Introduction

An eco-friendly substitute of the fossil fuels is hydrogen. Because the only product of hydrogen combustion in a gasoline engine is water. Furthermore, hydrogen has been proven to be the cleanest fuel and energy carrier on the planet due to its high energy density (142 kJ kg− 1), fuel efficiency (75%), heating value (52,000 Btu per lb), and largest range of flammability (4–75%)1. Although water is a good source of hydrogen but pure water cannot be broken. Therefore, to split water, the apt catalytic material (such as metal oxide, metal chalcogenide, and carbonaceous materials, etc.) is required. The metallic catalysts are either toxic or precious, or both. Thus, carbonaceous materials are better choice to serve as catalytst. The ingress of graphitic carbon nitride (g-C3N4) as a photocatalyst with moderate band-gap (∼2.7 eV), suitable band positions, high chemical/thermal stability, and other properties, will gear up the research in this field2. However, its performance is still restricted due to insufficient absorption of visible light, poor surface area, low electroconductivity, and high rate of recombination of photocarriers. Therefore, various modification strategies viz. elemental doping, defect creation, surface modification, composite formation, co-catalyst addition, and many more, are adopted to accelerate their photocatalytic performance3. The surface modification by composite formation is the best way to modify g-C3N4 for improving its photocatalytic properties. Several types of g-C3N4-based heterojunction composite have been developed by coupling g-C3N4 with other photocatalytic materials. But the good compatibility of g-C3N4 with graphene oxide (GO) with its electron-rich 2D-framework, free sp2 electron, low band-gap, high surface-area, good mechanical-stability, high electronic mobility and fascinating optoelectronic properties, will strongly advocate its candidature for its coupling with g-C3N44. The nitrogen conjoined into the carbon matrix enhances electron-donor properties, surface polarity, and electrical conductivity, while their improved stability in the electrolyte confirms the better charge transfer efficiency of the composite of g-C3N4 with GOs. That could result in unique structural, morphological, compositional, surface, and electrochemical properties by limiting the weaknesses of g-C3N4 and GO based composite materials. Some of the notable examples of GO and g-C3N4 coupled systems that used for H2 production through water splitting are: pPCN/2wt%rGO (715 µmolh− 1g− 1)5, Pt-rGO/g-C3N4(0.08%) (874.4µmolh− 1g− 1)6, GO/g-C3N4 (224.6 µmolh− 1)7, CNGO/CNQDs (1231 µmol h− 1)8, Ag20%@g-C3N4/r-GO (A20CNG) (954 µmol h− 1g− 1)9, 30%InVO4-g-C3N4/3%rGO (7449 µmolh− 1g− 1)10, H+/g-C3N4/GO-COOH (1091 µmolh− 1g− 1)11 and Z-scheme g-C3N4/rGO/ln2S3 (512.72 µmolh− 1g− 1)12.

The combination of GOs with g-C3N4 along with their high surface area, outstanding electrical conductivity, stability, and oxygen-containing functional groups make it a great material for supercapacitor electrodes. The material’s dispersibility and processability are further improved by their functional groups, which also increase ion-adsorption and raise capacitance. Ruoff et al.13 reported a graphene-based supercapacitor that demonstrated specific capacitances of 135 and 99 F g− 1 in aqueous and organic electrolytes, respectively. Wang et al.14 reported the reduced graphene as electrode materials and investigated their electrochemical performance. After 1200 cycle tests, a maximum specific capacitance of 205 Fg− 1 was achieved, along with a long cycle-life of almost 90% specific capacitance retention, which is far higher than that of carbon nanotube (CNT)-based supercapacitors. Further, GO and g-C3N4-based systems like: rGO/g-C3N4 /Ag2O/PANI15, Ni(OH)2/g-C3N4/RGO16, SnS2-gC3N4/rGO17, g-C3N4/rGO18, g-C3N4@RGO/ Ni(OH)2 composite19 and ternary rGO-pg-CN/PPyNTs nanocomposite20, are cited in Table 1 as supercapacitor material.

Table 1 GO and g-C3N4-based systems with their respective electrolyte used, specific conductance, current density, and retention power.
Full size table

The commercial potential of graphene-based supercapacitors as a futuristic affordable, effective, and environmentally friendly system, is used to establish them as a prominent candidate for electrical energy storage devices.

Here, we synthesise the g-C3N4, GO, g-C3N4@GO, g-C3N4@r-GO(AA), and g-C3N4@r-GO(NaBH4) systems, and apply them in electrochemical supercapacitor and water splitting for hydrogen generation. The above-mentioned systems were tested through Fourier emission scanning electron microscopy (FE-SEM), powder X-ray diffraction (PXRD), optical absorption spectroscopy, Fourier transmittance infra-red spectroscopy (FTIR), steady-state photoluminescence emission (PLE) spectroscopy, surface photovoltaic spectroscopy (SPV), cyclic voltammetry (CV), galvanostatic charge discharge (GCD), and electrochemical impedance techniques, etc. When we compare our studied composite g-C3N4@r-GO(AA) with the reported systems then it was found the performance of our prepared system as supercapacitor as well as photocatalytic material for water splitting is quite remarkable in terms of ease of composite synthesis, stability, biocompatibility, cost-effectiveness, hydrogen generation power by water splitting (339.82 µmolh− 1g− 1 with 2.52% apparent quantum efficiency at 420 nm in 20% CH3OH), and supercapacitor for electrical energy storage systems. As a supercapacitor it shows excellent specific conductivity (322.77 F/g) at reasonably low current density (2 A/g) along with good retention power (117.45% after 2000 cycles and 78.56% after 5000 cycles at 3 A/g) under the minimum concentration of electrolyte 2MKOH and current density.

