Introduction

Heterogeneous photocatalysis using semiconductor materials and solar energy offers an ideal green approach to the solution of serious energy shortages and environmental problems, where an active semiconductor material is undoubtedly an important key (O’Regan and Gratzel 1991; Hoffmann et al. 1995; Sushma and Girish Kumar 2017). To date, TiO2, with excellent performance, has long been the focus of semiconductor material. However, the wide band gap (Eg = 3.2 eV) and high recombination probability of photogenerated carriers have limited its practical application (Anpo and Takeuchi 2003; Choi et al. 1994). Compared with the conventional TiO2 photocatalyst, more attention has been paid to Cobalt-based oxides and hydroxide (e.g., Co3O4 and Co(OH)2) among the various transition-metal oxides (Li et al. 2015). Especially, Co3O4 has attracted increasing interest as a visible-light photocatalyst, because of its good sunlight utilization (Eg = 2.1 eV), remarkable oxidation ability (2.76 V vs. NHE), magnetic property, nontoxicity, chemical stability, and low cost (Zhang et al. 2014a, b). However, the serious aggregation and poor photoelectric conversion property of Co3O4 lead to low photocatalytic degradation activity; therefore, how to overcome the above disadvantages and improve the photocatalytic activity of Co3O4 is an urgent issue.

Recently, researchers found that the construction and preparation of semiconductor-heterojunction photocatalysts could effectively improve its photocatalytic activity, because the heterojunction with matched energy band gaps could form the internal in-built electric fields and provide a potential driving force to enhance the separation of photogenerated charge carriers, and restrain its recombination (Yu et al. 2008; Chen et al. 2005; Xiang et al. 2012). Generally, the semiconductor-heterojunction photocatalysts are mainly divided into the following four types: (1) the semiconductor/semiconductor heterojunction; (2) the metal/semiconductor heterojunction; (3) the carbon-based materials/semiconductor heterojunction (carbon-based materials mainly include the carbon nanotube and graphene); (4) the multicomponent heterojunction (Wang et al. 2014). The multicomponent heterojunction, which is composed of two or more semiconductor materials and an electronic transmission system, can effectively improve the utilization of solar light and drive separation and transportation of the photogenerated charge carriers. Among the various electronic transmission systems, graphene appears to be an ideal candidate, and its excellent electronic conductivity could enhance the separation efficiency of photogenerated charge carriers in the graphene-based nanocomposites. Moreover, graphene is also an attractive support material for the deposition and dispersion of nanoparticles due to its lamellar structure and high specific surface, leading to high photocatalytic activity (Zhang et al. 2010a; Lightcap et al. 2010). However, up to now, the synthesis of multicomponent heterojunction, in which cobalt oxides, cobalt hydroxide and graphene are spatially integrated, has never been reported. So, design and preparation of efficient and separable visible-light-driven Co3O4/Co(OH)2/RGO photocatalytic heterojunction, which conform to the needs of practical industry and life, is very meaningful and the aim of this study.

In this work, we report for the fabrication of a ternary heterojunction photocatalyst (Co3O4/Co(OH)2/RGO) integrated Co3O4, Co(OH)2 and reduced graphene oxide (RGO) through a simple one-pot hydrothermal approach. It is worth mentioning that the chemical reduction of graphene oxide (GO) and construction of Co3O4 and Co(OH)2 heterojunction operated at the same time during hydrothermal process, and RGO nanosheet plays a vital role in forming the heterojunction. The as-prepared Co3O4 and Co(OH)2 nanoparticles are scattered on the RGO nanosheet surface to avoid magnetic agglomeration of Co3O4 nanoparticles as well as the restacking of RGO nanosheets. The novel ternary heterojunction photocatalyst combine the merits of each component: Both Co3O4 and Co(OH)2 nanoparticles as visible-light-active photocatalysts can convert solar light into photogenerated electron-hole pairs, RGO nanosheets are used as electron collectors to transfer the photogenerated electrons from semiconductor photocatalyst, while magnetic Co3O4 nanoparticles are used to rapidly collect and recover photocatalysts using a powerful magnet. As expected, the as-prepared Co3O4/Co(OH)2/RGO ternary heterojunction photocatalyst exhibits higher catalytic ability in photodegradation of Congo red (CR) dyestuff, due to the synergistic effect between Co3O4, Co(OH)2, and RGO.

