Introduction

Although the Earth’s surface is covered with a huge amount of water (~ 71%), only about 0.3% is fresh and accessible to human beings. The rest, 99.7% fresh water, is trapped in the ice caps, soil, and in the atmosphere [1], and even the 0.3% accessible water is unattainable by humans [2]. Available fresh water for human consumption comes from lakes, ponds, reservoirs, streams, and rivers, hence maintaining these water bodies for use is a cause for concern for humanity. Dyes are the most important raw materials used in some of the industries which include paint, plastic, paper, pharmaceutical, textile, cosmetics, and food [3]. Wastewater from these industries usually contains dye residues which are not readily degradable. Wastewater containing these dye residues creates negative effects on the environment especially when they enter into the water bodies. Firstly, due to their color, they hinder the penetration of light into water bodies and therefore affecting the photosynthesis of aquatic systems [4]. Apart from the effect of their color, the effect of dyes on human health is also detrimental. The end products of dyes affect the formation of hemoglobin, skin, and lung [5]. Also, some of these dyes are carcinogenic and may cause dermatitis, allergies, and skin irritation [6]. Hence, it is of utmost importance to properly treat dye-contaminated wastewater before discharging into the environment.

Several conventional techniques have been used in remediation of wastewater problems caused by dye residues. The well-known techniques to flush out this issue include adsorption, reverse osmosis, electrodialysis, coagulation/flocculation, oxidation, irradiation, and aerobic and anaerobic microbial methods of degradation [7]. But, these conventional methods have shown some drawbacks such as incomplete treatment leading to the production of secondary pollutants (e.g., production of trihalomethanes which are carcinogenic, by the reaction of chlorine gas with some of the organic substances found in natural water during the chlorination process) [8], waste management arising from the non-biodegradability of most of the dyes, and high cost of operation [9].

Treatment of wastewater from dye-utilizing industries does not only rely on removal of the color but also on mineralization and degradation of the dye compounds. Advanced oxidation methods have proven to be more efficient because of their pollutant degradability capabilities [10]. The use of semiconductors as photocatalysts [11] is common and well known in advanced oxidation approaches for dye removal. ZnO semiconductors have shown to be effective when compared to other semiconductors used as photocatalysts and have been studied extensively due to its interesting features like wide band gap, weak toxicity, high electron mobility, and strong oxidizing power [12]. However, using ZnO as a photocatalyst exhibits some drawbacks which include its wide band gap (~ 3.3 eV) (this restricts its application in the UV range that is only ~ 5% of visible radiation), the low surface area of the bulk form, and high rate at which the photogenerated electron and hole recombine. In order to overcome these challenges and drawbacks, a good number of modifications have been made on ZnO which extends its light absorption to the visible light region and thus enhances its photocatalytic capability. These modifications include the preparation of the nanoparticles and nanocomposites with organic materials, mono- and co-doping methods, and coupling with other semiconductors [13,14,15]. Non-metal doping does not lead to the formation of recombination centers. Unlike metal doping, it narrows the band gap by creating a new valence band through raising of the upper valence band. Hence, non-metal doping is preferred in introducing impurities to semiconductors which enhance their photocatalytic activity in the visible region [16]. Among the non-metals, nitrogen is the most promising candidate for doping due to its similarity with oxygen with regard to ionic radii and electronegativity [17].

Graphene, an allotrope of carbon with intriguing properties such as large specific surface area, unique electronic properties, excellent mechanical strength, and high transparency, has extraordinary applications in many areas including hydrogen production, capacitor, energy storage devices, nanoelectronics, sensors, and photocatalysis [18]. The oxidized form of graphene, graphene oxide (GO), can undergo exfoliation easily and forms a stable dispersion because of the presence of oxygen functionalities (located on their basal planes and epoxide and hydroxyl groups and also on their sheet edge, carboxyl and carbonyl functional groups) [19]. Large surface-to-volume ratio and two-dimensional structures of GO sheets may act as a good support material where functional nanomaterials can be loaded while the oxygen functionalities may act as a point of nucleation to hold the active materials. Recently, nanocomposites comprising of metal oxide/metal/GO have been reported and their respective potentiality as photocatalysts, biosensors, and capacitors [20,21,22] identified. Additionally, GO plays a very crucial part in influencing the performance of semiconductors in GO/semiconductors composite as a photocatalyst because of its electron transport abilities, adsorption capability, and large surface area [23]. Hence, we envisaged that doping ZnO with nitrogen would reduce the band gap of the semiconductor with minimum rearrangement in the crystal structure [24] while the presence of GO in the system would increase the adsorption rate of the semiconductor and also increase electron mobility at the interface. The synergetic effects of N and GO on ZnO are anticipated to improve the photocatalytic ability of ZnO by extending its light absorption to the visible region while inhibiting the rate at which electron and hole recombine at the semiconductor surface.