Results

Synthesis

The GO was synthesized by using modified Hummer method21 and reduced by using the reducing agents i.e. ascorbic acid and NaBH4. Then g-C3N4 was synthesised using combustion of melamine at 550 °C for 4 h. Composite of above compounds was made when as-prepared g-C3N4 and GO was sonicated for 1 h.

Morphological investigation by FESEM

FE-SEM investigations are used to determine the surface morphology, and particle size of the investigated samples. The systems, GO (5.0 nm thick and 36–43 μm long), g-C3N4 (1.5 μm thick and 2.0 μm long), g-C3N4@GO (12.0 μm thick and 26.0 μm long) nanocomposite, g-C3N4@r-GO (AA) (10.0 μm thick and 22.0–31.0 μm long) nanocomposite, and g-C3N4@r-GO (NaBH4) (1.0–2.0 μm thick and 3.0–7.0 μm long) nanocomposite were crystallized in the form of cluster of the flakes, shallow-pored surface, multifaceted solid rod-like structures, hollow rods with facets inside and solid rods with outer facets like artifact (Fig. 1a-e; Table 2). Figure 1c illustrated the shallow cavities /pits (2.26 × 1.46 μm) of irregular shape on the surface of g-C3N4@GO composite, which might act as recombination centres for the photo carriers i.e. photoelectron and photohole.

Fig. 1
figure 1

FE-SEM images. (a)GO, (b)g-C3N4, (c) g-C3N4@GO, (d) g-C3N4@r-GO(AA), and (e) g-C3N4@r-GO(NaBH4) with scale in yellow color.

Full size image

XRD analysis

The crystalline phase and cell parameters of the as-synthesized systems (GO, g-C3N4, g-C3N4@GO, g-C3N4@r-GO(AA), and g-C3N4@r-GO(NaBH4)) were investigated by using their XRD patterns (Fig. 2a and b; Table 1). The studied systems exhibited two prominent diffractions around 13.0° (100 plane referred to oxidized sp3-carbon) and 27.7° (002 plane referred to non-oxidized sp2-carbon), which corresponds to the typical interplanar structural packing of s-triazine units and interlayer stacking of graphene-like carbon, respectively22,23,24. Few minor diffractions were also observed at 2θ∼ 20.73° (101 plane due to g-C3N4), and 2θ∼ 42.58° (300 plane due to GO25, referred to random rotations and translations between the pair of layers, known as turbostratic disorder26. The XRD profile of the GO-based composites (i.e. g-C3N4@GO, g-C3N4@r-GO(AA), and g-C3N4@r-GO(NaBH4)) demonstrate the blue shift in the diffraction (100) plane at 2θ = 11.85° with respect to GO, where the diffraction along 002 plane at 2θ = 27.7° remain intact that confirms the addition of the lighter elements (GO or r-GO) in g-C3N4 lattice. The diffraction patterns of the g-C3N4@r-GO(AA) nanocomposite displayed the highest (002) plane over minimum (100) plane that reflects the maximum reduction of GO by ascorbic acid.

Fig. 2
figure 2

X-ray diffraction patterns. (a) 2θ angle between 0° and 80° and (b) focused (002) plane around 11–13° and (001) diffraction and 27.7° angles of the as-studied samples: pristine GO, g-C3N4, g-C3N4@GO, g-C3N4@r-GO(AA) nanocomposite and g-C3N4@r-GO(NaBH4) nanocomposite. Inset of (b) focus on shifting of (100) diffraction plane around 6° to 15°. (c) FTIR spectra of the as-studied samples.

Full size image

The crystallinity of the solid solutions was determined by using the following formula (Eq. 1)27.

$${{\text{C}}_{{\text{rt}}}}={\text{ }}\left( {{{\text{I}}_{{\text{111}}}} - {{\text{I}}_{{\text{am}}}}} \right){\text{ }} \times {\text{1}}00/{{\text{I}}_{{\text{111}}}}\left( \% \right)$$
(1)

Where, Crt = the relative crystallinity (%); I111 = highest intensity of the (111) diffraction angle of the crystal lattice (arbitrary unit; au) and Iam= scattering strength diffracted by the non-crystalline (amorphous) environment (arbitrary unit; au). The X-ray profile was also used to determine the dislocation density (δ) of the samples using the following expression (Eq. 2)27.

$$\delta {\text{ }}={\text{15 }}\beta {\text{ cos}}\theta /{\text{4aD}}$$
(2)

Where, broadening of the diffraction line, i.e. full width at half maximum intensity is β (radian), Bragg’s diffraction angle is θ (degree), lattice constant is a (nm), and particle size is D (nm). The variation in dislocation density is related to the enhanced order in the crystal structure25. The strongest peak (100/002 plane) of the sample used to determine the degree of crystalline order27 with respect to the amorphous phase (Table 2). The highest crystallinity was observed for pristine g-C3N4, which was decreased on addition of GO/rGO. Further, Scherrer formula was used to determine the layer thickness (Table 2). Composite g-C3N4@r-GO(AA) exhibited the minimum layer thickness with highest dislocation density.

