Photocatalytic activity of CuInS2 nanoparticles synthesized via a simple and rapid microwave heating process

In this research, visible–light photocatalytic activities of CuInS2 nanoparticles for degradation of three organic dyes (rhodamine B; RhB, methylene blue; MB, and methyl orange; MO) were investigated. The CuInS2 nanoparticles were synthesized by a simple and rapid microwave heating process using sodium sulfide as a sulfur source and then characterized by x–ray diffraction (XRD), field emission scanning electron microscopy (FESEM), transmission electron microscopy (TEM), Brunauer–Emmett–Teller (BET), and UV–vis diffuse reflectance spectroscopy (UV–vis DRS) techniques. The synthesized CuInS2 nanoparticles exhibited excellent photocatalytic degradation activity to the cationic dyes (RhB and MB) when compared with that of anionic dye (MO). Zeta potential of the CuInS2 photocatalyst was measured to elucidate the adsorption ability toward dye molecules. A possible photocatalytic degradation mechanism was proposed based on active species quenching experiments and Mott–Schottky analysis.


Introduction
Semiconductor-mediated heterogeneous photocatalysis is a promising technology for energy and environmental applications [1][2][3][4]. Metal oxide semiconductors, such as TiO 2 and ZnO, have been widely used as photocatalysts for the degradation of organic contaminants in wastewater owing to their non-toxicity, and properties of chemical and thermal stability [5][6][7][8]. Unfortunately, these materials suffer from the poor utilization of visible light due to their large band gaps (3.2 eV for TiO 2 and 3.3 eV for ZnO) [5][6][7][8]. Binary metal sulfides are attractive candidates as highly efficient visible-light driven photocatalysts because the valence bands of these materials are constructed by 3p orbitals of S, which results in narrower band gaps when compared to those of the metal oxides, and thus make most of the metal sulfides sensitive to visible light [9]. However, these materials have lower charge separation efficiency and undergo photocorrosion, limiting their practical applications [10]. Recently, ternary metal sulfide-based photocatalysts have received increasing amounts of focus in photocatalysis research because of their large absorption coefficient over a wide spectral range, good thermal property, and long-term stability against photocorrosion when compared to relevant binary metal sulfide-based photocatalysts [11,12]. Interestingly, the nanostructured photocatalysts that have large specific surface area possess more active sites for the favorable transport of electrons and holes to the surface of the photocatalysts which are available for photocatalytic reactions. All of which results in an improvement in photocatalytic performance in comparison to bulk materials [13].
Copper indium sulfide (CuInS 2 ) is an interesting ternary metal sulfide that has a wide range of potential applications including photocatalysis [14][15][16][17][18], in photodetectors [19], bioimaging [20,21] and in solar cells Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI. [22,23]. Its conduction band is constructed from 5s orbital of In and its valence band consists of 3p orbitals of S [12]; thereby, a narrower band gap energy (1.53 eV for the bulk) can be achieved. In the photocatalysis, it has been explored as a photocatalyst for H 2 evolution [14,15,24,25], organic pollutant degradation [16,26,27], and nitrate ion reduction [17,18]. Hydro/solvothermal methods have been used to synthesize CuInS 2 nanoparticles [28][29][30][31][32]. However, these methods generally require expensive and complicated equipment, long periods of reaction time and high consumption of an electrical energy. Microwave heating synthesis is found to be an effective method for preparing nanomaterials over a short reaction time due to its high potential to pulse, which can generate a noticeable degree of heating inside the reaction system [33]. Our previous studies [23,34] have reported on the microwave synthesis of the CuInS 2 nanostructures using thioacetamide and L-cysteine as sulfur sources. Despite the fact that CuInS 2 nanostructures were obtained, long periods of irradiation time were required (22.5 min). This could be caused by the partial loss of energy for cleaving the C-S bond in the thioacetamide molecules [23] and the metal-organic bonds in the metal-L-cysteine complexes [34] to release the sulfide (S 2− ) ion to subsequently form the CuInS 2 nanoparticles.
In this research, we have reported on a cyclic microwave heating method for synthesizing CuInS 2 nanoparticles using sodium sulfide as a sulfur source. Since sodium sulfide is an ionic compound, it can more rapidly release sulfur ion rather than thioacetamide and L-cysteine which are a covalent compound. The required reaction time for the microwave synthesis of CuInS 2 nanostructures is expected to be relatively fast. The photocatalytic degradation activities of rhodamine B, methylene blue and methyl orange (representing organic pollutant) over the synthesized CuInS 2 photocatalyst were investigated. Mott-Schottky and active species quenching experiments were carried out in order to clarify a possible photocatalytic degradation mechanism.
