A Study of ZnâTi-Based H2S Removal Sorbents Promoted with Cobalt

Ind. Eng. Chem. Res. 2001, 40, 3547-3556

3547

A Study of Zn-Ti-Based H2S Removal Sorbents Promoted with Cobalt Oxides Hee Kwon Jun,† Tae Jin Lee,‡ Si Ok Ryu,‡ and Jae Chang Kim*,† Department of Chemical Engineering, Kyungpook National University, Taegu 702-701, Korea, and Department of Chemical Engineering, Yeungnam University, Kyongsan 712-749, Korea

Modification of conventional zinc titanate (ZT) sorbents to increase their reactivity and stability during multiple cycles of sulfidation/regeneration at high and middle temperatures was pursued by addition of Co3O4. Moreover, evaluation of the modified zinc titanates (ZTC) was conducted using several instruments such as an X-ray diffractometer, Fourier transform infrared spectrometer, scanning electron microscope, and so on. The ZTC sorbents prepared by the physical mixing of the ZT sorbent with 25 wt % of Co3O4 showed an excellent sulfur-removing capacity and no deactivation even after multiple cycles of sulfidation/regeneration in both high- and middle-temperature conditions. The cobalt permeated the Zn2TiO4 lattice, leading to a new spinel phase, ZnCoTiO4. It worked not only as an active site during the sulfidation process but also as a support to prevent zinc migration to the outside of the sorbents and to minimize volume expansion/contraction. Under middle-temperature conditions, the phase separation of the ZTC sorbent including a cobalt oxide intermediate in addition to the ZnCoTiO4 spinel structure was observed. Also, the cobalt additive increased the regeneration capacity of Zn-based sorbents because it played an important role as the catalyst for oxidation. Introduction The integrated gasification combined cycle (IGCC) is considered to be among the most efficient and environmentally acceptable technologies for power generation from coal.1 For the utilization of this technology, the removal of pollutants from the coal-derived fuel gas is needed. Among the pollutants, sulfur, which exists in the form of hydrogen sulfide (H2S) under the highly reducing conditions of the gasifier, must be removed from the hot coal gas not only to protect equipment against corrosiveness in the later stages of the process but also to meet strict government regulations for sulfur emissions. The hot-gas desulfurization (HGD) process to remove H2S by using single- or mixed-metal oxide sorbents has received a great deal of attention over the last 2 decades. Under the reducing condition, the metal oxide sorbents are converted into metal sulfides during a sulfidation process and are then restored to their initial state, metal oxides, by regeneration. Study on HGD for coal gas in IGCC has concentrated on the development of metal oxide sorbents having a high sulfur-removing capacity and long-term stability at both high and middle temperatures. For the removal of hydrogen sulfide in the coal-derived gas, several metal oxide materials have been studied to develop the regenerable sorbents in high-temperature ranges.2-21 Zinc-based sorbents are known to be among the best metal oxide sorbents that have the most favorable thermodynamics for H2S removal.22 Unfortunately, despite their favorable thermodynamics, it is wellknown that the vaporization of zinc from pure ZnO sorbent is a serious problem over many cycles of

sulfidation/regeneration at high temperatures. Thus, several supports such as TiO2, SiO2, and Al2O3 are added to the zinc-based sorbents to prevent zinc vaporization. Of these sorbents, zinc titanate sorbents have been considered to be among the most favorable sorbents. Nonetheless, this sorbent does have some disadvantages such as extensive spalling, cracking of pellets, and progressive loss of reactivity over multiple sulfidation/regeneration cycles above 650 °C. In addition to these supports, several other additives were introduced into the sorbent to overcome these problems, even though the roles of additives have not been clearly defined because of the lack of identification of new active sites and their mechanisms. One of the objectives of this work was to investigate the role of a cobalt additive and to propose sulfidation/ regeneration kinetics. For this purpose, the reactivities of Zn-Ti-based sorbents with/without a cobalt additive were studied in a fixed-bed reactor over multiple sulfidation/regeneration cycles at both high- and middletemperature conditions. In addition, the changes of physical properties for the sorbents before/after the reaction, the identification of a new active site, and a mechanism were investigated with the aid of Hg porosimetry (Micrometrics, Poresizer 9320), X-ray diffraction (XRD; Philips, X’PERT), Fourier transform infrared (FTIR; Mattson Instruments Inc., Galaxy 7020A), scanning electron microscopy-energy-dispersive spectrometry (SEM-EDS; Hitachi, S-4200), X-ray photoelectron spectroscopy (XPS; VG Microtech, MT 500/1), and inductively coupled plasma mass spectrometry (ICP-MS; Jobin-Yvon, Jobin-Yvon 38 plus). Experimental Section