Table 2 Crystalline parameters (particle size, particle thickness, band gap, dislocation density, and crystallinity) of the studied samples.
Full size table

FT-IR analysis

The FT-IR spectra were observed in the range of 4000 –500 cm− 1 wavenumber for the g-C3N4, GO, g-C3N4@GO, g-C3N4@r-GO(AA), and g-C3N4@r-GO(NaBH4), are illustrated in the Fig. 2(c). The corresponding functional groups present in samples are discussed in Table S2. The IR bands found at 523 cm− 1 due to the C–C = O bending vibration and IR bands at 800 cm− 1 owing to the out-of-plane bending vibrations of the six membered rings that either belongs to triazine or heptazine units within the structures28. The intense C—O stretching vibration of ether group was observed around 1079 cm− 1 wavenumber in the 1050–1200 cm− 1 IR spectrum region with their comparative intensity g-C3N4@r-GO(AA) > g-C3N4@r-GO(NaBH4) > g-C3N4@GO > GO > g-C3N4, The triazine or heptazine units are connected through the -NH groups that show IR bands in 1200–1400 cm− 1 that are the characteristic of the C-NH-C units in melam and melon29. Further, multiple bands found in the range of 1200–1600 cm− 1 are typical of C-N streching and bending vibrations of nitrogen heterocycles. The typical bending vibration band (sp3 C − H) found at 1384.03 cm− 1. Blue shift in IR bands as well as breaking of IR bands (1465 cm− 1 and 1417 cm− 1; 1384 cm− 1 and 1325 cm− 1 in g-C3N4@r-GO(AA) was observed, are attributed to the strong degree of reduction in g-C3N4@r-GO(AA)30. Bands observed at 2200 cm− 1 are resulted from the triple bond between C-N or CN = C = N present in the g-C3N4 or composites31,32. The g-C3N4@GO composite exhibited the IR peaks at 1777.89 cm− 1 and 2393.95 cm− 1 due to symmetrical carbonyl and C≡N streching, which confirms the association of g-C3N4 with GO bond in system.The IR band found at 2920.44 cm− 1 is attributed to the C-H(CH2) asymmetric stretching and band at 2851.66 cm− 1 allied with the C-H(CH2) symmetric stretching vibration. The OH- group of phenol or NH– group of secondary amides or amine corresponding to the IR band at 3278.76 cm− 1.

UV-Visible analysis

The UV-Visible optical absorption spectra (Figs. 3a, S1 and Table S2) were recorded for g-C3N4, GO, g-C3N4@GO, g-C3N4@r-GO(AA), and g-C3N4@r-GO(NaBH4) systems. All UV-Visible spectra show a maximum peak around the wavelength 230 nm. Optical spectra of the as-prepared samples were deconvoluted into the six prominent peaks at wavelength 204 nm (π→π transitions), 224–226 nm (π→π* transitions), 269–294 nm (sp2 conjugated double bonds), 312–378 nm (C: N-C charge transfer), 412 nm (n→π* electronic transitions with N lone pairs) and 715–806 nm (n→π* transitions with N lone pairs and multiple bands), are shown in Fig. 3b and f 33. The g-C3N4 attached to r-GO via the π→π* stacking mechanism that leads to the delocalization of the π-electrons in both aromatic rings (g-C3N4 and r-GO). The direct band and indirect band gap of the studied samples are measured through Tauc plots (Table S2) that indicates the minimum band gap (2.14 eV) for g-C3N4@r-GO(AA), which might contribute to high H2 production through water splitting and super conductance.

Fig. 3
figure 3

Optical absorbance spectra. (a) UV-Vis absorption spectrum of the all studied samples and deconvoluted UV spectra of (b) g-C3N4, (c) GO, (d) g-C3N4@GO, (e) g-C3N4@rGO(AA), and (f) g-C3N4@rGO(NaBH4).

Full size image

The deconvoluted UV-Vis peak around 204 nm attributed to π→π transitions. Electronic transitions between the basal plane of g-C3N4’s conjugated double bonds (sp2)34,35 and GO or r-GO contributed to the deconvoluted peak at 224–256 nm. The band between 269 and 294 nm assigned to π-π* transitions which commonly observed in conjugated ring systems including heterocycle aromatics. One prominent absorption band found at 312–378 nm linked to the transfer of charge from N to C. An extra absorption zone straddling between 412 nm and 715–806 nm were attributed to the n–π* electronic transitions involving the N lone pairs and multiple bonds. This absorption region was permitted for deforming polymeric units (due to coupling of GO with g-C3N4) but prohibited for planar symmetric s-triazine or heptazine structures36.

Photoluminescence (PL) study

Steady state PL spectra were recorded at excitation wavelength 238 nm (Fig. S2). Furthermore, the deconvoluted photoluminescence emission (PLE) spectra of the samples are shown in Fig. 4, where the order of the emission peak intensity of the observed samples is: g-C3N4 > GO > g-C3N4@GO > g-C3N4@r-GO(NaBH4) > g-C3N4@r-GO(AA). The excitonic PL phenomenon in composite might be originated from three different transitions: σ*–LP, π*–LP, and π*−π, electronic transition of lone pairs of nitrogen atoms in g-C3N4 with rGO. Rapid electron-hole pair recombination causes the high emission intensity because defects/pits in structure became centre of charge particle recombination. By modification of g-C3N4 with r-GO(AA), the fluorescence intensity decreases due to decrease in defect density in g-C3N4@r-GO(AA), associated with the effective photogenerated carrier separation and smooth charge transfer, which significantly increases the photocatalytic activity and specific conductance of g-C3N4@r-GO(AA)37.

Fig. 4
figure 4

Photoluminescence response of the samples. Deconvoluted photoluminescence emission (PLE) spectra of (a)g-C3N4, (b)GO, (c)g-C3N4@GO, (d) g-C3N4@r-GO(AA) and (d) g-C3N4@r-GO(NaBH4) systems observed at 238 nm excitation wavelength.