Experimental section

Preparation of GO

GO was prepared from natural graphite powder through a modified Hummers’ method (Hummers Jr and Offeman 1958; Li et al. 2012; Lu et al. 2011).

Preparation of Co3O4/Co(OH)2/RGO ternary heterojunction photocatalyst

A series of ternary Co3O4/Co(OH)2/RGO heterojunction photocatalysts with varying GO adding amounts were prepared by a hydrothermal process and labeled as x wt% Co3O4/Co(OH)2/RGO (x wt% is the weight ratio of GO to the resulting products). A certain amount of GO and 0.2 g of CTAB were dispersed in the mixture of distilled water (10 mL) and absolute ethanol (50 mL) with the aid of ultrasonication. Subsequently, 0.08 g of NaOH and 0.1455 g of Co(NO3)·6H2O were added into the above suspension, respectively. After magnetic stirring for 2 h, the suspension (pH 14) was loaded into a 100 mL of hydrothermal autoclave and kept the temperature at 160 °C for 4 h in an electric oven. After cooling to room temperature, the sample in autoclave was collected by high-speed centrifugation, then, washed with ethanol and water alternatively for several times until pH 7. Finally, the sample was dried at 50 °C for 24 h in an oven before characterization.

Preparation of Co3O4, Co(OH)2, RGO, Co(OH)2/RGO, Co3O4/Co(OH)2, and Co3O4/RGO photocatalysts

For control experiments, pure Co3O4 and β-Co(OH)2 samples were also prepared according to the previous report (Liang et al. 2006). 0.5 g of Co(CH3COO)2·4H2O was dissolved in the mixture of ethylene glycol (20 mL) and distilled water (0.4 mL), after microwave heating at 170 °C for 5 min; the cobalt acetate hydroxide precursor was obtained by centrifugation. Finally, the precursor was redispersed in 20 mL of distilled water and microwave heated at 100 °C for 20 min to obtain the β-Co(OH)2 sample. The pure Co3O4 was obtained by a hydrothermal method using the above cobalt acetate hydroxide precursor. The precursor was dispersed in 20 mL of distilled water, then, 1 mL of H2O2 (1 mol/L) was added, and the suspension was loaded into a 40 mL of hydrothermal autoclave and kept the temperature at 140 °C for 12 h to obtain the Co3O4 sample.

RGO was prepared through a facile hydrothermal treatment. GO was dispersed in the mixture of distilled water (60 mL) and absolute ethanol (30 mL) with the aid of ultrasonication, then, the suspension was loaded into a hydrothermal autoclave and kept the temperature at 120 °C for 10 h in an electric oven.

The Co(OH)2/RGO, Co3O4/Co(OH)2, and Co3O4/RGO composite photocatalysts were prepared by mixing the two corresponding powders at the desired ratio.

Characterizations

The crystal structure of ternary Co3O4/Co(OH)2/RGO heterojunction photocatalysts was characterized with the X-ray diffraction (XRD, Rigaku D/Max-2550) and Raman spectrometer (J-Y T64000). Its morphology and structure was measured on transmission electron micrographs (TEM, HITACHI H-8100 EM) and scanning electron microscopy (SEM, XL30 ESEM FEG). Photoluminescence (PL) spectra were recorded on a microplate reader (BioTek Synergy H1M) with the excitation wavelength of 320 nm. Thermogravimetric analysis (TGA) was operated on a TGA 4000 analyzer (PerkinElemer) in air, controlling the temperature from 30–900 °C with a rate of 10 °C/min. The UV-vis diffuse reflectance spectra were recorded with a Shimadzu 3600 UV-vis-NIR spectrophotometer equipped with an integrating sphere diffuse reflectance accessory, while BaSO4 was used as a reference. The total organic carbon (TOC) analysis was measured with a total organic carbon analyzer (TOC/L CPH).

