1082 Acta Chim. Slov. 2020, 67, 1082–1091 Tsvetkov et al.: Visible Light-Driven Photocatalytic Activity ... DOI: 10.17344/acsi.2020.5857 Scientific paper Visible Light-Driven Photocatalytic Activity of Magnetic Recoverable Ternary ZnFe2O4/rGO/g-C3N4 Nanocomposites Martin Tsvetkov,1,* Elzhana Encheva,1 Albin Pintar,2 and Maria Milanova1 1 Department of Inorganic Chemistry, Faculty of Chemistry and Pharmacy, St. Kl. Ohridski University of Sofia, J. Bourchier 1, 1164 Sofia, Bulgaria 2 Laboratory for Environmental Sciences and Engineering, National Institute of Chemistry, Hajdrihova 19, 1001 Ljubljana, Slovenia * Corresponding author: E-mail: mptsvetkov@gmail.com Received: 01-25-2020 Abstract ZnFe2O4/rGO/g-C3N4 ternary nanocomposite photocatalysts with different ZnFe2O4/g-C3N4 weight ratio (0.5, 0.75, 1) were prepared by a stepwise solvothermal method using ethylene glycol as the solvent. Physicochemical methods such as X-ray diffraction, UV-Vis diffuse reflectance spectroscopy and photoluminescence spectroscopy were applied in order to characterize the composites. The formation of a meso-/macroporous structure with specific surface area between 67 and 77 m2 g–1 was confirmed by N2 adsorption/desorption. The bandgap of the composites was found to be lower (2.30 eV) than that of g-C3N4 (2.7 eV). In contrast to pure g-C3N4, the composites showed no fluorescence, i.e. no recombination of e–/h+ took place. All samples, including pure g-C3N4 and ZnFe2O4, were tested for adsorption and photocatalytic deg- radation of aqueous malachite green model solutions (10–5 M) under visible light irradiation (λ > 400 nm). The results show that the prepared nanocomposites have higher absorption and photocatalytic activity than the pristine g-C3N4 and ZnFe2O4 and can be successfully used for water purification from organic azo-dyes. Keywords: Graphitic carbon nitride; reduced graphene oxide; zinc ferrite; photocatalysis; malachite green decomposition. 1. Introduction Synthetic organic dyes are severe water pollutants causing environmental problems. They are typically aro- matic compounds with structural variations, many of them resistant to degradation.1 Among them, malachite green is an organic water pollutant known to be harmful for living creatures because of its potential carcinogenicity, mutagen- icity and teratogenicity in mammals.2 Depending on the polluted water composition, different methods have been applied in order to solve water contamination problems, in- cluding biological reactions,3 sedimentation,4–6 coagula- tion,7,8 adsorption,5 reverse osmosis,9,10 membrane filtra- tion,11 ion exchange,12 etc. Photocatalytic processes have also been applied, and much effort has been spent on the development of different semiconductors as photocatalysts among which TiO2 and its modifications is well known.13-18 The aim of the research presented here is to develop new photocatalysts and to overcome the main limitation of TiO2, i.e. its wide band gap (3.2 eV) that makes it active under UV light irradiation only (about 5% of sunlight). Recently the graphite analogue, graphitic carbon nitride, g-C3N4, raised interest due to its unique electronic struc- ture. It is a non-metallic polymer with n-type semicon- ducting behavior and unique electrical, optical, structural and physicochemical properties. Like graphite, g-C3N4 has a two-dimensional planar π conjugation structure, able to enhance the electron transfer processes due to its excellent electronic conductivity.19 With its medium-sized band gap and its thermal and chemical stability in ambient environ- ment, it has become one of the most promising photocata- lytic materials.19 The interest in its application as a photo- catalyst increased after its photocatalytic properties were discovered by Wang et al..20 However, g-C3N4 also has some disadvantages such as a small specific surface area, a small number of active centers, quick recombination of the photo-induced e–/h+, low mobility of photoinduced e–/ h+21 and an wide band gap (2.7 eV).22 These shortcomings can be avoided by adding a co-catalyst to g-C3N4 to pre- 1083Acta Chim. Slov. 2020, 67, 1082–1091 Tsvetkov et al.: Visible Light-Driven Photocatalytic Activity ... pare nanocomposites. In recent years, particular interest has appeared in composites of g-C3N4 and reduced graph- ite oxide, rGO, due to the large specific surface area of rGO and