Efficient Waste to Energy Conversion Based on Co-CeO2 Catalyzed Water-Gas Shift Reaction

Waste to energy technology is attracting attention to overcome the upcoming environmental and energy issues. One of the key-steps is the water-gas shift (WGS) reaction, which can convert the waste-derived synthesis gas (H2 and CO) to pure hydrogen. Co–CeO2 catalysts were synthesized by the different methods to derive the optimal synthetic method and to investigate the effect of the preparation method on the physicochemical characteristics of Co–CeO2 catalysts in the high-temperature water-gas shift (HTS) reaction. The Co–CeO2 catalyst synthesized by the sol-gel method featured a strong metal to support interaction and the largest number of oxygen vacancies compared to other catalysts, which affects the catalytic activity. As a result, the Co–CeO2 catalyst synthesized by the sol-gel method exhibited the highest WGS activity among the prepared catalysts, even in severe conditions (high CO concentration: ~38% in dry basis and high gas hourly space velocity: 143,000 h−1).


Introduction
Economic development and population growth have increased the amount of globally generated waste, which is expected to rise from 2.0 billion tons per year in 2016 to 3.4 billion tons per year in 2050 [1,2]. Consequently, much attention has been directed at the development of waste to energy technologies such as waste gasification to reduce the extent of landfill depletion, environmental pollution, and waste treatment costs [3][4][5]. Notably, waste gasification can reduce waste mass (by~80%) and volume (by~90%), save landfill space, and decrease the emission of pollutants such as NO x and SO x [3].
Waste gasification typically affords synthesis gas (H 2 and CO), which can be used to generate value-added products such as synthetic crude oil, methanol, and dimethyl ether, and can also be employed as a substitute of reformed natural gas for pure H 2 production through the water-gas shift (WGS) reaction (CO + H 2 O → CO 2 + H 2 ) [4,[6][7][8][9].
The growing importance of fuel-cell-based vehicles and related devices has increased the demand for H 2 , used as a fuel in fuel cells [10][11][12][13]. However, more than 96% of H 2 is generated from natural gas-and petroleum-derived sources (i.e., from fossil fuels), which highlights the need for practical that the high diesel soot oxidation activity of a Co 3 O 4 -CeO 2 catalyst prepared by incipient wetness impregnation was due to the high dispersion of Co 3 O 4 nanoparticles on the CeO 2 support [40]. Although the preparation method strongly affects catalyst performance, no related research has been conducted in the case of Co-based catalysts for the HTS reaction using waste-derived synthesis gas.
Herein, we probe the effects of the preparation method (sol-gel, co-precipitation, incipient wetness impregnation, and hydrothermal) on the physicochemical characteristics of Co-CeO 2 catalysts and establish an optimal preparation method by comparing their activities for the HTS reaction using waste-derived synthesis gas. Moreover, correlations between catalyst physicochemical characteristics and activity are discussed. Table 1 lists the physicochemical properties of Co-CeO 2 catalysts synthesized by various methods, revealing that the Brunauer-Emmett-Teller (BET) area decreased after the reaction (30-82 m 2 g −1 for fresh catalysts vs. 24-73 m 2 g −1 for used catalysts). As a result of measuring the cobalt dispersion through CO-chemisorption (Table 1), the Co-CeO 2 (HT) catalyst featured the highest Co dispersion in both fresh and used forms. For all catalysts, dispersion decreased after the reaction, which was ascribed to active metal sintering [43]. In particular, the Co-CeO 2 (IWI) catalyst showed the highest decrease rate (91% ↓) among the prepared catalysts. The Co-CeO 2 (SG) catalyst and Co-CeO 2 (CP) catalyst showed a relatively higher decrease rate (SG: 74% ↓, CP: 86% ↓) compared to the Co-CeO 2 (HT) catalyst (68% ↓). Thus, even for catalysts with identical Co loadings, Co dispersion depended on the synthetic method, being the lowest for catalysts prepared using sol-gel or co-precipitation methods, which was ascribed to the partial coverage of metal on the surface [44]. Figure 1 presents the X-ray diffraction (XRD) patterns of different Co-CeO 2 catalysts, showing that all major diffraction peaks were consistent with the fluorite-type structure of CeO 2 [25]. Peaks at 2 theta = 31.2, 36.9, 45.0, and 65.2 • observed for all fresh samples ( Figure 1a) were attributed to the (220), (311), (400), and (440) reflections of Co 3 O 4 , respectively [45]. The patterns of all used catalysts featured a small diffraction peak of Co 0 at 2 theta = 44.2 • (Figure 1b), which confirmed the formation of Co 0 from Co 3 O 4 due to the pre-reaction reduction.