2. Experimental procedure 2.1. Synthesis of CuInS 2 nanoparticles All the chemicals were analytical grade and used without further purification. CuCl 2 ·2H 2 O (2 mmol; BDH Chemicals, Ltd, 98%), InCl 3 ·4H 2 O (2 mmol; Sigma-Aldrich, Co., 99.9%) and Na 2 S (4 mmol; Carlo Erba, 99.99%) were separately dissolved in 10 ml of ethylene glycol (Carlo Erba, 99.5%). These solutions were mixed together at room temperature for 30 min. The mixed solution was then irradiated using a domestic microwave oven operated at a frequency of 2450 MHz with a power of 600 W for 2-15 processing cycles. Each cycle lasted 30 s for every 60 s intervals (at 50% of microwave power). Black powder was collected and washed with deionized water and ethanol, respectively. Finally, the powder was dried at 80°C.

Characterizations
A Rigaku Miniflex II x-ray diffraction (XRD) spectrometer was used to determine the crystal structure, purity and crystallinity of the synthesized powders. Particle size and morphology of the powders were determined by a JEOL JSM-6335F field emission scanning electron microscope (FESEM) operating at 15 kV and a JEOL JEM-2010 transmission electron microscope (TEM) operating at 200 kV. The average diameter of the nanoparticles was determined using Image J software. Energy-dispersive x-ray (EDX) spectroscopy was used in conjunction with FESEM to analyte the elemental composition of the sample. A SHIMADZU UV-1800 UV-vis diffuse reflectance spectrophotometer (UV-vis DRS) was used to determine the optical properties of the sample. A specific surface area was obtained by using the Brunauer-Emmett-Teller method (BET, Autosorb1MP, Quantachrome).
A HORIBA nanopartica SZ-100 zeta potential analyzer was used to measure the zeta potential of the CuInS 2 nanoparticles. Briefly, 20 mg of CuInS 2 were dispersed in 100 ml of the ionic solution (1× 10 −3 molL −1 of KCl). The pH of the solution was adjusted to be within a range of 1 to 9 using 0.1 M HCl in the acidic range and 0.1 M NaOH in the basic range.
The Mott-Schottky plot was recorded on an electrochemical analyzer (Autolab PGSTAT, N-Series) equipped with a three-electrode system. Working electrode with 1 cm 2 of sample film was prepared as follows: 100 mg of the CuInS 2 powder was dispersed in a mixture of isopropanol and DI water for 20 min of sonication. After that, the mixed suspension was dropped on the FTO substrate and then dried at 70°C overnight. Pt sheet and Ag/AgCl (3.0 M KCl) were used as a counter electrode and a reference electrode, respectively. Na 2 SO 4 (0.5 M, pH∼7) was used as an electrolyte solution. The Mott-Schottky measurement was carried out at an applied frequency of 1000 Hz in the dark with an AC amplitude of 10 mV and potential ranged from −1.2 to +1.2 V (versus Ag/AgCl). The following equation was used to convert the measured potential versus Ag/AgCl to NHE scale;

Photocatalytic degradation experiment
CuInS 2 photocatalyst (100 mg) was dispersed in an aqueous solution of three organic dyes (rhodamine B, methylene blue, and methyl orange; 200 ml, 10 ppm). The suspension was magnetically stirred under dark conditions for 30 min to ensure that an equilibrium was reached between the adsorption and desorption on the surface of the CuInS 2 photocatalyst. The suspension was then irradiated under a 50 W light-emitting diode (LED) lamp for 0-360 min. The spectral irradiance of the LED light is shown in figure S1. The LED lamp exhibits a white light with the wavelength ranging from 400 to 700 nm, corresponding to a visible light region. The dye solution (5 ml) was taken after every 30 min of irradiation and then monitored using the UV-vis DRS spectrophotometer (SHIMADZU UV-1800) to determine the residual dye concentration. A blank experiment was carried out under visible light irradiation to verify that the decolorization of the dye was indeed due to the photocatalytic reaction. Decolorization efficiency of the photocatalyst was calculated from the changes in the intensity of the characteristic absorption peaks at the absorption maximum of the dyes (λ max =554, 664 and 464 nm for RhB, MB and MO, respectively) by applying the following equation; where C 0 represents the initial concentration of the maximum peak of the absorption spectrum, and C represents the residual dye concentration at each irradiated time interval.