its ability to efficiently separate photo-induced charges.23 The above effect can also be achieved by com- bining g-C3N4 with multi-wall carbon nanotubes.24 How- ever, these nanocomposites can only solve two of the above disadvantages of g-C3N4 as a photocatalyst. The wide band gap of 2.7 eV limits the application of visible light. In order to use g-C3N4 as a photocatalyst with visible light, it may be combined with other semiconductor materials with a nar- rower forbidden zone. Due to its narrow band gap of 1.9 eV, ZnFe2O4 as a co-catalyst can absorb a wider range of visible light wavelengths. It may also show improved separation of photogenerated electron-hole pairs. Its magnetic properties facilitate the removal of the composites from the reaction mixture, so they can be reused.25 In the literature available, there are publications presenting studies on triple nano- composites such as CoMoS2/rGO/C3N4 with visible light photocatalytic activity for hydrogen evolution,26 C3N4/ rGO/TiO227,28 for decomposition of methyl orange, rhodamine B, and phenol under visible light, and C3N4/ rGO/WO3 for degradation of methylene blue.29 Both the studies mentioned and our experience with ZnFe2O4 as photocatalyst suggested that ZnFe2O4/rGO/g-C3N4 would be a promising composition to study, as it potentially could overcome g-C3N4 shortcomings as photocatalyst as well as take advantage of rGO for the separation of photo-induced charges. The photocatalytic properties of the nanocompos- ites ZnFe2O4/rGO/g-C3N4 were tested for removal of mala- chite green as representative pollutant under visible light irradiation, showing better activity than the individual semiconductors. The work presented here on the prepara- tion and the properties of ZnFe2O4/rGO/g-C3N4 can con- tribute both to the knowledge of inorganic synthesis of such composites and to the improved photocatalytic removal of organic dyes from water. 2. Experimental 2. 1. Materials Chemicаls such аs urea (puriss. p.а., Flukа, Switzer- land), graphitе, Zn(NO3)2 · 6H2O, Fe(NO3)3 · 9H2O, аnd CH3COONa · H2O (all p.а., Sigma-Aldrich, USA) were used in this study. 2. 2. Synthesis of the Sаmples 2. 2. 1. Synthesis of Grаphitic Cаrbon Nitride, g-C3N4 Thermаl polycondensаtion of ureа in a closed cruci- ble at 550 °C for 5 h wаs аpplied. The powder wаs dispersed in wаter аnd homogenized by stirring for 1 h, fоllоwed by filtering, washing and drying at 50 °C оvernight. The suc- cessful synthesis was cоnfirmed by XRD and TEM analyses. 2. 2. 2. Synthesis of Reduced Graphene Oxide, rGO Graphene оxide was prepared by using the mоdified Hummer’s methоd starting frоm graphite flakes.30 In a typical prоcedure, 0.5 g of graphite was dispersed in 50 mL mixture of cоnc. H2SO4 and cоnc. H3PO4 (vоlume ratiо 9:1) and then ultrasоnicated for 1 hour. After that, 6 g of KMnO4 was added and magnetically stirred fоr 5 h fоllоwed by 12 h stirring at 50 °C. The sо prepared mixture was cооled to roоm temperature and transferred in a bea- ker containing 100 g of ice. After stirring and melting of the ice, 20 mL оf 30% H2O2 sоlution was added dropwise in оrder tо remоve the unreacted KMnO4. The suspension immediately changed its color from purple to yellow, indi- cating the formation of graphene oxide. The solid phase was separated by filtration and then dispersed in 100 mL of 5% HCl solution in order to remove all the metal cations and then separated again by centrifugatiоn and washing with water untill a pH = 7. The GO obtained was reduced further tо rGO by hydrоthermal treatment in a PTFE-lined autоclave at 180 °C for 12 h using hydrazine as a reducing agent. 2. 2. 3. Synthesis оf the Cоmposites ZnFe2O4/ rGO/g-C3N4 A sоlvоthermal methоd was used tо prepare the cоmposites. The metal salts Zn(NO3)2 · 6H2O and Fe(- NO3)3 · 9H2O were dissоlved in 50 mL оf ethylene glycоl, EG, with ratio n(Zn2+):n(Fe3+) = 1:2. The rGO was added and dispersed by 30 min magnetic stirring and 2 h of son- ication. After g-C3N4 was added, the suspension was stirred for 30 min by magnetic stirring, fоllоwed by 30 min sоnicatiоn in an ultrasonic bath. After adding 3 g of CH3COONa · 2H2O and stirring for 30 min, the metal ions were precipitated. The mixture was transferred tо a 75 mL PTFE autоclave and kept at 180 °C for 24 h. By varying the ZnFe2O4/g-C3N4 mass ratiо (0.5, 0.75, 1), three ZnFe2O4/rGO/g-C3N4 cоmposites cоntaining 5 wt% rGO were prepared. They are mentioned further in the text as CN50 (ZnFe2O4 : g-C3N4 = 0.5), CN75 (ZnFe2O4 : g-C3N4 = 0.75), and CN100 (ZnFe2O4 : g-C3N4 = 1). 