1. Introduction
Dyes, especially synthetic ones, are widely used in industries such as textiles, printing, rubber, cosmetics, plastics, and leather [1]. Azo dyes, which account for approximately 70% of synthetic dyes, are a typical recalcitrant substance because their azo bonds (N=N) connect substituted aromatic structures [2]. Chemical and biological degradation of azo dyes always produces aromatic amine compounds such as aniline (a grade 1 carcinogen) [3]. Azo dye wastewater has a detrimental impact on human health [4]. The purification of azo dye wastewater is a challenging task in environmental issues.
Current treatment methods for azo dye wastewater include adsorption [5,6], coagulation [7], membrane separation [8], photo/electro-catalysis [9,10], chemical oxidation [11] and biological degradation [2,12]. Adsorption is regarded as an effective and cost-efficient method for pollutant removal with minimal secondary pollution [13]. In contrast, photocatalysis has gained great interest due to its advantages such as short reaction time, high efficiency and low energy consumption [14]. The adsorption-photocatalysis facilitates the effective and environmentally-friendly removal of dye from wastewater. The dye adsorption increases its concentration on photocatalyst surface [15]. The photocatalytic degradation of dye keeps the photocatalyst maintain a continuous adsorption capacity. This technique is becoming the future direction of research in this area.
Graphitic carbon nitride (g-C3N4) is a nonmetallic n-type semiconductor with a unique two-dimensional structure. And it is an excellent photocatalyst with low bandgap of approximately 2.7 eV particularly under visible light [16]. The g-C3N4 preparation is relatively simple and inexpensive by thermal polymerization using single urea, cyanamide, dicyandiamide, melamine, or thiourea as precursors [17]. Those superior properties of g-C3N4 have attracted a considerable attention in many fields including hydrogen evolution reaction, photocatalysis, CO2 conversion, and dye wastewater degradation [18]. Nevertheless, the low solar light absorption of g-C3N4 and the high recombination of electron-hole pairs decreases its photocatalytic efficiency. Some attempts are made to modify g-C3N4. The g-C3N4 doped by metal atoms [16] can alter the band gap while some non-metal atoms (P, O, S and B) can modulate the morphology and porosity [20, 21]. Moreover, integrating with metal-based semiconductors (e.g. CuO, ZnO and TiO2) [19] and constructing target-specific morphology or heterojunction [22] were identified as potential avenues to enhance the light-trapping capacity of g-C3N4, accelerate the separation of electron-hole pairs, and boost the number of carriers, thereby enhancing its photocatalytic performance. Metal-organic frameworks (MOFs) exhibit high specific surface area and porosity as well as narrow band gaps. The combination of g-C3N4 with MOFs has recently drawn considerable research interest. Chen et al. [23] demonstrated that coupling g-C3N4 with the electron donor Co-MOF can enhance charge separation and transfer, and improve the photocatalytic activity of g-C3N4 for CO2 reduction, resulting in a great increase in the photocatalytic activity of g-C3N4 compared to that of single g-C3N4. Wang et al. [18] found 10%(wt) g-C3N4@Cu-MOF nanocomposites is 1.86 and 2.93 times as fast as that of Cu-MOF in photocatalytic antialgal activity and microcystin degradation, respectively. The combination of g-C3N4 with MOF in photocatalysis represents a promising method.
Cu/Zn-BTC is a bimetallic organic framework material with two catalytically active sites and its physicochemical properties are optimized by adjusting the ratio of two metal ions. DR28, a typical anionic diazo dye with high resistance and stability of light oxidation in water [22], was used as target pollutant. In this work, we doped g-C3N4 into Cu/Zn-BTC to prepare a new composite g-C3N4@Cu/Zn-BTC. The objective of this study was to investigate the structure, morphology, and the adsorption and visible photocatalytic properties of g-C3N4@Cu/Zn-BTC composite. Additionally, the effects of initial DR28 concentration, reaction time, and dosage on adsorption and catalysis of DR28 were studied. The synergistic adsorption and photocatalytic mechanism for DR28 removal and the reusability of g-C3N4@Cu/Zn-BTC composites were analyzed. This work provides an idea for the preparation and application of new functional composites in pollutant removal.