* To whom all correspondence should be addressed. E-mail: [email protected]. Phone: +82-53-950-5622. Fax: +82-53-950-6615. † Kyungpook National University. ‡ Yeungnam University.

Preparation of Sorbents. Zinc titanate (ZT) and modified zinc titanate (ZTC) sorbents used in this study were prepared by a physical mixing method. Zinc oxide, titanium dioxide, and cobalt oxide, of which particle size

10.1021/ie0011167 CCC: $20.00 © 2001 American Chemical Society Published on Web 07/03/2001

3548

Ind. Eng. Chem. Res., Vol. 40, No. 16, 2001

Table 1. Experimental Conditions for ZT-Based Sorbents sulfidation temperature (°C) high middle pressure (atm) flow rate (mL/min) gas composition (vol %) H2S H2 CO CO2 O2 N2

650 480 1 50

regeneration 800 580 1 50

1.5 11.7 9.6 5.2 balance

3-5 balance

is about 200-300 mesh, were sufficiently mixed with an inorganic binder, bentonite, for 1-2 h. Next, a liquid binder, ethylene glycol (EG), was added to the mixture in order to make the slurry of mixed-metal oxide. An extruder in our lab was used to formulate pellets to an outer diameter of 1 mm from the slurry. These wet pellets were dried for 4 h to remove moisture from the material in the temperature range of 250-300 °C. The dried pellets were then ground to particle size in the range of 250-300 µm diameter and calcined in a muffle furnace for 12 h in the temperature range of 900-1000 °C. The ramping rate of the temperature was maintained at 3 °C/min. The mole ratio of Zn to Ti was fixed to 1.5:1, and 0.34 g of Co3O4/unit g of ZT sorbent was added to make the ZTC sorbent. Apparatus and Procedure. Multiple cycles of sulfidation/regeneration were performed in a fixed-bed quartz reactor with a diameter of 1 cm placed in an electric furnace. One gram of sorbents was packed in this reactor, and the space velocity (SV) was maintained at 5000 h-1 to minimize the severe pressure drop and the channeling phenomena. All of the volumetric flows of gases were measured at STP (standard temperature and pressure) conditions. The temperature of the inlet and outlet lines of the reactor was maintained above 120 °C to prevent condensation of water vapor produced in the sulfidation processes. The outlet gases from the reactor were automatically analyzed every 8 min by a thermal conductivity detector (TCD) equipped with an autosampler (Valco). The column used in the analysis was a 1/8 in. Teflon tube packed with Porapak T. The conditions of sulfidation and regeneration and the composition of mixed gases are shown in Table 1. When the H2S concentration of the outlet gases reached 15 000 ppmv, the inlet stream of mixed gases containing 1.5% hydrogen sulfide was stopped and an inert nitrogen gas was introduced to purge the system until it reached the regeneration temperature. Finally, nitrogen gas mixed with 3 vol % of oxygen was introduced to regenerate the sulfidated sorbents until SO2 concentration reached a value less than 200 ppmv. Results and Discussion Comparison of Sulfur Removing Capacity. The best way to evaluate the absorption capacity of the sorbents is to determine the so-called breakthrough curves for H2S absorption. In a typical fixed-bed experiment, the H2S concentration of the outlet gas from the reactor is negligible until the entire bed is saturated with sulfur. At that point, the concentration of H2S rises rapidly up to the inlet H2S concentration. In this study, the breakthrough time is defined as the time necessary to detect the 200 ppmv H2S concentration in the outlet

Figure 1. Sulfur-removing capacity of ZT and ZTC sorbents at middle and high temperature: (a) ZT; (b) ZTC. Filled symbols: high temperature. Open symbols: middle temperature.