2. 3. Methоds for Characterization оf the Samples X-Ray Diffraction to determine the crystal structure оf the materials was performed using a PANalytical Em- pyrean X-ray diffractometer in the 2θ range of 15–80° using CuKα radiation (λ = 0.15405 nm) for the nano- composites and in 2θ range of 10–80° for the individual cоmponents, steps of 0.01° and 20 secоnds exposure time at each step. The average crystallite size was calculated using the well-known Scherrer’s equation.31 The micrо- structural infоrmation of the ZnFe2O4 was extracted by 1084 Acta Chim. Slov. 2020, 67, 1082–1091 Tsvetkov et al.: Visible Light-Driven Photocatalytic Activity ... full prоfile Rietveld methоd using the FullPrоf Suite soft- ware.32 UV-Vis absоrption spectroscopy was applied using an Evolution 300 UV-Vis spectrometer (Thermo Scientif- ic) for measuring the absorption of the samples in the range of 200-900 nm. Bandgap energies were calculated from the UV-Vis absоrption spectra in the range frоm 200 to 400 nm accоrding to Tauc’s equation αhν = A(hν − Eg)n/2, where A is a cоnstant independent of hν, Eg is the semicоnductor bandgap and n depends оn the type of transitiоn.33 Textural characteristics such as specific sur- face area, tоtal pоre vоlume, and pоre size distributiоn were determined at –196 °C using a TriStar II 3020 appa- ratus (Micromeritics). The tоtal pоre vоlume was esti- mated at a relative pressure P/P0 0.989. Transmission elec- trоn micrоscоpy (ТЕМ): a JEOL JEM 2100 micrоscоpe was used at 200 kV and up tо 100k magnificatiоn fоr characterizatiоn of the mоrpholоgy of the samples. Parti- cle size distribution analysis was performed by using Im- ageJ software.34 2. 4. Photocatalytic Tests The phоtоcatalytic tests were perfоrmed using a slurry of 0.5 g catalyst L–1 and a 10–5 M aqueous sоlutiоn of Malachite Green оxalate (MG), (Chroma GmbH) as a mоdel pollutant. The equipment and the procedure ap- plied were similar with those used in our previous stud- ies.14–18,35,36 For illumination 15W white LED (manufac- tured by V-TAC), 418–700 nm, situated at 10 cm distance abоve the slurry was used. 3. Results and Discussion 3. 1. Characterization of the Samples 3. 1. 1. Characterization of the Phase Composition by X-ray Diffraction The phase composition, cell parameters and crystallite size of the samples were determined using the X-ray diffrac- Figure 1. XRD patterns, from bottom to top (a) of the initial samples r-GO, g-C3N4, ZnFe2O4 and (b) of the nanocomposites CN50, CN75, CN100 Figure 2. Experimentally observed (dots), Rietveld calculated (continuous line) and difference (continuous bottom line) profiles, obtained after Ri- etveld analysis of the XRD data (a) ZnFe2O4 and (b) the composite CN100. Peak positions are shown at the base line as small markers. a) b) a) b) 1085Acta Chim. Slov. 2020, 67, 1082–1091 Tsvetkov et al.: Visible Light-Driven Photocatalytic Activity ... tion analysis. The composition of the initial substances r-GO, g-C3N4 and ZnFe2O4 were confirmed (Fig. 1, a). The strong diffraction peak observed at 27° 2theta in the pure g-C3N4 can be assigned to the (002) diffraction plane of lay- ered g-C3N4 (JCPDS 87-1526) (Fig. 1, a). It corresponds to the characteristic interlayer stacking оf arоmatic segments.37 The nanocomposites mainly show the presence of ZnFe2O4 (Fig. 1, b). The presence of g-C3N4 in the composites is de- tected below 30° 2theta shown by an inversed “Δ” (Fig. 1, b). Rietveld analysis of the XRD data of ZnFe2O4 and the cоmposite CN100 was performed (Fig. 2, a, b). The in- formation obtained by the Rietveld refinement was the crystallite size and microstrains as both are related (to ex- tend) to the catalytic properties of the materials.38,39 This is also proven by the BET measurements as it can be seen later in the text (Part 3.1.3). On the other side, the microstrains are related to the density of the defects in the crystal structure. The defects, known as active centres in catalysis, are places (especially on the surface) with low- er potential energy where the reaction between solid/liq- uid (or solid/vapor) occurs. Although this is not always true in context of photocatalysis as defects can also act as recombination centres for photogenerated e–/h+ pairs leading to lower activity. The lattice parameters, crystallite size, and the mi- crоstrain оf the cоmposites and the pure ZnFe2O4 are shown in Table 1. It can be seen that the increasing cоntent of ZnFe2O4 in the composites is causing changes in all the parameters mentioned. All the values are getting closer to those of the pure ZnFe2O4. The microstrain is decreasing in the line 0.0177 (CN50), 0.0147(CN75), 0.0116 (CN100), and 0.0053 for the pure ZnFe2O4, respectively. The struc- ture of the composites is more defective at lower zinc fer- rite content. The latter can be observed in the reductiоn оf the unit cell volume as a result of the micrоstrains. 