Graphene oxide promotes V-Cu-Ce-ZSM-5 to catalyze SO2 and NO at low temperature: performance and mechanism

A catalyst (V-Cu-Ce-ZSM-5) was explored to simultaneously remove the SO2 and NOx from flue gas by use of the ZSM-5 molecular sieve as the carrier, V and Cu as the active components, and Ce as the additive in low temperature of 150 °C. The performance of V-Cu-Ce-ZSM-5 was evaluated for the oxidation of NO and SO2 before and after the addition of graphene oxide (GO). The results showed that V-Cu-Ce-ZSM-5@GO0.5 had the best performance at a reaction temperature of 150 °C, and the oxidation efficiency of SO2 and NO was 94.60% and 83.64%, respectively. The multiple structural characterizations (BET, SEM, Raman, XRD, and XPS) revealed that the loading of V and Cu with the additive Ce expanded the specific surface area and pore volume of ZSM-5, provided more adsorption sites for SO2 and NO, and had good desulfurization and denitration activity. The addition of GO further improved the dispersibility of active components and auxiliaries, increased the number of active sites in the reaction process, and significantly improved catalytic activity.


Introduction
Fossil fuels such as coal, oil, and natural gas emit many SO 2 and NO x gaseous pollutants during combustion, causing serious harm to human health and ecosystems (Xiong et al. 2016;Adewuyi and Khan 2016;Li et al. 2019). Over the past few decades, several countries have taken steps to mitigate the increasingly serious consequences of air pollution. At present, as a mature method for NO x , and SO 2 removal in flue gas, the treatment techniques of selective catalytic reduction (SCR) and wet flue gas desulfurization (WFGD) have been widely applied (Yang et al. 2019b;Fujii et al. 2013). However, a high occupational area and high operation and maintenance costs make individual pollutant-treatment systems less suitable for popular application in small and medium-sized enterprises Phungrassami and Usubharatana 2021;Ssa et al. 2021). Therefore, great attention has been attached to developing new technologies for simultaneous removal of SO 2 and NO x from flue gas Yang et al. 2019a).
Currently, the oxidation-absorption process has been proved to have great advantages for simultaneous removal of NO x and SO 2 . Some common oxidants, including ozone, permanganate, persulfate, and hydrogen peroxide, oxidize NO and SO 2 to NO 2 and SO 3 , which are then removed in the wet flue gas desulfurization process (Hao et al. 2017;Du et al. 2020;Xiao et al. 2020). However, these reactions are limited to occurrence at lower pH values and production of acidic wastewater, making them difficult to use on a large scale (Ding et al. 2014;Ay and Kargi 2010;Zhao et al. 2014;Wu et al. 2018). Zhao (Zhao et al. 2006) et al. developed a gas catalytic system in which solid catalysts were placed in a fixed bed to react with vaporized oxidants to reduce wastewater generation and improve removal efficiency.
Catalytic removal of SO 2 and NO x has become one of the main research directions of air pollution control. In this technology, different types of catalysts were prepared by impregnation and hydrothermal synthesis with different elements on the carrier; SO 2 and NO x were converted into other harmless gas emissions through redox reactions to remove pollutants (Table 1) Liu et al. 2014;Zhao et al. 2016;Gao et al. 2018). Precious metal catalyst characteristics include high activity and good stability of precious metals, which removes pollutants, although expensive. Iron-based and manganese-based catalysts are cheaper than precious metals and have better catalytic activity only at high temperatures. Tian (Tian et al. 2019) et al. designed an Mn-Ce-Fe-Ti (CP-SD) catalyst with a good NO conversion efficiency and optimal reaction temperature of 250 °C.
Denitration reactions still need to be carried out at a high temperature, which is expensive and less efficient. This study used a ZSM-5 molecular sieve as the carrier, V and Cu as the active components, and Ce as the adjuvant to build upon previous research and find a low-temperature denitration catalyst with a lower cost. The catalytic activity and its influence on graphene oxide (GO) for SO 2 and NO oxidation were investigated by impregnation synthesis (Liu et al. 2019;Ding et al. 2021;Hou et al. 2015;Zhang et al. 2020;Qi et al. 2020). BET, SEM, Raman, XRD, and XPS characterization were used to study the internal structure and catalytic mechanisms. The catalyst provides a new method for the synergistic removal of multiple pollutants from coalfired flue gas.