Results and discussion
XRD patterns of the powders synthesized at 600 W using different processing cycles are shown in figure 1. No diffraction peak is observed from the XRD pattern of the powder synthesized at 2 cycles, implying that CuInS 2 crystals has not yet been formed. After being proceeded for 6 and 9 cycles, the diffraction peaks at 2θ of 27.87°, 46.37°and 55.01°, respectively, indexed to the (112), (204) and (312) planes of a tetragonal CuInS 2 structure (JCPDS database no. 32-0339), are found to be in a majority with the existence of a hexagonal CuS structure (JCPDS database no. 01-1281). It should be noted that measured temperatures of the reaction solutions irradiated at 6 and 9 cycles (3 and 4.5 min of irradiation, respectively) are 137 and 148°C (figure 2), respectively. Herein, the XRD results are consistent with the results of previous reports [23,35] which stated that CuS coexists with CuInS 2 when the temperature is lower than 160°C. However, the diffraction peak intensity of the CuInS 2 phase becomes gradually stronger while that of the CuS phase is decreased with increasing processing cycles, indicating the improved purity and crystallinity of the CuInS 2 crystals. After the reaction process has been prolonged to 15 cycles (7.5 min of irradiation, 163°C), pure CuInS 2 with highly crystalline is obtained. When compared with our previous reports [23,34], the synthesis of the CuInS 2 powder using Na 2 S as a sulfur source requires considerably lower processing times under the same microwave power (600 W). This outcome can probably be attributed to the rapid release rate of S 2− ion from the ionic Na 2 S compound. When Na 2 S is dissolved in the polar solvent, S 2− ion is immediately released [36] and subsequently reacts with Cu 2+ to form CuS nuclei, which is further reduced to Cu + by ethylene glycol under microwave heating. Finally, the Cu + reacts with In 3+ and S 2− in the solution to produce CuInS 2 crystals [23]. The size of the CuInS 2 crystals was calculated using the Debye-Scherrer's formula [37]; where D is the crystallite size in nanometer, λ is the wavelength of CuK α radiation, β is the full width at halfmaximum (FWHM) in radian, and θ represents the half-diffraction angle in degree. Scherrer's constant (K ) was derived in Scherrer's original paper with K=2[ln(2)/π] 1/2 [38,39], and a value of 0.94 was obtained for FWHM of the spherical crystals with cubic symmetry [39]. Although the Scherrer equation was derived for the material of the cubic crystals, it is usually used to estimate the crystallite sizes of both cubic and noncubic materials [38].
The calculated average crystallite size is 65 nm. Lattice parameters of CuInS 2 were calculated by assessing the relationship between the Bragg diffraction spacing and tetragonal cell parameters [40] as follows: where d hkl is the Bragg diffraction spacing for the (hkl) plane, h k l are the miller indices, and a and c are the lattice parameters. The calculated lattice parameters of the CuInS 2 nanoparticles are a=b=5.5310 Å and c=11.1104 Å, which agree well with those of the corresponding JCPDS database (a=b=5.5170 Å and c=11.0600 Å). FESEM image of the CuInS 2 powder synthesized at 15 cycles are shown in figure 3(a). The CuInS 2 powder is composed of nanoparticles with diameters in a range of 30-65 nm, and this is consistent with the TEM image presented in figure 3(b). The corresponding high-magnification TEM image ( figure 3(c)) shows a lattice fringe with a spacing of 0.20 nm, which corresponds to the (204) plane of the tetragonal phase of CuInS 2 . The SAED pattern ( figure 3(d)) shows bright concentric rings that are indicative of the tetragonal CuInS 2 phase, which corresponds well with the XRD pattern shown in figure 1. Moreover, the EDX spectrum of CuInS 2 (figure 2(e)) appears the characteristic signals of Cu, In, S, Al, Au and C elements. It should be noted that the signals of the Al, Au and C elements come from the aluminium supporting holder, coated gold, and environment, respectively. The results obtained from the SAED and EDX analyses further indicate the purity of the CuInS 2 crystals. Analyzed by the degree of BET adsorption-desorption of nitrogen gas at the temperature of liquid nitrogen, the specific surface area of the CuInS 2 nanoparticles is recorded at 31.64 m 2 g −1 .
