraction pattern of rGO2%@g-C3N4 shown in Figure 2 indicated that the peaks that appear at around 2θ = 13.26° and 27.46° are attributed to the (001) and (002) diffraction planes, respectively, of the hexagonal form of g-C3N4 (JCPDS # 01-087-1526) [31]. The diffraction plane (001) is associated with in-plane structural packing of tri-s-triazine with an interlayer spacing of 0.66 nm.

Figure 2

(a) Diffraction patterns of rGO2%@g-C3N4, HS-rGO2%@g-C3N4 and AgNPs-S-rGO2%@g-C3N4 whilst (b) enlarged spectra of AgNPs-S-rGO2%@g-C3N4 taken between 30 to 50°.

In comparison, (002) is related to interlayer aromatic stacking of C-N having the interlayer spacing of 0.32 nm [32]. However, the diffraction peaks of rGO could not be seen due to the reduction in functional moieties anchored with its basal plane. The absence of diffraction peaks for rGO is consistent with an earlier study [33]. Compared to rGO2%@g-C3N4, the peak (001) in the diffraction patterns of HS-rGO2%@g-C3N4 was shifted from 2θ = 13.26° to 12.99°, whereas (002) moved from 2θ = 27.46° to 27.5°. After thiolation, the shifting of the diffraction planes is attributed to incorporating the sulphur atoms into the copolymerised sheets of the g-C3N4 and within the layered assemblies of the rGO. This incorporation resulted in the reduction of the thickness of the layers in these composites [13]. In the case of AgNPs-S-rGO2%@g-C3N4, additional diffraction peaks at around 38.18° and 44.36° were observed. These were, respectively, attributed to the (111) and (200) planes of silver, as per JCPDS # 00-001-1164. Moreover, in AgNPs-S-rGO2%@g-C3N4, the small shifts of the (001) and (002) peaks towards lower diffraction angles are attributed to the successful interaction of AgNPs with HS-rGO2%@g-C3N4. These shifts are associated with the covalent interactions of AgNPs with the thiol moieties of HS-rGO2%@g-C3N4, which separates and keeps intact the layered assemblies of the composite material. This is due to the difference in atomic radii of silver, carbon and nitrogen [34]. Ag has a higher atomic radius (0.144 nm) in comparison to carbon (0.077 nm) and nitrogen (0.075 nm) [34]. Therefore, this difference was valuable for promoting successful interaction, which further enhanced light harvesting, resulting in higher photocatalytic H2 production.

3.2. Optical Properties

The optical properties give insight into the charge transfer from the VB to the CB, directly impacting the material’s ability to generate H2 from the water splitting. Therefore, UV-Visible diffuse reflectance spectroscopy was employed to explore the charge transfer conduction rate of rGO2%@g-C3N4, HS-rGO2%@g-C3N4, and AgNPs-S-rGO2%@g-C3N4. The absorption peaks of rGO2%@g-C3N4 were shifted to a higher wavelength after thiolation (HS-rGO2%@g-C3N4) and moved further following seeding with AgNPs (AgNPs-S-rGO2%@g-C3N4), as shown in Figure 3a. Compared to rGO2%@g-C3N4, an increase in the light absorption ability of HS-rGO2%@g-C3N4 is attributed to thiolation [35]. The associated thiol moieties are excellent hole quenchers and are well known to reduce the recombination rates of electron-hole pairs in the resulting composites [20]. Further, in the case of AgNPs-S-rGO2%@g-C3N4, the characteristic SPR of AgNPs enhances light absorption. However, the SPR is dependent on the size of the AgNPs. However, a non-observable SPR peak hump was noticed for AgNPs-S-rGO2%@g-C3N4 caused by two possible reasons (Figure 3a). First, the interaction of thiolated rGO with AgNPs can lead to quenching the SPR in AgNPs-S-rGO2%@g-C3N4 because of the difference in their dielectric constants [35]. A second possible reason may be low concentrations or average particle size (<10 nm). Thus, if the AgNPs have a size within a range of a few nanometres or the amount impregnated is low, their SPR will remain imperceptible [36]. In our case, the size of the AgNPs that could be seen in TEM analysis was found to be less than 10 nm, but their low concentration as determined by the XPS survey scan (Figure S2 and Table S1, Supplementary Materials) could be the second reason for the absence of the AgNPs SPR in the absorbance spectrum of AgNPs-S-rGO2%@g-C3N4.

