Regeneration Mechanism of Sulfur Absorption Via Samarium-doped Cerium Adsorbents in the Gas Atmosphere of O2/N2

Sulfides existing in many high-temperature gas mixtures have a negative effect on various industrial applications. Ce-based adsorbents are becoming a hotspot in the high-temperature desulfurization process owing to their excellent thermal stability at high temperatures and regeneration capacity. In this study, we investigate the regeneration path of samarium-doped cerium (SDC) sorbent at high temperature. The SDC adsorbent showed a good sulfur removal ability and excellent regeneration capacity. Ce2O2S and Ce(SO4)2 are observed in the used adsorbent, and Ce2O2S is the main sulfur-containing species. The regeneration path of the Ce2O2S is the key to the regeneration mechanism of the adsorbent. There are two regeneration paths for the Ce2O2S at high temperature in O2/N2 gas mixture. In air stream, the Ce2O2S is oxidized to Ce2O2SO4 and then decomposes into CeO2 and SO2. In a 2% O2/N2 gas condition, the Ce2O2S directly generates CeO2 and elemental sulfur with O2 assistance.


Introduction
With the rapid development of human society and fossil-fuel consumption increasing, serious environmental problems are gradually becoming significantly conflicted with economic development. Coal is one of the most sufficient fossil energy resources with inexpensive cost, and the consumption of coal has accounted for over 20% of consumed fossil fuels in the world [1,2]. Integrated gasification combined cycle (IGCC) power-plant is one of the strategies for efficient utilization of the coal. In the IGCC system, coal is first converted to CO and H 2 (synthesis gas) and then further reformed to H 2 and CO 2 via the water gas shift reaction [3,4]. In this process, synthesis gas can be considered a raw feedstock to produce the value-added chemicals and fuels, and H 2 as a high-density energy is used for producing energy. However, elemental sulfur in the coal is converted into hydrogen sulfide, and the presence of hydrogen sulfide seriously corrodes subsequent systems piping and catalytic systems. Therefore, it must be removed after the hot coal gas generation.
The H 2 S mixed in the high-temperature gas is often cleaned by a conventional amine solution method at a lower temperature (i.e., cold coal gas desulfurization) [5]. The operating temperature of the amine solution is below 150 • C. Thus, it will result in additional thermosteresis due to the cooling and heating process. Another method is to adsorb H 2 S by solid sorbents at high temperature (i.e., hot-gas desulfurization (HGD)) [6,7]. Therefore, HGD process (650~900 • C) is an efficient method for hot syngas purification. The overall thermal efficiency and power efficiency of the IGCC system will increase about

Desulfurization and Regeneration Assessments
The flow diagram of the desulfurization or/and regeneration system is presented in Figure 1. The sorbents were located in a quartz tube with an inner diameter of 10 mm at normal pressure using simulated hot coal gas (3000 mg/m 3 H 2 S, 10% H 2 , 20% CO, and N 2 balance gas). In each case, 500 mg sorbent was used for the test. The sorbents were heated to target temperature in nitrogen atmosphere at a rate of 5 • C min −1 . Subsequently, the simulated gas mixture was introduced into the quartz tube reactor for desulfurization test. The weight hourly space velocity (WHSV) was 12 L h −1 g −1 , which was controlled by mass flow controllers (D07-9F/YCM, produced by Seven-star Electronics Co., Beijing, China). H 2 S concentration was detected by the iodometric method.
The outlet changes of H 2 S concentration with time can be expressed by a breakthrough curve. The breakthrough sulfur capacity (BSC) was denoted as the content of sulfur removed by the adsorbent at the breakthrough point, which can be used to assess the sorbents ability for sulfur removal. It can be evaluated by the following formula: where SC is the effective sulfur capacity of adsorbent (g S/100 g adsorbent); WHSV is the weight per hour space velocity (L h −1 g −1 ); M S and M H 2 S are the molar weight of sulfur (32.06 g mol −1 ) and H 2 S (34.06 g mol −1 ), respectively; V m is the molar volume of H 2 S at 1 atm and 25 • C (24.5 L mol −1 ); t is the desulfurization reaction time (h); C in is the inlet concentration of sulfur dioxide (mg/m 3 ), while C out is the outlet concentration.