With respect to this study, we doped nitrogen into ZnO (N-ZnO) and immobilized the N-ZnO onto the GO surface via simple co-precipitation and wet chemical methods. The nanocomposites synthesized were further characterized in order to analyze the effect the modifications of ZnO nanoparticle with nitrogen and GO will have on its photocatalytic capability. The photocatalytic ability of the photocatalyst was graded through the photodegradation of brilliant green dye in water.

Experimental procedures

Materials

The reagents and chemicals used (zinc acetate dihydrate; Zn(O2CCH3)2(H2O)2, urea CO(NH2)2, sodium hydroxide (NaOH), brilliant green dye, graphite flakes, and potassium permanganate (KMnO4)) were of analytical standard and procured from Sigma-Aldrich, Germany, and used as it purchased.

Synthesis of ZnO nanoparticle

To synthesize ZnO nanoparticle, 10 g of zinc precursor (Zn(O2CCH3)2.(H2O)2) was dissolved in deionized (DI) water (250 mL) under continuous magnetic stirring at 25 °C. Thereafter, 3 M NaOH under continuous and vigorous stirring was added slowly. The addition of NaOH stopped when the resultant reaction mixture gets to a pH of 12; the reaction continued for about 30 min with stirring to obtain a homogeneous mixture. The precipitate formed was washed using DI H2O and C2H5OH simultaneously, dried for 12 h at a temperature of 50 °C, and calcined for 4 h at a temperature of 400 °C. The synthesized nanoparticle was stored as ZnO for further use.

Synthesis of N-doped-ZnO nanoparticles

Solvent-free reaction method was used in synthesizing N-ZnO nanoparticles (NPs) [25]. Un-calcined ZnO and urea were mixed and ground using agate mortar and pestle in a weight ratio of 1:2. Thereafter, the mixture was calcined for 4 h at a temperature of 400 °C. Thereafter, the mixture was taken out and allowed to stand for some time in other to cool, followed by grinding to get fine powder and stored for further use.

Graphene oxide synthesis

Graphene oxide synthesis was carried out using graphite flakes as a starting material by the modified “Hummers” method [26]. In a mixture of concentrated H3PO4 (40 mL) and concentrated H2SO4 (360 mL), a mixture of KMnO4 (18 g) and graphite powder (3 g) was slowly added with continuous stirring. The resulting solution was allowed to heat at 50 °C with continuous and vigorous stirring for 12 h. Thereafter, the temperature of the reaction mixture was brought to 25 °C, and 30% H2O2 (about 5 mL) was added in drops with continued stirring. The resulting solution mixture was centrifuged for 10 min at 7500 rpm and washed using deionized water, 30% HCl solution, and ethanol in succession followed by filtration using a Teflon membrane filter (1-μ pore size). The filtrate was finally coagulated with 200 mL of ether, filtered, and dried at 50 °C for overnight.

Synthesis of nitrogen-doped ZnO-immobilized GO composite

The N-ZnO-GO composite with 0.5 wt% of GO was prepared by using a wet chemical method similar to [27] starting with the dissolution of 0.04 g of GO in 60 mL of DI water and ultrasonicated for 30 min for exfoliation of GO. Thereafter, N-ZnO (0.4 g) was slowly added and the resulting solution was heated up to 3 h at a temperature of 70 °C under magnetic stirring. The reaction was cooled and centrifuged for 10 min at 7500 rpm followed by washing with DI water; a gray precipitate was obtained which was dried at 50 °C overnight and stored in desiccators for further use, labeled as NZGO-2. Other composites NZGO-1 representing 0.1 wt% of GO and NZGO-3 representing 1 wt% of GO were prepared using the same method by varying the GO ratio. A full schematic is provided in a three-stage preparation in Scheme 1.