Photoelectrochemical measurements

All the electrochemical measurements (transient photocurrents, current-voltage (I-V), electrochemical impedance spectroscopy (EIS)) were performed by using a CHI 660D electrochemical analyzer (CH Instruments, Shanghai), and a homemade three-electrode cell composed of a saturated calomel electrode, a platinum wire, and the as-prepared photocatalysts as the reference, counter, and working electrodes, respectively, and a 1-M NaOH solution was used as the electrolyte. The transient photocurrents were carried out using a 500-W xenon lamp (CHFXQ500W, Beijing) with a cutoff filter (λ > 420 nm). Electrochemical impedance spectroscopy (EIS) measurements were conducted in the frequency range from 100 kHz to 0.01 Hz at open circuit potential with an AC perturbation of 5 mV.

Photocatalytic test

The photocatalytic activity test is carried out in a set of photocatalytic devices purchased from Beijing Perfectlight Co. Ltd., which mainly includes the following three sections: (1) Pyrex reaction cell fitted with cooling-water interlayer and a quartz cover is used as a reactor; (2) Low-temperature thermostat bath (BILOW-W-501) is used to control the temperature of a photocatalytic reactor at 20 °C; (3) Xenon lamp (PLS-SXE300C, 500 W) fitted with an optical band-pass filter is used as a solar-light source to simulate visible light (λ > 420 nm). In a typical test, 0.05 g of ternary Co3O4/Co(OH)2/RGO heterojunction was first dispersed in 50 mL of 100 mg L−1 CR aqueous solution, and kept stirring in the dark for 60 min. After reaching the adsorption-desorption balance, the system was simulated with Xenon lamp, and the photodegradation efficiency of CR was monitored on a UV-vis spectrophotometer (PE, Lambda 25) in real time.

Active species test

The main active species during the photocatalytic process were detected by the scavenging experiments, p-benzoquinone (BQ, 0.0001 mol/L), disodium ethylenediaminetetraacetate (EDTA, 0.01 mol/L) and isopropanol (IPA, 0.01 mol/L) were used as ·O2− radicals, h+, and ·OH radicals scavenger, respectively. The scavengers were added to the CR solution prior to irradiation in three separated systems.

Results and discussion

Morphology of Co3O4/Co(OH)2/RGO ternary heterojunction

The morphology of obtained products was first monitored through the TEM, as shown in Fig. 1. Figure 1 shows TEM images of the as-prepared samples; Fig. 1a displays the well-dispersed GO nanosheets with many wrinkles. After the hydrothermal reaction of GO with Co(NO3)·6H2O, obviously, a large amount of nanoparticles with diameter at the range of 8 to 23 nm are scattered on the nanosheets surface (Fig. 1b, c). The HRTEM image in Fig. 1e and its corresponding electron diffraction pattern (Fig. 1d) suggest that the nanoparticles are highly crystallized; the measured interlayer spacing of 0.47 nm is same as the distance of (111) crystallographic planes of Co3O4 with a face-centered cubic structures (Nie et al. 2013) and that of 0.24 nm consistent with the (002) crystallographic planes of β-Co(OH)2 (Li et al. 2015; Zhou et al. 2016; Koza et al. 2013). These observations indicate that an intimate interface is built successfully between Co(OH)2 and Co3O4 nanoparticles at the nanoscale, and RGO as a support material avoids the agglomeration of Co3O4 and Co(OH)2 nanoparticles, making more surface active positions to participate in the photocatalytic process. For comparison, the 6-wt% Co3O4/Co(OH)2/RGO ternary heterojunction photocatalyst without CTAB was also prepared, and its TEM and HRTEM images were shown in Fig. 1f, g. An intimate interface of Co(OH)2 and Co3O4 nanoparticles is also observed; however, it is obviously that the RGO has a poor dispersion, indicating that the CTAB can play a role in dispersing GO nanosheets.