2. Experimental Materials and Methods
2.1. Experimental Reagents
Urea (CH4N2O) and DR28 were procured from Shanghai Myriad Biochemistry Co., Ltd. Zinc nitrate hexahydrate (Zn(NO3)2·6H2O) and copper nitrate trihydrate (Cu(NO3)2·3H2O) were sourced from Tianjin Jiangtian Chemical Technology Co., Ltd. 1,3,5-Benzenetricarboxylic acid (H3BTC), N,N-Dimethylformamide (DMF), and anhydrous ethanol (C2H5OH) were all acquired from Tianjin Biotest Technology Development Co., Ltd. Ultrapure water was utilized in the experiments. All chemicals mentioned were commercially available and of analytical reagent (AR) grade.
2.2. Material Preparation
The g-C3N4 and Cu/Zn-BTC was preparaed based on the thermal polymerization method by Guo et al. [24] and hydrothermal method by Kaur et al. [25], respectively, with some improvement.
Preparation of g-C3N4: After 10 g of urea was put in a capped crucible and sealed with high-temperature aluminum foil, it was heated to a temperature of 550°C at a rate of 5°C/min in a muffle furnace and maintained 2 h. And then the furnace was cooled down automatically to room temperature. The light-yellow block-like g-C3N4 sample obtained was ground to g-C3N4 powder.
Preparation of Cu/Zn-BTC: Firstly, 1.2 mmol Cu(NO3)2·3H2O and 0.6 mmol Zn(NO3)2·6H2O both were dissolved ultrasonically in 12 mL ultrapure water (Solution A). After 6 mL C2H5OH and 6 mL DMF were mixed well in a beaker, 1.0 mmol H3BTC was added and then sonicated to dissolve evenly (Solution B). Solution A was slowly injected into solution B, and the mixed solution was agitated for 30 min (solution C). Solution C was quickly transferred to a high-pressure reactor and heated at 120°C for 12 h in an oven. When the reactor was cooled to room temperature, the prepared blue crystals were collected by centrifuge and washed with ultra-pure water and C2H5OH three times respectively. Finally, the resulting Cu/Zn-BTC was dried overnight in a vacuum drying oven at 60°C and ground for further study.
Preparation of g-C3N4@Cu/Zn-BTC: The procedure was same as that of Cu/Zn-BTC except that x (x=0.1, 0.3, 0.5 and 1.0 g) g-C3N4 was added to solution A and then ultrasonicated for 20 min (Fig. 1). The product is denoted as x-g-C3N4@Cu/Zn-BTC.
2.3. Analytical Methods
The XRD patterns of the materials were analyzed by a Rigaku Ultima-V1 X-ray diffractometer (Japan) with Cu Ka (λ = 0.154178 nm) as a radiation source. The conditions were the voltage of 40 kV, the current of 40 mA, the scanning range from 5° to 80°, the scanning speed of 2°/min and the step width of 0.02°. The morphology and surface elements of the samples were analyzed in conjunction with a scanning electron microscope (SEM-EDS, TESCAN MIRALMS, Czech). The surface functional groups were analyzed using a Fourier Transform Infrared Spectrometer (FTIR, Nicolet iS10, USA). The FTIR spectrum was recorded in the wave number range between 4000 and 400 cm−1 after it was pressed tablets with KBr. An X-ray photoelectron spectrometer (XPS, Thermo Scientific K-Alpha, USA) was used to analysize the atomic valence state and molecular structure, as well as elemental composition. The specific surface area and pore size of the samples were analyzed by a fully-automated specific surface area and pore size analyzer (BET, TriStar II 3020, USA) under nitrogen adsorption-desorption (T=−195.85°C). A PerkinElmer Lambda 750 ultraviolet-visible-near infrared spectrometer (UV-Vis-NIR, USA) was employed to record the UV-Vis diffuse reflectance absorption spectra in a scanning range of 220–800 nm by using BaSO4 as a standard reflector.
2.4. Adsorption and Photocatalysis of DR28
Adsorption experiment: The set mass of g-C3N4@Cu/Zn-BTC was added to a conical flask containing 100 mL DR28 solution and then agitated in a constant temperature bath shaker for sampling at set time. After the sample was performed at high-speed centrifugation, the absorbance of the supernatant was measured at 488 nm wavelength using a UV-2600 UV-visible spectrophotometer (Shimadzu, Japan) for DR28 concentration.