Full size image

The PL spectra experienced at 238 nm excitation of complexes g-C3N4@r-GO(AA) and g-C3N4@r-GO(NaBH4) is deconvoluted into three major peaks positioned at 362, 391 and 434/443 nm are originated from three transitions: σ*–LP (lone pair), π*–LP and π*–π, respectively38. The blue shift in PL peak was observed after incorporating GO or r-GO into the g-C3N4 sample are attributed to the thermal expansion and exciton–phonon interaction39. Further, the planar structure of GO/r-GOs in composites consisting of delocalized π-electrons that enables its strong π-π interactions with π-conjugated aromatic moiety (g-C3N4)40 is an effective way to achieve the efficient photogenerated carrier separation. The IR peaks exhibited at 2977, 2211, 2111, 1979, 1384, 1255 and 1141 cm− 1 wave number, represent the high degree of π-conjugation in g-C3N4@r-GO(AA) over g-C3N4@r-GO(NaBH4), confirms the above statement.

Surface photovoltage study

Surface photovoltage (SPV) technique has been used to investigate carrier transfer dynamics. The SPV signal is a vector makeup of the value denoting the signal intensity and the phase denoting its direction, reflects the semiconductor’s surface-in-built electric field (p-n junction) generated on light irradiation. The SPV response (Fig. 5) of system with surface potential order are: g-C3N4 (n-type material with surface potential − 411.7 mV) < g-C3N4@GO (surface potential − 117.2 mV) < g-C3N4@r-GO(NaBH4) (surface potential 115.2 mV) < g-C3N4@r-GO(AA) (surface potential 597.2 mV) < GO (p-type material with surface potential 802.7 mV) that confirming a very low efficiency of the photovoltaic conversion that attributed to the Schottky interface contact between the g-C3N4 and GO in absence of electrolyte, which leads to the limited electron transfer process from the system to GO and approve the formation of p-n junction in composites between GO or r-GO and g-C3N4.

Fig. 5
figure 5

Surface photovoltaic spectra of g-C3N4, GO, g-C3N4@GO, g-C3N4@r-GO(AA) and g-C3N4@r-GO(NaBH4) systems.

Full size image

The samples GO, g-C3N4@r-GO(NaBH4) and g-C3N4@r-GO(AA) produce a positive voltage signal in surface photovoltage spectra is attributed to the electrons that excited by the groups –OH/ -COOH/ C-O-C and acting as the electron distribution points to flow away electrons. It confirms the p-type nature of GO and rest of the samples are either n-type semiconductor(g-C3N4) or having p-n junction41. This investigation elucidates the photochemical charge transfer through solid–solid interfaces at g-C3N4 and GO or r-GO junction. Here, the measured photovoltage is positive, where electrons are being carried away from the GO and moving in the direction of the Kelvin probe/ or the holes are being transported toward the g-C3N4. Where, the negative signal of n-type g-C3N4 promote the electrons to reach GO from the Kelvin probe/ or the electrons are being transported toward the g-C3N4. That confirm the formation of p-n junction in g-C3N4@r-GO(NaBH4) and g-C3N4@r-GO(AA) composites.

Electrochemical analysis

Cyclic voltammetry (CV)

Cyclic voltammetry (CV) is used to measure the electrochemical behavior of g-C3N4, GO, g-C3N4@GO, g-C3N4@r-GO(AA) and g-C3N4@r-GO(NaBH4), and their current-voltage characteristics (Fig. 6(a)). The capacity of CV analyses of the built electrode of studied samples was performed in 2 M KOH within a potential window of 0.2 V to 0.6 V at varied current densities. The specific capacitance (CSP) of above were calculated using their CV curves observed at various scan rate and illustrated by Table 3, which exhibit the max CSP value for g-C3N4@r-GO (AA) sample 322.77 F/g at scan rate of 5mVs− 1. The variation of CSP with scan rate for g-C3N4@r-GO (AA) sample was observed (Fig. 6(b)). The g-C3N4 exhibits strong chemical stability and pseudo-capacitive behaviour despite its moderate conductivity, whereas GO offers great electrical conductivity, a wide surface area, and outstanding flexibility42. But when GO and g-C3N4 are combined, the drawbacks of each compound are addressed; GO improves conductivity, while g-C₃N₄ provides redox-active sites43. Thus, g-C₃N₄@r-GO(AA) composites exhibit superior super-capacitive properties because of the synergistic effects and p-n junction build between the fundamental components, which are further reinforced by electrical conductivity, surface area, porosity, and electrochemical activity.

By offering a conductive pathway that increases charge-discharge rates and decreases internal resistance, the multilayered GO structure speeds up the transfer of electrons. In Faradic (redox) processes, the nitrogen functional groups of g-C₃N₄ facilitate ion transport44. Furthermore, it was revealed that increasing the scan rate from 5 mVs− 1 to 100 mVs− 1 results in a decrease in the CSP (from 322.77 F/g to 35.95 F/g) of the g-C3N4@r-GO (AA) electrode (Fig. 6c and Table 3). Interestingly, all the samples evaluated in the current investigation demonstrate a significant hysteresis loop, suggesting their potential application as a memory element. The r-GO exhibits the greatest hysteresis area compared to other samples because it contains more ionic functional groups than other nanocomposites45.