Fig. 1
figure 1

(a) TEM image of GO, (b) low, (c) high magnification TEM images, (d) electron diffraction pattern, and (e) HRTEM image of the as-prepared 6-wt% Co3O4/Co(OH)2/RGO ternary heterojunction photocatalyst with CTAB. (f) TEM and (g) HRTEM images of 6-wt% Co3O4/Co(OH)2/RGO ternary heterojunction photocatalyst without CTAB

Full size image

Structural analysis

The crystalline nature of the as-prepared samples was examined by X-ray diffraction (XRD) analysis, as shown in Fig. 2. Figure 2a shows the XRD patterns of the as-prepared samples with different hydrothermal temperature, at a lower reaction temperature of 100 °C and 120 °C; the main peaks, around 19.0°, 31.3°, 36.9°, 38.5°, 44.8°, 59.4°, and 65.2°, can be indexed to the (111), (220), (311), (222), (400), (511), and (440) planes of Co3O4 with face-centered cubic structures (JCDPDS no. 42-1467). In contrast, when the reaction temperature increased to 160 °C, the sample exhibits that additional peaks at 32.5°, 37.9°, 51.4°, 57.9°, and 59.6° can be indexed to the (100), (101), (102), (110), and (003) planes of β-Co(OH)2 phase with a hexagonal structure (JCDPDS no. 30-0443), indicating that the mixed-crystalline phases of Co3O4 and β-Co(OH)2 form under these reaction conditions (Li et al. 2015). Moreover, the crystallization degree of sample is further improved as reaction temperature increased to 180 °C. In the next work, we choose 160 °C as the reaction temperature to explore the effect of reaction time on the crystal phase of products, as shown in Fig. 2b. At a shorter reaction time of 2 h, only the face-centered cubic Co3O4 phase is obtained. However, when the reaction time is prolonged to 4 h, the diffraction peaks of β-Co(OH)2 phase with a hexagonal structure are observed besides Co3O4. Similarly, the crystallization degree of products is further improved as reaction time increased from 4 to 12 h. Figure 2c compares the XRD patterns of samples prepared at 160 °C for 4 h with different pH. The pH of precursor solution is adjusted by NaOH from 6.2 to 14; it is clearly seen that single face-centered cubic Co3O4 phase is obtained, when the pH is 6.2. When the pH increases from 8.6 to 14, both Co3O4 and β-Co(OH)2 phases are obtained. In contrast, the pH of precursor solution is adjusted to 11.6 using NH3·H2O instead of NaOH, but only face-centered cubic Co3O4 phase is produced. Based on the above results, we continue to explore the influence of GO adding amounts on the crystal phase of products prepared at 160 °C for 4 h, and pH 14. Without addition of GO, the diffraction peaks for 0-wt% sample can be ascribed to the well-crystallized Co3O4 with face-centered cubic structures, which is consistent with the above results of TEM images. After adding GO in the system, the diffraction peaks of β-Co(OH)2 phase with a hexagonal structure are observed besides Co3O4. In addition, the XRD peak of RGO nanosheets at approximately 26° is absent; this phenomenon may be due to the low amount of GO introduced in the system, or the weak peak intensity of RGO disguised by that of others, leading it invisible.

Fig. 2
figure 2

X-ray diffraction spectra of the as-prepared samples with (a) different hydrothermal temperature, reaction conditions: the amount of GO, 6 wt%; reaction time, 4 h; pH 14, (b) different reaction time, reaction conditions: the amount of GO, 6 wt%; reaction temperature, 160 °C; pH 14, (c) different pH, reaction conditions: the amount of GO, 6 wt%; reaction temperature, 160 °C; reaction time, 4 h, and (d) different GO adding amounts, reaction conditions: reaction temperature, 160 °C; reaction time, 4 h; pH 14

Full size image

Based on the above XRD results, we assume that the GO nanosheets play a vital role in forming the ternary Co3O4/Co(OH)2/RGO heterojunction. During the hydrothermal process, Co(NO3)2·6H2O is used as Co precursor, the hydration and hydrolysis reactions of Co2+ will occur to generate the hydroxo complex of Co2+ ([Co(OH)4]2−), and some [Co(OH)4]2− will be further oxidized and transformed into the hydroxo complex of Co3+ ([Co(OH)6]3−) with an increase of the system temperature and pressure. Finally, the Co3O4 formed through the fasciculation reaction of [Co(OH)4]2− and [Co(OH)6]3− with a mole ratio of 1:2. However, the mole ratio of [Co(OH)4]2− and [Co(OH)6]3− is above 1:2, due to the introduction of GO in this system, leading some Co(OH)2 obtained through the fasciculation reaction of [Co(OH)4]2−. Thus, the ternary Co3O4/Co(OH)2/RGO heterojunction photocatalysts are obtained by a simple one-pot hydrothermal method.