Catalyst Characterization
Catalyst structure and oxygen vacancy concentration were probed by Raman spectroscopy. As shown in Figure 2a, all catalysts showed peaks that correspond to Co 3 O 4 and CeO 2 [46]. In here, the peak at~457 cm −1 , which is attributed from the F 2g vibration mode of the fluorite structure of CeO 2 , was deconvoluted and magnified in Figure 2b to focus on the feature of this peak in each sample. Interestingly, in the case of the Co-CeO 2 (SG) catalyst, this peak was clearly shifted to lower wave numbers, which was indicative of CeO 2 structural distortion that generated lattice strain and defects in the CeO 2 lattice, and thus promoted the formation of oxygen vacancies [47,48]. This indicates that the largest amount of oxygen vacancy was formed in the Co-CeO 2 (SG) catalyst. In our previous research, we confirmed that oxygen vacancy concentration is related to catalytic performance in the WGS reaction [18]. The WGS reaction at the high temperature range occurs through the redox mechanism.
Catalysts 2020, 10, 420 4 of 14 In the high-temperature WGS reaction, CO molecules are adsorbed at catalyst active sites and oxidized to CO 2 by the lattice oxygen in CeO 2 , which results in the formation of oxygen vacancies. Subsequently, steam (H 2 O) dissociation on the catalyst surface regenerates lattice oxygen and affords H 2 . As high levels of oxygen vacancies facilitate the oxidation of CO to CO 2 , the Co-CeO 2 (SG) catalyst is expected to show the highest HTS activity [14]. Therefore, it is expected that the Co-CeO 2 (SG) catalyst will show higher activity than other catalysts for the HTS reaction due to the large number of oxygen vacancies. Oxygen vacancy formation was further confirmed by X-ray photoelectron spectroscopy (XPS) and oxygen storage capacity (OSC) measurements.
Catalysts 2020, 10  Catalyst structure and oxygen vacancy concentration were probed by Raman spectroscopy. As shown in Figure 2a, all catalysts showed peaks that correspond to Co3O4 and CeO2 [46]. In here, the peak at ~457 cm -1 , which is attributed from the F2g vibration mode of the fluorite structure of CeO2, was deconvoluted and magnified in Figure 2b to focus on the feature of this peak in each sample. Interestingly, in the case of the Co-CeO2 (SG) catalyst, this peak was clearly shifted to lower wave numbers, which was indicative of CeO2 structural distortion that generated lattice strain and defects in the CeO2 lattice, and thus promoted the formation of oxygen vacancies [47,48]. This indicates that the largest amount of oxygen vacancy was formed in the Co-CeO2 (SG) catalyst. In our previous research, we confirmed that oxygen vacancy concentration is related to catalytic performance in the WGS reaction [18]. The WGS reaction at the high temperature range occurs through the redox mechanism. In the high-temperature WGS reaction, CO molecules are adsorbed at catalyst active sites and oxidized to CO2 by the lattice oxygen in CeO2, which results in the formation of oxygen vacancies. Subsequently, steam (H2O) dissociation on the catalyst surface regenerates lattice oxygen and affords H2. As high levels of oxygen vacancies facilitate the oxidation of CO to CO2, the Co-CeO2 (SG) catalyst is expected to show the highest HTS activity [14]. Therefore, it is expected that the Co-CeO2 (SG) catalyst will show higher activity than other catalysts for the HTS reaction due to the large number of oxygen vacancies. Oxygen vacancy formation was further confirmed by X-ray photoelectron spectroscopy (XPS) and oxygen storage capacity (OSC) measurements.   