Fig. 6
figure 6

Electrochemical behavior measure by cyclic voltammetry. (a) CV curves of g-C3N4, GO, g-C3N4@GO, g-C3N4@r-GO (AA) and g-C3N4@r-GO (NaBH4) at a scan rate of 5 mV/s in 2 M KOH (b) CV curves of g-C3N4@r-GO (AA) at various scan rates (5, 10, 25, 50, 75, 100 mV/s). (c) Variation of specific conductance of samples at different scan rate.

Full size image
Table 3 Specific capacitance (CSP) of g-C3N4, GO, g-C3N4@GO, g-C3N4@r-GO (AA) and g-C3N4@r-GO (NaBH4) measured using the CV curve at different scan rates.
Full size table

Galvanostatic charge-discharge (GCD)

A galvanostatic charge-discharge (GCD) investigation was done for the manufactured electrodes to analyze the extent of discharge capacitance under the potential window of 0.4 V. The symmetric non-triangular GCD curves with the plateau at a voltage exhibited the capacitive behavior of the prepared electrodes, as shown in Fig. 7(a) and Table 4. The CSP has been calculated for the g-C3N4, GO, g-C3N4@GO, g-C3N4@r-GO (AA) and g-C3N4@r-GO (NaBH4) systems using the Eq. (2). Interestingly, the highest CSP value (117.7 F/g) is found for g-C3N4@r-GO (AA) composite in compared to the rest of the studied system. Therefore, the GCD curves for the g-C3N4@r-GO(AA) electrode were plotted by varying current density (Fig. 7b). As current density increases, capacitance decreases. The GCD curve for g-C3N4@r-GO(AA) electrode is asymmetric compared to its discharge counterpart46. Overall GCD data show significantly lower capacitance values than the CV results.

Fig. 7
figure 7

GCD curves. (a) Specific capacitance of g-C3N4, GO, g-C3N4@GO, g-C3N4@r-GO (AA) and g-C3N4@r-GO (NaBH4) at current density of 2 A/g (b) GCD curves of g-C3N4@r-GO (AA) at various current density (c) Variation in Csp at different current density for studied systems. (d) Capacitive retention of g-C3N4-rGO(AA) nanocomposite electrode for 5000 cycles at 3 A/g.

Full size image
Table 4 Specific capacitance of g-C3N4, GO, g-C3N4@GO, g-C3N4@r-GO (AA) and g-C3N4@r-GO (NaBH4) by varying current density.
Full size table

The retention of capacitance (%) as a function of cycle counts for the g-C3N4-rGO(AA) nanocomposite (Fig. 7d) exhibited excellent cyclic stability i.e. capacitive retention of 116.12% after 1000 cycles, 117.45% after 2000 cycles and 78.56% after 5000 cycles at a current density of 3 A/g. Also indicating activation behavior during the initial cycles measured with a minor drop, indicating that there is no major decline in GCD rather increase was observed till 2000 cycles. Even after 5000 cycles, the electrode retained 78.56% of its initial capacitance, demonstrating good long-term stability for these composite electrodes. The high retention of capacitance indicates about good GCD behavior and excellent mechanical stability of the composite electrode during continues charge/discharge cycles.

EIS analyses

Electrochemical impedance spectra of g-C3N4, GO, g-C3N4@GO, g-C3N4@r-GO (AA) and g-C3N4@r-GO (NaBH4) used to investigate the charge transfer and recombination processes at their solid/electrolyte interfaces of the photocatalyst47. The EIS parameters are determined by the well-fitted Nyquist plots of studied system (Fig. 8; Table 5). EIS spectra were appeared in arc-shape and smallest arc-radius found for g-C3N4@r-GO (AA) due to the lower electron transfer resistance, indicating the faster separation efficiency at the electrode interface48. The equivalent circuit (inset of Fig. 8) of the devices was constructed to analyze the impedance spectra, and the impedance spectra were fitted by the ZSimp Win software. The R1 is the series resistance of system. The first semi-circle (in high frequency region) can be assigned to the charge-transfer resistance (R2) of the Pt counter electrode/electrolyte interface. The second semi-circle (middle frequency) can be assigned to the charge-transfer resistance (R2) of as-prepared sample’s anode/electrolyte interface49. In Nyquist diagram, a smaller radius is an indication of an overall smaller charge transfer resistance or, equivalently, a more facile charge transfer process at the electrode/electrolyte interface.

Table 5 EIS parameters observed from the fitting curves and circuit diagrams of g-C3N4, GO, g-C3N4@GO, g-C3N4@r-GO (AA) and g-C3N4@r-GO (NaBH4).
Full size table

The electron lifespan is lowered that resulted in longer electron recombination times or shorter electron travel time via the anode’s pores would result in higher efficiency.

Fig. 8
figure 8

Nyquist plots of electrochemical impedance spectra and fitting curves along with circuit diagram. (a) Imaginary Z” range − 2.5 to 20 KOhm.cm2 (b-f) EIS fitting curves along with circuit diagram that used for determining the values of resistance, capacitance, chi square value, iteration and work load of of g-C3N4, GO, g-C3N4@GO, g-C3N4@r-GO (AA) and g-C3N4@r-GO (NaBH4). Red dots are experimental values and green triangles are fitted values.

Full size image

Hydrogen generation form water splitting

Comparative study of hydrogen generation capacity of the studied systems was measured in 120 mL of 20% methanolic solution of 0.3 g sample in Pyrex glass reactor under irradiation of 300 W Xe light, as shown in Fig. 9a-c. The H2 production through water splitting are reported for the g-C3N4, GO, g-C3N4@GO, g-C3N4@r-GO(AA), and g-C3N4@r-GO(NaBH4) nanocomposites. It was found that the composite g-C3N4@rGO (AA) exhibited the maximum efficiency for the H2 generation (339.82 µmol h− 1g− 1 with quantum yield 2.52% at 420 nm) with respect to their counterparts (GO, g-C3N4, g-C3N4@GO and g-C3N4@r-GO(NaBH4) (Fig. 9c).