Preparation of catalysts
First, GO powder was weighed and added to an empty beaker. Deionized water was added and ultrasonically dispersed for 20 min. Then 0.8037 g of ammonium metavanadate and 0.8662 g of oxalic acid were added into the dispersion, dissolved, and prepared for use. Second, 5.0 g of molecular sieve was added into the dispersion, stirred in a water bath at 60 °C for 3 h, mixed thoroughly, placed in a vacuum drying oven, dried at 105 °C for 12 h, and then roasted at 550 °C for 5 h in a muffle furnace to prepare a semi-finished catalyst with a 7% loading of V elements. Finally, 1.1806 g of copper nitrate and 1.2395 g of cerium nitrate was dissolved in a beaker, and fully mixed with the prepared semi-finished catalyst. The drying and roasting steps explained above were repeated to prepare V-Cu-Ce-ZSM-5@GO low-temperature combined desulfurization and denitration catalysts with graphene content of 0% wt, 0.1% wt, 0.3% wt, 0.5% wt, 0.7% wt, and 0.9% wt. The samples were marked as V-Cu-Ce-ZSM-5@GO 0.0 , V-Cu-Ce-ZSM-5@GO 0.1 , V-Cu-Ce-ZSM-5@GO 0.3 , V-Cu-Ce-ZSM-5@GO 0.5 , V-Cu-Ce-ZSM-5@GO 0.7 , and V-Cu-Ce-ZSM-5@GO 0.9 , respectively. Figure 1 shows the simulated flue gas system. The simulated flue gas consisted of N 2 , NO, SO 2 , and O 2 , with a total flow rate of 1.5 L/min, which was controlled by four mass flow controllers (MFCs). The concentration of SO 2 was 500-1300 mg/m 3 , NO 300-700 mg/m 3 , and O 2 6% (v/v). Nitrogen was used as the carrier gas with a flow rate of 0.5 L/min. The flue gas was first piped into the mixing simulator. After that, it enters the flue gas analyzer by bypass, and the concentration of SO 2 and NO was recorded as C in . Then, the mixed gas entered a heated catalytic reactor (1.0 cm in diameter and 15.0 cm in length), where the catalytic reaction was carried out. The gas after catalytic reaction was detected by a flue gas analyzer. At this time, the concentrations of SO 2 and NO were recorded as C out , and the exhaust gas was discharged after treatment. The oxidation efficiencies of SO 2 and NO were calculated using Eq. (1).

Catalytic performance evaluation
where η is the oxidation efficiency; C in and C out are the inlet and outlet concentrations of SO 2 and NO, respectively.

Gas product verification experiments
Following the reaction, the gas was detected by a flue gas analyzer. NO 2 was detected in the gas product, indicating that NO was oxidized to NO 2 by catalytic reaction. The reacted gas was introduced into the HCl+BaCl 2 solution, and over time, a white precipitate formed in the solution. These results indicated that SO 2 was oxidized to SO 3 by catalytic reaction, and SO 3 reacted with BaCl 2 solution to form BaSO 4 precipitate.

Characterization of the catalysts
BET analysis was performed at different partial pressures of nitrogen to determine the catalyst-specific surface area and pore distribution using the BELSORP MAX II analyzer from McKick Bayer Ltd. . Scanning electron microscopy (SEM) and energy-dispersive X-ray spectroscopy (EDS) analyses were performed using the Zeiss Sigma 500 scanning electron microscope and the Bruker XFlash 6130 equipped to observe the morphology, particle size, and dispersion of the catalyst (Lee et al. 2021). The carbon structures of the catalysts were studied by micro-Raman spectroscopy (LabRAM HR Evolution) supplied at 532-nm wavelength (Farah et al. 2020). X-ray diffraction (XRD) was performed using an Ultima IV with a scan range of 5°-90° to test the elemental composition on the catalyst (Cai et al. 2021). X-ray photoelectron spectroscopy (XPS) was performed using the Thermo Scientific ESCALAB 250Xi analyzer to observe the valence states of various elements in the catalyst before and after the reaction ) .