Catalyst structure and oxygen vacancy concentration were probed by Raman spectroscopy. As shown in Figure 2a, all catalysts showed peaks that correspond to Co3O4 and CeO2 [46]. In here, the peak at ~457 cm -1 , which is attributed from the F2g vibration mode of the fluorite structure of CeO2, was deconvoluted and magnified in Figure 2b to focus on the feature of this peak in each sample. Interestingly, in the case of the Co-CeO2 (SG) catalyst, this peak was clearly shifted to lower wave numbers, which was indicative of CeO2 structural distortion that generated lattice strain and defects in the CeO2 lattice, and thus promoted the formation of oxygen vacancies [47,48]. This indicates that the largest amount of oxygen vacancy was formed in the Co-CeO2 (SG) catalyst. In our previous research, we confirmed that oxygen vacancy concentration is related to catalytic performance in the WGS reaction [18]. The WGS reaction at the high temperature range occurs through the redox mechanism. In the high-temperature WGS reaction, CO molecules are adsorbed at catalyst active sites and oxidized to CO2 by the lattice oxygen in CeO2, which results in the formation of oxygen vacancies. Subsequently, steam (H2O) dissociation on the catalyst surface regenerates lattice oxygen and affords H2. As high levels of oxygen vacancies facilitate the oxidation of CO to CO2, the Co-CeO2 (SG) catalyst is expected to show the highest HTS activity [14]. Therefore, it is expected that the Co-CeO2 (SG) catalyst will show higher activity than other catalysts for the HTS reaction due to the large number of oxygen vacancies. Oxygen vacancy formation was further confirmed by X-ray photoelectron spectroscopy (XPS) and oxygen storage capacity (OSC) measurements.  The relative content of defect oxygen and the valence states of O, Ce, and Co in Co-CeO 2 catalysts were determined by XPS. The O 1s spectra of all samples ( Figure 3a) were deconvoluted into the peaks of lattice oxygen (528.9 eV), oxygen close to oxygen vacancy sites (530.3 eV), and surface-adsorbed oxygen species from hydroxyl groups or water (532.2 eV) [17]. The relative contents of these species were calculated from the areas of the corresponding peaks. As defect oxygen is generated from oxygen vacancies, the defect oxygen to lattice oxygen (O 2 2− /O 2− ) ratio is related to the concentration of oxygen vacancies. The obtained results are listed in Table 1 and demonstrate that the above ratio was the highest for the Co-CeO 2 (SG) catalyst, which was in good agreement with the results of the Raman spectroscopic characterization. The Ce 3d spectra of the Co-CeO 2 catalysts ( Figure 3b) were deconvoluted into 10 peaks corresponding to different Ce oxidation states (Ce 4+ and Ce 3+ ). According to the references, peaks marked as v 0 , v, v , v", and v and u 0 , u, u , u", and u were attributed to Ce 3d 5/2 and Ce 3d 3/2 , respectively. Peaks marked as v, v", v , u, u", and u were assigned to Ce 4+ , while those denoted as v 0 , v , u 0 , and u were attributed to Ce 3+ [25,49]. The generation of oxygen vacancies resulted in the reduction of Ce 4+ to Ce 3+ on the catalyst surface, providing the active sites for water dissociation. The relative content of Ce 3+ in Ce 3d spectra was calculated by considering the presence of Ce 3+ ions directly involved in the dissociation of water. As a result, the Co-CeO 2 (SG) catalyst showed the highest Ce 3+ concentration, in agreement with the results of the Raman spectroscopic analysis and the defect oxygen to lattice oxygen ratio (O 2 2− /O 2− ) [8]. The Co 2p 3/2 spectra of the Co-CeO 2 catalyst in the binding energy ranging from 776 to 786 eV are shown in Figure 3c. According to the literature, the spectra can be attributed to the Co 2+ and Co 3+ species, which were located at 781.0 ± 0.4 and 779.4 ± 0.4 eV, respectively [50][51][52][53]. These peaks revealed the coexistence of  (Table 1), increasing in the order of Co-CeO 2 (HT) < Co-CeO 2 (CP) < Co-CeO 2 (IWI) < Co-CeO 2 (SG). Thus, the Co-CeO 2 (SG) catalyst featured the highest OSC (and, hence, the largest oxygen vacancy concentration), in accordance with the Raman spectroscopy and XPS results.