After the adsorbent desulfurization, the used sorbent was regenerated at 800 • C with a heating rate of 10 • C/min. The gas atmosphere was an air stream with a WHSV of 12 L h −1 g −1 . The regeneration process was stopped until SO 2 could not be detected. The system was flushed by N 2 stream for 1 h after the regeneration process. Then, the new cycle process began. Each value of the sulfur capacity is the average of three measurements.
Materials 2020, 13, x FOR PEER REVIEW 3 of 13 sorbent was used for the test. The sorbents were heated to target temperature in nitrogen atmosphere at a rate of 5 °C min −1 . Subsequently, the simulated gas mixture was introduced into the quartz tube reactor for desulfurization test. The weight hourly space velocity (WHSV) was 12 L h −1 g −1 , which was controlled by mass flow controllers (D07-9F/YCM, produced by Seven-star Electronics Co., Beijing, China). H2S concentration was detected by the iodometric method. The outlet changes of H2S concentration with time can be expressed by a breakthrough curve. The breakthrough sulfur capacity (BSC) was denoted as the content of sulfur removed by the adsorbent at the breakthrough point, which can be used to assess the sorbents ability for sulfur removal. It can be evaluated by the following formula: where is the effective sulfur capacity of adsorbent (g S/100 g adsorbent); is the weight per hour space velocity (L h −1 g −1 ); and are the molar weight of sulfur (32.06 g mol −1 ) and H2S (34.06 g mol −1 ), respectively; is the molar volume of H2S at 1 atm and 25 °C (24.5 L mol −1 ); is the desulfurization reaction time (h); is the inlet concentration of sulfur dioxide (mg/m 3 ), while is the outlet concentration.
After the adsorbent desulfurization, the used sorbent was regenerated at 800 °C with a heating rate of 10 °C/min. The gas atmosphere was an air stream with a WHSV of 12 L h −1 g −1 . The regeneration process was stopped until SO2 could not be detected. The system was flushed by N2 stream for 1 h after the regeneration process. Then, the new cycle process began. Each value of the sulfur capacity is the average of three measurements.

Sulfur Collection Test
The sulfur collection experiment was conduct in a quartz tube. Air (10 ml/min, STP) or 2% O2/N2 gas mixture (5 ml min −1 ) was introduced to the quartz tube for the regeneration of the Ce-O-S powder. The sample with different weights (250, 500, 1000, or 1500 mg) was calcined at 800 °C for 6 h in air. The off-gas was finally immersed into an adsorption setup (two adsorption bottles loaded with cold water) to collect SO2 or elemental S. We use classic acid-base titration to measure the acid yield.

Characterizations
The structures of all the sorbents were analyzed by X-ray diffraction (XRD, Bruker D8) equipped with Copper-Ka radiation. The scan angle (2θ) was collected from 20° to 80° with a scan rate of 5° min −1 . X-ray photoelectron spectroscopy with an Al Kα X-ray (XPS, Perkin-Elmer model PHI 5600 system) analyzed the surface compositions of the samples. 500 mg Ce-O-S powder was heated up to 800 °C with a rate of 5 °/min in air atmosphere (10 ml min −1 air stream, STP), and the end-gas was tested by Fourier transform infrared spectroscopy (FTIR) (Thermo Fisher Scientific, Nicolet 6700)

Sulfur Collection Test
The sulfur collection experiment was conduct in a quartz tube. Air (10 mL/min, STP) or 2% O 2 /N 2 gas mixture (5 mL min −1 ) was introduced to the quartz tube for the regeneration of the Ce-O-S powder. The sample with different weights (250, 500, 1000, or 1500 mg) was calcined at 800 • C for 6 h in air. The off-gas was finally immersed into an adsorption setup (two adsorption bottles loaded with cold water) to collect SO 2 or elemental S. We use classic acid-base titration to measure the acid yield.

Characterizations
The structures of all the sorbents were analyzed by X-ray diffraction (XRD, Bruker D8) equipped with Copper-Ka radiation. The scan angle (2θ) was collected from 20 • to 80 • with a scan rate of 5 • min −1 .