Effect of reaction temperature
The effect of temperature variation on the efficiency for desulfurization and denitration of the catalyst was investigated by adjusting the temperature in the reactor. As shown in Fig. 2, the catalytic efficiency of the catalysts (V-Ce-ZSM-5 and Cu-Ce-ZSM-5) increased first and then decreased at the test temperature (50-300 °C), and the optimum performance was obtained at 150 °C. While reaction temperature increased, the oxidation groups and free radicals on the surface and inside of the catalyst were thermally activated (Bychkov et al. 2020;Hu et al. 2020), which significantly increases the number of activated molecules and the chemical reaction rate of the catalyst. However, with a further Fig. 1 Schematic diagram of the experiments. 1-4 -gas cylinders; 5 -reducing valve; 6 -mass flow controller; 7 -buffer bottle; 8 -electrical furnace; 9 -catalyst layer; 10 -flue gas analyzer; 11exhaust gas treatment device Fig. 2 Effect of the reaction temperature on the simultaneous removal increase in temperature, the recombination probability of free radicals would be expedited, which is an important reason for the decrease in removal efficiency (Liu et al. 2017). So the following catalyst activity tests were all tested at a temperature of 150 °C. Figure 3 shows the changes of desulfurization and denitration efficiency of different catalysts with the doping amount of V, Cu, and Ce. With the increase in V and Cu loading from 5 to 9%, the desulfurization and denitrification efficiency of the catalyst increased first and then decreased. When the loading of V was 7% and the loading of Cu was 8%, the efficiency was the best. After Ce doping, the efficiency of desulfurization and denitrification was significantly enhanced. When the Ce load was 7%, the efficiency was the best.

Effect of V, Cu, and Ce doping
With the increase in the doping amount of V, Cu, and Ce in the catalyst, part of the pore structure would be occupied, and the adsorption capacity of the catalyst for pollutants would be weakened. However, the oxidation capacity gradually increased, and the removal of pollutants changed to the oxidation process. The doping of multiple elements significantly increases the oxidation capacity of the catalyst, but does not always increase, because the pollutants will first adsorb on the adsorption site, and then the oxidation reaction occurs. With the progress of the reaction, the adsorption site gradually saturated, and desulfurization and denitrification efficiency will gradually decrease. The adsorption and oxidation performance of the catalyst was best when the loading of V was 7%, Cu 8%, and Ce 7%. Figure 4 shows the change in desulfurization and denitration efficiency of the catalyst with GO doping amount. With the increase in the amount of GO from 0 to 0.9%, the oxidation efficiency of the catalyst for SO 2 and NO increased gradually, and tends to be stable after the amount of GO was 0.5%. When the doping amount of GO was 0.5%, the catalyst showed the best catalytic effect, and the oxidation efficiencies of the catalyst for SO 2 and NO were 94.60% and 83.64%, respectively.

Effect of GO doping
It was evident that the addition of GO can effectively enhance the oxidation effect of gaseous pollutants, especially  Table 2 shows the BET characterization results of different catalysts. With the addition of V, Cu, and Ce elements, the total pore volume and the specific surface area of the catalyst increased gradually, while the pore size became smaller. This indicated that the infiltration of the elements improved the pore structure of the original carriers, resulting in a corresponding change in the specific surface area, thereby improving the performance of the catalysts. When the catalyst was doped with GO, ZSM-5 was favorable for element loading (Zheng et al. 2019). Compared with ZSM, the specific surface area of V-Cu-Ce-ZSM-5@GO0.5 increased by 42.53%, resulting in an increase in the number of active sites, thus improving the catalytic activity of the catalyst. Figure 5 shows the pore type of the catalyst which belongs to the slit pores, distinct from particle packing, which are similar to the pores created by a layered structure. Figure 6(a-c) shows the SEM images of ZSM-5, V-Cu-Ce-ZSM-5, and V-Cu-Ce-ZSM-5@GO 0.5 , respectively. Compared with ZSM-5, the surfaces of V-Cu-Ce-ZSM-5 and V-Cu-Ce-ZSM-5@GO 0.5 were agglomerated, and several particulates were uniformly dispersed on the surface of the catalyst. The results indicated that the ZSM-5 carrier could be stably loaded with V, Cu, and Ce elements during the catalyst synthesis. Furthermore, when the V-Cu-Ce-ZSM-5 catalyst was doped with GO, the porosity increased, and the particle division on the surface of the carrier was more uniform, which is consistent with the results of Zhang ) et al. These porous structures can significantly increase the diffusion resistance when the polluted flue gas passes through, which is beneficial to improving the effective contact area and contact time between the active components and the reactive gas, finally improving the catalytic efficiency of V-Cu-Ce-ZSM-5@GO 0.5 . Figure 7 shows the EDS mapping images of V-Cu-Ce-ZSM-5@GO 0.5 . V, Cu, and Ce elements were uniformly  distributed in the whole mixed catalyst. In addition, the results of EDS showed that the atomic ratios of V, Cu, and Ce were 2.79%, 2.53%, and 2.07%, respectively, which further confirmed the uniform fixation of V, Cu, and Ce nanoparticles on the surface of ZSM-5.