3. 1. 2. Characterization of the Sample Morphology by TE The mоrphology and the structure оf as-synthesized samples оbserved by TEM are shоwn in Figure 3. The lay- ered structure of the individual g-C3N4 can be seen in Fig. 3, a. The ZnFe2O4 particles are flower-shaped on the sur- face of g-C3N4 (Fig. 3, b). The electron diffraction of the samples g-C3N4 and ZnFe2O4 is shown in Fig. 3, c, d, re- spectively. It is used as supplementary analysis to the XRD and approves the successful preparation of ZnFe2O4 and g-C3N4. With increasing ZnFe2O4 content in the composites, polydispersed agglomerates are formed. The particle size distribution for CN50 is between 5–10 nm (Fig. 3, e), in accordance with the XRD data. 3. 1. 3. Textural Characterization Nitrogen adsorption – desorption isotherms mea- sured at −196 °C on powdered samples (Fig. 4, a) showed that the samples are of type IV, which is the typical charac- teristic of mesoporous materials according to the IUPAC classification.40 The isotherm of ZnFe2O4 with Н1 loop is typical for well-defined cylindrical pores or agglomerates of approximately uniform spheres (Fig. 4, a). The Н3 loop for the g-C3N4 and the composites are distinctive for non-rigid aggregates of plate-like particles with slit-shaped pores. The hysteresis loops observed are characteristic of mesoporous solids and their shape exhibits a change in the pore structure. Macropores may be present as well, based on the shape of the hysteresis loops near P/P0 = 1.41 The average pore size is rather close for the samples g-C3N4, CN50 and CN75 (Table 2), while that of ZnFe2O4 is larger and that of CN100 smaller. The composites show a maxi- mum in the pore size distribution at about 25–50 nm, Table 2. Textural characteristics of the samples studied. Sample Specific surface Total pore volume, Average pore size, area, SBET, m2 g–1 Vtotal, cm3 g–1 Daverage, nm g-C3N4 88 0.47 22 ZnFe2O4 34 0.27 32 CN50 72 0.44 25 CN75 77 0.44 23 CN100 67 0.23 14 Table 1. Lattice parameters, crystallite size and microstrain; CN50 (ZnFe2O4 : g-C3N4 = 0.5), CN75 (ZnFe2O4 : g-C3N4 = 0.75), and CN100 (ZnFe2O4 : g-C3N4 = 1). Sample Unit cell, Å Crystallite size, nm Microstrain, % Rwp, % χ2 CN50 8.4312 ± 0.0007 8 ± 0.2 1.77 9.1 1.74 CN75 8.4321 ± 0.0005 9 ± 0.4 1.47 8.4 1.71 CN100 8.4325 ± 0.0003 11 ± 0.4 1.16 7.9 1.68 ZnFe2O4 8.4342 ± 0.0002 23 ± 0.3 0.53 7.2 1.53 1086 Acta Chim. Slov. 2020, 67, 1082–1091 Tsvetkov et al.: Visible Light-Driven Photocatalytic Activity ... while CN100 shows a broad polydispersed pore size distri- bution (Fig. 4, b). The pure g-C3N4 sample has the largest specific sur- face area, 88 m2 g–1, while pure ZnFe2O4 with 34 m2 g–1 has the lowest one among the samples tested (Table 2). In spite of the statement that g-C3N4 exhibits low SBET,42 88 m2 g–1 is a reasonably good value, comparing for example with 9.6 m2 g–1 reported in ref.42 for g-C3N4 obtained by the same hydrothermal method for 48 h/180 °C (12 h/180 °C in present work). Apparently, the duration of the hydro- thermal treatment is influencing the agglomeration of the sample. The addition of rGO and ZnFe2O4 caused a reduc- tion of the specific surface area leading to composites with 77, 72, and 67 m2 g–1 surface area, which could be due to Figure 3. TEM micrographs of (a) g-C3N4 and (b) CN50 presented along with the electron diffraction (c) and (d), respectively. The particles size distribution for CN50 is shown in (e). a) b) c) d) e) 1087Acta Chim. Slov. 2020, 67, 1082–1091 Tsvetkov et al.: Visible Light-Driven Photocatalytic Activity ... their deposition on the pores of carbon nitride. Quite like- ly the presence of g-C3N4 inhibits the agglomeration of ZnFe2O4 particles and makes them uniformly dispersed. 3. 