Photocatalytic experiment: A quantity of g-C3N4@Cu/Zn-BTC was added to a conical flask with 100 mL DR28 solution. The reactor was transferred to a reaction apparatus equipped with a visible LED lamp as the light source after 30 min dark adsorption. The lamp (the voltage of 220 V, the power of 45 W, and the wavelength range of 400–830 nm) was positioned 18 cm away from the liquid surface. At the set time, the sample was taken to determine DR28 concentration. The adsorption rate (or photodegradation rate) and adsorption capacity were calculated according to Eq. (1) and (2).
where α is adsorption/photodegradation rate, C0 (mg/L) is the initial concentration, Ct (mg/L) is the concentration at time t, Qt (mg/g) is the adsorption capacity at time t, V (L) is the solution volume, and m (g) is the amount of adsorbent dosage.
3. Results and Discussion
3.1. Catalyst Characterization
3.1.1. SEM-EDS
The images of SEM and EDS were shown in Fig. 2. The g-C3N4 appeared a thin sheet-like structure Fig. 2a. The Cu/Zn-BTC presented the octahedral structure of Cu/Zn-BTC with obtuse top vertexes (Fig. 2b), being in agreement with previous literatures [26]. The 0.5-g-C3N4@Cu/Zn-BTC also showed the octahedral structural morphology (Fig. 2c). The Cu, Zn, C, N, and O elements uniformly dispersed on g-C3N4@Cu/Zn-BTC surface (Fig. 2d~h) and every element content was provided in Fig. 2i, being consistent with the constituent elements of the material. The results indicated that the g-C3N4@Cu/Zn-BTC composite was successfully prepared.
3.1.2. XRD
The XRD patterns of g-C3N4, Cu/Zn-BTC and 0.5-g-C3N4@Cu/Zn-BTC, are shown in Fig. 3a. The two strong characteristic peaks of g-C3N4 at diffraction angles (2θ) of 13.0° and 27.5° corresponded to its (100) and (002) crystal planes, showing that the interlayer stacking of the homo triazine structure and conjugation of the exotic π bonds in g-C3N4. This result is consistent with the previous report [27]. The g-C3N4@Cu/Zn-BTC showed 11 strong typical characteristic peaks at diffraction angles 2θ of 6.77°, 9.52°, 11.67°, 13.48°, 14.73°, 16.55°, 17.53°, 19.10°, 20.28°, 26.03°, and 27.73° which corresponded to (200), (220), (222), (400), (420), (422), (511), (440), (442), (731), and (002) crystal planes. Those peaks were concordance with the principal diffraction peaks and crystalline planes in the XRD pattern of Cu/Zn-BTC in previous literatures [26, 28]. The peak intensities of 0.5-g-C3N4@Cu/Zn-BTC were lower than those of Cu/Zn-BTC. The 0.5-g-C3N4@Cu/Zn-BTC also appeared the characteristic diffraction peaks of the (002) and (100) facet in g-C3N4. Furthermore, the diffraction peak intensity initially increases and then decreases as the g-C3N4 doping amount increases (Fig. S1) and the peak intensities of 0.3-g-C3N4@Cu/Zn-BTC was the highest, showing that g-C3N4 doping did not disrupt the structure of Cu/Zn-BTC and there existed the effective interaction between Cu/Zn-BTC and g-C3N4.