The reduction properties of the Co-CeO 2 catalysts were probed by H 2 -temperature programmed reduction (H 2 -TPR) measurements ( Figure 4), with two main reduction peaks observed for all samples. The first peak at 211-327 • C was attributed to the reduction of Co 3 O 4 to CoO, while the second one at 262-414 • C represented the reduction of CoO to Co 0 . The temperature of the first peak increased in the order of Co-CeO 2 (HT) < Co-CeO 2 (SG) < Co-CeO 2 (CP) < Co-CeO 2 (IWI), while that of the second peak increased in the order of Co-CeO 2 (HT) < Co-CeO 2 (CP) < Co-CeO 2 (IWI) < Co-CeO 2 (SG) (i.e., the reduction of CoO to the Co 0 in Co-CeO 2 (SG) catalyst occurred at the highest temperature). This means that the active species in HTS, Co 0 (metallic cobalt), is formed at higher temperature than other catalysts in the case of the Co-CeO 2 (SG) catalyst. Thus, we concluded that the use of the sol-gel method for Co-CeO 2 catalyst preparation may improve the interaction between Co 0 and the CeO 2 support. The strong metal-support interaction (SMSI) is an important factor that affects both catalytic activity and stability (e.g., catalysts with SMSI show high performance) [54][55][56]. Interestingly, the CoO reduction temperature of the prepared catalysts (described as the SMSI) followed the same trend as the defect oxygen to lattice oxygen ratio, showing that the SMSI induces the generation of oxygen vacancies. Similar findings were reported by Fichtl et al., which describe that the formation of oxygen vacancy is related to the interaction between active metal and support [57,58]. As a result, it is in-line that the Co-CeO 2 (SG) showed the SMSI estimated by H 2 -TPR result, with the aforementioned description in the XPS and Raman results, which proves that Co-CeO 2 (SG) showed a large number of oxygen vacancies.
According to the literature, the spectra can be attributed to the Co 2+ and Co 3+ species, which were located at 781.0 ± 0.4 and 779.4 ± 0.4 eV, respectively [50][51][52][53]. These peaks revealed the coexistence of Co3O4 and CoO species on the surface of the catalysts. Based on the above XPS results, we expected the Co-CeO2 (SG) catalyst to show the highest activity among the prepared catalysts due to its larger number of oxygen vacancies.   OSC was calculated by performing pulse reactions with H2 as the reductant and O2 as the oxidant. O2 oxidized the catalyst oxygen vacancies, while Ce 4+ was reduced to Ce 3+ by H2. The amount of consumed O2 coincided with the OSC (Table 1), increasing in the order of Co-CeO2 (HT) < Co-CeO2 (CP) < Co-CeO2 (IWI) < Co-CeO2 (SG). Thus, the Co-CeO2 (SG) catalyst featured the highest OSC (and, hence, the largest oxygen vacancy concentration), in accordance with the Raman spectroscopy and XPS results.
The reduction properties of the Co-CeO2 catalysts were probed by H2-temperature programmed

Reaction Results
The catalytic activity test was conducted in the temperature range of 400-550 °C to determine the most effective method of Co-CeO2 catalyst synthesis ( Figure 5). Among the prepared catalysts, the Co-CeO2 (SG) catalyst showed the highest CO conversion (~90%) across the entire tested temperature range, whereas the Co-CeO2 (HT) catalyst showed the lowest CO conversion (~39%). The activity of the Co-CeO2 (IWI) catalyst at 400 °C was close to that of the Co-CeO2 (SG) catalyst, but decreased at higher temperatures. To sum up, catalyst performance was strongly related to the reduction properties of CoO, which can be described in terms of the SMSI and oxygen vacancy concentration. Oxygen vacancies are known to actively participate in the WGS reaction and promote the diffusion of lattice oxygen [17,59]. Hence, the SMSI effect-related generation of oxygen vacancies resulted in increased catalytic activity. In addition, the catalyst oxygen storage capacity depended on the preparation method and also affected catalytic activity, whereas physicochemical properties such as BET surface area and catalyst dispersion were believed to have less influence. To identify the methanation reaction (CO + 3H2 → CH4 + H2O), which is a side reaction of the WGS reaction, the selectivities to CO2 and CH4 were calculated and shown in Figure 6. Except for Co-CeO2 (HT) catalyst (96% CO2 selectivity), the catalysts showed a CO2 selectivity of 100%, which indicated that methanation did not occur. Thus, the Co-CeO2 (SG) catalyst showed the highest

Reaction Results
The catalytic activity test was conducted in the temperature range of 400-550 • C to determine the most effective method of Co-CeO 2 catalyst synthesis ( Figure 5). Among the prepared catalysts, the Co-CeO 2 (SG) catalyst showed the highest CO conversion (~90%) across the entire tested temperature range, whereas the Co-CeO 2 (HT) catalyst showed the lowest CO conversion (~39%). The activity of the Co-CeO 2 (IWI) catalyst at 400 • C was close to that of the Co-CeO 2 (SG) catalyst, but decreased at higher temperatures. To sum up, catalyst performance was strongly related to the reduction properties of CoO, which can be described in terms of the SMSI and oxygen vacancy concentration. Oxygen vacancies are known to actively participate in the WGS reaction and promote the diffusion of lattice oxygen [17,59]. Hence, the SMSI effect-related generation of oxygen vacancies resulted in increased catalytic activity. In addition, the catalyst oxygen storage capacity depended on the preparation method and also affected catalytic activity, whereas physicochemical properties such as BET surface area and catalyst dispersion were believed to have less influence.