Scheme 1
scheme 1

Schematic diagram illustrating the three stages involved in the synthesis of N-ZnO-GO nanocomposite (GO = graphene oxide; N = nitrogen; ZnO = zinc oxide; DI = deionized water)

Full size image

Evaluation of photocatalytic activity

The photocatalytic performance of the nanocomposites synthesized was evaluated by making use of the photodegradation of BG dye under simulated visible light irradiation using 300-W Xe lamp (100 mW/cm2) with a 420-nm dichroic UV filter at 25 °C. Fifty milligrams of the photocatalyst in 100 mL of BG (20 ppm) was used for all the photodegradation experiments and the distance between the experimental setup and the light source was maintained at 10 cm in order to produce a light irradiation which is equivalent to 1 sun. Before exposure to the light, the dye and the photocatalyst adsorption-desorption equilibrium was obtained by stirring the mixture in the absence of light for 30 min. Thereafter, the mixture was exposed to light with continuous magnetic stirring and 3 mL of mixture withdrawn after 15-min intervals for a 90-min duration to monitor the concentration changes after light irradiation. The changes in the BG dye concentration in the solution before and after irradiation were determined using a UV–vis spectrophotometer (Shimadzu UV2450) at λ = 623 nm. The chemical oxygen demand (COD) of the dye solution before and after degradation was analyzed using the Spectroquant Pharo 300 and the COD percentage was calculated as per Eq. 1:

$$ COD\%=\frac{COD_i-{COD}_f}{COD_i}\times 100\kern0.5em $$
(1)

where CODi is the COD before (initial) degradation and CODf is the COD after (final) degradation.

Characterization

The surface (external and internal) morphologies of the photocatalysts synthesized were analyzed on scanning electron microscopy (SEM) on a TESCAN (Vega 3 XMU) and TEM on a JEM-2100F, respectively. The elemental composition determination was done on energy-dispersive X-ray spectroscopy (EDX) which was coupled to the SEM. The crystallinity of the synthesized photocatalysts was analyzed by XRD at 40 kV, 30 mA with Cu Kα radiation (λ = 0.15406) and K-beta filter. The presence of the functional groups in the synthesized photocatalyst was confirmed by using FTIR spectrometer (PerkinElmer FTIR spectrum-100). The photoluminescence (PL) spectra of the photocatalysts were generated using the PerkinElmer LS45 fluorescence spectrometer at room temperature with a Xe lamp and at 325-nm excitation wavelength while the optical properties of the photocatalysts were analyzed by using the UV–vis DRS (Shimadzu UV–2450) with barium sulfate as the reference.

Results and discussions

SEM, TEM, and EDS

The SEM image of N-ZnO is depicted in Fig. 1a while Fig. 1b depicts the SEM image of N-ZnO-GO. N-ZnO nanoparticles which have a relatively uniform size are densely packed on the surface of GO nanosheets illustrating the homogenous mixing of N-ZnO with the GO sheet. Figure 1c, d depicts the TEM image of sheet-like GO nanosheets and N-ZnO-GO nanocomposite, respectively. It is obvious from Fig. 1d that N-ZnO nanoparticles which have particle sizes between 15 and 35 nm are successfully immobilized over the surface of GO nanosheets. The coupling of N-ZnO onto the surface of GO nanosheets may be attributed to the presence of epoxide together with hydroxide groups on the GO sheets that act as sites of attachment for N-ZnO nanoparticles [28]. The elements present in the composite were analyzed using SEM-coupled EDS. Figure 1e showed the EDS spectrum of N-ZnO-GO materials to contain carbon, nitrogen, oxygen, and zinc element. The obtained results served as proof that the synthesized photocatalyst was uncontaminated with impurities during the synthesizing process.