3.1.3. FTIR
The FTIR spectra of g-C3N4, Cu/Zn-BTC and 0.5-g-C3N4@Cu/Zn-BTC are shown in Fig. 3b. g-C3N4 existed three distinct absorption regions [29]: 1) the peak at 3418.62 cm−1 is caused by the stretching vibration of N-H in g-C3N4 or O-H of water molecules adsorbed on g-C3N4 surface, 2) the peaks at 1200–1700 cm−1 (1635.90 cm−1 (C=N), 1574.30 cm−1 (C=N), 1420.65 cm−1 (C-N), 1324.53 cm−1 (C-N) and 1254.46 cm−1 (C-N)) were attributed to the stretching vibrations of the C-N aromatic heterocycles, and 3) the strong peak at 812.70 cm−1 corresponded to the outward bending vibrations of the C-N heterocycles. In the FTIR spectrum of Cu/Zn-BTC the three absorption peaks at 490.01, 730.02, and 1110.33 cm−1 corresponded to the stretching vibration of the Cu-O bond, the bending vibration of the Zn-O bond, and the stretching vibration of the C-O-Cu bond, respectively. The strong absorption peak at 1374.39 cm−1 was due to the symmetric vibration of the C-O bond, the peaks at 1573.17 and 1620.84 cm−1 were caused to the asymmetric stretching of the C=O bond, and the broad peak at 3419.16 cm−1 was mainly related to the water adsorbed on the surface of Cu/Zn-BTC [27]. The FTIR spectrogram of 0.5-g-C3N4@Cu/Zn-BTC exhibited those peaks of both g-C3N4 and Cu/Zn-BTC including the characteristic peaks at 3424.71, 1624.63, 1572.36, 1420.14, 1324.95, 1044.03, 813.33 and 727.87 cm−1, indicating the functional group structures such as C-N, Cu-O, Zn-O, and C-O bonds existed in the composite.
3.1.4. XPS
In order to gain insight into the surface chemical composition and electronic states of g-C3N4, Cu/Zn-BTC, and 0.5-g-C3N4@Cu/Zn-BTC, the XPS data were monitored (Fig. 4). The total XPS spectrum (Fig. 4a) revealed that the C, N, O, Cu, and Zn elements were present in the 0.5-g-C3N4@Cu/Zn-BTC (All elements were calibrated with the standard charge (284.80 eV) of C element), which is consistent with the EDS results. The C1s XPS binding energy peaks of 0.5-g-C3N4@Cu/Zn-BTC appeared at 284.80, 286.02, and 288.19 eV (Fig. 4b) which were the characteristic ones of g-C3N4 and Cu/Zn-BTC, corresponding to the C-C bond of the exogenous carbon, the C-NHx (x=1, 2) bond at the edge of the heptazine unit or the C-O bond, and the C=O/N-C=N bond of the triazine ring structure. The XPS spectrum of N1s showed (Fig. 4c) that both 0.5-g-C3N4@Cu/Zn-BTC and g-C3N4 had the two characteristic peaks which corresponded to the C=N-C bond of the triazine ring and the nitrogen atom at the ring edge of the tertiary nitrogen N-(C)3 group. The two energies of the former were 398.57 and 400.29 eV, while those of the latter were 398.62 and 400.86 eV. A slight decrease in the outer electron energy of nitrogen is observed after the formation of the composite. The XPS energy spectrum of O1s appeared two main peaks (Fig. 4d) due to the O2 molecules adsorbed on the sample surface and the O-H bonds bound to water molecules [30]. The peaks of 0.5-g-C3N4@Cu/Zn-BTC was less than the other two materials. The binding energies of Cu2p in 0.5-g-C3N4@Cu/Zn-BTC at 934.88 and 954.64 eV corresponded to Cu2p3/2 and Cu2p1/2 orbitals (Fig. 4e), being consistent with the presence of Cu2+ in the Cu/Zn-BTC structure. The binding energy at 932.66 eV is predominantly attributed to the Cu+ species and the peak band in the range of 936 eV to 946.6 eV can be attributed to the “jittery satellite” band of Cu2+ [31]. The Zn 2p1/2 orbital binding energies of 0.5-g-C3N4@Cu/Zn-BTC and Cu/Zn-BTC were 1022.24 and 1022.33 eV, while the Zn 2p3/2 orbital binding energies were 1045.21 and 1045.44 eV (Fig. 4f), respectively. The electronic binding energies of both Cu2+ and Zn2+ in the composites were lower than Cu/Zn-BTC, which can be attributed to the physicochemical reactions that occur during the composite process of g-C3N4 and Cu/Zn-BTC. The XPS analysis further showed the g-C3N4@Cu/Zn-BTC composite are not simply physically mixed by g-C3N4 and Cu/Zn-BTC, but rather form a lower-energy dense structure.