The successful one-pot reduction of GO and the formation of Co(OH)2 and Co3O4 were also verified by Raman spectroscopy. Figure 3 shows the Raman spectra of GO and 6-wt% Co3O4/Co(OH)2/RGO ternary heterojunction. For GO, there are two obvious peaks at ~ 1355 (D-band) and ~ 1601 cm−1 (G-band). After the hydrothermal reaction, it is obvious that the as-prepared sample shows relative higher intensity ratio of D to G-band (ID/IG = 0.98) than that of GO (0.82), indicating the reduction of GO during the hydrothermal reaction (Chen and Yan 2010). In addition, other peaks at approximately 485, 524, 692, and 1127 cm−1 are detected, which are consistent with the reported E2g, 2F2g, and A1g phonon modes of Co3O4 and the OH deformation modes of β-Co(OH)2, respectively (Liu and Patzke 2014; Abd El-sadek et al. 2010; Yang et al. 2010). All the above characterization results further proved that the chemical reduction of GO and the formation of Co3O4 and Co(OH)2 nanoparticles occurred simultaneously during one-pot hydrothermal process.

Fig. 3
figure 3

Raman spectra of GO and the 6-wt% Co3O4/Co(OH)2/RGO ternary heterojunction

Full size image

In order to obtain the content information of Co(OH)2, Co3O4 in the Co3O4/Co(OH)2/RGO ternary heterojunction, thermogravimetric analysis (TGA) was used to characterize the samples, as shown in Fig. 4. For the Co3O4/Co(OH)2/RGO ternary heterojunction, the mass loss (~ 2 wt%) below 100 °C is caused by the loss of adsorbed water from sample (Wang et al. 2011). There are two rapid mass loss at 200–400 °C and 400–700 °C, the former is attributed to the thermal decomposition of Co(OH)2 into Co3O4, and the latter is attributed to the combustion of carbon skeleton of RGO nanosheets (Li et al. 2010). According to the TGA results, we can calculate that the content of Co(OH)2, Co3O4, and RGO is ~ 7 wt%, 85 wt%, and 6 wt%, respectively. The calculated content of RGO is consistent with the adding amount of RGO.

Fig. 4
figure 4

TGA curves of (a) pure Co3O4, (b) Co3O4/RGO, (c) 6-wt% Co3O4/Co(OH)2/RGO ternary heterojunction, and (d) pure Co(OH)2

Full size image

Optical absorption characteristic

The optical absorption characteristic is an important parameter to evaluate photocatalytic performance of semiconductors. Figure 5 shows the UV-vis diffuse reflectance spectra of pure Co(OH)2, Co3O4, and Co3O4/Co(OH)2/RGO ternary heterojunction. The pure Co(OH)2 shows an obvious visible-light absorption (400–700 nm), which is ascribed to the d-d transition of Co(II) (Zhang et al. 2014a). The pure Co3O4 shows a single broad absorption band across the entire range (200–800 nm), which is the O → Co2+ charge transfer, d-d transition of Co2+ and Co3+ (Liu and Patzke 2014). For the Co3O4/Co(OH)2/RGO ternary heterojunction, their absorption curves are similar to those of pure Co3O4 because of the higher content of Co3O4 in ternary heterojunction.