Materials 2020, 13, 1225 4 of 13 X-ray photoelectron spectroscopy with an Al Kα X-ray (XPS, Perkin-Elmer model PHI 5600 system) analyzed the surface compositions of the samples. 500 mg Ce-O-S powder was heated up to 800 • C with a rate of 5 • /min in air atmosphere (10 mL min −1 air stream, STP), and the end-gas was tested by Fourier transform infrared spectroscopy (FTIR) (Thermo Fisher Scientific, Nicolet 6700) with scanning range from 4000 to 600 cm −1 . The thermodynamic behaviors were conducted by thermo gravimetric analysis (TGA, NETZSCH STA 449 F3). Around 14.5 mg spent adsorbent or pure Ce(SO 4 ) 2 ·4H 2 O was analyzed by TGA. The sample was heated in air condition from room temperature to 900 • C with a heating rate of 8 • C/min. Figure 2 shows the H 2 S removal capacity and breakthrough curves of pure CeO 2 and SDC sorbents at 800 • C. As shown, compared with pure CeO 2 sorbent, SDC sorbent shows a long breakthrough time, suggesting that doping Sm can improve the sulfur adsorption capacity of CeO 2 . The maximal sulfur capacity of CeO 2 and SDC is 7.9 and 12.1 g S/100g sorbent, respectively. Figure 3 shows the regenerated ability of the SDC sorbent at 800 • C. As shown, after six continuous cycles, the sulfur capacity of regenerated adsorbent is similar to the fresh sorbent. It can be seen that there is a slight decline in desulfurization capacity after the first cycle. The tiny loss of sulfur capacity can be ascribed to sintering of the sorbent and the active components aggregation, as a large amount of heat is released during the regeneration process [18]. The breakthrough sulfur capacity changes ranging from 10.1 to 11.9 g S/100g. In addition, the deactivation curves after the breakthrough point have a similar tendency. This suggests that SDC is a thermal-stable desulfurization sorbent. Table 1 shows the microstructural properties of sorbents. As shown, the BET surface area of the CeO 2 is only 64 m 2 g −1 . The total pore volume and pore size values of the CeO 2 are 0.117 cm 3 g −1 and 9.86 nm, respectively. As for fresh SDC sorbent, the surface area can reach up to 271 m 2 g −1 , and the total pore volume is 0.362 cm 3 g −1 , causing a superior desulfurization performance ( Figure 2). This suggests that the doped Sm plays a positive role in improving the CeO 2 surface area. In addition, the microstructural properties of the regenerated SDC sample are also investigated. As shown, an obvious phenomenon could be found for the BET results. Basically, the surface area has a decreasing tendency, which is from 269 to 245 m 2 g −1 after six cycle tests. However, for the crystallite size (D) and average pore size (P), an increasing tendency could be found, suggesting that the SDC sorbent exhibits a slight sintering phenomenon during the six regenerations process at high temperature. with scanning range from 4000 to 600 cm −1 . The thermodynamic behaviors were conducted by thermo gravimetric analysis (TGA, NETZSCH STA 449 F3). Around 14.5 mg spent adsorbent or pure Ce(SO4)2·4H2O was analyzed by TGA. The sample was heated in air condition from room temperature to 900 °C with a heating rate of 8 °C/min. Figure 2 shows the H2S removal capacity and breakthrough curves of pure CeO2 and SDC sorbents at 800 °C. As shown, compared with pure CeO2 sorbent, SDC sorbent shows a long breakthrough time, suggesting that doping Sm can improve the sulfur adsorption capacity of CeO2. The maximal sulfur capacity of CeO2 and SDC is 7.9 and 12.1 g S/100g sorbent, respectively. Figure 3 shows the regenerated ability of the SDC sorbent at 800 °C. As shown, after six continuous cycles, the sulfur capacity of regenerated adsorbent is similar to the fresh sorbent. It can be seen that there is a slight decline in desulfurization capacity after the first cycle. The tiny loss of sulfur capacity can be ascribed to sintering of the sorbent and the active components aggregation, as a large amount of heat is released during the regeneration process [18]. The breakthrough sulfur capacity changes ranging from 10.1 to 11.9 g S/100g. In addition, the deactivation curves after the breakthrough point have a similar tendency. This suggests that SDC is a thermal-stable desulfurization sorbent. Table 1 shows the microstructural properties of sorbents. As shown, the BET surface area of the CeO2 is only 64 m 2 g −1 . The total pore volume and pore size values of the CeO2 are 0.117 cm 3 g −1 and 9.86 nm, respectively. As for fresh SDC sorbent, the surface area can reach up to 271 m 2 g −1 , and the total pore volume is 0.362 cm 3 g −1 , causing a superior desulfurization performance ( Figure 2). This suggests that the doped Sm plays a positive role in improving the CeO2 surface area. In addition, the microstructural properties of the regenerated SDC sample are also investigated. As shown, an obvious phenomenon could be found for the BET results. Basically, the surface area has a decreasing tendency, which is from 269 to 245 m 2 g −1 after six cycle tests. However, for the crystallite size (D) and average pore size (P), an increasing tendency could be found, suggesting that the SDC sorbent exhibits a slight sintering phenomenon during the six regenerations process at high temperature.

