Catalytic Conversion for Hydrogen Sulfide

In this study, a series of transition metal mono-substituted heteropoly compounds H7PMo11MO39 (M = Co2+, Mn2+, Ni2+ and Zn2+) (HPMo11M) and single-absent heteropoly compounds H3PMo11O39 (HPMo11) were prepared for highly effective removal of hydrogen sulfide (H2S) from gas stream. The heteropoly compounds were characterized Fourier transform infrared spectroscopy (FT-IR), elemental analysis and scanning electron microscopy (SEM). The results confirmed that the transition metal ions successfully replaced the Mo atom. H7PMo11CoO39 showed that the outstanding desulfurization capacity and the H2S removal efficiency can reach more than 90% for 3 h. Besides, after regeneration, the desulfurization capacity of H7PMo11CoO39 towards H2S only a drop of 5.11% of the initial desulfurization capacity. Optimization experiments demonstrated that H7PMo11CoO39 had the ideal desulfurization performance under the condition of low H2S concentration or high dosage of H7PMo11CoO39. An appropriate temperature of 25°C is necessary for high removal efficiency. The optimum pH value for desulfurization is 5. The kinetic data can be well described by pseudo-first-order kinetic model. The desulfurization products were proved to be S and SO4 based on X-ray photoelectron spectroscopy (XPS) characterization results.


INTRODUCTION
Heteropoly compounds (HPCs) have been widely used in various organic and catalytic synthesis reactions. This is because the HPCs have high catalytic activity, stable structure, redox and redox reversibility, excellent selectivity, and can be used as homogeneous and heterogeneous catalysts. When the central heteroatoms in the HPCs are replaced by transition metal ions, a wide variety of substituted HPCs can be generated. This expands the field of catalytic reduction of HPCs. In recent years, transition metal-substituted heteropoly compounds have attracted the attention of researchers at home and abroad and become a hot research field (Weakley and Malik, 1967). The vacant heteropoly compound anions are similar to multidentate ligands, with two or more atoms that can provide lone electron pairs. Therefore, people can synthesize substituted heteropoly compounds with transition metal ions (Bi et al., 2004;Proust et al., 2008). The transition metalsubstituted heteropoly compound has the same crystal water as the heteropoly compound, and at the same time, there is stable structural water directly connected to the skeleton structure inside the anion. Therefore, the transition metal-substituted heteropoly compounds have a stable structure and strong thermal stability. In addition, they can also store a large number of electrons, thereby increasing the conductivity and enhancing the catalytic activity and are more widely used in various catalytic oxidation reactions (Santos et al., 2015).
According to relevant literature reports, heteropoly compounds have been involved in various reactions as acidic and redox catalysts. Transition metal-substituted heteropoly compounds have been widely used in the field of catalytic oxidation of organic substances such as alkanes, olefins, alcohols and sulfur-containing compounds, and the catalytic oxidation desulfurization of fuel oil . However, for the desulfurization reaction to remove hydrogen sulfide, most of the catalysts used are phosphotungstates replaced by vanadium (V). Liu et al. (2017a) and Lütkehoff et al. (1995) found that the efficiency of H2S removal of the pure phosphotungstic acid (H3PW12O40) and phosphomolybdic acid (H3PMo12O40) were 20% and 80%, respectively. Ma et al. (2014) and Wang et al. (2003) substituted the part elements of phosphomolybdic acid with V and W to form H7[P2Mo17VO62]·39H2O and H3PW10Mo2O40 to remove H2S. The results showed that the efficiency of H2S removal of them were only 85%. Few studies have used transition metal (Co, Ni, Zn, etc.) substituted phosphomolybdate to remove hydrogen sulfide in mixed gas.
In summary, we used alkaline degradation to synthesize a series of transition metal monosubstituted heteropoly compounds H7PMo11MO39 (M = Co 2+ , Mn 2+ , Ni 2+ and Zn 2+ ) (HPMo11M) and single-absent heteropoly compounds H3PMo11O39 (HPMo11). They were used to absorb hydrogen sulfide in acid gas. We used Fourier transform infrared spectroscopy (FT-IR), elemental analysis and scanning electron microscopy (SEM) to characterize and analyze the morphology and structure of HPMo11M. The dynamic absorption experiment was used to test the removal and regeneration performance of HPMo11M for hydrogen sulfide. At the same time, conditions optimization experiments were conducted to explore the optimal values of temperature, hydrogen sulfide concentration, heteropoly compound concentration and solution pH. Finally, based on Xray photoelectron spectroscopy (XPS) characterization results, the desulfurization products and desulfurization mechanism were discussed.