Fig. 9
figure 9

H2 production performance and mechanism. Comparative hydrogen generation capacity (a) with time, (b) stability test after I, II and III usages for 3 h, (c) H2 production rate of g-C3N4, GO, g-C3N4@GO, g-C3N4@r-GO(AA), and g-C3N4@r-GO(NaBH4) nanocomposites and (d)electron transfer mechanism of g-C3N4@r-GO(AA).

Full size image

The sustainability of the g-C3N4@r-GO(AA) sample was checked in three cycles (Fig. 9c) and it was found that the corresponding 95.22%, 91.19% and 88.19% H2 generation capacity of the system was retained. When we compared our results with the reported systems5,6,7,8,9,10,11,12. Our studied system exhibits good results even without containing any metallic content under the low electrolyte concentration and less amount of cocatalyst (1.5%Pt). The proposed charge transfer mechanism for photocatalytic water splitting is illustrated in Fig. 9d. Here, g-C3N4 might act as n-type material and GO/r-GO act as a p-type material and a p-n junction formed at the boundary of the g-C3N4 and GO or r-GO component. Under exposure of light tends to jump the electron from valance band (VB) to conductance band (CB) of composite. The as-approached CB electrons at the GO site, and directed to be moved towards CB of g-C3N4, where it reduced water into hydrogen gas. And the hole resides at the VB sites of g-C3N4 and GO oxidized the CH3OH into diverse products (HCOOH, HCHO, CO2, etc.).

When g-C3N4 and GO/r-GO are combined to generate a g-C3N4@GO or g-C3N4@r-GO composite then the photocatalytic (339.82 µmolh− 1 g− 1 that retained 89.80% after third cycle) as well as super capacitance (retained 78.56% after 5000 cycles) powers of composite are significantly enhanced in compared to the pure g-C3N4 or GO alone. The higher visible light absorption and interfacial interaction, low rate of charge recombination, minimum defect, and effective charge transfer from g-C3N4 (n-type material) to GO (p-type material) are responsible for the improved performance of the composite50. Interfacial interaction always related to the synthetic approach, such a role reducing agent (AA or NaHBH4) in reduction of GO is crucial in deciding the electron acceptor/transporter role of GO through the rational design of composite. In conclusion, CN-GO composites are useful materials for photocatalytic H2 production because they take use of the distinct structural and electrical characteristics of both constituents to overcome their individual drawbacks.

System exhibits high thermal and chemical stability and non-toxicity such that it has been considered as the most promising photocatalyst for environmental improvement and energy conservation. Hence, it is of great importance to obtain high-quality g-C3N4 and gain a clear understanding of its optical properties.

Discussion

To overcome the limitations of the n-type material g-C3N4, we empowered it either by coupling with p-type GO or r-GO (reduced by ascorbic acid and NaBH4). It was observed that the composite g-C3N4@r-GO(AA), and g-C3N4@r-GO(NaBH4) are more efficient for H2 production than their counterparts. Further among the composites, g-C3N4@r-GO(AA) is the most efficient because ascorbic acid is a better reducing agent then NaBH4 because NaBH4 is selective in reduction, which is also proven by the XRD, FTIR, CV and GCD analyses. The cavities/pits observed in the morphology of g-C3N4@GO, might be act as a recombination centres, which can reduce the activity of the composite, where the rod like artifacts attributed more efficiency in samples. Where, no cavities are observed in the surface morphology of either GO or g-C3N4, g-C3N4@r-GO(AA), and g-C3N4@r-GO(NaBH4). Chemical / thermal stability and biocompatibility of g-C3N4 @ r-GO (AA) and g-C3N4 @ r-GO (NaBH4) as well as H2 production can be reused due to its high sustainability and recyclability confirm its industrial usages. Regarding material toxicity, more research on this required.

Positive surface potential (supported by) of GO and negative surface potential of g-C3N4 and the surface potential of the rest samples straddle between both limits. That proves the formation of p-n junction in composites. The nanocomposite g-C3N4@r-GO (AA) show the maximum rate of hydrogen generation i.e. 339.82 µmolh− 1g− 1 for photocatalytic water splitting with 2.52% apparent quantum efficiency at 420 nm. which is 6.2, 3.5, 2.9, and 1.5 times higher than their counterparts viz. g-C3N4, GO, g-C3N4@GO, and g-C3N4@GO(NaBH4), respectively. Similarly, the composite g-C3N4@r-GO (AA) demonstrated the highest specific capacitance of 322.77 Fg− 1 (scan rate: 5 mVs− 1) in aqueous electrolyte in basic medium (2 M KOH). The effective interface with layered structure, p-n junction, minimum band gap, and minimum charge-recombination rate in nanocomposite are significantly supported the reasonably high H2 generation rate and good capacitive behaviour. The study unfolds the way to improve the activity of the material for industrial scale hydrogen production from composite and capacitive investigations.

Methods

Experimental material

The chemicals such as: graphite powder (Alfa Aesar, 200mesh, purity 99.9995%), chloroplatinic acid (H2PtCl6,Sigma Aldrich), graphite (Sigma-Aldrich), NaOH (Sigma Aldrich), H2SO4(Merck), NaNO3(Sigma-Aldrich), KMnO4 (Sigma-Aldrich), H2O2 (Sigma-Aldrich), NH2CH2COOH (Sigma-Aldrich), FTO glass (Sigma-Aldrich), and solvents are applied as-received in synthesis, and water splitting applications.