Regeneration Mechanism of the SDC Sorbent
The SDC sorbents show a typical fluorite structure, and the main component of SDC is CeO2. Thus, the investigation of the regeneration mechanism of the SDC sorbent can be simplified into the cerium oxide regeneration according to some reports [31,32]. Figure 4 shows the XRD patterns of different powders. Ce2O2S diffraction peaks are clearly observed after desulfurization process of the SDC sorbent as shown in Figure 4a. The main sulfur-containing phase in the Ce-O-S powder is Ce2O2S phase. Apart from Ce2O2S, sulfate can also be found, which is ascribed to the transformation between the Ce2O2S and Ce(SO4)2 after H2O producing. The various Ce-O-S phases come from the following reactions [33]: The diffraction peak of the Re1 powder is the same with pure CeO2 pattern as shown in Figure  4b,c, suggesting that Ce-O-S powder could be regenerated to the CeO2 powder after regeneration process in air. To observe clearly, pure Ce(SO4)2 was treated by the same regeneration process. As shown in Figure 4d,e, the Re 2 powder is confirmed to be CeO2 phase, suggesting that Ce(SO4)2 could also be regenerated to CeO2 after calcination in air stream. Figure 5 displays the XPS spectrum of O 1s of pure CeO2, Ce-O-S, and Re1 powder. Oxygenrelated peaks located at 527.5 ~ 530.0 eV have an approached peak position that belongs to crystal lattice oxygen (OL). The adsorbed oxygen (OA) peaks are observed at 530.0 ~ 535.0 eV, in line with the literature [34]. The contents of the OL of different powders are showed in Table 2. Compared with pure CeO2, the OL/(OL + OA) ratio of the Ce-O-S powder decreases significantly, which is from 91.5% to 4.7%, accompanied by an incremental peak area of adsorbed oxygen (OA). This indicates that OL is the active composition for the H2S removal. As shown in reaction (2), the OL is consumed by the hydrogen sulfide and hydrogen to generate water during the desulfurization process. However, after calcination of Ce-O-S powder in air (Re1 powder), the OL/(OL + OA) ratio is recovered to 75.9% (Table  2), accompanied by an incremental peak area of the OL. Meanwhile, the different sulfur valences are found on the surface of the Ce-O-S powder and Re1 powder, as shown in Figure 6. The representative S peaks of the Ce-O-S powder located at 166.0 ~ 173.0 eV are associated with SO3 2-and SO4 2-species, while the peaks at approximately 162.0 ~ 166.0 eV are attributed to S 2-species [31,35]. After calcination process in air, the representative peaks of S 2− of the Re1 powder disappeared, and only little sulfur

Regeneration Mechanism of the SDC Sorbent
The SDC sorbents show a typical fluorite structure, and the main component of SDC is CeO 2 . Thus, the investigation of the regeneration mechanism of the SDC sorbent can be simplified into the cerium oxide regeneration according to some reports [31,32]. Figure 4 shows the XRD patterns of different powders. Ce 2 O 2 S diffraction peaks are clearly observed after desulfurization process of the SDC sorbent as shown in Figure 4a. The main sulfur-containing phase in the Ce-O-S powder is Ce 2 O 2 S phase. Apart from Ce 2 O 2 S, sulfate can also be found, which is ascribed to the transformation between the Ce 2 O 2 S and Ce(SO 4 ) 2 after H 2 O producing. The various Ce-O-S phases come from the following reactions [33]: The diffraction peak of the Re1 powder is the same with pure CeO 2 pattern as shown in Figure 4b,c, suggesting that Ce-O-S powder could be regenerated to the CeO 2 powder after regeneration process in air. To observe clearly, pure Ce(SO 4 ) 2 was treated by the same regeneration process. As shown in Figure 4d,e, the Re 2 powder is confirmed to be CeO 2 phase, suggesting that Ce(SO 4 ) 2 could also be regenerated to CeO 2 after calcination in air stream. attributed to SO 4 2− is found, as shown in Table 2. Thus, the above results suggest that the CeO 2 -based adsorbent (e.g., SDC in this work) after sulfur adsorption can be regenerated by calcination in air stream. The lattice oxygen is regenerated accompanied by the removal of the sulfur. (1.7%) attributed to SO4 2− is found, as shown in Table 2. Thus, the above results suggest that the CeO2based adsorbent (e.g., SDC in this work) after sulfur adsorption can be regenerated by calcination in air stream. The lattice oxygen is regenerated accompanied by the removal of the sulfur.    (1.7%) attributed to SO4 2− is found, as shown in Table 2. Thus, the above results suggest that the CeO2based adsorbent (e.g., SDC in this work) after sulfur adsorption can be regenerated by calcination in air stream. The lattice oxygen is regenerated accompanied by the removal of the sulfur. Table 1. Surface area (S), total pore volume (V), crystallite size (D), and average pore size (P) of sorbents.     Thermal behaviors of Ce-O-S powder and fresh Ce(SO4)2·4H2O were investigated by TGA. The thermal behaviors of the Ce(SO4)2·4H2O are generally displayed as below [36].

