3.1. Characterization of the Zn-MOF
XRD pattern was used for characterization of the structure and crystallinity of the samples and the related patterns are shown in Fig. 1. The peaks at 9.8, 15.5, 16.8 and 20.8 indicated the miller facet of (220), (400), (420), and (531), respectively. These peaks approve the MOF structure of Zn(BDC) and the pattern of NH2-modified-Zn(DBC) did not differ so much. The lack of some crystal faces indicates that the morphology of Zn (BDC) MOF is a random slab of a certain thickness rather than a cubic crystal.
Figure 2 displays the FTIR spectra of the prepared zinc metal complex in the range of 400–4000 cm− 1. In Figure, the zinc–metal complex showed an absorption frequency of 3430 cm− 1 which confirms the presence of the hydroxyl group in the complex. Peaks appearing at 2987 cm− 1 indicate C–H stretching present in the aromatic ring and 1612 cm− 1 were assigned due to the linker and C = O stretching lower than those of terephthalic acid, respectively which confirmed the coordination of Zn metal. The absorption band at 1544 cm− 1 and 1502 cm− 1 confirmed the aromatic C = C bending vibrations. Asymmetric and symmetric stretching vibrations of − COO − are appeared at 1388 cm− 1. The absorptions at 927 cm− 1, 792 cm− 1 and 648 cm− 1 were due to the presence of a small amount of DMF as residual solvent. The absorption peaks at 522 and 591 cm− 1 correspond to Zn–O stretching, approving the structure of the metal coordination bond in MOF. The absence of a peak at 1700 cm− 1 implies the removal of the carboxyl group of terephthalic acid during the reaction. The FTIR spectrum of Zn(BDC)-NH2 (not presented) was similar to that of Zn(BDC) sample, just the peak of region 3200–3400 cm− 1 was broader due to overlapping of NH2 and O-H peaks.
Figure 1
Figure 2
The results of SEM of the synthesized Zn-MOF are shown in Fig. 3. Due to different multiple coordination modes in the Zn-MOF crystals, multiple structures have been formed.
Figure 3
The bandgap of the synthesized MOF was calculated at 1.93 eV based on the E = 1243/λ. The calculated band gap is in agreement with the literature(Tiwari, Sagara et al. 2019). The MOF showed a lower bandgap compared to the NiCo2O4 (2.7 eV) and layered double hydroxides(Hosseini, Majidi et al. 2018, Masoumi and Hosseini 2020, Nazari and Hosseini 2021)
The chemical-physical properties of the synthesized sample are presented in Table 1. The BET surface of the photocatalyst was 19.04 m2.g− 1 followed by isotherm type II, indicating a microporous MOF.
Table 1
The physical-chemical properties of the Zn(BDC) MOF-NH2
Sample name | BET surface area (m2.g− 1) | C value in BET formula | Pore volume(m3.g− 1) | Pore mean diameter (nm) |
Zn(BDC)-MOF-NH2 | 19.04 | 4.82 | 4.37 | 4.13 |
Table 1
3.2. Mechanism of photocatalytic desulfurization
Figure 4. shows the proposed mechanism for photocatalytic degradation of DBT over MOFs. The propulsion of the process of photo-desulfurization using Zn-MOF photocatalyst is the presence of hydroxyl radicals. The photocatalytic mechanism following the degradation of DBT at the Zn-MOF level is described. Experiments were also performed in the absence of the photocatalyst. Experiments performed in the absence of photocatalysts showed that dibenzothiophene showed high stability under visible light without Zn-MOF photocatalysts and was not degraded. The result is in agreement with the literature(Zarrabi, Entezari et al. 2015).
As mentioned above, the promoting factor for the desulfurization reaction in the presence of photocatalyst is based on the formation of radicals, especially hydroxyl radicals. The hydrogen peroxide (0.1mL) was used as a hydroxyl radical source.
During degradation of one-mole hydrogen peroxide in the presence of light on the photocatalyst, 2 mol of hydroxyl radical is produced. As a result, a chemical reaction occurs between the active hydroxyl radicals and the dibenzothiophene molecules. First, the oxidation process of the sulfur heteroatom in DBT is carried out by a hydroxyl radical. Hydroxyl radicals attack the sulfur atom of DBT, thus breaking the covalent bond between the carbon atom and the sulfur in the dibenzo-thiophene molecule.