Fig. 1
figure 1

SEM images for a N-ZnO and b NZGO-1, and TEM images of c GO and d NZGO-1 and e EDS spectrum of NZGO-1

Full size image

FTIR, XRD, and BET analyses

The FTIR spectra of graphene oxide, NZGO-1, N-ZnO, and zinc oxide are illustrated in Fig. 2a. The characteristic bands of GO were identified at 3421, 1612, 1731, 1047, 1417, 1221, and 857 cm−1, corresponding to OH stretchings, aromatics C=C stretchings, carbonyl CO stretching, alkoxyl C–O stretching, carboxyl OH stretchings, and symmetrical and asymmetrical ether group stretching, respectively [29].

Fig. 2
figure 2

a FTIR of GO, ZnO, N-ZnO, and NZGO-1 and b XRD pattern of the as-prepared photocatalysts

Full size image

ZnO presence in the composite is confirmed by the typical vibration of Zn–O band elongation at 490 cm1. Additionally, the peaks observed around 1654 and 1388 cm−1 in the nanocomposite are linked to the stretching vibrations of carboxylic of acetate groups (asymmetrical and symmetrical) attached to the ZnO [30] while the peak observed around 2050 cm−1 is related to the vibration of –OH groups [31]. Also, a blue shift in the wavelength around the absorption peak 1657 cm−1 of the nanocomposite (NZGO-1) was observed compared to N-ZnO and ZnO and may be ascribed to complete exfoliation of GO nanosheets by N-ZnO. Hence, N-ZnO is successfully immobilized over the GO nanosheets.

XRD analysis gives an insight on the arrangement of atoms in its crystal structure. The XRD patterns of the photocatalysts prepared (i.e., ZnO, N-ZnO, NZGO-1, NZGO-2, and NZGO-3) are illustrated in Fig. 2b. Observation from the figure shows that the diffraction peak patterns of N-ZnO, NZGO-1, NZGO-2, and NZGO-3 all match the hexagonal wurtzite structure of ZnO as depicted in ICDD: 04-008-8198 [32]. The similarity in the diffraction peaks of these as-prepared photocatalysts indicates that the modifications made on ZnO did not alter its structural pattern and orientation. Also, the absence of the typical GO diffraction peaks in the composite can be linked to its low diffraction intensity and its small amount which is an indication of complete exfoliation and dispersion of GO nanosheets in the nanocomposite [33].

The particle sizes of the nanocomposites were computed based on the diffraction peaks at 101 orientation plane using Debye–Scherrer’s formula. While the lattice constants a (Eq. 2) and c (Eq. 3) were calculated based on 100 and 002 orientation planes, respectively, using modified Bragg’s law [34], lattice constants

(2)

and

(3)

The unit cell volume of the as-prepared photocatalyst was calculated using the following formula [35].

$$ \sqrt{3}\times {a}^2\times \frac{c}{2} $$
(4)

where a and c stand for the lattice constants of the cell.

As observed in Table 1, the crystalline size, the cell volume, and the lattice constants of ZnO decrease after modifications with N, presumed to link to the fact that N occupies interstitial positions in the ZnO crystal lattice.

Table 1 Lattice parameters, cell volume, and crystallites of the as-prepared photocatalysts (standard deviation in parentheses)
Full size table

The introduction of GO in the nanocomposite increased the specific surface area (SBET) of ZnO from 4.4561 m2 g−1 in pristine ZnO to 22.123 m2 g−1 in 1%wt GO (N-ZnO-GO) as obtained from nitrogen adsorption/desorption process. The result is recorded in Table 1 and the observed changes in the surface area are linked to a higher surface area provided by graphene sheets (about 2630 m2 g−1). The largest increase in surface area was from the urea reaction to produce the N-ZnO (4.4 to 12.7 m2 g−1). The subsequent increases from adding GO are more incremental. The higher surface area is anticipated to facilitate adsorption of the dye molecules from the dye solution and thus an enhanced photodegradation. Standard deviation of the surface area results was also calculated (Table 1). It can be found that the surface area results are consistent, varying only by a maximum of 4% thus making this result reliable.