3.1.5. Specific surface area and pore analysis
The specific surface area, pore volume and pore size distribution of material are important for its properties. The N2 adsorption and desorption curves and pore size distributions of the three materials are shown in Fig. S2. The Cu/Zn-BTC followed type I isotherm while both the g-C3N4 and the 0.5-g-C3N4@Cu/Zn-BTC were type IV isotherms with H3 hysteresis loops, indicating that the two latter were mesoporous materials with slit pores formed by the stacking of lamellar particles [32]. The average pore sizes of Cu/Zn-BTC, g-C3N4 and 0.5-g-C3N4@Cu/Zn-BTC were 1.91, 3.86 and 12.13 nm, respectively. The BET specific surface areas of Cu/Zn-BTC, g-C3N4 and 0.5-g-C3N4@Cu/Zn-BTC were 445.48, 40.21 and 187.52m2/g, respectively, which are consistent with literature reports [33,34]. The 0.5-g-C3N4@Cu/Zn-BTC composite decreased 2.37 times in comparison to Cu/Zn-BTC and enlarged 4.66 times of g-C3N4. The combination of g-C3N4 with Cu/Zn-BTC lead to the alteration of the pore structure, thereby reducing the surface area. The composite could potentially increase the adsorption and photocatalytic capacity due to the introduction of g-C3N4 lamellar structure and the enlargement of its average pore size.
3.1.6. UV-Vis-NIR
The light absorption ability exerts a significant influence on the photocatalysis. The UV-Vis-NIR spectrometer was employed to identify the light absorption of the three materials (Fig. S3a). g-C3N4@Cu/Zn-BTC performed the best absorbance in the three materials in the ultraviolet range, but in the visible region (400~780 nm) the absorbance was Cu/Zn-BTC>g-C3N4@Cu/Zn-BTC> g-C3N4. The band gaps (Eg) of the three materials are calculated by Tauc plot by Eq. (3) and (4) [35].
where α is the absorption index, h is Planck’s constant, ν is the incident photon frequency, λ is the absorption wavelength, A is a constant and n is the type of semiconductor (1/2 for direct semiconductors and 2 for indirect ones). The Eg values of g-C3N4, Cu/Zn-BTC and 0.5-g-C3N4@Cu/Zn-BTC were 2.63, 2.39 and 2.17 eV (Fig. S3b), respectively. g-C3N4 is basically consistent with the reported results [36]. The results demonstrate that Cu/Zn-BTC attached to g-C3N4 can effectively reduce the band gap and increase the visible light response intensity in comparison with g-C3N4, which enhances the visible light photocatalytic activity.
3.2. DR28 Adsorption
The DR28 adsorption on g-C3N4, Cu/Zn-BTC and 0.3-g-C3N4@Cu/Zn-BTC (all dosage of 0.01 g) at its initial concentration of 20 mg/L over time is shown in Fig. 5a. The DR28 adsorption on g-C3N4, Cu/Zn-BTC and 0.3-g-C3N4@Cu/Zn-BTC reached to the adsorption equilibrium at 6, 2 and 2 h, respectively. The adsorption efficiencies were 39.63%, 44.10% and 82.53% with the corresponding adsorption capacities of 79.259, 89.753, and 165.370 mg/g. The removal efficiency of 0.3-g-C3N4@Cu/Zn-BTC is 2.1 times that of g-C3N4 and 1.87 times that of Cu/Zn-BTC, indicating that g-C3N4@Cu/Zn-BTC with the multilayer pore structure is conducive to DR28 adsorption [37].
The removal efficiency of DR28 by g-C3N4@Cu/Zn-BTC with different doping contents of g-C3N4 is shown in Fig. 5b. The DR28 removal efficiency gradually increases with the increase of g-C3N4 doping from 0.1 to 0.5 g. The adsorption efficiency of DR28 was 88.24% for 0.5-g-C3N4@Cu/Zn-BTC and 87.0% for 1.0-g-C3N4@Cu/Zn-BTC with a slight decrease. There is an interaction between g-C3N4 and Cu/Zn BTC in g-C3N4@Cu/Zn-BTC and an optimal value exists at 0.5 g g-C3N4 doping. Therefore, 0.5-g-C3N4@Cu/Zn-BTC was used in all the subsequent adsorption and catalytic experiments.
The experimental data fitted the pseudo-second-order kinetic model better than the pseudo-first-order model (Fig. S4 and Table S1), showing that the chemisorption was the dominant process. The adsorbate molecules were first adsorbed rapidly to the particle surface by chemisorption and then diffused to the interior of the particle by intra-particle diffusion basing on the intra-particle diffusion model fitting results (Fig. S4).