Characterization Analytical Methods
The characterization methods used for the catalysts are shown in Table 1.

H2S Absorption and Regeneration Experiments
The H2S absorption experiment was performed at standard atmospheric pressure. The desulfurization reagent was obtained by dissolving a certain amount of HPMo11M in 50 mL of deionized water and transferring it to a cylindrical glass reactor. The mixture of N2 and H2S was introduced into the glass reactor at a flow rate of 100 mL min -1 . The H2S concentration was detected by a TH-990s H2S gas analyzer (Liu et al., 2017). The residual H2S gas was absorbed by NaOH solution. The temperature of the solution during the reaction was controlled by a thermostatic water bath. The effects of different operating parameters on desulfurization were studied by controlled experiments, under HPMo11M concentration range of 0.0005 to 0.002 mol L -1 and pH range of 1 to 9. The reaction device diagram is shown in Fig. 1. The desulfurization performance of HPMo11M solutions was mainly evaluated by H2S removal efficiency (η, %): where C0 and Ct represent the inlet and outlet concentrations of H2S, respectively. After the reaction, the desulfurizing agent was regenerated and recovered in a water bath at 90°C with 500 mL min -1 of air continuously for 6 h. The regenerated desulfurizing agent solution can continue to perform H2S desorption experiment.
The Fourier transform infrared spectra (FT-IR) of HPMo12, HPMo11 and HPMo11M (M = Co 2+ , Mn 2+ , Ni 2+ and Zn 2+ ) are shown in Fig. 2, which are consistent with the characteristic peaks of Keggintype molybdenum heteromeric compounds (Xie et al., 2008). In the FT-IR spectrum of HPMo12, the characteristic peaks at 1059, 953, 881, and 747 cm -1 correspond to the stretching vibration peaks of P-Oa, Mo = Od, Mo-Ob-Mo, and Mo-Oc-Mo, respectively. After the transformation from HPMo12 to HPMo11, the four characteristic peaks of P-Oa, Mo = Od, Mo-Ob-Mo and Mo-Oc-Mo changed significantly. In particular, the P-Oa bond stretching vibration peak at 1059 cm -1 was split into two vibration absorption peaks (1077 and 1030 cm -1 ) in the FT-IR spectrum of HPMo11. This phenomenon indicates that after losing a Mo atom, the symmetry of the vibration mode of the heteroatom is reduced due to the destruction of the symmetry around it. From the FT-IR spectrum of HPMo11M, it can be seen that when the transition metal is bound to HPMo11, the stretching vibration peak of the P-Oa bond of HPMo11M appears near the original position without splitting, indicating that the structure has been restored, that is, the symmetry around the heteroatoms is restored by the metal ions to some extent. The P-Oa bonds of HPMo11M (M = Co 2+ , Mn 2+ , Ni 2+ and Zn 2+ ) FT-IR spectra show stretching vibration peaks at 1049, 1090, 1097 and 1092 cm -1 , respectively. The stretching vibration peaks of Mo = Od, Mo-Ob-Mo and Mo-Oc-Mo also have displacement changes. These results showed that the transition metal ions successfully replaced the Mo atom.

ICP of HPMo12 and HPMo11Co
The contents of P, Mo, and Co in HPMo12 and HPMo11Co were determined by an inductively coupled plasma spectrometer (ICP-AES). The results of elemental analysis are shown in Table 2.
The experimental values are basically consistent with the theoretical calculations, indicating that Co in HPMo11Co replaces a Mo.

Catalyst scanning electron microscope (SEM) analysis
For the morphological analysis of the HPMo12 and HPMo11M (represented by HPMo11Co), their SEM images are shown in Fig. 3. As can be seen in Fig. 3(a), the HPMo12 is a flat layer structure connected together. After the adding of transition metal cobalt (Co), the morphology becomes prismatic (as shown in Fig. 3(b)), which indicates that the transition metal Co successfully replaced Mo.