Samples S (m 2 g −1 ) D (nm) V (cm 3 g −1 ) P (nm)
Ce3O2(SO4)4 → 3CeO2 + 4SO3 As shown in Figure 7, the weight loss of Ce(SO4)2·4H2O comes from the H2O removal of the endothermic powder below 300 °C, forming the crystalline Ce(SO4)2. The further weight loss of 13.5% is ascribed to the change from Ce(SO4)2 to Ce3O2(SO4)4 accompanied by the SO3 releasing when the temperature is heated up to 520 °C. The Ce2O(SO4)3 is an intermediate species during the heating process. The final products of CeO2 and SO3 are obtained above 860 °C as shown in reaction (7). However, the TGA result of the Ce-O-S powder is interesting. As shown, the endothermic process of the Ce-O-S powder displays two steps during the whole temperature-rise period. An incremental weight of around 11.3% is observed from 230 to 600 °C, and then, 11.1% weight loss can be found from 650 to 800 °C. From the results of the XRD and XPS, the sulfur species can be removed from the sorbent after calcination in the air. Thus, the weight loss (11.1%) of the Ce-O-S sample is due to the sulfur removal. It can be seen from the results that Ce2O2S is the dominant S-containing species after desulfurization. Thus, the regeneration path of the Ce2O2S is the key to the regeneration mechanism of the sorbent. However, the regenerated path of the Ce2O2S during the heating process in oxidizing atmosphere is relatively complicated [37][38][39][40]. In this study, the paths most likely are two processes, as follows.
Case 1 [37,38]: Ce2O2S(s) + O2(g) → 2CeO2(s)+ 1/2S2(g) (9) As shown in Figure 7, the weight loss of Ce(SO 4 ) 2 ·4H 2 O comes from the H 2 O removal of the endothermic powder below 300 • C, forming the crystalline Ce(SO 4 ) 2 . The further weight loss of 13.5% is ascribed to the change from Ce(SO 4 ) 2 to Ce 3 O 2 (SO 4 ) 4 accompanied by the SO 3 releasing when the temperature is heated up to 520 • C. The Ce 2 O(SO 4 ) 3 is an intermediate species during the heating process. The final products of CeO 2 and SO 3 are obtained above 860 • C as shown in reaction (7). However, the TGA result of the Ce-O-S powder is interesting. As shown, the endothermic process of the Ce-O-S powder displays two steps during the whole temperature-rise period. An incremental weight of around 11.3% is observed from 230 to 600 • C, and then, 11.1% weight loss can be found from 650 to 800 • C. From the results of the XRD and XPS, the sulfur species can be removed from the sorbent after calcination in the air. Thus, the weight loss (11.1%) of the Ce-O-S sample is due to the sulfur removal. It can be seen from the results that Ce 2 O 2 S is the dominant S-containing species after desulfurization. Thus, the regeneration path of the Ce 2 O 2 S is the key to the regeneration mechanism of the sorbent. However, the regenerated path of the Ce 2 O 2 S during the heating process in oxidizing atmosphere is relatively complicated [37][38][39][40]. In this study, the paths most likely are two processes, as follows.