gas. Amounts of sulfur absorbed were calculated from the H2S breakthrough curves, and those for the ZT and ZTC sorbents are shown in Figure 1. When both sulfidation and regeneration are considered as a onecycle process, the horizontal axis indicates the number of cycles repeated. The vertical axis indicates the percent of sulfur absorbed per 1 g of sorbent until the H2S concentration in the outlet gas of the reactor reached 200 ppmv. The percents of sulfur absorbed into the ZTC sorbent were 22% and 16%, respectively, at high and middle sulfidation temperatures. The sulfurremoving capacities were maintained during the multiple sulfidation/regeneration cycles. However, in the case of the ZT sorbent, the sulfur absorption capacities between the second and third cycles were 19% and 12% at the high and middle sulfidation temperatures, respectively, but were rapidly deactivated at both high and middle temperatures. Horizontal solid lines in Figure 1a,b show theoretical values for two sorbents, that is, 23.1% for ZT and 14.5% for ZTC, when only the metal zinc is considered to be an active site and the ratio of metal zinc to sulfur is 1:1. Figure 1b shows that the sulfur-removing capacity for the ZTC sorbent is higher than the theoretical value. Therefore, it is thought that a new active site, in addition, to zinc exists within the ZTC sorbent. Because titanium is known to be a nonactive component for H2S absorption,4,17 cobalt was supposed to work as another new active site for H2S absorption. Change of Structure. As was mentioned previously, the ZTC sorbent, unlike the ZT sorbent, showed a very high sulfur-removing capacity and no deactivation of sorbent from results obtained in the multiple-cycle tests at both high and middle temperatures, even after 1015 cycles. To investigate these properties, the structure change of sorbents before/after the reaction was examined by a powder XRD. In Figure 2, XRD patterns of the two different fresh sorbents at high temperatures indicated a perfect spinel structure of Zn2TiO4 and the difference in XRD patterns between two sorbents was

Ind. Eng. Chem. Res., Vol. 40, No. 16, 2001 3549

Figure 2. XRD patterns of ZT and ZTC sorbents before/after reaction at high-temperature conditons: (a) ZT; (b) ZTC; (I) fresh; (II) sulfurized; (III) regenerated.

not found. After a sulfidation process, the peaks of TiO2 separated from the spinel structure and sulfides were observed in both sorbents, while the peaks of Zn2TiO4 disappeared in XRD. The sulfide peaks of the ZT sorbent at 28° and 47.5° were reported to be assigned to ZnS in the literature [JCPDS No. 39-1363]. In the case of the ZTC sorbent, other sulfide peaks being defined as Co9S8 were also observed at 30° and 52° [JCPDS No. 19-0363]. These results indicate that a cobalt additive would contribute to an increase in the desulfurization capacity and work as another active site. The XRD patterns after the regeneration process under high-temperature conditions illustrate that most sulfides were converted into the initial phases of Zn2TiO4 without sulfate formation. Figure 3 shows XRD patterns for the two sorbents at middle temperatures. After a sulfidation process, the XRD patterns of two sorbents at middle temperatures were almost identical to those treated at high-temperature conditions. However, after 15 cycles of regeneration at middle temperatures, the XRD patterns of two sorbents are very different from those treated at high temperatures. ZnO and TiO2 were observed in the XRD of the ZT sorbent, while new peaks assigned to Co3O4 in the literature were also observed in the ZTC sorbent [JCPDS No. 43-1003]. From results obtained in XRD, therefore, it was concluded that the ZTC sorbent after 15 cycles of the regeneration process at middle temperatures had not only a Zn2TiO4 spinel structure but also another Co3O4 phase. It also indicates the fact that sulfides of ZnS and Co9S8 produced during the sulfidation process were not completely restored to their initial phase. XRD measurement at various calcination temperatures was conducted on the ZTC sorbent to illustrate the tendency of Co3O4 formation with respect to the calcination temperatures. Figure 4 shows XRD peaks of fresh ZTC sorbents prepared at three different calcination temperatures. When XRD patterns were compared at 700 and 850 °C, it was noticed that the intensity of the Co3O4 peaks gradually decreased with an increase in the calcination temperature. At 1000 °C, however, the Co3O4 peaks almost disappeared. From the experiments performed at high and middle tempera-

Figure 3. XRD patterns of ZT and ZTC sorbents before/after reaction at middle-temperature conditions: (a) ZT; (b) ZTC; (I) fresh; (II) sulfurized; (III) regenerated.