Effect of transition metal species on H 2S removal
The hydrogen sulfide absorption of HPMo12, HPMo11 and HPMo11M (M = Co 2+ , Mn 2+ , Ni 2+ and Zn 2+ ) were compared under the same conditions, as shown in Fig. 4. Compared with HPMo12 and vacant HPMo11, the heteropoly phosphomolybdate substituted by transition metal elements had better desulfurization effect and longer duration. Therefore, the focus is on the desulfurization and regeneration performance of heteropoly phosphomolybdate substituted by transition metal elements. The hydrogen sulfide absorption and regeneration results of HPMo11Co, HPMo11Mn, HPMo11Ni and HPMo11Zn are shown in Fig. 5. In comparison, HPMo11Co and HPMo11Zn had better removal effect on hydrogen sulfide, reaching more than 90%. Comparing the desulfurization performance after regeneration, it was found that HPMo11Co had the best effect. According to the comprehensive comparison of desulfurization and regenerated desulphurization effects, the desulfurization ability of heteropoly phosphomolybdate was ranked as follows: HPMo12 < HPMo11 < HPMo11Zn < HPMo11Ni < HPMo11Mn < HPMo11Co. Therefore, HPMo11Co was selected as the desulfurizer for the following condition optimization experiments in this study.

Effect of temperature on H2S removal
The temperature of the desulfurizer has a significant effect on H2S removal, and the results are shown in Fig. 6. It can be observed that the desulphurization efficiency of heteropoly   phosphomolybdate decreased significantly with the increase of temperature within the investigated range (25-75°C). The desulfurizer at room temperature (25°C) showed the best removal performance, and the desulfurization efficiency can be maintained above 90% within 350 min. This study also exhibited that higher temperature was not conducive to the removal of H2S mainly in two aspects: on the one hand, the oxidation process of H2S was exothermic, so high temperature was not conducive to the absorption process (Huang et al., 2017); on the other hand, the increase of temperature tended to retard the dissolution of H2S in aqueous solutions. Therefore, the degree of solubility played a leading role in this process. The room temperature (25°C) was selected as the best desulfurizer temperature, and subsequent experiments were carried out accordingly.

Effect of HPMo11Co concentration on H2S removal
The effect of HPMo11Co concentration on desulfurization performance at 25°C is shown in Fig. 7. It can be observed that the higher the concentration of HPMo11Co, the higher the desulfurization efficiency and the longer the reaction time. At the concentration of 0.0005 mol L -1 for the HPMo11Co, the desulfurization efficiency decreased rapidly in a short time. When the concentration was 0.001 mol L -1 , the time for the H2S removal efficiency to remain above 92% was reduced to 360 min. When the concentration was 0.0015 mol L -1 , the desulphurization efficiency dropped below 50% rapidly after 400 min. At the concentration of HPMo11Co was 0.002 mol L -1 , the H2S removal efficiency can be maintained above 92% within 400 min, and slowly decreased to less than 70% after 450 min. The desulfurization rate decreased significantly because HPMo11Co was consumed with the extension of the reaction time. In short, the higher the concentration of HPMo11Co, the more favorable the desulfurization reaction. Therefore, the subsequent experiments were implemented with 0.002 mol L -1 of HPMo11Co.

Effect of H2S concentration on H2S removal
The effect of different concentrations of H2S on the desulfurization performance of HPMo11Co solution at 25 °C was studied (Fig. 8). The results showed that the lower the H2S concentration, the more favorable the desulfurization reaction. The desulfurization efficiency can be maintained above 90% within 350 min at most H2S concentrations. Especially when the concentration of H2S was 1500 mg m -3 , the desulfurization efficiency reached above 95% and the time lasted for 400 min. The reason for this phenomenon is that low concentration of H2S will consume less HPMo11Co at the same time. The sudden drop in the desulfurization efficiency at the later stage of the reaction was due to the exhaustion of HPMo11Co. Combined with actual conditions such as research time and gas flow rate, the subsequent experiments were implemented with H2S concentration of 1750 mg m -3 .

Effect of desulfurizer pH on H2S removal
The pH of desulfurizer is an important factor affecting the removal efficiency of H2S (Zheng et al., 2018). During the research, the pH of the desulfurizer was adjusted between 1 and 9 with 1 M NaOH and 1 M HCl, and the H2S concentration was maintained at 1750 mg m -3 . As shown in Fig. 9, the removal efficiency of H2S decreased rapidly at the pH value was 1 and 3. The removal efficiency of H2S can be maintained above 90% within 360 min at the pH value was 5. When the pH value increased from 5 to 9, the removal efficiency of H2S decreased significantly. This phenomenon may occur because the alkaline environment will destroy the structure of the heteropolyanion, while the heteropolyanion in the acidic environment can maintain a stable keggin-type structure.
Therefore, the HPMo11Co solution showed the best desulfurization performance under the optimal conditions (25°C, the 0.002 mol L -1 concentration of HPMo11Co, the H2S concentration of 1750 mg m -3 and desulfurizer pH of 5). As shown in Fig. 9, the H2S removal efficiency remained above 90% within 360 min.