Figure 3

(a) UV-Visible absorption spectra (b) bandgap analysis and (c) PL spectra for rGO2%@g-C3N4, HS-rGO2%@g-C3N4, and AgNPs-S-rGO2%@g-C3N4.

The change in the absorbance of light can further transform the conduction behaviour of the subjected material, which negatively affects hydrogen production. Therefore, the Kubelka–Munk function was used to estimate the conduction behaviour of rGO2%@g-C3N4, HS-rGO2%@g-C3N4 and AgNPs-S-rGO2%@g-C3N4 (Figure 3b) [37]. The bandgap calculated for rGO2%@g-C3N4 was 2.50 eV, which is less than that of g-C3N4 [38]. Further, reduction in bandgap value of g-C3N4 after the addition of rGO is ascribed to the interactions by covalent bonding between carbon atoms in g-C3N4 and rGO [39]. This interaction leads to the reduction of the bandgap for rGO2%@g-C3N4. The bandgap of rGO2%@g-C3N4 was further reduced after thiolation and was about 2.46 eV for HS-rGO2%@g-C3N4. This decrease in the bandgap after thiolation is attributed to the enhanced strong affinity between the functional groups of g-C3N4 and rGO, leading to improved absorption of light with enhanced charge transfer behaviour between their CB the VB [40]. Further, the bandgap was reduced to 2.42 eV AgNPs-S-rGO2%@g-C3N4. However, during this transfer, these charge carriers may recombine, which hinders H2 production. Therefore, the charge recombination ratio of rGO2%@g-C3N4, HS-rGO2%@g-C3N4, and AgNPs-S-rGO2%@g-C3N4 was estimated using the PL spectra (Figure 3c). The intensity of the PL spectra is directly proportional to the ratio of charge recombinations. The PL spectra of pure g-C3N4 (Figure S1) showed strong emission, attributed to the rapid recombinations of charge carriers. However, the PL intensity was reduced with the addition of rGO to the g-C3N4. The addition of rGO to g-C3N4 behave as shuttle for excited electrons to move from g-C3N4 to rGO at the interface between them and it deters the charge recombination [41]. Therefore, the intensity of the PL spectra for rGO1%@g-C3N4 (Figure S1) was less than that of the pristine g-C3N4 and was further reduced for rGO2%@g-C3N4. However, as the percentage of rGO increased from 2 wt.% to 3 wt.% and 4 wt.% (Figure S1), the PL intensity increased, which indicates the rapid recombination of charged particles in rGO3%@g-C3N4 and rGO4%@g-C3N4. The increase in PL intensity at a higher percentage (above 2% in our case) is attributed to the negative shielding effect of rGO, which enhanced the charge-recombination ratios for rGO3%@g-C3N4 and rGO4%@g-C3N4 [42]. Therefore, we selected rGO2%@g-C3N4 for thiolation. The PL intensity of HS-rGO2%@g-C3N4 was reduced than that of rGO2%@g-C3N4 and was further reduced (Figure 3c) after adding AgNPs. The reduction of the PL intensity after thiolation and seeding with AgNPs (AgNPs-S-rGO2%@g-C3N4) showed that the charge recombination ratio has reduced to the point desired for enhanced H2 production.

3.3. Surface Chemical State Investigation

The C1s spectrum of rGO2%@g-C3N4 (Figure 4a) revealed three peaks at around 284.9, 285.8, and 288.6 eV accredited to the C=C, C=N, and C=O moieties, respectively [43,44], with contributions of 75.28%, 14.55%, and 10.17%. However, in thiolated HS-rGO2%@g-C3N4 and AgNPs coated composites (AgNPs-S-rGO2%@g-C3N4) (Figure 4b,c), the bonding of detected elements was changed as reflected by changes in their spectra. The percentage contributions of C=C, C=N, and C=O changed to 70.38%, 18.69%, and 10.93%, respectively, in HS-rGO2%@g-C3N4, while for AgNPs-S-rGO2%@g-C3N4, they were 58.77%, 29.80% and 11.43%, respectively. The increase in C=O suggests that more anchoring sites are available to enhance catalytic activity [45]. Therefore, AgNPs-S-rGO2%@g-C3N4 also has a high photocatalytic H2 production rate in comparison to rGO2%@g-C3N4 and HS-rGO2%@g-C3N4 (Section 3.5). Further, Figure 4d shows the changes in C=C, C=N, and C=O of rGO2%@g-C3N4, HS-rGO2%@g-C3N4, and AgNPs-S-rGO2%@g-C3N4.