2. Optical and Photocatalytic Properties 3. 2. 1. Optical Properties The UV/Vis spectra of g-C3N4, ZnFe2O4 and the compоsites are presented in Fig. 5, a, clearly shоwing en- hanced light absоrption оf the composites, prоbably due tо interfacial interactiоn between g-C3N4 and ZnFe2O4.43 It can be expected that the enhanced light absоrption cоuld lead tо higher phоtоcatalytic activity by generating mоre photоinduced charge carriers under visible light. Based on these UV/Vis spectra, the band gap energy was calculated for all the samples (Fig. 5, b). The values for the similar band gaps of the composites with energy of 2.30–2.31 eV (538–536 nm), between the values of g-C3N4, 2.7 eV (458 a) b) Figure 4. Adsorption-desorption isotherms of the pure g-C3N4 and ZnFe2O4, and the composites CN50, CN75, CN100 (а) and BJH pore diameter distribution, determined from the desorption branch of the isotherm (b); V- pore volume, D – pore diameter. Figure 5. (a) UV/Vis spectra and (b) the energy of the forbidden zone, Eg, for all the samples studied a) b) Figure 6. Typical photoluminescence of g-C3N4, compared with the absence for CN50. 1088 Acta Chim. Slov. 2020, 67, 1082–1091 Tsvetkov et al.: Visible Light-Driven Photocatalytic Activity ... nm) and ZnFe2O4, 2.06 eV (600 nm), confirm their pro- spective for photocatalytic activity higher than that of g-C3N4. Such a prospective is also indicated by the absence of fluorescence in the composites, which provides evidence for efficient inhibition of radiative recombination of pho- togenerated e–/h+ (Fig. 6). The strong fluorescence of the pure g-C3N4 related to strong e–/h+ recombination (Fig. 6), may explain the low photocatalytic activity of the pristine sample. 3. 2. 2. Degradation of Malachite Green Under Visible Light Irradiation The photocatalytic performance of the samples for degradation of malachite green under visible light illumi- nation is shown in Fig. 7. In the given range of reaction conditions, adsorption of malachite green on the catalyst surface cannot be neglected (Table 3). However, this was well considered in the subsequent interpretation of col- lected experimental data. The relevant data for the rate constants are summarized in Table 3. The rate constant obtained for the photolysis was 0.6 × 10–3 min–1. The pure samples g-C3N4 and ZnFe2O4 showed low values for their rate constants: 2.9 and 4.6 × 10–3 min–1, respectively. The rate constants of the compos- ites were higher, and showed increasing values with in- creasing ZnFe2O4/g-C3N4 ratio (0.5, 0.75, 1), i.e. 4.0 × 10–3, 5.1 × 10–3 and 7.7 × 10–3 min–1, respectively. Apparently, ZnFe2O4 and g-C3N4 show a synergetic effect, which is best demonstrated for the composite CN100. The highest degradation of malachite green achieved was 63 % for 150 min illumination with visible light. In Table 3, the data for the ratio k/SBET (min–1 g m–2) are presented, showing the best activity for ZnFe2O4, followed closely by the compos- ite CN100. The observed photocatalytic activity may be cor- related to the physical properties of the catalysts, such as: (i) Surface area: the largest SBET surface area of g-C3N4 among the samples tested could provide more active sites to adsorb and convert MG molecules in comparison with the ZnFe2O4 and the composites. However, this is not ob- served; g-C3N4 may be less active than expected because of its strong e–/h+ recombination shown by the fluorescence (Fig. 6). Among the composites, CN100 has less than aver- age SBET but showed the best photocatalytic activity. (ii) Large pore volume: it would favor the diffusion of MG molecules within the pores towards the active sites on the surface of the photocatalysts. However, g-C3N4 with the largest pore volume shows the lowest activity. (iii) Pore size distribution: The composite CN100 has a very broad pore size distribution showing best activity i.e. positively influ- encing the activity (Fig. 4, b). (iv) Bandgap energy: among all samples tested, the composites have the lowest and equal value for Eg but show different activity. Thus the band gap energy alone cannot explain all differences; the activity is determined by a combination of factors. From this it can be concluded that the most active composite CN100 has an optimum combination of band gap value, ZnFe2O4/g-C3N4 ratio and absence of e–/h+ recombina- tion. The rGO, being present in equal amounts for all the composites, has