Raman analysis
The structure of graphene is usually analyzed using Raman spectroscopy, and the structural defects of graphene (D-peak), the sp2 carbon atom in-plane vibrations of graphene (G-peak), and the interlayer stacking pattern of carbon atom (2D-peak) information are well represented in the Raman spectra (Jiang et al. 2022). The Raman spectra of ZSM-5 and V-Cu-Ce-ZSM-5@GO 0.5 are shown in Fig. 8(a). Compared with ZSM-5, the Raman spectra of V-Cu-Ce-ZSM-5@GO 0.5 showed distinct characteristic peaks at 1355 cm −1 , 1586 cm −1 , and 2708 cm −1 , respectively. The analysis shows that these three characteristic peaks are attributed to the D-peak, G-peak, and 2D-peak of graphene, respectively (Akhavan et al. 2016). Furthermore, the 2D/G ratio is 1.23, indicating that the loaded graphene is a single layer (Akhavan 2015). The 2D peak of V-Cu-Ce-ZSM-5@GO 0.5 was fitted, as shown in Fig. 8(b), and only one Lorentz peak appears in the figure, which again verifies that the graphene on V-Cu-Ce-ZSM-5@GO 0.5 is a single-layer structure.

XRD analysis
According to Fig. 9(a) displaying the XRD patterns of ZSM-5, V-Cu-Ce-ZSM-5, and V-Cu-Ce-ZSM-5@GO 0.5 , the diffraction peaks of ZSM-5 located at 13.24° and 15.64° disappeared with the loading of V, Cu, and Ce and produced obvious diffraction peaks of V 2 O 5 and CuO at 32.520° and 48.320°, indicating that V, Cu, and Ce interacted chemically with the ZSM-5 carrier; these elements were immobilized on the ZSM-5 carrier. Meanwhile, the position of the characteristic spectral peaks did not change with the addition of GO. However, the intensity of the characteristic peaks was enhanced, indicating that the addition of GO increased element loading and enhanced the catalytic reaction intensity of the catalyst. Figure 9 (b) shows the XRD magnification patterns of V-Cu-Ce-ZSM-5@GO0.5. After 2θ = 25°, all the diffraction peaks of V, Cu, and Ce elements appear. The diffraction peaks of 2θ at 28.549°, 33.083°, 46.486°, 56.346°, and 59.093° belong to the characteristic peaks of CeO 2 (Marzouk et al. 2018); the diffraction peaks at 26.759°, 32.520°, 36.230°, and 41.289° were assigned to V 2 O 5 (Seo and Bae 2015); and the peaks at 29.847°, 38.568°, 43.848°, 48.320°, and 55.257° were the main peak patterns of CuO (He et al. 2013). The diffraction peaks at 25.513°, 38.251°, and 41.805° were the main peak patterns of the oxygencontaining functional group of GO. Figure 10 shows the high-resolution XPS spectra of V-Cu-Ce-ZSM-5@GO 0.5 (a). The V, Cu, and Ce elements supported on the catalyst mainly exist in the form of oxides of V 2 O 5 , CeO 2 , and CuO. According to the oxidation binding energy data, fine spectra of V2p (b), Cu2p (c), and Ce3d (d) were simulated. Compared with the concentration of oxygen, when V exists at a low concentration, it is often necessary to find the vanadium binding energy with a lower spin-orbit splitting peak on one side of the O1s peak and collect V2p and O1s regions together (507-540 eV) to achieve accurate peak fitting. V produced an asymmetric peak shape in the V2p region, while the vanadium oxide had a symmetric peak shape and a significantly split spin orbit. Two prominent peaks of the V metal were located at 517.2 eV and 524.4 eV, respectively; those of Si2p peak at 100 eV and 150 eV which have no effect on this experiment, as shown in Fig. 10(b). The higher peak intensity indicated that vanadium oxide was dominated by V 2 O 5 on the catalyst surface. Figure 10 (c) shows the energy spectrum of Cu2p, which could be split into two peaks, Cu2p 3/2 (peak at 934.7 eV) and Cu2p 1/2 (peak at 953.7 eV); these were assigned to Cu(II) (Arul et al. 2020). Figure 10 (d) shows that the Ce3d region had well-separated spin-orbit components, each of which was further split by multiple splitting. Ce(IV) had a peak at 917 eV, which was not