Reaction Results
The catalytic activity test was conducted in the temperature range of 400-550 °C to determine the most effective method of Co-CeO2 catalyst synthesis ( Figure 5). Among the prepared catalysts, the Co-CeO2 (SG) catalyst showed the highest CO conversion (~90%) across the entire tested temperature range, whereas the Co-CeO2 (HT) catalyst showed the lowest CO conversion (~39%). The activity of the Co-CeO2 (IWI) catalyst at 400 °C was close to that of the Co-CeO2 (SG) catalyst, but decreased at higher temperatures. To sum up, catalyst performance was strongly related to the reduction properties of CoO, which can be described in terms of the SMSI and oxygen vacancy concentration. Oxygen vacancies are known to actively participate in the WGS reaction and promote the diffusion of lattice oxygen [17,59]. Hence, the SMSI effect-related generation of oxygen vacancies resulted in increased catalytic activity. In addition, the catalyst oxygen storage capacity depended on the preparation method and also affected catalytic activity, whereas physicochemical properties such as BET surface area and catalyst dispersion were believed to have less influence. To identify the methanation reaction (CO + 3H2 → CH4 + H2O), which is a side reaction of the WGS reaction, the selectivities to CO2 and CH4 were calculated and shown in Figure 6. Except for Co-CeO2 (HT) catalyst (96% CO2 selectivity), the catalysts showed a CO2 selectivity of 100%, which indicated that methanation did not occur. Thus, the Co-CeO2 (SG) catalyst showed the highest catalytic activity without any side reactions, and its potential was further probed by an additional stability test performed at 450 °C and a GHSV of 143,000 h −1 for 20 h. To identify the methanation reaction (CO + 3H 2 → CH 4 + H 2 O), which is a side reaction of the WGS reaction, the selectivities to CO 2 and CH 4 were calculated and shown in Figure 6. Except for Co-CeO 2 (HT) catalyst (96% CO 2 selectivity), the catalysts showed a CO 2 selectivity of 100%, which indicated that methanation did not occur. Thus, the Co-CeO 2 (SG) catalyst showed the highest catalytic activity without any side reactions, and its potential was further probed by an additional stability test performed at 450 • C and a GHSV of 143,000 h −1 for 20 h. In the case of the primary stability test of Co-CeO2 (SG) catalyst, the test was performed under the reaction conditions of 450 °C and GHSV = 143,000 h −1 to examine the stability of Co-CeO2 (SG) catalyst in the HTS reaction. Figure 7 indicates that the Co-CeO2 (SG) catalyst showed excellent stability, almost reaching the equilibrium CO conversion and maintaining activity up to 20 h, even in the very high GHSV of 143,000 h −1 . However, we tried to check the potentially higher activity and stability that may be limited by the thermodynamic equilibrium. Due to this reason, we repeated the stability test in the condition where the catalyst may show lower activity than the thermodynamic equilibrium (T = 400 °C, GHSV = 203,000 h −1 ). Although the Co-CeO2 (SG) catalyst showed the most outstanding activity among the prepared catalysts, it was found that the stability of the catalyst gradually decreased within 20 h in the newly established reaction condition. From this result, further study on the development of catalysts for HTS reaction using waste-derived synthesis gas with improved stability seems to be necessary for its application in the industrial field. In order to confirm the catalytic performance of the Co-CeO2 (SG) catalyst, the performance was compared with the pre-developed Co and Ce based catalysts for HTS reaction through the literature survey in Table 2 [60][61][62][63][64]. The Co-CeO2 (SG) catalyst developed in this study showed superior In the case of the primary stability test of Co-CeO 2 (SG) catalyst, the test was performed under the reaction conditions of 450 • C and GHSV = 143,000 h −1 to examine the stability of Co-CeO 2 (SG) catalyst in the HTS reaction. Figure 7 indicates that the Co-CeO 2 (SG) catalyst showed excellent stability, almost reaching the equilibrium CO conversion and maintaining activity up to 20 h, even in the very high GHSV of 143,000 h −1 . However, we tried to check the potentially higher activity and stability that may be limited by the thermodynamic equilibrium. Due to this reason, we repeated the stability test in the condition where the catalyst may show lower activity than the thermodynamic equilibrium (T = 400 • C, GHSV = 203,000 h −1 ). Although the Co-CeO 2 (SG) catalyst showed the most outstanding activity among the prepared catalysts, it was found that the stability of the catalyst gradually decreased within 20 h in the newly established reaction