Fig. 5
figure 5

UV-vis diffuse reflectance spectra of pure (a) Co(OH)2, (e) Co3O4 and Co3O4/Co(OH)2/RGO ternary heterojunction with varying GO adding amounts, (b) 4 wt%, (c) 6 wt%, (d) 8 wt%

Full size image

Photocatalytic activity of Co3O4/Co(OH)2/RGO ternary heterojunction

Congo red (CR), a benzidine-based, direct, anionic diazo dyestuff, is mainly used as dyeing agent for cotton, hemp, silk, and other textile products to give a red color. The effluents containing CR dyestuff are highly colored, resulting in major environmental problems. Thus, we choose CR as the target dyestuff to evaluate the visible-light photocatalytic activity of ternary Co3O4/Co(OH)2/RGO heterojunction. Figure 6a compares the kinetics of photodegradation of CR by using pure Co(OH)2, Co3O4, 6-wt% Co(OH)2/RGO, Co3O4/Co(OH)2, 6-wt% Co3O4/RGO, and ternary Co3O4/Co(OH)2/RGO photocatalysts with varying amounts of RGO. Obviously, the decomposition of CR is negligible in the absence of photocatalyst in the control experiment. For comparison, the photocatalytic activity of pure Co(OH)2 and Co3O4, 6 wt% Co(OH)2/RGO, 6 wt% Co3O4/RGO, and Co3O4/Co(OH)2 are also checked, and about 15%, 27%, 20%, 10%, and 37% of CR, respectively, had been degraded under the same testing conditions. However, it is obvious that the Co3O4/Co(OH)2/RGO ternary heterojunction shows significant improvement of photodegradation efficiency of CR compared to the above photocatalysts, and about 73%, 90%, and 63% are eliminated from the solution within 60 min in the presence of 4%, 6%, and 8% Co3O4/Co(OH)2/RGO ternary heterojunction photocatalysts, indicating the presence of synergistic effect between Co3O4, Co(OH)2, and RGO in photocatalytic reaction. The highest photocatalytic efficiency was obtained from 6% Co3O4/Co(OH)2/RGO as 90%; when the RGO content is increased to 8%, the photocatalytic activity is reduced, indicating that there is an optimal value for the adding amount of RGO. When the RGO content is lower than 6%, the photogenerated electrons in ternary heterojunction cannot be effectively transferred by a small amount of RGO. When the RGO content is higher than 6%, a large amount of RGO will directly affect the incident light transmittance of photocatalyst, preventing the generation of photoinduced electron-hole pairs in the photocatalyst.

Fig. 6
figure 6

(a) The kinetics of photodegradation of CR (■) by using pure Co(OH)2 (◄), pure Co3O4 (▼), 6-wt% Co(OH)2/RGO (►), 6-wt% Co3O4/RGO (▲), Co3O4/Co(OH)2 (), and Co3O4/Co(OH)2/RGO ternary heterojunction with different amounts of RGO (● = 4 wt%; = 6 wt%; ♦= 8 wt%). (b) UV-vis absorption spectra of CR using 6-wt% Co3O4/Co(OH)2/RGO ternary heterojunction photocatalyst under visible light (λ > 420 nm). (c) The TOC content and CR removal efficiency at different irradiation times using 6% Co3O4/Co(OH)2/RGO ternary heterojunction photocatalysts. (d) Kinetic fit for the photodegradation of CR (■) by using pure Co(OH)2 (◄), Co3O4 (▼), and Co3O4/Co(OH)2/RGO ternary heterojunction with different amounts of RGO (●= 4 wt%; ♦= 6 wt%; ▲ = 8 wt%)

Full size image

In addition, the UV-vis absorption spectra of CR, the TOC content, and CR removal efficiency at different irradiation times using 6-wt% Co3O4/Co(OH)2/RGO ternary heterojunction photocatalyst during the photocatalytic process were recorded and shown in Fig. 6b, c. As the illumination time is prolonged, the absorption intensity of CR aqueous solution gradually decreased, but, the maximum absorption wavelength of CR aqueous solution keeps at 496 nm, and no new absorbance peaks are observed. Furthermore, it can be seen that the CR removal efficiency dramatically decreases in the first 60 min, and reaches 99% after 150 min. Correspondingly, the TOC content shows a similar trend, and the TOC removal efficiency reaches 95% after 150 min (Fig. 6c), indicating that the CR dyestuff has been completely degraded, and the photocatalytic activity of Co3O4/Co(OH)2/RGO ternary heterojunction is better than that of many previous reports on Cobalt-based photocatalysts (Bin and Hui 2015; Shi et al. 2012). Moreover, according to the kinetic study results in Fig. 6d, the photodegradation process of CR complies with pseudo-first-order kinetics, and the 6-wt% Co3O4/Co(OH)2/RGO ternary heterojunction shows a higher kinetic constant (0.0329 min−1) than others (Table 1).