the function of solid-state electron medi- atоr,28, 29 adsorbent, photosensitizer and electron accep- tor.28 For the discussion of the mechanism of the photo- catalytic reaction, the values of the band edges i.e. the po- tentials of the current band (CB) and the valence band (VB) of the semiconductors ZnFe2O4 and g-C3N4 should be considered. Some of the literature data are summarized Figure 7. Photocatalytic performance of as-prepared samples for degradation of malachite green under visible-light illumination. Table 3. Rate constants and extent of malachite green removal based on adsorption on the catalyst surface and degradation. Sample Rate constant, Rate constant to SBET, Adsorption, Degradation after 150 min, × 10–3 min–1 × 10–4, min–1 g m–2 % % g-C3N4 2.9 0.395 47 35 ZnFe2O4 4.6 1.353 63 44 CN50 4.0 0.556 76 41 CN75 5.1 0.66 78 48 CN100 7.7 1.149 86 63 1089Acta Chim. Slov. 2020, 67, 1082–1091 Tsvetkov et al.: Visible Light-Driven Photocatalytic Activity ... and presented in Table 4 along with data from our study. It can be seen that in the literature for g-C3N4 similar values were reported, i.e. CB -1.03 and VB 1.64 eV44 as well as –1.26 and 1.34 eV42. This is in good agreement with the value for the VB of g-C3N4 (1.54 eV) determined by X-ray photoelectron spectroscopy.45 The literature data for Zn- Fe2O4 are less consistent: values observed include –0.06 and 1.8 eV44 as well as -1.54 eV and 0.38 eV.42 Taking into account the literature data for the current and valence band it should be mentioned that the CB and VB values for ZnFe2O4 are lying over those for g-C3N4 according to ref.44 but under values of g-C3N4 according to ref.42 i.e. inconsis- tency in the data is observed. This can lead to a different way of the interpretation of energy transfer during the photocatalytic process, particularly the migration of elec- trons and holes between the current band and the valence band of the semiconductors ZnFe2O4 and g-C3N4. For the samples synthesized ZnFe2O4 and g-C3N4 the band edge pоsitiоns were evaluated applying the sim- ple equatiоns EVB = X–E0+0.5Eg and ECB = EVB–Eg. The symbols used ECB, EVB, and X are showing the pоtentials of the conduction band, of the valence band and the electro- negativity of the semiconductors ZnFe2O4 or g-C3N4 de- fined as the geometric average of the absolute electronega- tivity of the constituent atoms.46 According the literature data the energy of the free electrons on the hydrogen scale E0 is about 4.5 eV.46 For the semiconductors ZnFe2O4 and g-C3N4 the X values were calculated to be 5.82 and 4.73 eV, respectively. Following this, the bottom of current band and the top of valence band were calculated to be –1.08 eV and 1.54 eV for g-C3N4, and 0.29 eV and 2.35 eV for Zn- Fe2O4, respectively (Table 4). The data for ZnFe2O4 are in good agreement with data in ref.47 in spite of the different synthetic method used, influencing the value. Based on these results, a mechanism for photodegra- dation of MG over ZnFe2O4/r-GO/g-C3N4 composites can be proposed (Fig. 8). When ZnFe2O4/r-GO/g-C3N4 com- posites are exposed to visible light, both ZnFe2O4 and g-C3N4 are excited. The photogenerated holes and elec- trons are in the valence band and conduction band, re- spectively. g-C3N4 can effectively absorb visible light to form photoexcited charge carriers. Because the current band of g-C3N4 is more negative than that of ZnFe2O4, the electrons migrate into the current band of ZnFe2O4; holes in the valence band of ZnFe2O4 si- multaneously migrate to the VB of g-C3N4. By this the pho- togenerated electrons are accumulated on ZnFe2O4 and holes accumulated on g-C3N4. This in turn with water-dis- solved oxygen and adsorbed water molecules causes the for- mation of radicals. These are well known as oxidizing spe- cies and as a result MG degradation takes place. The rGO is improving the photocatalytic properties of the composites obtained by efficient separation of photo-induced charges.23 4. Conclusions Nanocomposites of the type ZnFe2O4/r-GO/g-C3N4, based on coupling of two semiconductors, were success- fully prepared by applying solvothermal synthesis, where ethylene glycol was used as a solvent. All of the compos- ites, including the stand alone components, were tested and showed activity for photocatalytic degradation of mal- achite green in aqueous solution under visible light irradi- ation. The composites