UV-vis DRS spectrum of the CuInS 2 powder synthesized at 600 W for 15 cycles is shown in figure 4(a). The band gap energy of CuInS 2 can be calculated by the plot of transformed Kubelka−Munk function (F(R).hν) 2 versus photon energy hν [41], where R represents the proportion of light reflected, h is the Planck's constant and ν is the photon frequency. As illustrated in the inset of figure 4, the band gap energy of CuInS 2 was calculated as 1.40 eV. This implies that the CuInS 2 nanoparticles have an ability to absorb visible light, suggesting a promising use of this nanomaterial as a visible light active photocatalyst. To determine the flat-band potential (E f ) of CuInS 2 , Mott-Schottky measurement was performed at a frequency of 1000 Hz and potential ranged from −1.2 to +1.2 V (versus Ag/AgCl). Figure 4(b) shows the negative slope of the Mott-Schottky plot of the CuInS 2 electrode, indicating that CuInS 2 is a p-type semiconductor. In addition, the flat band potential (E f ) of CuInS 2 is −0.10 V (versus Ag/AgCl), which corresponds to 0.11 V (versus NHE). Generally, the flat band potential of the p-type semiconductors is about 0.1 V more negative than the valence band (VB) potential [42]. Therefore, the VB potential of CuInS 2 was estimated to be 0.21 V versus NHE. Consequently, based on its E g ( figure 4(a)), the conduction band (CB) potential was calculated to be −1.19 V versus NHE. Notably, the Mott-Schottky plot was also carried out at a frequency of 100 Hz on the same electrochemical analyzer. The Mott-Schottky plot shown in the supporting information (figure S2) reveals the negative slope with the flat band potential (E f ) of −0.15 V (versus Ag/AgCl), which is close to the E f value obtained from the Mott-Schottky plot at 1000 Hz.
Decolorization efficiencies of RhB, MB and MO with and without the CuInS 2 photocatalyst are shown in figure 5(a). In the absence of the CuInS 2 photocatalyst, the photolysis of RhB, MB, and MO dyes by visible light is  only 7.67, 2.96, and 11.94%, respectively, indicating that the self-photodegradations of these dyes are very slow and negligible under visible light. In the presence of the CuInS 2 photocatalyst, the CuInS 2 photocatalyst degrades MB and RhB by up to 64.08% and 74.29% within 180 min, respectively, and the decolorizations of MB and RhB are almost completed when the irradiation is continued up to 360 min (91.21% for MB and 97.80% for RhB). Surprisingly, only 12.96% of MO is decolorized, which is found to be quite close to its level of photolysis suggesting the high selectivity of the CuInS 2 photocatalyst toward cationic dyes. The changes in the absorption intensity of RhB, MB and MO dyes that occur as a function of time in the presence of the CuInS 2 photocatalyst under visible light irradiation are shown in figures 5(b)-(d), respectively. Under visible light irradiation, two competitive processes take place simultaneously during the photocatalytic degradations of the dyes; N-dealkylation and destruction of the chromophore structure. The absorption intensity of the peaks of the dyes gradually decreases upon increasing irradiation time without detection of any new peaks and shifts of the absorption maximum. This implies that the dyes are continuously degraded during the photocatalytic processes through the chromophore destruction [43][44][45].