Figure 4

Chemical state analysis of C1s (a) rGO2%@g-C3N4, (b) HS-rGO2%@g-C3N4, (c) AgNPs-S-rGO2%@g-C3N4 and (d) apparent change in functional groups for rGO2%@g-C3N4, HS-rGO2%@g-C3N4, and AgNPs-S-rGO2%@g-C3N4.

The O1s spectra of the prepared materials provided evidence of oxygen intercalations. The peaks that appeared at around 529.2, 531.0 and 532.3 eV are attributed to lattice oxygen (Oi), carbon to oxygen bonding (C=O), and hydroxyl functional groups (OH), respectively [46,47]. The rGO2%@g-C3N4 spectra showed the contribution of Oi to be around 38.82%, whereas the HS-rGO2%@g-C3N4 and AgNPs-S-rGO2%@g-C3N4 spectra contributed 19.49% and 36.78%, respectively (Figure 5a–c). The contribution of C=O was found to be maximised for AgNPs-S-rGO2%@g-C3N4 and was 54.11%, which is according to the spectrum of C1s (Figure 5d). The percentage of OH was around 22.55%, 35.34% and 9.12% for rGO2%@g-C3N4, HS-rGO2%@g-C3N4 and AgNPs-S-rGO2%@g-C3N4, respectively.

Figure 5

O1s-attached functional groups (a) rGO2%@g-C3N4, (b) HS-rGO2%@g-C3N4 (c) AgNPS-S-rGO2%@g-C3N4 and (d) increment in C=O bonding of rGO2%@g-C3N4, HS-rGO2%@g-C3N4, and AgNPS-S-rGO2%@g-C3N4.

The S2p spectrum analysis of HS-rGO2%@g-C3N4 (Figure 6a) shows two peaks at around 164.17 and 165.33 eV, attributed to the 2p3/2 and 2p1/2, respectively, with a separation of 1.18 eV. However, after depositing the AgNPs on HS-rGO2%@g-C3N4, the 2p3/2 and 2p1/2 peaks were shifted to 163.98 and 165.19, increasing the separation of 1.21 eV (Figure 6b). This change in the binding position of the 2p3/2 and 2p1/2 peaks revealed the interaction of the AgNPs with the sulphur group of HS-rGO2%@g-C3N4.

Figure 6

(a,b) S2p chemical state analysis of HS-rGO2%@g-C3N4 and AgNPS-S-rGO2%@g-C3N4 (c) Ag3d analysis for AgNPs-S-rGO2%@g-C3N4.

Furthermore, the Ag3d spectrum of AgNPs-S-rGO2%@g-C3N4 (Figure 6c) was measured to explore the chemical nature of AgNPs. The peaks, which appeared at 367.93 and 373.93 eV, are attributed to Ag3d5/2 and Ag3d3/2, respectively, which confirms the existence of zerovalent silver [48]. Moreover, zerovalent silver in AgNPs is highly desirable as it enhances the rate of photocatalytic reactions. In the present work, this property of silver is also responsible for accelerating hydrogen production [49].

3.4. Surface Morphology

The well-dispersed aqueous droplets of rGO2%@g-C3N4, HS-rGO2%@g-C3N4, and AgNPs-S-rGO2%@g-C3N4 were drop-casted onto copper stubs and air-dried before scanning electron microscope (SEM) analysis was conducted. Likewise, for transmission electron microscope (TEM) analysis, the sample of AgNPs-S-rGO2%@g-C3N4 was prepared by drop-casting it onto a copper-grid followed by air-drying. The morphology appearing in the SEM images of rGO2%@g-C3N4, HS-rGO2%@g-C3N4, and AgNPs-S-rGO2%@g-C3N4 revealed that there is no interlayer stacking between g-C3N4 and rGO layers in their thiol-enriched (HS-rGO2%@g-C3N4) and AgNP-decorated (AgNPs-S-rGO2%@g-C3N4) composites compared to the non-functionalised (rGO2%@g-C3N4) composite (Figure 7a–c).

Figure 7

(a–c) SEM analysis and (d–f) EDS analysis of rGO2%@g-C3N4, HS-rGO2%@g-C3N4, and AgNPs-S-rGO2%@g-C3N4.