Case 1 [37,38]: Considering Case-1, as shown in (8) and (9), one oxygen molecule will be captured by Ce 2 O 2 S to form CeO 2 -molecules, accompanied by the release of the elemental S during the oxidation process. Because the oxygen molecule weight is equivalent to the weight of half S 2 (molecular weight: O 2 = 1/2 S 2 ), the weight of the powder could be kept stable after the regeneration process. Therefore, for the path of the Case-1, the formed elemental sulfur will firstly be adsorbed on the surface of the powders below 650 • C. After 650 • C, the formed S will be desorbed with continuous flushing air and increasing temperature. For Case-2, some studies reported that Ce 2 O 2 S could combine with the oxygen to form Ce 2 O 2 SO 4 , which is relatively stable below 700 • C [40]. The Ce 2 O 2 SO 4 will decompose into CeO 2 and SO 2 after the temperature is above 700 • C. Thus, Case-2 is also most likely to be the path of the Ce 2 O 2 S regeneration. Case 2 [39,40]: Ce2O2SO4 → (s) 2 CeO2(s)+ SO2(g) (11) Considering Case-1, as shown in (8) and (9), one oxygen molecule will be captured by Ce2O2S to form CeO2 molecules, accompanied by the release of the elemental S during the oxidation process. Because the oxygen molecule weight is equivalent to the weight of half S2 (molecular weight: O2 = 1/2 S2), the weight of the powder could be kept stable after the regeneration process. Therefore, for the path of the Case-1, the formed elemental sulfur will firstly be adsorbed on the surface of the powders below 650 °C. After 650 °C, the formed S will be desorbed with continuous flushing air and increasing temperature. For Case-2, some studies reported that Ce2O2S could combine with the oxygen to form Ce2O2SO4, which is relatively stable below 700 °C [40]. The Ce2O2SO4 will decompose into CeO2 and SO2 after the temperature is above 700 °C. Thus, Case-2 is also most likely to be the path of the Ce2O2S regeneration. To confirm the regeneration route of the Ce2O2S, XPS was characterized to investigate the sulfur valence of the Re3 powder. The Ce-O-S powder was firstly heated to 690 °C in air conditions and then was fleetly cooled down to form Re3 powder. If Case-1 is the real regeneration path, the elemental sulfur will be observed on the surface of the Re3 powder. As shown in Figure 8, the S 2p spectra is close to the result of Figure 6. S 0 was not found and only S 6+ was observed, indicating that the regeneration path is not Case-1. Thus, Case-2 should be confirmed during the heating process. Fourier transform infrared spectroscopy (FTIR) result confirmed the speculation of Case-2. As shown in Figure 9, the peaks of SO2 located at 1000 ~ 1200 cm −1 [41,42] and the intense peaks of SO3 located at 1300 ~ 1500 cm −1 were observed [43,44]. The SO3 is obtained by the further oxidation of the SO2. Therefore, from the above results, the regeneration path of the Ce2O2S species follows Case-2, when the powder is calcined in air conditions. Figure S1 shows SEM images of the morphology of the fresh SDC (a), the surface of the fresh SDC (b), the surface of the used SDC (c), and the surface of the Re1 powder (d). As shown, the whole morphology of the fresh SDC presents a flake-like structure. The surface of the SDC is pretty smooth. However, as shown in Figure S1c, agglomerations, holes, and bubble-like structures are observed on the surface after the desulfurization process. After regeneration, the SDC surface becomes relatively smooth again and there are some particles corresponding to the active species or deciduous CeO2, as shown in Figure S1d. After desulfurization and regeneration, many flake-like and rectangular particles corresponding to SDC are still present, suggesting that this adsorbent has high thermal stability and regeneration capacity. To confirm the regeneration route of the Ce 2 O 2 S, XPS was characterized to investigate the sulfur valence of the Re3 powder. The Ce-O-S powder was firstly heated to 690 • C in air conditions and then was fleetly cooled down to form Re3 powder. If Case-1 is the real regeneration path, the elemental sulfur will be observed on the surface of the Re3 powder. As shown in Figure 8, the S 2p spectra is close to the result of Figure 6. S 0 was not found and only S 6+ was observed, indicating that the regeneration path is not Case-1. Thus, Case-2 should be confirmed during the heating process. Fourier transform infrared spectroscopy (FTIR) result confirmed the speculation of Case-2. As shown in Figure 9, the peaks of SO 2 located at 1000~1200 cm −1 [41,42] and the intense peaks of SO 3 located at 1300~1500 cm −1 were observed [43,44]. The SO 3 is obtained by the further oxidation of the SO 2 . Therefore, from the above results, the regeneration path of the Ce 2 O 2 S species follows Case-2, when the powder is calcined in air conditions. Figure S1 shows SEM images of the morphology of the fresh SDC (a), the surface of the fresh SDC (b), the surface of the used SDC (c), and the surface of the Re1 powder (d). As shown, the whole morphology of the fresh SDC presents a flake-like structure. The surface of the SDC is pretty smooth. However, as shown in Figure S1c, agglomerations, holes, and bubble-like structures are observed on the surface after the desulfurization process. After regeneration, the SDC surface becomes relatively smooth again and there are some particles corresponding to the active species or deciduous CeO 2 , as shown in Figure S1d. After desulfurization and regeneration, many flake-like and rectangular particles corresponding to SDC are still present, suggesting that this adsorbent has high thermal stability and regeneration capacity.  The gas mixture of the SO2 and SO3 can be collected as acid during the adsorbent regeneration in the air. According to the TGA result, 11.1% weight loss is attributed to the sulfur removal, and thus, theoretical yields (TY) of the acid (mol) are the same with the S loss (mol). However, the actual yields (AY) of acid can be calculated through the classical acid-base titration. Additionally, the ratio (mol%) of the AY and the consumed NaOH is 1:2. Thus, the AY/TY can be described as follows.