Discussion
The reaction mechanism diagram of the low-temperature desulfurization and denitration catalyst is shown in Fig. 11. During the reaction, the V=O bond present in V 2 O 5 serves as the active site for adsorption and oxidation of SO 2 , which can effectively oxidize SO 2 to SO 3 . When the roasting temperature exceeds 500 °C, part of CeO 2 will react with V 2 O 5 to form CeVO 4 with higher oxidation activity, which makes more unsaturated chemical bonding groups generated on the catalyst surface and improves the ability of the catalyst to adsorb oxygen. CeVO 4 can promote SO 2 to SO 3 oxidation and improve the desulfurization efficiency. Denitrification reaction mainly uses the oxidation activity of the metal oxide CuO to oxidize NO to higher valence NO x , and when CeO 2 is added, CeO 2 will form solid solution through roasting, which can stabilize the catalyst lattice and allow more active sites to be retained. In the process of co-impregnation, vanadium and copper will form V-O-Cu bonds; during the reaction, the highly active CuVO x will be oxidized to produce Cu(I)V(4+)O x-1 , and the reduction of vanadium can be promoted in the presence of copper, thus improving the overall effectiveness of the catalyst for desulfurization and denitrification (BYesen and Mathisen 2014). Density functional theory studies have shown that the hydroxyl groups on the surface of graphene oxide can enhance the adsorption of SO 2 and NO. After the dissociation of molecular O 2 , highly reactive epoxy groups will be formed on the surface of graphene oxide, and hydroxyl SO 2 and epoxy groups will form a unique charge transfer channel. The electrons are transferred from the hydroxyl groups to the adsorbed SO 2 and then to the epoxy groups, so that the epoxy groups are preactivated to enhance their oxidative capacity. For the oxidation of NO, the addition of graphene oxide can reduce the reaction barrier, which is beneficial to the oxidation reaction .

Conclusions
This study synthesized a stable V-Cu-Ce-ZSM-5@GO 0.5 catalyst by an impregnation method with good catalytic activity for SO 2 and NO in the flue gas. Under experimental conditions, when the catalytic reaction temperature was 150 °C, the oxidation efficiency of SO 2 and NO were 94.60% and 83.64%, respectively, which significantly reduced the reaction temperature of the traditional flue gas desulfurization and denitration system. The BET characterization results showed that the loading of V, Cu, and Ce components expanded the specific surface area and the pore volume of the carrier, and the addition of GO further promoted the loading of active components on ZSM-5. The SEM and EDS characterization results showed that V, Cu, and Ce were uniformly loaded on the surface of the ZSM-5 carrier, and the addition of GO further increased the loading and dispersion of elements. Raman spectroscopy analysis showed that graphene oxide was loaded on ZSM-5 as a single layer. The results of XRD and XPS analyses showed that element V was present in the catalyst in the form of V 2 O 5 , and Cu and Ce were present in the form of CuO and CeO 2 , respectively. This study provides a new idea for the low-temperature removal of SO 2 and NO from flue gas.

Declarations
Ethics approval and consent to participate Not applicable.

Competing interests
The authors declare no competing interests.

Fig. 11
Reaction mechanism diagram of low-temperature desulfurization and denitration catalyst