3.3. DR28 Photocatalytic Degradation
The solution with an initial DR28 concentration of 20 mg/L was subjected to visible light irradiation following 30 min dark adsorption at 0.01 g dosage for all the three catalyst (Fig. 6a). The 0.5-g-C3N4@Cu/Zn-BTC catalyst has the best catalytic performance for the three catalysts. The removal rate of the 0.5-g-C3N4@Cu/Zn-BTC catalyst was 99.3% at 40 min, which was 1.14 and 2.42 times that of g-C3N4 (87.3%) and Cu/Zn-BTC (41.1%), respectively. The adsorption-photocatalytic synergistic degradation of DR28 exhibited a higher removal rate (100%) at shorter reaction time (50 min) in comparison to the adsorption test (Fig. 5) at a initial 20 mg/L DR28 concentration and the same dosage for 0.5-g-C3N4 @Cu/Zn-BTC.
In order to investigate the degradation ability of 0.5-g-C3N4 @Cu/Zn-BTC catalyst on DR28, 0.01 g catalyst were added to different DR28 concentration solution. As the initial DR28 concentration increased, the removal rate became lower and the removal rate of DR28 remained relatively high (Fig. 6b). The DR28 photocatalytic degradation reached to 96.3% at 40 min at its concentration of 40 mg/L. The effect of dosages on the DR28 removal efficiency (the initial concentration of 40 mg/L) is shown in Fig. 8c. The DR28 removal efficiency increased with an escalation in catalyst dosage but the augmented value gradually declined due to the light transmittance declines and the decrease of DR28 molecules adsorbed per active site. The appropriate amount of catalyst is crucial for the reaction rate and operation cost. Table S2 presents the DR28 removals reported in literatures, showing that g-C3N4@Cu/Zn-BTC exhibits both high adsorption efficiency and excellent catalytic activity.
The catalytic process was fitted with first-order reaction kinetics model as Eq. (5).
where k1 (min−1) is the kinetic rate constant. The fitting results are shown in Fig. 6d and Table S3. The correlation coefficients R2 of the DR28 catalytic degradation by the 0.5-g-C3N4@Cu/Zn-BTC at different concentrations exceeded 0.95, indicating that the reaction conformed to the first-order kinetic model. The photocatalytic reaction rate constants k1 reduced from 0.1248 to 0.03348 min−1 as the DR28 concentration changed from 20 to 50 mg/L. The DR28 molecules at lower initial concentrations had a greater opportunity to interact with the catalyst, resulting in a higher removal rate. The zero-order reaction kinetic model fitting (Table S4 and Fig. S5) of the data in Fig. 8c showed the degradation rate had an important relationship with 0.5-g-C3N4@Cu/Zn-BTC at low dosages of 0.005 and 0.01 g, showing that the catalysis was decisive.
3.4. Possible Mechanism of DR28 Removal
The g-C3N4@Cu/Zn-BTC exhibit efficient DR28 removal. The Mott Schottky curve is shown in Fig. 7. The slopes of the tangent lines of g-C3N4 and Cu/Zn-BTC are positive, indicating that the two materials both are n-type semiconductors. The flat band potentials of g-C3N4 and Cu/Zn-BTC (i.e. the intersection of the tangent line and the X-axis) were −1.25 and −0.62 eV vs. saturated Ag/AgCl reference electrode (Ag/AgCl), respectively, which were converted to be −0.64 and −0.01 eV vs. normal hydrogen electrode (NHE) by Eq. (7). The Cu/Zn-BTC is molecular photocatalytic material, the frontier molecular orbital potential is calculated by Eq. (8).
where E0 (Ag/AgCl) is 0.1976 eV at 25°C, E(Ag/AgCl) is the operating voltage at pH=7, and E(HOMO) and E(LUMO) (eV) are the highest and the lowest occupied molecular orbital potential, respectively. The flat band potential is usually below 0.1–0.3 eV than the conduction band (CB) and the 0.1 eV value was adopted for safety. The CB and the VB (valence band) of g-C3N4 were −0.74 and 1.89 eV, and the E(LUMO) and E(HOMO) of Cu/Zn-BTC was −0.11 and 2.28 eV, respectively. Therefore, the possible mechanism of g-C3N4@Cu/Zn-BTC adsorption-photocatalytic synergistic degradation of DR28 was deduced.
