Research into the reaction process and the effect of reaction conditions on the simultaneous removal of H2S, COS and CS2 at low temperature

In this work, tobacco stem active carbon (TSAC) catalysts loaded on to CuO and Fe2O3 were prepared by a sol–gel method and used for the simultaneous removal of hydrogen sulfide (H2S), carbonyl sulfide (COS) and carbon disulfide (CS2). The influences of the operating conditions such as reaction temperature, relative humidity (RH), O2 concentration, and gas hourly space velocity (GHSV) were discussed. DRIFTS results showed that the deactivation was attributed to the generation of S and sulfates. H2O promoted the generation of sulfate. The enhancement of the hydrolysis of COS/CS2 was due to the promotion of H2S oxidation by O2. A high GHSV decreased the contact time between the gases and the catalyst. Meanwhile, a high GHSV was not conducive to the adsorption of gases on the surface of the catalyst. XPS results indicated that the deactivation of the catalyst was attributable to the formation of S containing components, such as thiol/thioether, S, –SO– and sulfate. BET results indicated that the adsorptive ability of the catalyst was related to the microporous volume and surface area.


Introduction
H 2 S, COS and CS 2 are major sulfur-containing compounds produced from industrial tail gas, such as closed carbide furnace tail gas. 1-5 H 2 S, COS and CS 2 not only pollute the natural environment but also cause corrosion to reactors, and poison catalysts. 6,7 Meanwhile, there is a small amount of vapor and oxygen in the closed carbide furnace tail gas. A catalytic hydrolysis method has been widely developed to remove COS and CS 2 from industrial tail gas. 3,8,9 The hydrolysis product H 2 S can be removed by a catalytic oxidation method. Raw materials that have undergone desulfurization are less hazardous and corrosive, and thus can be used to produce other products. Therefore, it is of great signicance to remove H 2 S, COS and CS 2 .
In previous studies, many researchers focused on the inuence of catalyst characterization on the removal of H 2 S, COS and CS 2 . 1,10-14 However, the inuence of reaction conditions is also important for the removal of H 2 S, COS and CS 2 . Song et al. investigated the inuence of reaction conditions on the hydrolysis of COS and CS 2 . 15,16 The results showed that reaction conditions affected the reaction rate and the reaction product species. Furthermore, the reaction conditions directly affected the reaction process, such as the change of surface functional groups. In our previous studies, it could be found that -OH, -COO, and -C]O groups played important roles in the desulfurization process. [17][18][19] -OH promoted the hydrolysis of COS and CS 2 , and -COO and -C]O groups promoted the oxidation of H 2 S. [20][21][22] Furthermore, CO 2 could be converted into -COO and -C]O groups during this process, which enhanced the removal of H 2 S. However, there are few studies on the simultaneous removal of H 2 S, COS and CS 2 , and the detailed changes in surface functional groups during the desulfurization process under different reaction conditions were unknown. Therefore, this study is important and valuable.
Yunnan province is the main area that produces tobacco. However, a large number of tobacco stems are discarded every year. Previous studies showed that tobacco stems could be used for the preparation of biochar, and this showed a high adsorption ability. 23,24 Therefore, the preparation of tobacco stem biochar could solve the disposal problems of waste tobacco stems. In this work, the inuence of reaction conditions (reaction temperature, relative humidity (RH), O 2 content, and gas hourly space velocity (GHSV)) on the removal of H 2 S, COS and CS 2 was investigated. Meanwhile, the inuence of reaction conditions on the change of surface functional groups was analyzed by DRIFTS (diffuse reectance infrared Fourier transform spectroscopy), BET (surface area and pore structure analysis) and XPS (X-ray photoelectron spectroscopy).