Figure 4. XRD patterns of the ZTC sorbent to various calcination temperatures.

tures, it was concluded that a higher calcination temperature than 850 °C is desirable to make spinel structures of ZnO, TiO2, and Co3O4 in the sorbent. As the calcination temperature increased, it is believed that metal cobalt permeated the lattice of the spinel structure of zinc titanate and then was replaced by one of the metal zincs in Zn2TiO4, followed by ZnCoTiO4 formation. This is a general phenomenon of metals that have the same oxidation number and nearly the same size atomic radius. Unfortunately, it was impossible to distinguish the difference between the spinel structures of Zn2TiO4 and ZnCoTiO4 by XRD analysis. The change of structure of sorbents before/after the reaction was also examined by using FT-IR. For comparison purposes, the FT-IR spectra of pure components are shown in Figure

3550


Figure 5. FT-IR spectra of various single oxides.

5. The absorption peaks of ZnO appeared at 437, 498, and 534 cm-1, those of TiO2 at 510-710 cm-1, and those of Co3O4 at 570 and 662 cm-1. These spectra agreed well with the results of the works of Ibarra et al. and Nyquist and Kagel.23,24 Absorption peaks for ZT and ZTC sorbents before/after sulfidation and regeneration under high-temperature conditions are shown in Figure 6. No differences between the two sorbents (ZT vs ZTC) were observed. The strong absorption peak at 600 cm-1, which was the average value of the representative absorption peaks of pure ZnO and TiO2, and the quartet peaks at 880-980 cm-1 were observed in both the ZT and ZTC sorbents after regeneration. These results agree with the studies of Ibarra et al. and Siriwardane et al.23,25 The weak FT-IR absorption peaks for both sorbents after sulfidation imply that the oxides within sorbents were almost completely converted into sulfides

such as ZnS and Co9S8 and that these sulfide peaks were nonactive in this range. After regeneration at middle temperatures, FT-IR spectra for the ZT sorbent were very different from those of ZTC. A difference between fresh and regenerated sorbents in FT-IR spectra for ZT was not found as shown in Figure 7a, while in Figure 7b new peaks at 680 cm-1 assigned to Co3O4 and at 1100 cm-1 designated to sulfate were observed in FT-IR spectra for the ZTC sorbent after regeneration. These results obtained in FT-IR spectra imply that a sulfate might be formed in the ZTC sorbent during regeneration in the middletemperature ranges. Because CoSO4 is known to be a very stable sulfate, it cannot be removed from the sorbent by treatment at middle temperatures and it can cause SO2 slippage in subsequent cycles because of incomplete regeneration. However, we already confirmed that these sulfate peaks disappear during regeneration at high temperatures. From these XRD and FT-IR data, it can be concluded that Zn2TiO4 was formed in ZT sorbents over calcination at high temperatures and the ZnS and TiO2 produced after sulfidation at high/middle temperatures were almost completely regenerated under high- or middle-temperature conditions. However, in the case of the ZTC sorbent, it was expected that a metal cobalt permeated the Zn2TiO4 lattice to form a new spinel structure, ZnCoTiO4, at high calcination temperatures. ZnCoTiO4 was converted to ZnS, Co9S8, and TiO2 during the high or middle sulfidation process. Next, these sulfides and TiO2 were restored to the initial spinel structure phase over hightemperature regeneration. However, after the regeneration cycle at middle temperatures, the separated Co3O4 phase and the sulfate in addition to the spinel structure

Figure 6. FT-IR spectra of ZT and ZTC sorbents before/after reaction at high temperature: (a) ZT; (b) ZTC.


Figure 7. FT-IR spectra of ZT and ZTC sorbents before/after reaction at middle temperature: (a) ZT; (b) ZTC.