Absorption kinetics
The absorption kinetics of the selected optimum conditions was studied. According to the original data of pH = 5 curve in Fig. 9, the absorption curve of desulfurizer (the abscissa was time t, the ordinate was the outlet concentration Ct) was made, and it was integrated. The H2S absorption capacity Qt of the desulfurizer HPMo11Co at time t was determined by the above integration results, the H2S flow rate, the H2S inlet concentration C0, and the quality of the desulfurizer HPMo11Co:   10 showed kinetic absorption and fitting curves. It could be observed that the H2S absorption of HPMo11Co rose rapidly in a short period of time, then the growth rate gradually slowed down, and finally reached saturation. This was consistent with the curve results of the H2S removal efficiency. The concentration of the desulfurizer in the later test decreased, resulting in the reduction of H2S removal efficiency. Finally, the H2S absorption of the desulfurizer reached 11.5 mol-H2S g -1 -HPMo11Co.
In order to further explore the relationship between the H2S absorption capacity Qt and the time t, pseudo-first-order (Eq. (3)) and pseudo-second-order (Eq. (4)) kinetic models were used to fit the experimental data: where Qe (mol-H2S g -1 -HPMo11Co) is the equilibrium absorption capacity of HPMo11Co; Qt (mol-H2S g -1 -HPMo11Co) is the absorption capacity at different time t (min); k1 and k2 are the parameters of pseudo-first-order and pseudo-second-order kinetic models, respectively. The pseudo-first-order and pseudo-second-order fitting curves of HPMo11Co are displayed in Fig. 10, and the fitting constants are given in Table 3. It can be seen that the experimental data can be well described by pseudo-first-order kinetic model with the correlation coefficients (R 2 = 0.9963). In addition, the fitted absorption capacity of the pseudo-first-order model is 32.89 mol-H2S g -1 -HPMo11Co, which is closer to the experimental absorption capacity (11.5 mol-H2S g -1 -HPMo11Co) than that of the pseudo-second-order model (60.93 mol-H2S g -1 -HPMo11Co). Therefore, pseudo-first-order kinetic model is predominant in the absorption process of H2S in HPMo11Co solution.