Catalyst preparation
The raw material of walnut shell biochar was from Yunnan province. The main preparation parameters were rstly that the tobacco stem was washed twice with water and smashed to 4 mesh size for use in this study. Then, the walnut shell was calcined at 700 C for 1 h under nitrogen (N 2 ) conditions, and sieved to 40-60 mesh size. Aer that, the carbonized material and the activator (CO 2 ) were mixed together, and calcined at 800 C for 1 h.
Secondly, a colloidal solution was made with certain amounts of Cu(NO 3 ) 2 $3H 2 O solution, Fe(NO 3 ) 3 $6H 2 O solution and K 2 CO 3 solution. The activated carbon catalysts were supported by the desired proportions (mass fraction of CuO was 10% and Cu/Fe ¼ 10/1). Then, the samples were dipped into ultrasonic conditions for 30 min, dried at 100 C in the drying oven and calcined at 400 C at a heating rate of 5 C min À1 for 4 h under nitrogen (N 2 ) conditions. Lastly, the catalysts were impregnated by 5% (mass fraction) KOH, and kept under ultrasonic conditions for 10 min, then dried for 6 h at 100 C in the drying oven to get the catalyst (Cu-Fe/TSAC) needed for the experiments. Cu-Fe/TSAC showed a high desulfurization efficiency, and the sulfur capacity was 231.28 mgS g À1 .

Catalytic activity measurements
Desulfurization tests were performed in a xed-bed quartz reactor (3 mm inside diameter, 140 mm length) under atmospheric pressure (Fig. 1). H 2 S, COS and CS 2 from gas cylinders (1% H 2 S in N 2 ; 1% COS in N 2 ; 0.3% CS 2 in N 2 ) were diluted with N 2 (99.99%) to the required concentrations (H 2 S: 500 ppm; COS: 400 ppm; CS 2 : 60 ppm). The gas hourly space velocity (GHSV) of the reaction mixture was standardized at 10 000-20 000 h À1 . The water comes from a saturator system, and the relative humidity (RH) was 0-60%. The reaction temperature of this reactor was controlled at 50-70 C by a water-bath with a circulating pump, with an accuracy of AE0.1 C. FULI 9790II gas chromatography was used to analyze the total H 2 S, COS and CS 2 concentrations of the gaseous feed and effluent from the reactor. The conversion rates of H 2 S, COS and CS 2 are achieved according to eqn (1).
The sulfur capacity (mgS g À1 catalyst) is dened as the sulfur deposition per unit mass of desulfurizer agent in H 2 S, COS and CS 2 between time points (ending at 85% conversion).

Characterization
DRIFTS spectra were collected using a Nicolet iS50 FTIR spectrometer equipped with a smart collector. Mass ow controllers were used to control the volume ow of different gases to the required concentrations. The heating cable controlled the temperature (70 C) of mixed gas until it entered the reactor. A reactor heater controlled the temperature (70 C) of the reactor in the DRIFTS experiments. In this case, it ensured that the reaction temperature of the gas phase and the solid phase are the same. IR spectra were recorded by accumulating 100 scans at a resolution of 4 cm À1 . Nitrogen adsorption-desorption isotherms were obtained by a Quantachrome surface area analyzer instrument. Before the measurement, the samples were outgassed under vacuum at 393 K for 24 h. Specic surface areas, and mesoporous and micropore adsorption-desorption isotherms were calculated by Brunauer-Emmett-Teller (BET), Barret-Joyner-Halenda (BJH) and Horvath-Kawazoe (HK) methods, respectively. XPS (ESCALAB 250) analysis was performed using Al Ka radiation, where the energy of the Al target powered was 200 W.