Figure 8. SEM morphologies of ZT and ZTC sorbents before/after reaction at high temperature: (a) ZT; (b) ZTC.

were observed because of insufficiently high enough temperatures for regeneration. Change of Physical Properties. SEM morphologies of fresh ZT and ZTC sorbents and of those after multiple cycles of sulfidation/regeneration at high temperatures are shown in Figure 8. The average particle sizes of two fresh sorbents, ZT and ZTC, were observed at 400 and 650 nm, respectively, and the particle sizes were uniformly distributed on both sorbents. However, after 10cycle regeneration at high temperatures, the particle size of the ZT sorbent decreased by about 75% in comparison with fresh material. In the case of the ZTC

sorbent, only 33% of the particle size decreased in comparison with fresh sorbent. It was expected that the change of physical properties for ZT sorbent during sulfidation/regeneration cycles was more severe than that of the ZTC sorbent. To show the change of physical properties of the sorbents before/after the reaction, the results obtained from Hg porosimetry are also shown in Table 2. In the case of the ZT sorbent, its pore volume and porosity decreased approximately 50% and 30%, respectively. On the contrary, however, its density and surface area increased compared to the fresh materials. Also, after 10 cycles of sulfidation/regeneration, its

3552


Table 2. Hg Porosimetry Analysis for ZT and ZTC Sorbents before/after Reaction at High Temperature no. of cycles

sorbent ZT ZTC

raw regenerated raw regenerated

pore vol. (cm3/g)

surface area (m2/g)

average pore diameter (4V/A)

bulk density (g/mL)

skeletal density (g/mL)

porosity (%)

0.3071 0.1571 0.2225 0.3301

3.242 8.862 1.742 2.629

3789 709 5107 5023

1.2465 1.7220 1.8150 1.6781

2.0196 2.3609 3.0442 3.7626

38.28 27.06 40.38 55.40

10 10

Table 3. EDS Analysis for ZT-Based Sorbents before/ after Reaction at High Temperature EDS atomic (%) no. of cycles ZT ZTC


Zn

Ti

O

Co

Zn/Ti ratio

44.29 21.94 31.22 42.57 16.97 37.21 24.8 14.91 34.47 23.71 23.65 13.24 37.45 23.51

10 10

2.02 2.51 1.66 1.79

Table 4. XPS Analysis for ZT-Based Sorbents before/ after Reaction at High Temperature XPS atomic (%) no. of cycles ZT ZTC


10 10

Zn 2p

Ti 2p

9.4 13.5 3.0 4.2

7.6 1.6 4.8 2.7

Co 2p

Zn/Ti ratio

4.0 3.8

1.24 8.44 0.63 1.56

average pore diameter seriously decreased from 3789 to 709 Å. These results can be explained by the fact that the structure and pores of the sorbent were broken and the smaller pores were created during multiple cycles of sulfidation/regeneration at high temperatures. For this reason, the pore volume and porosity seriously decreased while the density and surface area increased. On the other hand, the pore volume and porosity of the ZTC sorbent increased considerably while the surface area and average pore diameter did not change much in comparison with the fresh materials. The surface and bulk composition of sorbents before/ after the reaction were examined using SEM-EDS, XPS, and ICP-MS. The results of SEM-EDS for the ZT and ZTC sorbents before/after the reaction are tabulated in Table 3. The Zn/Ti ratios of two sorbents before the reaction were 2.02 and 1.66, respectively. After 10 cycles of regeneration at high temperatures, the Zn/Ti ratio for the ZT sorbent was 2.51 and that of ZTC, 1.79. The same tendency was observed in the XPS experiments. As shown in Table 4, the Zn/Ti ratios of two sorbents before the reaction were 1.24 and 0.63, respectively. After the sulfidation/regeneration process, their values changed to 8.44 and 1.56, respectively. The increase in the Zn/Ti ratio of the ZT sorbent after multiple cycles of sulfidation/regeneration was more than twice that of the ZTC sorbent. The reason metal zinc existed excessively on the surface of the ZT sorbent was because the metal zinc contained in the ZT sorbent gradually diffused to the surface of the sorbent during sulfidation/ regeneration cycles at high temperatures. These results are general phenomena observed for ZT-based sorbents, which are known as one of main reasons that give rise to the evaporation of zinc. On the other hand, analysis using ICP-MS to examine the bulk composition of the ZT sorbent before/after the reaction reveals that the Zn/ Ti ratio before/after the reaction did not change significantly and their values were 1.45 and 1.50, respectively. These results indicate that the metal zinc did not