To elucidate the surface charge of photocatalyst nanoparticles, the zeta potential of the CuInS 2 photocatalyst at different pH solutions was measured for an isoelectric point determination. As is shown in figure 6, the pH value at the isoelectric point, pH (IEP) , of the CuInS 2 photocatalyst is 3.27. Therefore, under our testing conditions in which the pH values of the RhB, MB and MO solutions in the presence of CuInS 2 are 3.86, 3.95, and 3.85, respectively, the CuInS 2 photocatalyst has a negative charge on its surface. As a result, the cationic dyes (RhB and MB) could be easily attached on the negatively charged surface of the CuInS 2 photocatalyst due to the electrostatic attraction force between them, thereby could promote photodegradation activity more effectively than the anionic dye (MO). The kinetics of the RhB and MB degradations by the CuInS 2 photocatalyst followed the pseudo-first-order model as is shown in the following equation [46]; where C 0 is the initial concentration of dye and C is the concentration at the irradiation time (t). As is shown in figure 7(a), the initial rate constants (k) with the coefficient of determination (R 2 ), calculated from the initial slope of the fitted line between ln(C 0 /C) and irradiation time (0-180 min), are 1.24× 10 −2 min −1 for RhB with R 2 =0.96, and 0.59× 10 −3 min −1 for MB with R 2 =0.99. These results confirm that the degradation of RhB is faster than that of MB. Under the same photocatalytic degradation pathway of RhB and MB, the faster photocatalytic degradation of RhB is probably due to the different frontier molecular orbital energies of RhB and MB [47]. The lowest unoccupied molecular orbitals (LUMOs) energy level of RhB is more positive than that of MB. Therefore, the higher the LUMOs energy level is favorable for the excited electrons to transfer from the LUMOs energy level of the RhB molecules to the conduction band of the semiconductor and thus, facilitate the photodegradation reaction process [48].   Figure 7(b) shows the quenching experiments used to investigate the active species involved in the photodegradation process of RhB over the CuInS 2 photocatalyst. Disodium ethylenediaminetetraacetate (EDTA-2Na, 1 mM), potassium dichromate (K 2 Cr 2 O 7 , 1 mM), dimethyl sulfoxide (DMSO, 1 mM), and ascorbic acid (AA, 1 mM) were used to trap the hole (h + ), electron (e − ), hydroxyl radical (·OH) and superoxide radical (· -O 2 ), respectively. The additions of these quenchers decrease the kinetic rates of RhB, implying that h + , e − , ·OH and · -O 2 are the active species in the photodecolorization of RhB. Based on the experimental results, the energy band positions and possible photocatalytic degradation mechanism have been proposed in figure 7(c). After the CuInS 2 photocatalyst is exposed to visible light, electrons (e − ) in the valence band (VB) of the CuInS 2 photocatalyst are excited to the conduction band (CB) leaving holes (h + ) in the VB. The photogenerated e − and h + then migrate to the surface of the photocatalyst in order to undergo photocatalytic reactions. The e − accumulated at CB of CuInS 2 efficiently reduce dissolved O 2 in the solution to · -O 2 radical because the CB potential of the CuInS 2 (−1.19 V) is more negative than the reduction potential of O 2 /· -O 2 (−0.33 V versus NHE). In addition, since the reduction potential of O 2 /H 2 O 2 is 0.695 V (versus NHE), the e − can react with O 2 and H + to produce H 2 O 2 which would then further produce ·OH radicals in the photocatalytic reaction [49][50][51][52]. These explanations correspond with the significant decreases that take place in the kinetic rate of RhB after the addition of K 2 Cr 2 O 7 quencher into the reaction solution. The h + that occurred on the VB of CuInS 2 cannot oxidize the H 2 O molecules or OH − to produce ·OH radicals since the VB potential of CuInS 2 (+0.21 V) is lower than the standard potentials of ·OH,H + /H 2 O (2.72 V versus NHE) and ·OH/OH − (1.99 V versus NHE). Therefore, ·OH radicals can be generated by further reduction of · -O 2 radicals, as has been mentioned above. Finally, the · -O 2 and ·OH radicals oxidize the dye molecules to yield the final products such as CO 2 , H 2 O and etc [50]. Simultaneously, the h + in the VB of CuInS 2 can directly degrade the dye molecules. The mechanism for the degradation of dye molecules via · -O , 2 ·OH, or the direct h + oxidation pathway is given by the following equations;

Conclusions
Pure and highly crystalline CuInS 2 nanoparticles were successfully synthesized through a simple and rapid microwave heating process using sodium sulfide as a sulfur source, which only required 7.5 min of microwave irradiating. This photocatalyst showed higher photocatalytic activity on the cationic dyes (RhB and MB) than on the anionic dye (MO), which was attributed to the electrostatic attraction between the cationic dye molecules and the negative charge on the photocatalyst's surface. The quenching tests indicated that h + , e − · -O 2 and ·OH were responsible for the photodegradation process. The results suggested that the CuInS 2 nanoparticles could be a potential photocatalyst for the degradation of organic pollutants in wastewater under visible light. In addition, this synthesis method is expected to be applicable to the preparation of other ternary chalcogenide nanomaterials.