Table 1 Pseudo-first-order rate constants and R2 of different photocatalysts photodegraded on CR
Full size table

Recycling ability is also an important factor to estimate the photocatalyst performance besides the photocatalytic efficiency, and Fig. 7a shows the recycling photocatalytic test of 6-wt% Co(OH)2/RGO, Co3O4/RGO, and Co3O4/Co(OH)2/RGO ternary heterojunction photocatalyst. Obviously, there is no significant change in photocatalytic activity during the five cycles, implying that a good stability of Co(OH)2/RGO, Co3O4/RGO, and ternary Co3O4/Co(OH)2/RGO photocatalysts. Moreover, the ternary heterojunction can be rapidly collected from the suspension using a powerful magnet, as shown in Fig. 7b, which is very meaningful in the practical industry and life applications. Thus, all above-mentioned results indicate that the Co3O4/Co(OH)2/RGO ternary heterojunction is a promising candidate for photocatalysts in practical application.

Fig. 7
figure 7

(a) Recycling photocatalytic test of 6-wt% Co(OH)2/RGO (►), 6-wt% Co3O4/RGO (▲), and 6-wt% Co3O4/Co(OH)2/RGO (♦) ternary heterojunction photocatalyst, (b) a digital photo illustrates the separation process of ternary heterojunction from the suspension by using a powerful magnet

Full size image

Synergistic effect of Co3O4/Co(OH)2/RGO ternary heterojunction

The photoluminescence (PL) spectrum is an effective means to investigate the microscopic behavior of photoinduced charge in semiconductor, including generation, transmission, and recombination, because the PL emission is derived from the recombination of photoinduced electron-hole pairs. Figure 8a compares the PL spectra of pure Co(OH)2, Co3O4, and 6-wt% Co3O4/Co(OH)2/RGO ternary heterojunction; there is a broad band in range 380–600 nm for all the samples. Obviously, the emission intensity of 6-wt% Co3O4/Co(OH)2/RGO ternary heterojunction is much weaker than that of pure Co(OH)2 and Co3O4, indicating a high charge-separation rate (Wu et al. 2015; Hou et al. 2013).

Fig. 8
figure 8

(a) Photoluminescence spectra, (b) transient photocurrents, (c) EIS Nyquist plots, and (d) I-V characteristics of pure Co(OH)2, Co3O4, and 6-wt% Co3O4/Co(OH)2/RGO ternary heterojunction

Full size image

The transient photocurrent of ternary Co3O4/Co(OH)2/RGO photocatalyst under irradiation with visible light (λ > 420 nm) for 20 s is shown in Fig. 8b. As can be seen, the pure Co3O4 (0.43 μA) and Co(OH)2 (0.64 μA) show a small anodic photocurrent, which means that the photogenerated electrons move to the bulk and subsequently transfer to the ITO electrode, leaving holes on the outer surface. However, compared with the pure Co3O4 and Co(OH)2, a significant enhancement in the photocurrent is observed for ternary Co3O4/Co(OH)2/RGO photocatalyst (1.5 μA), indicating that the synergistic effect between Co3O4, Co(OH)2, and RGO can enhance the separation of charge carriers.

Figure 8c shows the EIS Nyquist plots of pure Co(OH)2, Co3O4, and 6-wt% Co3O4/Co(OH)2/RGO ternary heterojunction. Generally, the smaller arc in an EIS Nyquist plot indicates a smaller charge-transfer resistance on the electrode surface (Lim et al. 2014). The Co3O4/Co(OH)2/RGO ternary heterojunction shows a relative smaller arc size than that of pure Co(OH)2 and Co3O4, manifesting the higher efficiency of charge separation and transfer. The above experimental results indicate that the ternary heterojunction could significantly improve the separation of photoinduced electron-hole pairs and restrain its recombination, which is very beneficial to improve the photocatalytic activity.