UV–vis analysis

The higher the optical absorbance of a photocatalyst, in the visible region which spreads across over about 43% of the visible spectrum, the more effective it is for photocatalysis. Figure 3a depicts the UV–visible DRS of ZnO, N-ZnO, NZGO-1, NZGO-2, and NZGO-3, as measured using the UV–vis spectrophotometer set to a wavelength range between 200 and 800 nm. Figure 3a shows sharp absorption peak for ZnO around 388 nm which is a sign impurities suppressed ZnO nanoparticle and of good crystalline. As can be observed from the UV results, N-ZnO shows a slight red shift in wavelength compared to the pristine ZnO which can be attributed to the modifications made on the electronic levels of ZnO by incorporating nitrogen into its crystal structure. This observation may be related to the fact that there are more 2p π-electrons of nitrogen compared to the 2p π-electrons of those of the replaced oxygen in the lattice structure which decreases the band gap by creating another energy state between the VB of ZnO and its CB [34]. In comparison with the absorbance of ZnO and N-ZnO, the absorption of NZGO-1, NZGO-2, and NZGO-3 composites is more intensive in the whole visible light spectrum. A similar observation has also been documented in literature [36]. Again, Li et al. [37] noted a similar phenomenon where the modification of ZnO with GO improved the absorption of ZnO in the whole visible region. The increment in absorbance over the visible range may be credited to graphene oxide introduction in the N-ZnO-GO composite which not only increases the surface electrical charge of the oxides in the composite but also suppresses electron-hole formation process during irradiation [38]. Also, Fig. 3b which is a plot of Kubelka–Munk [39] function of energy against photon energy shows lowering of the band gap of the synthesized photocatalysts from 3.30 to 2.83 eV, affirming the reason for the increased absorbance in the visible region. Similar band gaps for ZnO (3.20 eV) and N-ZnO (2.98 eV) have been reported elsewhere in literature of [40, 41] and [25, 42], respectively. Hence, this increase in the absorbance of light intensity and range has the ability to increase the performance of the modified zinc oxide (with nitrogen and GO) compared to unmodified ZnO photocatalytically.

Fig. 3
figure 3

a UV–vis DRS of the photocatalysts synthesized. b Tauc plots for the as-prepared photocatalysts

Full size image

Photoluminescence spectra

One of the essential factors affecting the photocatalytic ability of a semiconductor used as a photocatalyst is the rate at which photogenerated electron and hole recombine. The higher the recombination, the lower the photocatalytic activity of the respective semiconductor, due to less availability of carriers for generation of radicals which react and degrade organic dyes. Photoluminescence (PL) spectrum was used in this study to analyze the effect of nitrogen and GO in the electron-hole recombination rate of ZnO. Figure 4 depicts the PL spectra of ZnO, N-ZnO, NZGO-1, NZGO-2, and NZGO-3 measured using a set excitation wavelength of 325 nm. While a higher PL emission intensity suggests a faster electron-hole recombination rate, a lower PL emission intensity, on the other hand, indicates a slower electron-hole recombination rate. The peaks observed at 360 nm were linked to near band edge emission of ZnO; at 395 nm, it was linked to near UV emission which arises from the photogenerated electron-hole recombination, while the peak at 420 nm was linked to blue emission arising from the Zn interstitial donor level to Zn vacancy acceptor energy level; the peak at 482 nm is linked to blue-green emission and may be ascribed to the defects created by oxygen vacancies. In comparison, the PL spectrum intensities of ZnO and N-ZnO against the PL spectrum of the composite with 0.1%wt GO show the latter with the lowest intensity which is an indication of electron-hole recombination suppression and may be attributed to GO introduction in the composite. The lower intensity observed at the PL emission spectrum of the nanocomposite indicates an enhancement in the transfer charges taking place at the interface and thus suppressing the electron-hole recombination rate taking place at the defect sites, implying that the generation of photo-induced electrons in the CB of N-ZnO nanoparticles is conveyed rapidly to GO sheet. From PL spectrum results, NZGO-1 is expected to have a higher photocatalytic activity as compared to the respective ZnO and N-ZnO photocatalysts.