In parallel (Figure 7d–f), the thiol peaks indicated in the EDX spectrum have also confirmed the thiol enrichment in HS-rGO2%@g-C3N4,. According to XPS studies, the engagement of these thiol groups with NaBH4 reduced the silver ions under these experimental conditions, especially continuous sonication. It ensured the formation of AgNPs in the resulting AgNPs-S-rGO2%@g-C3N4 composite. Briefly, small-sized AgNPs (less than 10 nm) were formed between the interlayer stacking of g-C3N4 and rGO in AgNPs-S-rGO2%@g-C3N4, as seen in the TEM analysis shown in Figure 8.

Figure 8

(a) Surface morphology of AgNPs-S-rGO2%@g-C3N4 via TEM and (b) the estimated size of the AgNPs was 7.5 ± 2.3 measured using imageJ software.

3.5. Hydrogen Production and Proposed Mechanism for AgNPs-S-rGO2%@g-C3N4

The photocatalytic H2 production by g-C3N4 (Figure 9a) had a production rate of 555-μmol g−1 h−1, and this low rate was accredited to the fast recombinations of charge carriers (electron-hole pair). This H2 production rate of g-C3N4 was further improved with the addition of rGO. In rGO1%@g-C3N4, the reaction rate increased to 886.66 μmol g−1 h−1, which is 1.59 fold of g-C3N4. This increment in rate exhibited the contribution of rGO to accelerating the charge transfer efficiencies by tumbling the speedy recombinations of charge carriers. The role of rGO in increasing the H2 production rate is consistent with the previous literature [50]. However, in these measurements, the H2 production rate was reached 1369.86 μmol g−1 h−1 with an increased weight percentage of rGO from 1% to 2% (rGO2%@g-C3N4), which is 2.4-fold higher than the rate of g-C3N4. This was the maximum H2 production rate achieved compared to 1, 3 and 4 wt.% loading of rGO and confirms the positive role of rGO as a co-catalyst in nanocomposites in enhancing catalytic performance. Thus, it is imperative to adjust the weight percentage of rGO in the subjected nanocomposites as its higher ratio can produce shielding effects, which ultimately reduces catalytic performance [51].

Figure 9

(a) H2 production and (b) H2 production rate per hour under the same experimental conditions for g-C3N4 and its composites with rGO at a concentration of 1 to 4 wt.%.

Therefore, the catalytic performance was reduced by further increasing the weight percentage of rGO to 3% and 4%. The rGO3%@g-C3N4 showed H2 production rate of 1220 μmol g−1h−1, whilst rGO4%@g-C3N4 had a rate of 1093.33 μmol g−1h−1. This reduced hydrogen production above a specific weight percentage of added rGO is attributed to the light-shielding effect. This kind of result for a particular weight percentage of rGO has also been reported earlier [52]. Therefore, rGO2%@g-C3N4 was selected as the most optimised sample for maximum hydrogen production under set experimental conditions. Figure 9b shows the change in the H2 production rate with rGO to g-C3N4 at various concentrations of rGO. The introduction of thiol groups (HS) could further increase the H2 production rate of rGO2%@g-C3N4. In addition, the groups can act as hole quenchers which also reduce the charge recombination ratio [53]. Therefore, HS-rGO2%@g-C3N4 (after the thiolation of rGO2%@g-C3N4), Figure 10a shows the H2 production rate of 1745.23 μmol g−1h−1, which was 1.27 times higher than the rate of rGO2%@g-C3N4. The HS groups have a strong affinity for noble transition metals; therefore, AgNPs were grown, using a chemical reduction method, over the surface of thiolated sheets of rGO2%@g-C3N4, resulting in AgNPs-S-rGO2%@g-C3N4. This AgNPs-S-rGO2%@g-C3N4 (Figure 10a) was found to generate hydrogen at a rate of 3772.5 μmol g−1h−1, which is 2.16- and 2.76-fold higher than HS-rGO2%@g-C3N4 and rGO2%@g-C3N4, respectively. Moreover, Table 1 shows the ternary composites is better than binary composites for photocatalytic hydrogen production. These results also revealed the potential use of thiolated reduced graphene-oxide and graphitic carbon nitride with silver nanoparticles for enhanced photocatalytic hydrogen production from water.