At the end of the reaction, sulfur is released as SO2 gas and the biphenyl molecule is the product of this reaction.
The reason for proposing the mechanism for this process is that no trace of sulfur and its compounds remains after the end of the reaction and no compound was observed on the surface of the Zn-MOF catalyst by FTIR analysis in which sulfur atoms are presented (Zarrabi, Entezari et al. 2015).
Figure 4
3.3. Modelling and optimization of the process
As mentioned in the experimental section, the photocatalytic removal of DBT over MOF catalysts was studied by RSM and it was modelled and optimized.
The range of effective parameters in the reaction was as follows: the reaction time range was 60–180 minutes, the amount of Zn-MOF photocatalyst was 0.02–0.06 g and the concentration of dibenzothiophene was 20–80 ppm.
A total of 15 experiments were designed by the Box-Behnken method considering three levels for each parameter (-1, 0, + 1). The equation created to respond to the process is as follows:
Response = 57.0–0.315 iradiation time(min) – 8.85 photocatalyst(g) – 0.207 C(ppm) + 0.000618 iradiation time(min)*irradiation time(min) – 0.275 photocatalyst(g)*photocatalyst(g) – 0.00431 C(ppm)*C(ppm) + 0.0569 iradiation time(min)*photocatalyst(g) + 0.001667 iradiation time(min)*C(ppm) + 0.1271 photocatalyst(g)*C(ppm) (1)
The model (Eq. 1) indicates the effect of the independent parameters and the interaction of the parameters with the response. The coefficient of determination between the predicted responses and the experimental responses by the model is 97.2, which indicates that the model fits the data well.
Also, two-dimensional contour diagrams and three-dimensional surface diagrams examined and presented the interaction between the parameters (time, amount of photocatalyst, and concentration of dibenzothiophene). Figure 5 shows the two-dimensional contour and three-dimensional surface diagrams of the amount of Zn- MOF photocatalyst (gram) and the reaction time (minutes) on the desulfurization efficiency. It is deduced from Fig. 5 that the optimal response is when the reaction time is 165–180 minutes and the amount of nano photocatalyst is about 0.055–0.060 g.
Also in the two-dimensional and three-dimensional diagrams of Fig. 6, the interaction between the concentration of dibenzothiophene and the Zn- MOF photocatalyst on the response was investigated. The results of the graphs showed that the system response is most effective from the concentration of di-benzothiophene 50–75 ppm and the photocatalyst term with the value of 0.05–0.060 g. In general, hydroxyl radicals have a rapid action in the degradation of dibenzothiophene molecules due to the high number of active centers present at the photocatalytic surface.
At higher than optimal values, the rate of degradation is low and a limited number of photocatalysts can be activated due to the increased transparency of the suspension, which increases light scattering and reduces the penetration depth of light photons.
Also, at high concentrations of dibenzothiophene, the number of active photocatalytic sites decreases due to nanoparticle occupation (Seshadri, Chitra et al. 2008, Zhang, Wu et al. 2008). Increasing the initial concentration of dibenzothiophene reduces the efficiency of the photocatalytic reaction for two reasons:
First, at high concentrations of dibenzothiophene, more of its molecules are adsorbed on the photocatalyst surface and its active sites.
Second, the number of visible light photons that reach the photocatalyst level at high concentrations of di-benzothiophene decreases, and the catalyst loses its activity significantly due to the occupation of its sites by di-benzothiophene molecules. (Mai, Lu et al. 2008, Sobana, Selvam et al. 2008)
An optimal value was then determined for the amount of Zn-MOF photocatalyst and the concentration of dibenzothiophene.
Two-dimensional and three-dimensional diagrams of Fig. 7 show the correlation between the concentration of benzothiophene and irradiation time on the response and show that the concentration of dibenzothiophene is in the range of 50–75 ppm and reaction time is in the range of 165–180 minutes has maximum value.
The optimal removal conditions of dibenzothiophene from the fuel model (gasoline) on the Zn-MOF photocatalyst were predicted by the model, which occurs in the concentration of dibenzothiophene, the amount of photocatalyst, and the reaction time of 50 ppm, 0.06 g and 180 minutes, respectively. According to the above conditions, the predicted removal efficiency of DBT was 50.68%, while the experimental experiment resulted in a removal efficiency of 50.41% of dibenzothiophene from the fuel model.
Figure 5
Figure 6
Figure 7