volatilize yet. Rather, it only diffused from inside the sorbents to outside through 10 cyclic treatments. It is considered that the deactivation of the sulfur-removing capacity for the ZT sorbent without a cobalt additive at high-temperature conditions is caused by the changes of these physical properties including pore volume, surface area, density, and porosity. However, these changes in physical properties can be minimized as a cobalt additive is added to the ZT sorbent. Molar Volume Effect. The change of physical properties by expansion/contraction of sorbents during sulfidation/regeneration is considered to be another reason to cause spalling or cracking of sorbents. That is to say, volume expansion generally occurred during the sulfidation process as the metal oxide was converted into metal sulfide while contraction occurred during the regeneration process as the sorbents returned to their original oxide phase. In the case of the ZT sorbent, only metal zinc works as an active site for the sulfidation process. Thus, when ZnO (molar volume: 14.34 mL/mol) is converted into ZnS (molar volume: 23.85 mL/mol), the percent of its molar volume expansion is 66.2%. The molar volume could be obtained from the density calculated from X-ray measurements by the NBS EXAIDS83 program. The same degree of volume contraction occurs during the regeneration process. Volume expansion/contraction of the ZT sorbent is sufficient for sorbents to be broken during multiple cycles of sulfidation/regeneration. As stated in the discussion of XRD, sulfide peaks of the ZTC sorbent after the sulfidation process were ZnS and Co9S8. The molar volumes for Co3O4 and Co9S8 are 39.767 and 147.49 mL/mol, respectively. When the oxide phases of cobalt were converted into the sulfide phase, the percent of volume expansion of the cobalt was only 23.2%. This value is 3 times lower than that of metal zinc (66.7%). As a result, it is expected that the changes of physical properties due to the volume expansion/contraction were suppressed by adding cobalt additives. Figure 9 shows SEM morphologies for two sorbents before/after the reaction at middle-temperature conditions. In the case of the ZT sorbents, the particle size of sorbents after sulfidation is 2 or 3 times larger than that of the fresh material. It was also observed that the particle size of sorbents after a 15-cycle regeneration is larger than that of the fresh sample, even though the size is smaller than that of sorbents after sulfidation. The experimental results from porosimetry before/after the reaction at middletemperature conditions are shown in Table 5. After 5 and 15 cycles of regeneration, it was found that the physical properties of the ZT sorbents were continuously and seriously changed, which also produced very similar results at high temperatures. However, in a separate experiment using EDS and XPS, it was found that the zinc was not diffused out from the bulk to the exterior surface of the sorbents at these middle-temperature conditions, which indicates that deactivation or a change in the physical properties of the ZT sorbent was due to the expansion/contraction of the sorbents rather than zinc mobility.


Figure 9. SEM morphologies of ZT and ZTC sorbents before/after reaction at middle temperature: (a) ZT; (b) ZTC. Table 5. Hg Porosimetry Analysis for ZT and ZTC Sorbents before/after Reaction at Middle Temperature no. of cycles

sorbent ZT ZTC


5 15 5 15

pore vol. (cm3/g)

surface area (m2/g)


bulk density (g/mL)


porosity (%)

0.3071 0.2524 0.1338 0.2225 0.1565 0.1753

3.242 7.081 4.687 1.742 3.239 2.649

3789 1426 1141 5107 1933 2647

1.2465 1.3841 2.5619 1.8150 1.8425 1.7573

2.0196 2.1272 3.8975 3.0442 2.5889 2.5395

38.28 34.94 34.27 40.38 28.83 30.80

The results from porosimetry for the ZTC sorbent before/after the reaction at middle-temperature conditions are also shown in Table 5. They were very different from those at high-temperature conditions. For the experiments performed at high-temperature conditions, changes in the physical properties of the sorbents during the multiple sulfidation/regeneraton cycles were not observed. However, in middle-temperature conditions, particle sizes measured by SEM morphology, pore volume, average pore diameter, and surface area measured by porosimetry did considerably change, especially during initial sulfidation/regeneration cycles. Because these severe changes in the physical properties of the