Firstly, the CR molecules were adsorbed on the surface of the material in dark adsorption. When g-C3N4@Cu/Zn-BTC excited by visible light, the CB electrons (e−) are effectively separated from the holes (h+) (Eq. (9) and (10)). The VB of Cu/Zn-BTC is more positive than that of g-C3N4 while the CB of g-C3N4 is more negative than that of Cu/Zn-BTC. The photo-excitation shall make the e− produced in g-C3N4 transfer to the CB of Cu/Zn-BTC and make h+ in Cu/Zn-BTC transfer to the VB of g-C3N4 (Eq. (11) and (12)), generating an intrinsic magnetic field. Concurrently Cu2+ and Zn2+ can capture e− (Eq. (13)), enhancing further the magnetic field intensity. Thus, it has been hypothesised that a Z-type heterojunction may be formed inside the composite [3], which increases the separation efficiency of electrons and holes and improves the efficiency of the photocatalytic reaction. Since the standard oxidation potential of •OH/OH− is 1.99 eV, h+ on the VB of g-C3N4 cannot oxidize water to generate •OH. Since the standard reduction potential of O2/•O2 − is −0.33 eV [37], the eon CB of g-C3N4 can react with O2 to form •O2 − (Eq. (14) and (15)), and •O2 − can be further converted into •OH (Eq. (16) and (17)). The generated •O2 − and •OH species with strong oxidizing ability, serving as the primary active species for the DR28 degradation into H2O and CO2 (Eq. (18)). DR28 continuous and efficient adsorption increases its concentration on the catalyst surface, greatly accelerating DR28 photocatalytic degradation. The photoelectrochemical (PC) response of g-C3N4, Cu/Zn-BTC and 0.5-g-C3N4@Cu/Zn-BTC composite exhibits that the composite enhances the photogenerated charge-carrier separation efficiency single g-C3N4 and Cu/Zn-BTC (Fig. S6a), although the charge-carrier transport speed of 0.5-g-C3N4@Cu/Zn-BTC is between g-C3N4 and Cu/Zn-BTC. The results provide a compelling evidence that the DR28 degradation is a synergistic process of adsorption and catalysis.
3.5. Repeatability Experiments
The removal of DR28 by 0.5-g-C3N4@Cu/Zn-BTC adsorption (2 h) and adsorption (0.5h)-photocatalytic degradation (50 min) under five cycles are compared in Fig. 8. The material regenerated in 0.1 mol/L NaOH solution (20 mg/L) and washed by ultrapure water for next-cycle adsorption but was directly used for next-cycle photocatalytic degradation. The adsorption performance of 0.5-g-C3N4@Cu/Zn-BTC gradually decreased as the number of cycles increased, and the DR28 removal rate decreased from 88.24% to 74.01% at equilibrium. This phenomenon may be attributed to the chemical saturation of active sites. The photocatalytic degradation rate only decreased from 99.3% to 93.0% at 400 min for five cycles, which may be caused by the adsorption decrease. The experimental results indicated that 0.5-g-C3N4@Cu/Zn-BTC was both an adsorbent and a photocatalyst with reusable potential.
4. Conclusion
The g-C3N4@Cu/Zn-BTC composite easily prepared in this work achieved the effective removal of DR28 by adsorption and visible light photocatalytic degradation, being a promising adsorbent and photocatalyst. The composites formed a dense lower energy structure. The dark adsorption removal of DR28 by 0.5-g-C3N4 @Cu/Zn-BTC reached to 88.24% at 120 min and its catalytic efficiency rose to 100% in 50 min under visible light irradiation for 20 mg/L DR28 solution. The DR28 photocatalytic degradation efficiency reached 99.3% in 40 min for the 40 mg/L DR28 solution. The adsorption of DR28 conforms to the pseudo-second-order kinetic model, and the catalytic degradation conforms to the first-order kinetic model. 0.5-g-C3N4@Cu/Zn-BTC can still maintain a high removal rate for multi-cycles, which shows its practical value in industrial wastewater treatment and potential commercialization.
PDF Links
PubReader
Full text via DOI
Download Citation 
Supplement
Print 