Fig. 4
figure 4

Photoluminescence spectra of the as-prepared photocatalysts

Full size image

Photocatalytic degradation and mineralization of BG

The photocatalytic ability of the photocatalysts synthesized ZnO, N-ZnO, NZGO-1, NZGO-2, and NZGO-3; nanocomposites were investigated through the photodegradation of BG dye under simulated visible light after 90 min of irradiation. Figure 5a is a representation of the UV–vis absorption spectrum of BG dye dissolved in water (starting dye concentration, 20 ppm in 100 mL of DI) with 0.05 g of the composite with NZGO-1 obtained at 25 °C after 90 min of irradiation time. As seen, the notable absorption peak of the dye at 623 nm decreases as the irradiation time increases and almost vanishes after 90 mins which is a promising indication that the concentration of BG decreases as the time at which it is exposed to irradiation increases.

Fig. 5
figure 5

a UV-absorption spectrum showing the photocatalytic degradation of BG under simulated visible light for NZGO-1, b illustration of the photodegradation of BG without a catalyst and with the as-prepared photocatalysts, c the ln(Co/Ct) versus time graph for the as-prepared photocatalysts, and d a bar graph showing the amount of BG in milligrams per gram absorbed after stirring in the dark for 15 min

Full size image

The photodegradation efficiency of BG with the modified systems (i.e., N-ZnO, NZGO-1, NZGO-2, and NZGO-3) is much better than that of the unmodified ZnO as shown in Fig. 5b. The optimum degradation was observed with the nanocomposite containing 0.1%wt GO, while increasing the amount of GO beyond this was observed to inhibit the photocatalytic activity. Similar observations have been reported [29] with reasons related to the presence of excess GO in the system causing increased light irradiation absorption as well as scattering. This scattering effect leads to lower photocatalytic efficiency because only a small amount of light irradiation passes through the reaction mixture and thus a lower number of active sites of the catalyst are activated for photodegradation. The percentage removal of BG as illustrated in Fig. 5b was increased from 66 to 83, 88, 96, and 100% for the synthesized photocatalysts ZnO, N-ZnO, NZGO-1, NZGO-2, and NZGO-3, respectively, after 90 min of light irradiation. From the results above, it can be deduced that the increased photocatalytic capacity of the nanocomposites was because of the presence of GO in the system and the reduced band gap of the composites.

Introducing the GO to the nanocomposite makes it possible for more BG molecules to be adsorbed on the surface of the semiconductor as compared to the unmodified ZnO nanoparticles. In order to ascertain the adsorption abilities, the as-prepared photocatalyst was investigated at equilibrium and computed by using Eq. 5;

$$ {A}_t=\frac{\left({C}_0-{C}_t\right)V}{m} $$
(5)

where At measured in milligrams per gram represents the amount of BG adsorbed per gram of the catalyst with time (t). C0 and Ct are the concentration (initial) at t = 0 and the concentration (final) of BG at time t in parts per million, respectively. V represents the volume of BG solution in milliliters and m stands for the mass of photocatalyst in grams. From the bar graph in Fig. 5d, it can be seen that the adsorption of BG increased from 20 to 26 mg g−1, 28, 29, and 31 mg g−1 for the catalysts ZnO, N-ZnO, NZGO-1, NZGO-2, and NZGO-3, respectively. The increase in the adsorption ability of the composite on the introduction of GO may be related to a larger surface area of exposure and oxygen functionalities possessed by GO and the π − π stacking interaction which exists between GO and BG which implies an increase in the surface electric charge of ZnO. Figure 5c shows that the photodegradation of BG under simulated visible light using the as-prepared photocatalysts obeyed Lagergren [43] pseudo first-order kinetics with respect to BG concentration:

$$ \ln \left(\frac{C_0}{C_t}\right)=k\times t $$
(6)

where K is pseudo first-order rate constant and C0 is the initial concentration at t = 0, i.e., after stirring in the dark for 30 min. Ct stands for concentration at t = time. This is an indication that the dye is adsorbed onto the semiconductor surface occurs through. The rate constant of the composite (NZGO-1) is about five times greater than that of the unmodified ZnO as analyzed from Fig. 5c. This enhancement of the rate constant is observed because of the modifications made on ZnO using N and GO.