Synthesis of synthesis of g-C3N4

Graphitic carbon nitride was synthesized by 10 g of urea is mixed with 10 g of thiourea and crystallized in 60 mL of ethanol then the mixture of urea and thiourea were kept in tubular furnace for 2 h at 400 °C and as-obtained product was further washed with DIW and dried in an oven at 60 °C for overnight.

Synthesis of graphene oxide (GO)

Graphene oxide (GO) was fabricated by acclimatizing the modified Hummer’s method21. In which, the graphite powder (2 g) and NaNO3(1 g) were added to 36 N H2SO4 (46mL) in a 500mL beaker and stirred for15min in an ice-bath, followed by the slowly addition of KMnO4 (6 g) to the above mixture. The resulted solution was stirred and mixed well with distilled water (92mL) for 2 h at RT and allowed to age for 30 min, followed by stirring for 1 h after addition of distill water (280mL) and 30%H2O2 (20mL). Then the suspension was washed several times with water till the pH = 7 attained. The GO exfoliation was made by mild sonication of solution for 30 min at RT. As-synthesized dark brown powder GO was dried in an air-oven at 60 °C for 12 h.

Synthesis of g-C3N4-GO nanocomposite

As-synthesized 0.5 g g-carbon nitride was added to 50 mL of DIW in one beaker and ultrasonicated for 15 min and in second beaker 0.5 g GO in 50 ml of DIW was ultrasonicated for 15 min and thereafter the suspensions of g-C3N4 and GO were mixed and ultrasonicated for 1 h then stirred for 2 h then resultant product was washed several times with DIW and ethanol and dried in an oven for 24 h at 60 °C. As-synthesized g-C3N4-GO nanocomposite was used for water splitting after characterization.

Synthesis of r-GO(AA)

Reduced graphene oxide with ascorbic acid was synthesized by using 2 g of graphene oxide suspension in 50 mL of water and 2 g of ascorbic acid solution in 50 mL of water was sonicated separately for 1 h and after mixing this solution again sonicated for one hour and vigorously stirred for 2 h and filtered and several times washed with DIW and ethanol and dried in an oven at 60 °C for 12 h.

Synthesis of g-C3N4@r-GO (AA) nanocomposite

As-synthesized 0.5 g Carbon nitride was added to 50 mL of DIW in one beaker and ultrasonicated for 15 min and in second beaker as synthesized 0.5 g RGO (AA) in 50 mL of DIW was ultrasonicated for 15 min. then both suspensions of g-C3N4 and RGO (AA) were mixed and ultrasonicated for 1 h then stirred for 2 h then resultant product was washed several times with DIW and ethanol and dried in an oven for 24 h at 60 °C.

Synthesis of r-GO(NaBH4)

Graphene oxide was reduced with reducing agent NaBH4.Where, 2 g of graphene oxide suspended in 50 mL of water and then 2 g of sodium borohydride solution in 50 mL of water was sonicated separately for 1 h and after mixing the solution, it is again sonicated for 1 h and vigorously stirred for 2 h then filtered and washed several times with DIW and ethanol. Thereafter, the sample was dried overnight in an oven at 60 °C.

Synthesis of g-C3N4@r-GO (NaBH4) nanocomposite

As-synthesized 0.5 g g-C3N4 was added to 50 mL of DIW in one beaker and ultrasonicated for 15 min and in second beaker as synthesized 0.5 g r-GO(NaBH4) in 50 mL of DIW was ultrasonicated for 15 min then both suspensions of g-C3N4 and r-GO(NaBH4) were mixed and ultrasonicated for 1 h then stirred for 2 h then resultant product was washed several times with DIW and ethanol separately, then dried in an oven for 24 h at 60 °C.

As-synthesized GO, g-C3N4, g-C3N4@GO, g-C3N4@r-GO(AA), and g-C3N4@r-GO(NaBH4) nanocomposite were used in water splitting after characterization.

Characterization

Energy dispersive X-ray (EDX) spectroscopic analyses (Carl Zeiss, Germany, Model: Supra 55 with GEMINI Technology) supported by FE-SEM and FE-SEM technique was used to observe the morphology and elemental composition of the systems under study in a voltage range of 0.02-30 kV at Central University of Gujarat (India). Powder X-ray diffraction (XRD) was used to analyze the phase and crystallinity of the GO, g-C3N4, g-C3N4@GO, g-C3N4@r-GO(AA), and g-C3N4@r-GO(NaBH4) systems. The diffractometer (Rigaku Ultima IV, Japan, floor model with CuKα radiation (λ = 0.15418 nm)) with the fast detector (one-dimensional LYNXEYE 0.002) is situated at the Department of Pure and Applied Chemistry University of Kota, Kota (India) and had an accuracy of ± 0.02° angle over the whole measuring range of 2θ = 0°–80° at 40 mA current and 40 mV voltage. UV-Vis-NIR spectrophotometer (Shimadzu, Model: UV-3600 plus) equipped with double grating monochromators of the University of Kota, Kota (India) was used to record UV-Vis absorbance spectra in the wavelength range 200–800 nm. To measure the FTIR spectra of the samples, an IR spectroscope (ALPHA-II, platinum-ATR, BUKER) situated at Sambalpur University in Odisha was used. Using a spectrofluorometer (FluoroMax-4 model, HORIBA Scientific) installed at Banasthali Newai, Tonk (India), the steady-state photoluminescence emission (PLE) spectra were captured at an excitation energy of 227 nm. The electrochemical impedance spectroscopy (EIS) study of the studied systems was performed in the frequency range of 1,000,000 Hz to 0.01 Hz by applying 0.01 V voltage with in ± 10 mV perturbation by using the electrochemical work station (CH Instrument CHI 7087E electrochemical Analyzer) at MNIT, Jaipur (India) using a conventional three electrodes cell. Where, the GO, and nanocomposites were casted on FTO glass was used as working electrodes, graphite as counter electrode, and Ag/AgCl electrode were used as reference electrode that dipped in an 20% aqueous CH3OH. The SPV instrument (make: KP Technology, Scotland; model: APS04-N2-RH) of NRF centre of Indian Institute of Technology, New Delhi was used to measure surface photovoltaic spectroscopy (SPS) at room temperature while exposed to visible light. A homemade vacuum chamber, equipped with a vibrating gold Kelvin probe (Delta PHI Besocke) used for the analyses. To prepare the SPV samples, an aqueous dispersion of 0.5 mg catalyst was drop-cast onto 1 × 1 cm2 glass substrates that were air-dried and annealed for a full night at 70 °C. An Oriel Cornerstone 130 monochromator was used to filter the 175 W Xe lamp’s monochromatic light, which had an intensity range of 0.1 to 0.3 mW cm2.