where C(NaOH) is the molar concentration of NaOH, V(NaOH) is the consumed volume of NaOH, m is the weight of the Ce-O-S, and M is molar mass of S. Figure 10 and Table 3 show the AY, TY, and AY/TY versus the weight of the Ce-O-S powder regeneration. As shown, the AY is about 74% ~ 82% of the theoretical yield; the AY increases with the rise of the Ce-O-S powder weight, but the AY/TY decreases. This is because Ce-O-S powder is not completely regenerated, which also can be seen from the results of Figure 6 and Table 2.  The gas mixture of the SO2 and SO3 can be collected as acid during the adsorbent regeneration in the air. According to the TGA result, 11.1% weight loss is attributed to the sulfur removal, and thus, theoretical yields (TY) of the acid (mol) are the same with the S loss (mol). However, the actual yields (AY) of acid can be calculated through the classical acid-base titration. Additionally, the ratio (mol%) of the AY and the consumed NaOH is 1:2. Thus, the AY/TY can be described as follows.
where C(NaOH) is the molar concentration of NaOH, V(NaOH) is the consumed volume of NaOH, m is the weight of the Ce-O-S, and M is molar mass of S. Figure 10 and Table 3 show the AY, TY, and AY/TY versus the weight of the Ce-O-S powder regeneration. As shown, the AY is about 74% ~ 82% of the theoretical yield; the AY increases with the rise of the Ce-O-S powder weight, but the AY/TY decreases. This is because Ce-O-S powder is not completely regenerated, which also can be seen from the results of Figure 6 and Table 2. The gas mixture of the SO 2 and SO 3 can be collected as acid during the adsorbent regeneration in the air. According to the TGA result, 11.1% weight loss is attributed to the sulfur removal, and thus, theoretical yields (TY) of the acid (mol) are the same with the S loss (mol). However, the actual yields (AY) of acid can be calculated through the classical acid-base titration. Additionally, the ratio (mol%) of the AY and the consumed NaOH is 1:2. Thus, the AY/TY can be described as follows.