show better activity than the pris- tine g-C3N4 and ZnFe2O4, with the CN100 sample in which g-C3N4 and ZnFe2O4 were present in equal amount showing the highest activity. The improved photocatalytic activity was due to the synergy and the charge transfer be- tween g-C3N4 and ZnFe2O4 as well as the efficient separa- Table 4. Potentials of current band (CB) and valence band (VB) of ZnFe2O4 and g-C3N4. ZnFe2O4 g-C3N4 № CB, VB, Eg, CB, VB, Eg, Ref. eV eV eV eV eV eV 1 0.29 2.35 2.06 –1.08 1.54 2.62 Present work 2 0.41 2.38 1.97 – – – 45 3 –0.06 1.8 1.76 –1.03 1.64 2.67 42 4 –1.54 0.38 1.92 –1.26 1.34 2.60 40 5 – – – – 1.54 – 43 Figure 8. Illustration of the mechanism of the photocatalytic activ- ity of as prepared ZnFe2O4/GO/g-C3N4 samples. 1090 Acta Chim. Slov. 2020, 67, 1082–1091 Tsvetkov et al.: Visible Light-Driven Photocatalytic Activity ... tion of photo-induced charges by rGO. More research has to be done to find the optimum ZnFe2O4/g-C3N4 ratio. The examined composites show potential for degradation of water-dissolved organic pollutants. Acknowledgements. M.T. acknowledges the financial support from the program “Young scientists and Postdoctoral candidates” of the Bulgarian Ministry of Education and Science, MCD № 577/2018. The textural measurements were done in Prof. Albin Pintar’s Laboratory, National Institute for Chemis- try, Slovenia. 5. References 1. J. Longbo, Y. Xingzhong, P. Yang, L. Jie, Z. Guangming, W. Zhibin, W. Hou, Appl. Catal. B-Environ. 2017, 217, 388–406. 2. E. Sudova, J. Machova, Z. Svobodova, T. Vesely, Veterinarni Med. 2007, 52, 527–539. DOI:10.17221/2027-VETMED 3. N. J. Karrer, G. Ryhiner, E. Heinzle, Water Res. 1997, 31, 1013–1020. DOI:10.1016/S0043-1354(96)00356-9 4. L. Ying-yu, L. Lin, L. Xiao-yan, J. Clean. Prod. 2020, 244, 118705. DOI:10.1016/j.jclepro.2019.118705 5. L. Boutilier, R. Jamieson, R. Gordon, C. Lake, W. Harte, Water Res. 2009, 43, 4370–4380. DOI:10.1016/j.watres.2009.06.039 6. S. Bhattacharjee, S. Datta, C. Bhattacharjee, Desalination 2007, 212, 92–102. DOI:10.1016/j.desal.2006.08.014 7. X. Zi-Peng, S. De-Zhi, J. Hazard. Mater. 2009, 168, 1264–1268. DOI:10.1016/j.jhazmat.2009.03.008 8. S. Zodi, O. Potier, F. Lapicque, J.-P. Leclerc, Desalination 2010, 261, 186–190. DOI:10.1016/j.desal.2010.04.024 9. S. S. Mansouri, I. A. Udugama, A. Mitic, A. Rubin, L. Ru- dolfsson, K. V. Gernaey, Comput. Aided Chem. Eng. 2017, 40, 391–396. DOI:10.1016/B978-0-444-63965-3.50067-2 10. A. F. S. Foureaux, E. O. Reis, Y. Lebron, V. Moreira, L. V. San- tos, M. S. Amaral, L. C. Lange, Sep. Purif. Technol. 2019, 212, 171–179. DOI:10.1016/j.seppur.2018.11.018 11. W. Pronk, A. Ding, E. Morgenroth, N. Derlon, P. Desmond, M. Burkhardt, B. Wu, A. G. Fane, Water Res. 2019, 149, 553– 565. DOI:10.1016/j.watres.2018.11.062 12. R. Bochenek, R. Sitarz, D. Antos, Chem. Eng. Sci. 2011, 66, 6209–6219. DOI:10.1016/j.ces.2011.08.046 13. C. Guillard, H. Lachheb, A. Houas, M. Ksibi, E. Elaloui, J. M. Hermann. J. Photochem. Photobiol. A Chem. 2003, 158, 27–36. DOI:10.1016/S1010-6030(03)00016-9 14. R. Kralchevska, M. Milanova, M. Tsvetkov, D. Dimitrov, D. Todorovsky, J. Mater. Sci. 2012, 47, 4936–4945. DOI:10.1007/s10853-012-6368-4 15. R. Kralchevska, M. Milanova, D. Hristov, A. Pintar, D. Todor- ovsky, Mater. Res. Bull. 2012, 47, 2165–2177. DOI:10.1016/j.materresbull.2012.06.009 16. R. Kralchevska, M. Milanova, P. Kovacheva, J. Kolev, G. Avdeev, D. Todorovsky, Cent. Eur. J. Chem. 2011, 9, 1027– 1038. DOI:10.2478/s11532-011-0089-4 17. M. Uzunova-Bujnova, R. Kralchevska, M. Milanova, R. Todorovska, D. Hristov, D. Todorovsky, Catal. Today 2010, 151, 14–20. DOI:10.1016/j.cattod.2010.02.058 18. M. Uzunova-Bujnova, R. Todorovska, M. Milanova, R. Kral- chevska, D. Todorovsky, Appl. Surf. Sci. 2009, 256, 830–837. DOI:10.1016/j.apsusc.2009.08.069 19. G. Dong, Y. Zhang, Q. Pan, J. Qiu, J. Photochem. Photobiol. C Photochem. Reviews 2014, 20, 33–50. 20. X. Wang, K. Maeda, A.Thomas, K. Takanabe, G. Xin, J. M. Carlsson, K. Domen, M. Antonietti, Nat. Mater. 2009, 8, 76– 80. DOI:10.1038/nmat2317 21. J. Wen, J. Xie, X. Chen, X. Li, Appl. Surf. Sci. 2017, 391, 72– 123. DOI:10.1016/j.apsusc.2016.07.030 22. L. Shi, F. Wang, J. Zhang, J. Sun, Ceram. Int. 2016, 42, 18116– 18123. DOI:10.1016/j.ceramint.2016.08.124 23. V. S. Amrutha, K. S. Anantharaju, D. S. Prasanna, Arabian J. Chem. 2017, 10 DOI:10.1016/j.arabjc.2017.11.016 24. K. Woan, G. Pyrgiotakis, W. Sigmund, Adv. Mater. 