Specific capacitance measurement

The working electrodes were prepared by combining the as-synthesized composite with carbon black, and polyvinylidene fluoride (PVDF) (Mol. wt.–534,000 by GPC) (Sigma Aldrich) in 8:1:1 ratio. Then the slurry was prepared by mixing the above mixture with a polar solvent N-methyl-2-pyrrolidone (NMP). A drop-cast method was used to pour the slurry onto 1.0 cm x 0.5 cm nickel foam. Then these electrodes were dried in a hot air oven set at 80 °C for 12 h. The electrochemical workstation (Make: Biologic and Model VSP-300) of UPES, Dehradun was used to conduct the cyclic voltammetry (CV), galvanostatic charge discharge (GCD), and electrochemical impedance spectroscopy (EIS) studies by using three-electrode systems (as prepared electrode as-working electrode, Pt as-counter an electrode and Ag/AgCl as-reference electrode). EIS fitting was performed by using software (EIS Spectrum Analyser). A solution of 1.0 M KOH was used as an electrolyte. In the CV experiments, the applied potential range was + 0.2 V vs. Ag/AgCl to + 0.6 V vs. Ag/AgCl, while in the GCD analysis, the voltage window was + 0.2 V vs. Ag/AgCl to + 0.6 V vs. Ag/AgCl. The following relation (Eq. 3) is used to calculate the specific capacitance of the prepared samples51.

$${C_{sp}}=~\frac{1}{{2mVk}}\mathop \smallint \limits_{{{v^ - }}}^{{{v^+}}} I\left( v \right)dV$$
(3)

In Eqs. (3, 4), m(g) represents the active material’s mass, \(\:\varDelta\:\)V is the applied voltage, k(V/s) the scan rate, C(F/g) the specific capacitance, and I(A) the discharge current. The CSP has been calculated using the following relation (4–6)51.

$${C_{sp}}=~\frac{{I\Delta t}}{{m\Delta v}}$$
(4)
$${\text{Energy Density}}\left( {\text{E}} \right)=\frac{{{C_{sp}}\Delta {V^2}}}{2}$$
(5)
$$Power Density=\frac{{3600x E}}{{\Delta t}}$$
(6)

Where, I referred to the applied current to discharge, Δt is time for discharge, m, and ΔV are as above.

Photocatalytic hydrogen production

In this investigation 0.30 g of powder photocatalyst was precisely dispersed in 120 mL of an aqueous methanol (20% CH3OH at pH = 7.0). Within a 150 mL reaction cell made of Pyrex glass, the reaction took place. The water jacket around the reaction cell help to keep the reaction temperature constant at 25 °C. The reaction vessel was hermetically sealed using a plastic wire lock and a rubber septum to guarantee the isolated cell environment. A 300 W xenon lamp (Xe lamp-HX1, ISS-Model PE300UV) was used to analyze the nanocomposite at 420 nm for photocatalytic hydrogen production. An intensity of 3 × 1021 photons per hour was attained by carefully adjusting the irradiation conditions. Over the course of 3.0 h time, the gaseous products that released during the photocatalytic process were observed in every 30 min intervals. A sophisticated gas chromatograph (manufactured by Shimadzu) was used to analyze the gaseous products using gas chromatograph well equipped with Thermal Conductivity Detector (TCD) and 5 A molecular sieve columns. A precise and contextually relevant assessment of the apparent quantum yields of hydrogen production in the studied systems was calculated AQE by utilizing following Eqs. (7) and (8)26.

$$AQY=\frac{{\left[ {{\text{Number of reacted electrons or holes}}} \right]{\text{}}}}{{{\text{}}\left[ {{\text{Number of incident photons}}} \right]{\text{}}}}$$
(7)
$$AQY=~\frac{{\left[ {{\text{Number of hydrogen molecules evolved}} \times {\text{}}2{\text{}}} \right]{\text{*}}100{\text{}}}}{{{\text{}}\left[ {{\text{Number of incident photons}}} \right]{\text{}}}}$$
(8)

Where, energy E = nhν, n = number of photons, h = Planck’s constant and ν= f requency. The number of photons per second can be measured by dividing the energy with Planck’s constant.