Effect of reaction temperatures on the simultaneous removal of H 2 S, COS and CS 2
The inuence of reaction temperatures on the catalytic performance of the Cu-Fe/TSAC catalyst is illustrated in Fig. 2. The conversion of H 2 S, COS and CS 2 rst increased and then decreased with increasing temperature, and was highest at 60 C. The H 2 S, COS and CS 2 conversion was 100% in the initial 600, 150 and 180 min respectively. As shown in Fig. 2(d), the sulfur capacity rst increased and then decreased with increasing reaction temperatures. The highest sulfur capacity (231.28 mgS g À1 ) was achieved at 60 C.
The reaction rates of simultaneous catalytic hydrolysis of COS and CS 2 and catalytic oxidation of H 2 S were poor at low temperatures. Therefore, with increasing reaction temperature, the sulfur capacity was increased and the reaction rate of catalytic hydrolysis can be increased, and the hydrolysis reaction could occur more easily. However, the conversion of H 2 S to S or sulfate on Cu-Fe/TSAC involves parallel reactions. With increasing reaction temperatures, the yield rate of sulfuric acid increases faster than that of sulfur. At higher temperatures, H 2 S can be oxidized to sulfate more easily, and the higher concentration of SO 4 2À poisons the hydrolysis activity. 25 The majority of the products on the exhausted Cu-Fe/TSAC were S/SO 4 2À species which accumulated on the active carbon's surface and had a negative effect on the hydrolysis activity. Thus, the removal efficiency of H 2 S, COS and CS 2 declined sharply at 70 C. DRIFTS results were used to further study the catalytic reaction of Cu-Fe/TSAC at different temperatures. As shown in Fig. 3, the Cu-Fe/TSAC surface generates CO 2 (2363 cm À1 ), C]O groups (1604 cm À1 ), C-S groups (2080 cm À1 ) and S-O groups (1140 cm À1 and 1307 cm À1 ) as the reaction proceeds. 1,26,27 Furthermore, CO 2 was produced as the reaction time progressed, which can prove that the reaction is indeed the hydrolysis of COS and CS 2 . The formation of S-O groups can prove that the H 2 S was oxidized. Compared with reactions at 60 C and 70 C, fewer S-O groups were generated at 50 C. This indicated that a temperature of 50 C was not conducive to the oxidation of H 2 S. Compared with reactions at 50 C and 60 C, more S-O groups and fewer C-S groups were generated at 70 C. This indicated that a temperature of 70 C enhanced the  hydrolysis of COS/CS 2 and the oxidation of H 2 S. However, excessive oxidation of H 2 S could generate more sulfate, which could lead to the deactivation of the catalyst. At the temperature of 60 C, the number of C]O groups increased and the amount of CO 2 decreased, which indicated that CO 2 could be converted into C]O groups in the reaction. As a result, the catalyst has a good adsorptive ability for COS/CS 2 /H 2 S and a good oxidation ability for H 2 S over time. Therefore, a temperature of 60 C is conducive to the hydrolysis of COS/CS 2 and the oxidation of H 2 S. The result was in accordance with the activity experiment.

Effects of RH on the simultaneous removal of H 2 S, COS and CS 2
The effects of RH on H 2 S, COS and CS 2 removal were studied by introducing feed gas through a humidier. Inuences of different RHs on the catalytic performance are plotted in Fig. 4. Removal efficiency for H 2 S, COS and CS 2 rst increased and then decreased with increasing RH. Low RH should benet the hydrolysis and oxidation activities. When the RH was 49%, the catalyst showed the best activity, as 100% H 2 S, COS and CS 2 conversion was maintained for about 600 min, 150 min and 180 min respectively. As shown in Fig. 4(d), the sulfur capacity rst increased and then decreased with increasing RH. The sulfur capacity was highest (231.28 mgS g À1 ) when the RH was 49%. The sulfur capacity decreased to 176.50 mgS g À1 at the RH of 60%. The selective catalytic oxidation of H 2 S to S or HS À will be easier in the presence of less vapor. The fact that excessive water could restrain catalytic activity might be due to competition between H 2 S (COS or CS 2 ) and vapor for the same active sites of the catalyst. 28 Another reason is that the pores of the catalyst's surface will form water lms when the RH reaches a certain amount. Although the formation of water lms would provide more accommodating spaces for the product, excessive water lms may stop H 2 S, COS and CS 2 diffusing on the hydrolysis center and inhibit the catalytic hydrolysis reaction. 29 In order to further investigate the effect of different RHs on the conversion of H 2 S, COS and CS 2 over Cu-Fe/TSAC, DRIFTS measurements of the catalytic reactions over Cu-Fe/TSAC were taken. As shown in Fig. 5, there was no obvious CO 2 peak when the reaction was performed at 0% RH. This indicated that the removal of COS/CS 2 was an adsorption process without H 2 O. 1 Aer introducing H 2 O (49% RH), a peak due to CO 2 appeared over time. This proved that the removal of COS/CS 2 with H 2 O was due to a hydrolysis process. Furthermore, more S-O groups were generated when the RH was 49%, which indicated that H 2 O promoted the generation of sulfate. Meanwhile, the decrease in CO 2 and the increase in C]O groups indicated that H 2 O promoted the conversion of CO 2 . It can be deduced that the catalytic hydrolysis reaction occurs on the surface of Cu-Fe/TSAC, where the hydrolysis of COS and CS 2 produces CO 2 and H 2 S.