2009, 21, 2233–2239. DOI:10.1002/adma.200802738 25. Q. Ming, S. Qin, W. Guanglei, Zh. Bohan, W. Zhengdong, W. Hongjing, Mater. Sci. Eng. B 2017, 224, 125–138. DOI:10.1016/j.mseb.2017.07.016 26. X. Xuejun, S. Zhichun, L. Liping, W. Zehao, C. Ze, R. Rui, H. Yonghong, W. Duan, Appl. Surf. Sci. 2018, 435, 1296–1306. 27. M. Huang, J. Yu, Q. Hu, W. Su, M. Fan, B. Li, L. Dong, Appl. Surf. Sci. 2016, 389, 1084–1093. DOI:10.1016/j.apsusc.2016.07.180 28. F. Wu, X. Li, W. Liu, S. Zhang, Appl. Surf. Sci. 2017, 405, 60– 70. DOI:10.1016/j.apsusc.2017.01.285 29. G. Zhao, X. Huang, F. Fin, G. Zhang, J. T. S. Irvine, Catal. Sci. Technol. 2015, 6, 3416–3422. DOI:10.1039/C5CY00379B 30. W. S. Hummers, R. E. Offeman, J. Am. Chem. Soc. 1958, 806, 1339–1339. DOI:10.1021/ja01539a017 31. M. Tsvetkov, M. Milanova, L. C. J. Pereira, J. C. Waerenborgh, Z. Cherkezova-Zheleva, J. Zaharieva, I. Mitov, Chem. Papers, 2016, 70, 1600–1610. 32. J. Rodriguez-Carvajal, Newsletter, 2001, 26, 12–19. 33. J. Tauc, R. Grigorovici, A. Vancu, Phys. Status Solidi 1966, 15, 627–637. DOI:10.1002/pssb.19660150224 34. C. A. Schneider, W. S. Rasband, K. W. Eliceiri, Nature meth- ods, 2012, 9, 671–675, PMID 22930834. DOI:10.1038/nmeth.2089 35. M. Tzvetkov, M. Milanova, Z. Cherkezova-Zheleva, I. Spasso- va, E. Valcheva, J. Zaharieva, I. Mitov, Acta Chim. Slov. 2017, 64, 299–311. DOI:10.17344/acsi.2016.3049 36. K. Zaharieva, V. Rives, M. Tsvetkov, Z. Cherkezova-Zheleva, B. Kunev, R. Trujillano, I. Mitov, M. Milanova, Mater. Chem. Phys. 2015, 160, 271–278. DOI:10.1016/j.matchemphys.2015.04.036 37. G. Tzvetkov, M. Tsvetkov, T. Spassov, Superlattice Microst. 2018, 119, 122–133. DOI:10.1016/j.spmi.2018.04.048 38. I. N. Leontyev, A. B. Kuriganova, M. Allix, A. Rakhmatul- lin, P. E. Timoshenko, O. A. Maslova, A. S. Mikheykin, N. V. Smirnova, Phys. Status Solidi B 2018, 1800240. DOI:10.1002/pssb.201800240 1091Acta Chim. Slov. 2020, 67, 1082–1091 Tsvetkov et al.: Visible Light-Driven Photocatalytic Activity ... 39. W. L. Smith, J. Appl. Crystalography 1972, 5, 127–130. DOI:10.1107/S0021889872008921 40. K. S. W. Sing, D. H. Everett, R. A. W. Haul, L. Moscou, R. A. Pierotti, J. Rouquerol, T. Siemieniewska, Pure Appl. Chem. 1985, 57, 603–604. 41. J. G. Yu, Q. J. Xiang, M. H. Zhou, Appl. Catal. B-Environ. 2009, 90, 595–602. DOI:10.1016/j.apcatb.2009.04.021 42. L. Chen, W. Ma, J. Dai, J. Zhao, C. Li, Y. Yan, J. Photochem. Photobiol. A 2016, 328, 24–32. DOI:10.1016/j.jphotochem.2016.04.026 43. S. Borthakur, L. Saikia, J. Environ. Chem. Eng. 2019, 7, 103035. DOI:10.1016/j.jece.2019.103035 44. S. Patnaik, K. K. Das, A. Mohanty, K. Parida, Catal. Today, 2018, 315, 52–66. DOI:10.1016/j.cattod.2018.04.008 45. W. Hu, J. Yu, X. Jiang, X. Liu, R. Jin, Y. Lu, L. Zhao, Y. Wu, Y. He, Colloids Surf. A Physicochem. Eng. Asp. 2017, 514, 98–106. DOI:10.1016/j.colsurfa.2016.11.058 46. R. Beranek, Adv. Phys. Chem. 2011, 2011, ID 786759, 20 pag- es, DOI:10.1155/2011/786759 47. L. Jing, Y. Xu, C. Qin, J. Liu, S. Huang, M. He, H. Xu, H. Li, Mater. Res. Bull. 2017, 95, 607–615. DOI:10.1016/j.materresbull.2017.06.003 Povzetek Nanokompozitni fotokatalizatorji ZnFe2O4/rGO/g-C3N4 z različnimi masnimi razmerji ZnFe2O4/g-C3N4 (0,5; 0,75; 1) so bili pripravljeni z večstopnejsko solvotermalno metodo ter uporabo etilen glikola kot topila. Za karakterizacijo kom- pozitov so bile uporabljene različne metode, kot so rentgenska difrakcija, UV-Vis spektroskopija in fotoluminiscenčna spektroskopija. Nastanek mezo-/makroporozne strukture s specifično površino med 67 in 77 m2 g–1 je bil potrjen z ad- sorpcijo/desorpcijo N2. Ugotovljeno je bilo, da je v primerjavi z g-C3N4 (2,7 eV) širina prepovedanega pasu kompozitov manjša (2,30 eV). V nasprotju s g-C3N4, kompoziti niso izkazovali fluorescence, torej ni prišlo do rekombinacije e–/h+. Vsi vzorci, vključno s g-C3N4 in ZnFe2O4, so bili testirani za adsorpcijo in fotokatalitično razgradnjo vodnih raztopin zelenega malahita (10–5 M) pri obvsevanju z vidno svetlobo (λ > 400 nm). Rezultati kažejo, da imajo pripravljeni na- nokompoziti večjo absorpcijo in fotokatalitično aktivnost kot nemodificirana g-C3N4 in ZnFe2O4 in so zato potencialni kandidati za razgradnjo organskih azobarvil v vodi. Except when otherwise noted, articles in this journal are published under the terms and conditions of the  Creative Commons Attribution 4.0 International License