where C (NaOH) is the molar concentration of NaOH, V (NaOH) is the consumed volume of NaOH, m is the weight of the Ce-O-S, and M is molar mass of S. Figure 10 and Table 3 show the AY, TY, and AY/TY versus the weight of the Ce-O-S powder regeneration. As shown, the AY is about 74%~82% of the theoretical yield; the AY increases with the rise of the Ce-O-S powder weight, but the AY/TY decreases. This is because Ce-O-S powder is not completely regenerated, which also can be seen from the results of Figure 6 and Table 2.  During the investigation process, we find that elemental S can be obtained during the regeneration process in a 2% O2/N2 gas mixture. After the regeneration process of Ce-O-S powder (i.e., 2% O2/N2 (5 ml min -1 , STP) for 6 h), the adsorption bottle precipitates some particles in the water and on the wall and bottom. Figure 11 shows the XPS and XRD result of these particles. The representative doublet peak of the S 2p located at 162 ~ 166 eV is associated with elemental sulfur [45,46], and only S 0 valence is observed, as shown in Figure 11a. Additionally, the diffraction peaks of the particles in the XRD pattern are attributed to the elemental sulfur, as shown in Figure 11b, indicating that these precipitates are elemental S. The elemental sulfur can be precisely obtained through the regeneration path of Case 1 in a 2% O2/N2 gas atmosphere. Some studies claimed that Ce2O2S could react with SO2 in high temperature (500 ~ 700 °C), resulting in the production of elemental sulfur (20% yield). However, in this study, we can obtain the elemental sulfur through precisely controlling the oxygen content during the regeneration process. This could provide a new idea for the adsorbent regeneration.  During the investigation process, we find that elemental S can be obtained during the regeneration process in a 2% O 2 /N 2 gas mixture. After the regeneration process of Ce-O-S powder (i.e., 2% O 2 /N 2 (5 mL min −1 , STP) for 6 h), the adsorption bottle precipitates some particles in the water and on the wall and bottom. Figure 11 shows the XPS and XRD result of these particles. The representative doublet peak of the S 2p located at 162~166 eV is associated with elemental sulfur [45,46], and only S 0 valence is observed, as shown in Figure 11a. Additionally, the diffraction peaks of the particles in the XRD pattern are attributed to the elemental sulfur, as shown in Figure 11b, indicating that these precipitates are elemental S. The elemental sulfur can be precisely obtained through the regeneration path of Case 1 in a 2% O-2 /N 2 gas atmosphere. Some studies claimed that Ce 2 O 2 S could react with SO 2 in high temperature (500~700 • C), resulting in the production of elemental sulfur (20% yield). However, in this study, we can obtain the elemental sulfur through precisely controlling the oxygen content during the regeneration process. This could provide a new idea for the adsorbent regeneration.  5.20 × 10 -3 74.6 During the investigation process, we find that elemental S can be obtained during the regeneration process in a 2% O2/N2 gas mixture. After the regeneration process of Ce-O-S powder (i.e., 2% O2/N2 (5 ml min -1 , STP) for 6 h), the adsorption bottle precipitates some particles in the water and on the wall and bottom. Figure 11 shows the XPS and XRD result of these particles. The representative doublet peak of the S 2p located at 162 ~ 166 eV is associated with elemental sulfur [45,46], and only S 0 valence is observed, as shown in Figure 11a. Additionally, the diffraction peaks of the particles in the XRD pattern are attributed to the elemental sulfur, as shown in Figure 11b, indicating that these precipitates are elemental S. The elemental sulfur can be precisely obtained through the regeneration path of Case 1 in a 2% O2/N2 gas atmosphere. Some studies claimed that Ce2O2S could react with SO2 in high temperature (500 ~ 700 °C), resulting in the production of elemental sulfur (20% yield). However, in this study, we can obtain the elemental sulfur through precisely controlling the oxygen content during the regeneration process. This could provide a new idea for the adsorbent regeneration. Figure 11. XPS spectra (a) and XRD pattern (b) of the particles after adsorbing the off-gas from the heated process of the Ce-O-S powder under a 2% O 2 /N 2 (10 mL/min, STP) for 6 h at 800 • C.

Conclusions
The SDC adsorbent showed a good sulfur removal ability and excellent regeneration capacity. The maximal sulfur capacity of SDC sorbent reaches up to 12.1 g S/100g sorbent, while the BSC of the CeO 2 is 7.9 g S/100g sorbent. The regeneration capacity of the adsorbents occurs because the Ce-O-S species can easily be regenerated to CeO 2 in oxidizing atmosphere. Ce 2 O 2 S is the main sulfur-containing species, and thus, the key of the regeneration mechanism of the adsorbent is the regeneration path of the Ce 2 O 2 S. There are two regeneration paths for the Ce 2 O 2 S at high temperature in O 2 /N 2 gas mixture. The end products of the Ce 2 O 2 S-containing powder are the CeO 2 and SO 2 after the regeneration process in air conditions. The Ce 2 O 2 SO 4 is an intermediate product during the heating process. However, the Ce 2 O 2 S directly generates CeO 2 and elemental sulfur in a 2% O 2 /N 2 gas condition.
Supplementary Materials: The following are available online at http://www.mdpi.com/1996-1944/13/5/1225/s1, Figure S1: SEM images of the morphology of the fresh SDC (a), the surface of the fresh SDC (b), the surface of the used SDC (c) and the surface of the Re1 powder (d)