Figure 10

(a) Hydrogen production and (b) rate of hydrogen production per hour under the same set experimental conditions for rGO2%@g-C3N4, HS-rGO2%@g-C3N4 and AgNPs-S-rGO2%@g-C3N4.

Table 1

Ternary composites of g-C3N4/rGO is better than binary composites for photocatalytic hydrogen production.

Binary Photocatalysts	Additive Co-Catalyst	Condition of Sacrificial Agent	Radiation Source	Maximum Hydrogen Yield(µmol h−1 g−1)	Reference with Year
Thiolated g-C3N4/rGO	Ag	Triethanolamine 15%	300 W Xenon lamp	3772.5	This work
g-C3N4/rGO	MoS2	Triethanolamine 0.1%	300 W Xenon lamp	317	[50] 2018
g-C3N4/rGO	MoS2	Sodium sulfite 0.25 Molar	450 Xenon lamp	1650	[54] 2017
g-C3N4/rGO	NiS2	Triethanolamine 1%	300 W Xenon lamp	1555.34	[25] 2019
g-C3N4/rGO	CoMoS2	Triethanolamine 1%	300 W Xenon lamp	684	[55] 2018
g-C3N4/rGO	CdS	Lactic acid 10%	350 W Xenon lamp	1000.5	[56] 2017

The stability of AgNPs-S-rGO2%@g-C3N4 was also investigated by repeating the experiment under the same conditions for four cycles (Figure 11a). The AgNPs-S-rGO2%@g-C3N4 was washed multiple times with ethanol and deionised water before starting each cycle. The efficiency of the AgNPs-S-rGO2%@g-C3N4 (Figure 11b) remained at 96% even after four cycles, supporting the reusability of the synthesised AgNPs-S-rGO2%@g-C3N4 catalyst.

Figure 11

(a,b) Reusability and stability of AgNPs-S-rGO2%@g-C3N4 for four consecutive cycles.

The photocatalytic production of H2, illustrated in Figure 12, elaborates that, in AgNPs-S-rGO2%@g-C3N4, the electrons (e−) in the VB of g-C3N4 were excited by visible light to the CB, leaving behind holes (h+) in the VB. The incorporation of AgNPs in this composite also helped absorb light due to their characteristic SPR and improved the catalytic activity by reducing the recombination of photogenerated electrons and holes. The photoexcited e− in CB of g-C3N4 reacted with water molecules adhered to its surface and generated hydrogen gas. Meanwhile, some of the e− in the CB of the g-C3N4 were transferred to the CB of rGO. Similar to g-C3N4, these electrons (e−) in the CB of the rGO also reduced water to H2. At the same time, the h+ generated in the g-C3N4 oxidised the existing species of hydroxides to water and oxygen molecules.

Figure 12

Proposed enhanced hydrogen production mechanism of AgNPs-S-rGO2%@g-C3N4.

3.6. Photocatalytic CIP Degradation

The UV-absorption spectra (Figure 13a) of CIP antibiotics (25 mg/L and pH 7) showed the appearance of the maximum absorption at 274 nm, which is the characterises peak of CIP [57]. This absorption continued down with the increase in the irradiation time. The % removal of CIP with and without a catalyst is shown in Figure 13b. The results revealed 2% removal of CIP in the absence of rGO2%@g-C3N4 (photolysis) which also indicated the stability of CIP [58]. However, after adding rGO2%@g-C3N4, its degradation was started and reached 83.29% within 70 min by irradiating under visible light (400 nm, 150 watts). These results well exposed the potential benefit of rGO2%@g-C3N4 photocatalyst for the removal of CIP from contaminated water. The CIP degradation was further enhanced to 90.29% and 95.90% with HS-rGO2%@g-C3N4 and AgNPs-S-rGO2%@g-C3N4, respectively. The reaction rate was 0.02555, 0.03331 and 0.04564, respectively, for rGO2%@g-C3N4, HS-rGO2%@g-C3N4 and AgNPs-S-rGO2%@g-C3N4. It is inferred that the reaction rate has reached a maximum value for AgNPs-S-rGO2%@g-C3N4 (Figure 13c,d). Generally, the degradation ability by catalytic systems depends on various factors such as its band gap, charge carries recombination ratio etc.

Figure 13

(a) UV-Absorption spectra of CIP (b) percentage removal of CIP with and without a catalyst (c,d) reaction rate constant for CIP degradation.