Effect of O 2 content on the simultaneous removal of H 2 S, COS and CS 2
The curves plotted in Fig. 6 show the effect of O 2 content on the removal efficiency of H 2 S, COS and CS 2 . It is very difficult to control the oxygen content below 0.5%, although we wanted to investigate the lower O 2 content. So the non-oxygen conditions were investigated. What is more, the O 2 content is extremely low, even close to zero, in closed carbide furnace tail gas which is a reductive atmosphere. The catalytic removal efficiency initially increased and then decreased with increasing O 2 content. When the O 2 content was 0.5%, the removal efficiency of H 2 S, COS and CS 2 was highest. 100% H 2 S, COS and CS 2 conversion was maintained for about 600 min, 150 min, 180 min respectively. As shown in Fig. 6(d), the sulfur capacity rst increased and then decreased with the increase in O 2 content. The sulfur capacity was 231.28 mgS g À1 when the O 2 content was 0%. When the O 2 content was 0.5%, the sulfur capacity increased to 239.18 mgS g À1 . With a further increase in O 2 content, the sulfur capacity decreased. The sulfur capacity was only 133.98 mgS g À1 when the O 2 content was 5%.
It is clear that the addition of a little O 2 enhanced the catalytic activity in the COS and CS 2 hydrolysis. The reason may be that sufficient oxygen can increase the oxidation of H 2 S and promote the catalytic hydrolysis of COS and CS 2 . The oxidation rate of H 2 S speeds up with the increasing oxygen content. This will lead to the generation of more sulfate, with the inhibition effect greater than the promotion effect.
In order to further investigate the effect of different amounts of O 2 on the conversion of H 2 S, COS and CS 2 from Cu-Fe/TSAC, DRIFTS measurements of the catalytic reaction over Cu-Fe/ TSAC were taken. As shown in Fig. 7, more S-O groups (related to C]O groups) appeared when the reaction was performed below 5% O 2 . This indicated that O 2 was conducive to the removal of H 2 S due to the oxidation of H 2 S. 27 Meanwhile, fewer C]O groups appeared when the O 2 content was 5%, which indicated that O 2 mainly promoted the oxidation of H 2 S but not the conversion of C]O groups.

Effect of the gas hourly space velocity on the simultaneous removal of H 2 S, COS and CS 2
As shown in Fig. 8, the removal efficiency for H 2 S, COS and CS 2 decreases with increasing GHSV. 100% H 2 S, COS and CS 2 conversion was maintained for about 600 min, 150 min and 180 min respectively when the GHSV was 10 000 h À1 . As shown in Fig. 8(d), the sulfur capacity was 231.28 mgS g À1 when the GHSV was 10 000 h À1 . The sulfur capacity decreased with increasing GHSV, and the capacity was 165.55 mgS g À1 when the GHSV was 20 000 h À1 . At low GHSV, more gases could be adsorbed on the surface of the catalyst. As a result, the catalytic hydrolysis and catalytic oxidation reactions could occur fully. However, a high GHSV decreased the contact time between the gases and the catalyst. This led to the decrease of reaction time and conversion efficiency. Meanwhile, a high GHSV was not conducive to the adsorption of gases on the surface of the catalyst. This further decreased the catalytic efficiency.