Photocatalytic mechanism of Co3O4/Co(OH)2/RGO ternary heterojunction

The main active species during the photocatalytic process were detected by the scavenging experiments, as shown in Fig. 9a. The photocatalytic degradation of CR is slightly inhibited in the presence of EDTA (h+ scavenger), while it is intensively suppressed by the addition of IPA (·OH scavenger) or p-BQ (·O2− scavenger) into this system. Thus, it can be inferred that the main active species affect the photocatalytic activity in the following order: ·OH > ·O2− > h+.

Fig. 9
figure 9

(a) Photocatalytic activity of the 6-wt% Co3O4/Co(OH)2/RGO ternary heterojunction for degradation of CR in the presence of different scavengers, EDTA (●), p-BQ (■), IPA (▲), and no scavenger (♦). (b) Microscopic mechanism of the Co3O4/Co(OH)2/RGO ternary heterojunction during the photocatalytic process

Full size image

The flat band potentials of photocatalysts were recorded by the I-V test (Fig. 8d). The flat band potential represents the apparent Fermi level of a semiconductor in equilibrium with a redox couple, and the Fermi level lies close to the conduction band (CB) of the n-type semiconductor (Zhang et al. 2010a, b). The pure Co(OH)2 and Co3O4 show flat potential at − 1.60 V and 0.42 V vs NHE, according to the band gap energy of Co(OH)2 (Eg = 2.9 eV) and Co3O4 (Eg = 2.1 eV); the calculated valence band (VB) is 1.30 V and 1.68 V vs NHE for pure Co(OH)2 and Co3O4, respectively.

We analyzed the test results mentioned above, and concluded the operating of Z-scheme charge carrier mechanism in the Co3O4/Co(OH)2/RGO ternary structure (Fig. 9b). According to the band theory of semiconductors, both Co(OH)2 and Co3O4 are visible-light active components; they can be excited simultaneously, using the visible light (λ > 420 nm). Their photogenerated electrons can be transferred from each VB to CB. On the one hand, some electrons in the CB of Co3O4 easily transfer to the VB of Co(OH)2 (electron transfer path I) and recombine with the holes (Maeda et al. 2016). On the other hand, other electrons in the CB of Co(OH)2 can transfer to the Fermi level of graphene (0 V vs. NHE) (electron transfer path II) that is an excellent electron conductor (Iwase et al. 2011; Hayashi et al. 2011). The processes of electron transfer paths I and II could effectively enhance the separation of photogenerated charge carriers and restrain its recombination. As a result, a lot of photogenerated electrons on graphene and Co(OH)2 nanoparticles surface as well as holes on Co3O4 nanoparticles surface that participate in the photocatalytic reactions can directly and indirectly decompose CR dye, respectively. Thus, the ternary Co3O4/Co(OH)2/RGO heterojunction photocatalyst could exhibit a better photocatalytic activity.

Conclusions

In conclusion, ternary Co3O4/Co(OH)2/RGO heterojunction photocatalysts were constructed by a simple one-step synthetic strategy, and the RGO plays a vital role in constructing the ternary heterojunction. The obtained ternary Co3O4/Co(OH)2/RGO heterojunction showed an excellent photocatalytic performance stimulated by visible light, which should be contributed to the cooperative effects and cooperation of two aspects: (1) RGO as a support material avoids the agglomeration of Co3O4 and Co(OH)2 nanoparticles, making more surface active positions to participate in the photocatalytic process; (2) The synergistic effect between Co3O4, Co(OH)2, and RGO enhance the separation of photogenerated charge carriers and restrain its recombination. Furthermore, the ternary Co3O4/Co(OH)2/RGO heterojunction photocatalysts can be rapidly collected from the suspension using a powerful magnet and recycled with good stability, which is very meaningful in the practical industry and life. So, this research makes sense to construct graphene-based multicomponent nanocomposites for photocatalysis and other applications.