The bandgap of AgNPs-S-rGO2%@g-C3N4 was (2.42 eV) which is less in comparison to HS-rGO2%@g-C3N4 (2.46 eV) and rGO2%@g-C3N4 (2.50 eV). This indicates that the conduction of charge carrier is high for AgNPs-S-rGO2%@g-C3N4 in comparison to HS-rGO2%@g-C3N4 and rGO2%@g-C3N4. Moreover, PL analysis (Figure 3C) also indicated the separation of charge-recombination ratios for AgNPs-S-rGO2%@g-C3N4 while moving towards the conduction band. This supports accelerating the conduction of charge carriers, ultimately producing the hydroxyl, reactive oxygen spices and super radical oxides supporting CIP’s enhanced degradation [59].

Effect of AgNPs-S-rGO2%@g-C3N4 Dosage over CIP Degradation

The effect of heterogeneous catalyst dosage may change the reaction rate and efficiency of the photocatalytic process. Therefore, further investigation was carried out at different catalyst dosages of AgNPs-S-rGO2%@g-C3N4 while keeping the other reaction parameters such as irradiation time, pH and dye concentrations constant. In this regard, 15 mg/L, 25, 35 and 45 mg/L of AgNPs-S-rGO2%@g-C3N was selected while keeping the CIP concentration fixed, i.e., 25 mg/L and pH was kept constant at 6. The results indicated (Figure 14) the enhanced photocatalytic degradation of CIP, when the dosage of AgNPs-S-rGO2%@g-C3N4 increased from 15 mg/L to 25 mg/L. However, after increasing the dosage of AgNPs-S-rGO2%@g-C3N4 from 25mg/L, 35, and 45 mg/L, a decline in CIP degradation was noticed. Therefore, maximum CIP degradation was found 84.87%, 95.90%, 88.22% and 87.27%, respectively, for 15 mg/L, 25, 35, and 45 mg/L. These results indicated that the optimal value of AgNPs-S-rGO2%@g-C3N4 was 25 mg/L necessary to obtain the maximum degradation of CIP at pH 6. With an increase in the photocatalyst dosage, an obvious increase in the reactive/reaction sites occurred, leading to a high CIP’s photocatalytic degradation rate. This enhanced degradation was attributed to the increase in hydroxyl radical while irradiation. However, a high catalyst dosage of AgNPs-S-rGO2%@g-C3N4 beyond its optimal value might lead to suspension opacity, which possibly increased the light scattering inside the reactor and low infiltration of phonons. These factors will eventually decrease the catalytic rate due to the less active catalyst present in the reactor. It can be stated that when present in high concentrations, the catalyst nanoparticles become agglomerated, leading to a decrease in the available reactive active sites. Further deactivation of active catalyst might occur due to the bombardment of activating molecules with molecules at the ground state, which resulted in the decrease of photocatalytic efficiency [60].

Figure 14

(a,b) Effect of AgNPs-S-rGO2%@g-C3N4 dosage over CIP degradation at constant CIP concentration (25 mg/L) and constant pH-6.

4. Conclusions

The ratio of rGO to g-C3N4 has a fundamental impact on the photocatalytic CIP degradation and H2 production rate. The optimal concentration of rGO will curb the charge recombination ratio of the charged particles at the interface of their heterojunctions. Therefore, the optimum concentration of rGO, i.e., 2 wt.%, in the composite rGO2%@g-C3N4, exhibited the maximum H2 production rate. Furthermore, the thiolation of rGO and g-C3N4 in HS-rGO2%@g-C3N4 also overcame the negative shielding effect of rGO. It decreased the bandgap, building the capability to absorb more photons, which is favourable for enhanced catalytic activity. Moreover, shifting the diffraction peaks for planes of thiolated rGO and g-C3N4 indicated a successful interaction between thiolated functional groups and AgNPs, as confirmed by XPS studies. This interaction also formed smaller AgNPs, keeping the layered assemblies of AgNPs-S-rGO2%@g-C3N4 intact, stable and exfoliated. Moreover, C1s and O1s spectra revealed the fundamental role of C=O in the enhanced CIP and hydrogen production activity of AgNPs-S-rGO2%@g-C3N4 compared to HS-rGO2%@g-C3N4 and rGO2%@g-C3N4, which were 3772.5, 1745.2, and 1369.8 μmol g−1h−1, respectively, under the experimental conditions.