condition. From this result, further study on the development of catalysts for HTS reaction using waste-derived synthesis gas with improved stability seems to be necessary for its application in the industrial field. In the case of the primary stability test of Co-CeO2 (SG) catalyst, the test was performed under the reaction conditions of 450 °C and GHSV = 143,000 h −1 to examine the stability of Co-CeO2 (SG) catalyst in the HTS reaction. Figure 7 indicates that the Co-CeO2 (SG) catalyst showed excellent stability, almost reaching the equilibrium CO conversion and maintaining activity up to 20 h, even in the very high GHSV of 143,000 h −1 . However, we tried to check the potentially higher activity and stability that may be limited by the thermodynamic equilibrium. Due to this reason, we repeated the stability test in the condition where the catalyst may show lower activity than the thermodynamic equilibrium (T = 400 °C, GHSV = 203,000 h −1 ). Although the Co-CeO2 (SG) catalyst showed the most outstanding activity among the prepared catalysts, it was found that the stability of the catalyst gradually decreased within 20 h in the newly established reaction condition. From this result, further study on the development of catalysts for HTS reaction using waste-derived synthesis gas with improved stability seems to be necessary for its application in the industrial field. In order to confirm the catalytic performance of the Co-CeO2 (SG) catalyst, the performance was compared with the pre-developed Co and Ce based catalysts for HTS reaction through the literature survey in Table 2 [60][61][62][63][64]. The Co-CeO2 (SG) catalyst developed in this study showed superior catalytic performance to reach the equilibrium CO conversion, even at a relatively high CO concentration (17.02 vol.%) and GHSV (143,000 h −1 ). As a result, it was confirmed that the synthesis of a Co-based catalyst for HTS reaction using waste gasification synthesis gas was successfully achieved. In order to confirm the catalytic performance of the Co-CeO 2 (SG) catalyst, the performance was compared with the pre-developed Co and Ce based catalysts for HTS reaction through the literature survey in Table 2 [60][61][62][63][64]. The Co-CeO 2 (SG) catalyst developed in this study showed superior catalytic performance to reach the equilibrium CO conversion, even at a relatively high CO concentration (17.02 vol.%) and GHSV (143,000 h −1 ). As a result, it was confirmed that the synthesis of a Co-based catalyst for HTS reaction using waste gasification synthesis gas was successfully achieved. In the present study, the Co-CeO 2 catalyst activity result was influenced by the synthetic method. The high performance of the Co-CeO 2 (SG) catalyst can be explained as follows. First, oxygen vacancy concentration affects the HTS reaction. Raman spectroscopy, XPS, and H 2 -O 2 pulse reaction results showed that the Co-CeO 2 (SG) catalyst had the highest concentration of oxygen vacancies among the prepared catalysts. As the WGS reaction primarily proceeds through a redox mechanism, it is strongly influenced by the concentration of oxygen vacancies. Liotta et al. developed the Co 3 O 4 -CeO 2 catalysts for the total oxidation of propene. The prepared catalysts showed increased surface oxygen vacancies through the application of a CeO 2 support and optimization of Co loading [32]. Reina [34]. In line with the reported studies, we have confirmed that the catalytic activity of Co-CeO 2 catalysts were improved through the enhancement of OSC. In this study, the OSC of the catalyst was critically affected by the catalyst preparation method. Second, the SMSI effect is important for the generation of oxygen vacancy, which is directly related to the catalytic activity. According to the TPR results, the CoO reduction temperature was highest for the Co-CeO 2 (SG) catalyst, which was ascribed to the SMSI of this catalyst. Similar findings were reported dealing with the relation between the redox properties of the catalyst and catalytic activity. Reina et al. studied Au/CeO 2 /Al 2 O 3 catalysts doped with ZnO for the WGS reaction and reported that the catalysts exhibited improved CeO 2 redox properties with the presence of ZnO [33]. Additionally, Reina et al. investigated Au/MO x /CeO 2 -Al 2 O 3 catalysts and found that the interaction between gold and Co 3 O 4 increased the reduction degree in the Co doped Au/CeO 2 -Al 2 O 3 systems [34]. As a result, the Co-CeO 2 (SG) catalyst showed the most outstanding catalytic activity in the HTS reaction due to the large number of oxygen vacancies induced by the strong interaction between cobalt and CeO 2 support, despite the extremely high GHSV and CO concentration.