COD, which is a measure of the amount of dissolved oxygen gobbled by chemical oxidation of the organic matters in water, was used to investigate the extent of mineralization of the dye solution. With the optimum conditions, the COD obtained before degradation was 187 mg L −1 and was reduced to 37 mg L −1 after 90 min of irradiation. This corresponds to 80% degradation of BG.

Photodegradation kinetics and mechanism

Based on the fact that effluents from the industries can hardly be neutral, one important parameter not to be ignored in regard to wastewater treatment is the effect the pH of the dispersions has on the reaction rate. pH of the dissolution affects the charge on the semiconductor’s surface hence influencing how fast reactions take place on the semiconductor’s surface. Hence, pH effect on the photodegradation of BG was evaluated in this study and the results are displayed in Fig. 6a. Based on the issues of cost and the feasibility of carrying out pH adjustments at large scale, extreme pH values were not considered in this study; pH values in the range of 4–11 were used. From Fig. 6a, it is obvious that the photodegradation rate of BG increases as the pH increases up to a maximum pH of 9. Thereafter, a further increment in pH results in the retardation of the reaction rate. This observation may be ascribed to the fact that the surface of the semiconductor becomes negatively charged as pH increases making it easier for the cationic dye to be adsorbed onto the surface of the catalyst [44]. Since photodegradation process begins with adsorption of the dye, enhanced adsorption resulted in high degradation process. Also, increasing the pH leads to the generation of more active species in the system (e.g., ·OH radicals) which results from the reaction between the photogenerated holes (h+) of the semiconductor and the OH− of the base [45]. Hence, an intimate interaction between the adsorbed dye and.OH radicals leads to a higher rate of degradation on increasing the pH value.

Fig. 6
figure 6

a Effects of pH on the photodegradation of BG, b effect of photocatalyst dosage, c photostability of NZGO-1 after three times of use, and d photocatalytic mechanism

Full size image

On further increasing the pH above the optimum value, the cationic dye changes into a neutral molecule hence making adsorption onto the surface of the photocatalyst difficult and therefore leading to the retardation in the rate of dye degradation. The effect of the amount of catalyst used in degradation of BG was studied by varying the amount of NZGO-1 from 0.01 to 0.1 g, while other parameters are kept constant (see illustration in Fig. 6b). As can be observed from Fig. 6b, the rate of degradation of BG increased when the amount of the catalyst increases up to a maximum amount of 0.05 g. Thereafter, a further increase directly results in the retardation of the degradation rate.

This result may be ascribed to the fact that there are more numbers of active sites as well as active surface area of the catalyst when increasing the catalyst amount which may lead to more generation of superoxide and hydroxide radicals that enhance the photodegradation of BG. But on further increasing the catalyst amount above its optimum dose, there is a decrease in the photocatalyst’s surface area (which would facilitate pollutant adsorption and light absorption) caused by particle–particle interaction (agglomeration). Also, the turbidity of the suspension increases resulting in less light penetration through the suspension and hence the decrease in the dye molecules’ degradation rate.