Desulfurization Products and Desulfurization Mechanism
FT-IR spectroscopy was used to analyze functional groups of HPMo11Co before and after desulfurization and regeneration. As shown in Fig. 11, the presence of four characteristics peaks of the heteropoly acid anion after absorption and regeneration, indicating that the compound had stable Keggin-type structure.
The solid powder obtained by evaporating the HPMo11Co before and after absorption and regeneration was used for XPS analysis. Fig. 12 showed the XPS spectra of the Mo 3d5/2 orbital on the surface of the HPMo11Co compound before absorption as well as before and after regeneration. It can be seen that the valence of heteropoly compounds did not change after the whole process of absorption and regeneration, but it did not mean that Mo atoms did not participate in the reaction (Zhao et al., 1996(Zhao et al., , 1998. The whole process was as follows: before absorption, the valence of Mo atom in HPMo11Co was +VI, and the binding energy of Mo 3d5/2 was 233.09 eV; after absorption, the binding energy of Mo 3d5/2 changed and increased by 230.72 eV, indicating that Mo became +IV valence. Whereas after high temperature air blow-off regeneration, the binding energy of Mo 3d5/2 recovered to 233.42 eV, and Mo recovered to +VI Fig. 11. FT-IR spectra of PrPMo HPMo11Co before and after desulfurization and regeneration. Fig. 12. The XPS spectra of Mo 3d5/2 in HPMo11Co before absorption as well as before and after regeneration. valence. XPS spectral characterization results showed that during the absorption of H2S, Mo (+VI) in HPMo11Co was reduced to Mo (+IV), and then oxidized to Mo (+VI) in the process of hightemperature air stripping.
The XPS patterns of the Co 2p orbital of the HPMo11Co compound surface before absorption as well as before and after regeneration are shown in Fig. 13. Before absorption, the XPS spectra of Co 2P of the sample had two absorption characteristic peaks with binding energies of 783.27 eV and 799.33 eV respectively, indicating that there are two different valence states of Co in the sample before absorption, namely Co (+II) and Co (+III). After H2S absorption, the binding energy of Co 2p became 780.78 eV, indicating that the valence of Co in this compound was +II. After high temperature air blow-off regeneration, the XPS spectra of Co 2P of the sample recovered two absorption characteristic peaks with the binding energies of 783.32 eV and 799.04 eV, indicating that Co in the sample was oxidized and its valence state was restored. Based on the above characterization results, Co (+III) in HPMo11Co was reduced to Co (+II) during the absorption process, and then regenerated to Co (+III) through high-temperature air stripping (Khassin et al., 2001;Le et al., 2016).
The XPS spectra of the S 2p orbital of solid powder obtained by evaporating HPMo11Co before and after regeneration are shown in Fig. 14. Among them, a characteristic absorption peak with the binding energy of 162.32 eV illustrated that H2S had been absorbed. At this time, the valence of the element S was -II. The characteristic absorption peak with the binding energy of 163.76 eV indicated that S (-II) had changed into S (0). The characteristic absorption peak at 169.07 eV indicated the presence of S (+VI) in the solution. After high-temperature air stripping regeneration, the characteristic absorption peak areas of S and S (+VI) increased significantly, indicating that the desulfurization product contained S elementary substance and S (+VI) (Liu et al., 2017;Liu and Wang, 2017b, c). The BaCl2 solution was added to the regenerated solution to prove that S (+VI) was SO4 2-, since white precipitation was observed. The calculation of the peak areas showed that elemental sulfur S accounted for 55%, and SO4 2accounted for 45%. Therefore, the peak area ratio was 11:9, so that the molar ratio of S and SO4 2was estimated to be approximately 11:9. S was not oxidized to produce SO4 2-. At the beginning, the concentration of transition metal monosubstituted heteropoly compounds (M = Co 2+ , Mn 2+ , Ni 2+ and Zn 2+ ) was high and the oxidation was strong, which could convert hydrogen sulfide into SO4 2-. As the reaction proceeded, the concentration of transition metal mono-substituted heteropoly compounds decreased and the oxidation became weak, and hydrogen sulfide could be converted into S. In addition, carbon disulfide (CS2) is known to be a very good solvent of sulfur, which can extract elemental sulfur from desulfurization products at room temperature, so as to achieve the purpose of separating S and SO4 2-.   We used XPS to analyze surface element compositions of HPMo11Co before absorption as well as before and after regeneration, including Co, Mo, P, and S elements, and calculated the relative composition content as shown in Table 4. It can be seen that the surface of HPMo11Co did not contain S element before the absorption of H2S, but after the absorption of H2S, the content of S element on the surface of HPMo11Co increased to 48.7%, and the content of S element on the surface of HPMo11Co was 15.1% after regeneration. This indicated that the surface of the regenerated HPMo11Co was covered by a small amount of S, leading to a decrease in the desulfurization rate after regeneration.
The absorption performance of H7PMo11CoO39 for H2S was compared with the previous desulfurizers as listed in Table 5. H7PMo11CoO39 exhibited higher efficiency of H2S removal than other desulfurizers, illustrating that the absorption performance of metal mono-substituted heteropoly compounds had been improved in this work.

CONCLUSION
We studied the desulfurization performance of a series of transition metal monosubstituted heteropoly phosphomolybdate and mono-vacancy heteropoly phosphomolybdate under different reaction conditions. The major results were concluded as follows: (1) Multiple characterization results confirmed the successful synthesis of H3PMo12O40, H7PMo11O39 and H7PMo11MO39 (M = Co 2+ , Mn 2+ , Ni 2+ and Zn 2+ ). (2) H7PMo11CoO39 had the best desulfurization performance and regeneration effect and was selected as the most suitable desulfurizer. Its solution had the best desulfurization capability at 25°C. The removal efficiency of hydrogen sulfide by the desulfurizer could reach 90% within 400 min. The higher the concentration of H7PMo11CoO39, the higher the removal efficiency of hydrogen sulfide. (3) Low H2S concentration was conducive to the progress of the desulfurization reaction.
H7PMo11CoO39 showed the highest desulfurization efficiency at pH 5. The study on the desulfurization absorption kinetics under optimal conditions showed that the process fitted into the pseudo-first-order kinetic model, and the correlation coefficient (R 2 ) was 0.9963. (4) The final products of desulfurization were S and SO4 2-. H7PMo11CoO39 played a redox role in the desulfurization process, which could convert H2S into sulfur resources.