Preparation of Catalysts
Co-CeO 2 (85 wt% CeO 2 support, 15 wt% Co) was prepared by sol-gel, co-precipitation, incipient wetness impregnation, and hydrothermal methods (Co-CeO 2 (SG), Co-CeO 2 (CP), Co-CeO 2 (IWI), and Co-CeO 2 (HT), respectively) using stoichiometric quantities of Co(NO 3 ) 2 ·6H 2 O (Sigma-Aldrich, St. Louis, MO, USA) and Ce(NO 3 ) 3 ·6H 2 O (Sigma-Aldrich, St. Louis, MO, USA). To obtain Co-CeO 2 (SG) catalyst, stoichiometric quantities of cobalt, cerium precursors, and citric acid (Sigma-Aldrich, St. Louis, MO, USA) were dissolved in deionized water (20 mL), and the solution was stirred at 80 • C for 24 h. As the water evaporated, citric acid acted as a chelating agent to result in gel formation. After gelation, the product was dried at 100 • C for 4 h to afford a fluffy mass that was then calcined at 500 • C for 6 h. The Co-CeO 2 (CP) catalyst was prepared by treating a solution of Co(NO 3 ) 2 ·6H 2 O and Ce(NO 3 ) 3 ·6H 2 O with aqueous (15 wt%) KOH (Duksan, Ansan, Korea) to reach pH 10.5 at 80 • C. The precipitate was stirred for 72 h, washed with deionized water for five times, dried overnight at 100 • C, and calcined at 500 • C for 6 h. In the case of the Co-CeO 2 (IWI) catalyst, a solution of Co(NO 3 ) 2 ·6H 2 O in deionized water (0.5 mL) was used to impregnate CeO 2 prepared by precipitation, as described above. The impregnation process was carried out by dropping one or two drops until the solution (0.5 mL) was completely consumed. The solid products were dried at 100°C for 12 h in air and then calcined at 500 • C for 6 h in a furnace. To prepare the Co-CeO 2 (HT) catalyst, aqueous (15 wt%) NaOH (Duksan, Ansan, Korea) was injected into a solution of Co(NO 3 ) 2 ·6H 2 O and Ce(NO 3 ) 3 ·6H 2 O to reach pH 10 at 80 • C, and the mixture was stirred for 30 min for the precipitate to form. The suspension was held at 180 • C for 24 h in a hydrothermal reactor, washed with deionized water, dried overnight at 100 • C, and then calcined at 500 • C for 6 h.