The reusability of NZGO-1 in the photodegradation of BG was also investigated by using the photocatalyst three times consecutively in the photocatalytic experiment. As shown in the bar graph (Fig. 6c), 94% degradation of BG was still achieved after three cycles of photodegradation which illustrates the photocatalyst is photostable and can be reused during its application. The observed decrease in the photodegradation rate may be due to the loss of some photocatalyst during sample withdrawal for analysis, filtrations, and drying for reuse in the experiment. Figure 6d is a schematic illustration of the mechanism and charge transfers taking place during the photocatalytic activity of ZnO, N-ZnO, and NZGO-1 nanocomposites. As shown in Fig. 6e, on exposure to sunlight, there is the excitement of electrons to the CB of ZnO from its VB. With respect to the energy gap of unmodified ZnO, it is obvious that a greater amount of irradiation energy is needed to excite electrons to the CB from the VB of ZnO. Also, electron-hole recombination rate is high due to the fact that electrons are unstable in the CB and tend to drop back quickly to the VB. On introduction of nitrogen, as shown in Fig. 6e, there is a decrease in the energy gap caused by the formation of a new energy state which results from hybridizing O 2p and N 2p energy states. Although the band gap was reduced, the recombination rate of electron hole was not greatly influenced and that is in agreement with the PL results in Fig. 5. On introduction of GO to the system as illustrated in Fig. 6e, the excited electrons are transferred to the GO nanosheets from the CB of ZnO. This transformation of excited electrons enhances charge separation and thus inhibits electron-hole recombination rate. Similar results have been previously reported [46, 47]. Superoxide radicals are produced from photogenerated electrons upon reaction with oxygen while the holes produce hydroxide radicals on reaction with water. These radicals degrade the dye molecules and organics in water into H2O, CO2, and non-toxic minerals. Therefore, the enhancement of the photocatalytic efficiency of NZGO-1 in the visible range of the visible spectrum is perhaps associated with the synergetic effects of N and GO in the synthesized nanocomposites.

The comparison of different photocatalysts used in the photodegradation of BG with the newly prepared photocatalyst presented in this work is displayed in Table 2. With respect to the light source, UV light forms only about 5% of the visible radiation while visible light covers up to 46% of the spectrum. Our newly synthesized photocatalyst used visible light in degradation of BG, hence a suitable candidate in terms of harvesting the more abundant visible radiation for photocatalysis. Also, referring to the length of degradation (in time), amount of catalyst used, dye concentration, and the degradation percentage, our newly prepared photocatalyst achieved 100% degradation of BG (20 ppm) dye within 90 min using as low as 500 mg L −1 of the catalyst. These results infer that the as-prepared N-ZnO-GO is a viable photocatalyst in photodegradation of BG under visible light irradiation.

Table 2 Comparison of the results of some of the recent works based on the photodegradation of brilliant green dye in wastewater using different light sources with the present work
Full size table

The tendency of the most efficient catalyst (NZGO-1) to mineralize other dyes was assessed by using it to degrade 20 ppm methylene blue (MB) dye at the optimum pH (pH 9) by following the same procedure as described in the “Evaluation of photocatalytic activity” section. The result indicates that the catalyst performed impressively well against the dye with 98% degradation efficiency (Fig. 7). Thus, the NZGO-1 nanocomposite is effective against the degradation of other dyes in water apart from brilliant green.

Fig. 7
figure 7

a UV-absorption spectrum showing the photocatalytic degradation of MB under simulated visible light for NZGO-1. b Photodegradation profile of MB using NZGO-1 photocatalysts

Full size image

Conclusion

N-ZnO immobilized on GO nanosheets with different GO weight percentages (0.1, 0.5, and 1%) was successfully synthesized and characterized. The composites N-ZnO-GO, when compared to ZnO and N-ZnO, show an enhanced photocatalytic capability under visible region and may be ascribed to the combined effects of nitrogen and GO on ZnO nanoparticles which include increased adsorption rate, reduced band gap, and the slower rate at which electron and hole recombine. The photodegradation rate of BG by the composite was observed to be dependent on the amount of GO and the optimum degradation was obtained with the composite having NZGO-1. At higher GO content, there is lighter scattering than penetration through the solution which then decreases the photocatalytic ability of the composites containing NZGO-2 and NZGO-3. Also, a higher amount of GO in the composite affects the crystallite size of ZnO. The results also showed that the degradation of BG using NZGO-1 with an optimum amount of 0.05 mg occurs faster in basic medium (pH 9) compared to that in acidic medium. Photostability test of NZGO-1 analyzed after three cycles of exposure gave 94% degradation of BG which implies that the composite is also highly stable and can be recovered and reused. From the results, it can be concluded that N-ZnO-GO is an appropriate photocatalyst for efficient photodegradation and mineralization of BG dye in water and future work in this project will investigate photodegradation of other forms of organic dyes in wastewater.