BET results and reaction process analysis
Nitrogen adsorption isotherms and microporous size distribution for fresh and deactivated Cu-Fe/TSAC are shown in Fig. 9 and Table 1. As shown in Fig. 9 and Table 1, Cu-Fe/TSAC had the characteristics of surface area (554 m 2 g À1 ), microporous volume (0.21 cm 3 g À1 ) and total pore volume (0.29 cm 3 g À1 ). The N 2 adsorption quantity, microporous volume and surface area in the deactivated catalyst obviously decreased, which indicated that the adsorptive ability of the catalyst decreased over time. This affected the desulfurization ability of the catalyst. The XPS characterization results and the data of fresh and deactivated Cu-Fe/TSAC (S2p) are shown in Fig. 10 and Table 2. From the XPS results, it could be found that the thiol/thioether, S, -SO-, and sulfate amounts increased from 0% to 1.92%, 2.90%, 0.83% and 1.25% respectively. This indicated that the deactivation of  the catalyst was attributed to the formation of S containing components, such as thiol/thioether, S, -SOand sulfate. Combined with the BET results, the formation of Cu 2 O led to the decrease of the microporous volume and surface area.
According to previous experimental results, the removal processes of COS, CS 2 and H 2 S could be divided into two parts: the catalytic hydrolysis reaction and the catalytic oxidation reaction. In the removal processes, COS and CS 2 were rst hydrolyzed into H 2 S, and then H 2 S was oxidized into S/sulfates. Combined with previous DRIFTS results, the removal processes of COS and CS 2 were different. For the removal of CS 2 , CS 2 and H 2 O were rst adsorbed on the surface of the catalyst, and then the hydrolysis reaction occurred under the effect of surface functional groups and CuO. However, COS had not been found on the deactivated catalyst, which indicated that gaseous COS directly reacted with adsorbed H 2 O under the effect of surface functional groups and CuO. Furthermore, the conversion process for H 2 S could be regarded as H 2 S / S / SO 2 / SO 4 2À . As mentioned above, the possible reaction process diagram is shown in Fig. 11. As shown in Fig. 11, the main reactions on the surface of the catalyst were as follows.
In these reactions, * represents the hydrolysis activity center and Q represents the oxidation activity center. Eqn (2)-(9) are the catalytic hydrolysis processes and eqn (10)-(12) are the catalytic oxidation processes.

Conclusions
Modied tobacco stem active carbon (Cu-Fe/TSAC) was prepared by a sol-gel method, and tested for the simultaneous removal efficiency of H 2 S, COS and CS 2 . The inuences of reaction conditions for the removal of H 2 S, COS and CS 2 were investigated. A high reaction temperature improved the hydrolysis and oxidation efficiency, but accelerated the deactivation of the catalyst. DRIFTS results indicated that the deactivation was attributed to the generation of S and sulfates. Excessive water lms decreased the diffusion of H 2 S, COS and CS 2 on the hydrolysis center, and inhibited the catalytic hydrolysis reaction. DRIFTS results indicated that H 2 O promoted the generation of sulfate. Appropriate O 2 content directly promoted the oxidation of H 2 S, and indirectly promoted the hydrolysis of COS and CS 2 . DRIFTS results indicated that the enhancement of hydrolysis of COS/CS 2 was attributable to the promotion effect of O 2 for H 2 S oxidation. A high GHSV decreased the contact time between the gases and the catalyst. Meanwhile, a high GHSV was not conducive to the adsorption of gases on the surface of the catalyst. XPS results indicated that the deactivation of the catalyst was attributed to the formation of S containing components, such as thiol/thioether, S, -SO-, and sulfate. BET results indicated that the adsorptive ability of the catalyst was related to the microporous volume and surface area.

Conflicts of interest
There are no conicts to declare.