Catalyst Characterization
Brunauer-Emmett-Teller (BET) surface areas were determined from N 2 adsorption/desorption isotherms recorded at 77 K using a ASAP 2010 device (Micromeritics, Norcross, GA, USA). Co dispersion was probed by CO-chemisorption at 50 • C (Autochem 2920, Micromeritics, Norcross, GA, USA) using 10% CO/He. X-ray diffraction (XRD) patterns were recorded for 2 theta = 20-80 • using a Ultima IV diffractometer (Cu Kα radiation, 40 kV, 40 mA, Rigaku, Tokyo, Japan). Raman spectra (LabRam Aramis spectrometer, Horiba Jobin Yvon, Longjumeau, France) were recorded using the 532 nm excitation line of a Nd-YAG laser. X-ray photoelectron spectroscopy (XPS) measurements were performed on a K-Alpha spectrometer (Thermo-Scientific, Waltham, MA, USA) using monochromated Al Kα radiation. All spectra were adjusted using the reference C 1s binding energy of 284.6 eV. H 2 -O 2 pulse reactions were carried out to calculate OSC. Specifically, catalysts were pretreated by heating for 2 h in He at 400 • C and exposed to an H 2 pulse (10% H 2 /Ar) at the same temperature to consume mobile oxygen. The reduced catalysts were exposed to an O 2 pulse (10% O 2 /Ar) to oxidize mobile oxygen, and OSC was calculated from the amount of consumed O 2 . H 2 -temperature programmed reduction (H 2 -TPR) was performed in 10% H 2 /Ar from 100 to 800 • C at a heating rate of 10 • C min −1 using a Micromeritics Autochem 2920 instrument. Detector sensitivity was calibrated by reduction of a known amount of NiO.

Catalyst Performance Test
Catalytic performance was tested at 400-550 • C in a fixed-bed microtubular quartz reactor at atmospheric pressure. Catalyst powder (0.03 g) was placed on quartz wool (catalyst bed) in a quartz reactor with an inner diameter of 4 mm, a thermocouple for measuring the reaction temperature was inserted into the catalyst bed, and the reactor was heated in a furnace. Before measurement, the catalyst was reduction-activated by heating for 2 h in 5% H 2 /N 2 from room temperature to 400 • C, and the temperature was further held for 1 h, and then increased from 400 to 550 • C in 50 • C increments. The reactant gas comprised CO (17.02 vol%), CO 2 (9.55 vol%), CH 4 (1.03 vol%), H 2 (13.14 vol%), H 2 O (55.20 vol%), and N 2 (4.06 vol%), which is typical of synthesis gas produced through waste gasification. Water was injected into the reactor using a syringe pump and vaporized at 180 • C upstream of the reactor. After the reaction, reactant and product gases were analyzed online by an Agilent 3000-micro gas chromatograph. The reaction temperature and inlet gas flow rate were maintained for 40 min during each temperature step in consideration of the time required for the reaction gas from the catalyst bed to reach the micro gas chromatograph. The outlet gas was chilled to 2.2 • C and passed through a trap to remove residual water before analysis by the micro gas chromatograph. The CO conversion of product gas for each temperature was determined as the five measurements performed in 3 min intervals. The feed H 2 O/(CH 4 + CO + CO 2 ) ratio was set to 2.0 to avoid coke formation. The GHSV equaled 143,000 h −1 . The calculation of equilibrium conversion of CO was carried out using HSC 6.0 software (Version 6.0, Outotec, Espoo, Finland, 2006). The Gibbs free energy minimization method was applied, and the parameters related to the reaction including temperature, pressure, and reactant composition were altered. CO conversion and CO 2 and CH 4 selectivities were calculated as follows.

Conclusions
Co-CeO 2 catalysts prepared by various synthetic methods were applied for the HTS reaction of waste-derived synthesis gas. The Co-CeO 2 (SG) catalyst showed the best performance, even under extremely harsh conditions (GHSV = 143,000 h −1 , CO conc. = 38.2% dry basis), when compared to the industrial HTS reaction condition (GHSV = 3000 h −1 , CO conc. = 10%). The outstanding performance of the developed catalyst was ascribed to the high concentration of oxygen vacancies, which is also related to the strong metal-support interaction. Raman spectroscopy, XPS, and H 2 -O 2 pulse reaction data confirmed that the Co-CeO 2 (SG) catalyst had the high concentration of oxygen vacancies. Due to the formation of a larger number of oxygen vacancies induced by the strong interaction between the Co and CeO 2 support, the Co-CeO 2 (SG) catalyst showed the highest catalytic activity among the prepared catalysts. However, the stability test result conducted at 400 • C and accelerated GHSV (203,000 h −1 ) showed that the developed catalyst gradually deactivated within 20 h. Based on the stability test results, a subsequent study will be conducted to design a catalyst capable of maintaining high stability and elucidate the mechanism of the catalyst deactivation.