ZTC sorbent at middle temperatures could not be explained well, a new mechanism or an obvious interpretation was required to describe the facts that differed from those at high temperatures. To explain these phenomena, multicyclic tests of sulfidation/regeneration, in which the sorbents calcined at 700 °C were used, were carried out at middle-temperature conditions. The sulfur removal capacity of the ZTC sorbent calcined at 700 °C was similar to that of ZTC calcined at 1000 °C, while the deactivation of sorbents was not observed. From XRD analysis as shown in Figure 6, it is known that the structure of the sorbent calcined at 700 °C was different from that of sorbents calcined at 1000 °C. It is

3554


Table 6. Hg Porosimetry Analysis for ZT and ZTC Sorbents before/after Reaction at Middle Temperature no. of cycles

sorbent ZT (700 °C)a ZTC (700 °C)a a


10 6

pore vol. (cm3/g)

surface area (m2/g)


bulk density (g/mL)


porosity (%)

0.2622 0.1828 0.2667 0.2262

8.436 19.209 7.139 5.514

1243 381 1494 1641

1.4198 1.4460 1.4048 1.5429

2.2619 1.9657 2.2463 2.3698

37.23 26.44 37.46 34.89

Calcination temperature.

Figure 11. SO2 breakthrough curves of ZT and ZTC sorbents at high and middle temperature: (a) regenerated at high temperature; (b) regenerated at middle temperature.

Figure 10. SEM morphologies of the ZTC sorbent calcined at 700 °C before/after reaction at middle temperature.

assumed that this experimental result for the sorbents calcined at 700 °C may be because the cobalt added to the sorbent did not enter the lattice of the Zn2TiO4 spinel structure, but instead existed separately as a pure single oxide state of Co3O4. As shown in Figure 10 and Table 6, differences in the SEM morphologies and physical properties between the regenerated sorbent and the fresh ZTC sorbent calcined at 700 °C were not observed even after 6 cycles of regeneration at middle temperatures. This indicates that severe changes in the physical properties over multiple cycles of sulfidation/ regeneration performed under middle-temperature conditions for ZTC sorbents calcined at 1000 °C are not due to volume expansion/contraction like ZT sorbents but rather only to the separation of phases caused by insufficient regeneration temperatures.

Comparison of Regeneration Capacity. In addition to the sulfur capacity and structural stability, regeneration characteristics are among the most important factors to be considered. The SO2 breakthrough curves during the test at high/middle-temperature conditions are plotted in Figure 11. From Figure 11, it was found out that the two sorbents were perfectly regenerated at high temperatures. However, on the other hand, different regeneration characteristics were shown at middle temperatures. It was observed for ZT sorbents that the concentration of SO2 had a maximum value at approximately 30 min and then the outlet concentration of SO2 slowly decreased gradually to the end. In the case of the ZTC sorbent, however, we found that most of the sulfur absorbed during sulfidation was desorbed over the process of regeneration both under middle and high conditions for a short time. It was considered that a cobalt additive played an important role for the regeneration of metal sulfide. To identify the reason of that stated above, a TPD test was performed after sulfidation at middle temperatures. The test was carried out by measuring the concentration of SO2 desorbed through the introduction of nitrogen gas containing 3 vol % of oxygen when the ramping rate of temperature was 1 °C/min. The TPD experimental results are shown in Figure 12. In the case of the ZT sorbent, no SO2 was desorbed during the initial period


Conclusions

Figure 12. SO2 TPD results of various metal oxides.

Compared to conventional ZT-based sorbents, cobaltcontaining zinc titanate sorbents have shown an excellent sulfur-removing capacity and no deactivation even after 10 cycles of sulfidation/regeneration in a fixed-bed reactor at both high and middle temperatures. At hightemperature conditions, cobalt permeated the Zn2TiO4 lattice, leading to a new spinel phase, ZnCoTiO4. It worked not only as an active site during the sulfidation process but also as a support to prevent zinc migration to the outside of the sorbents. It also minimized volume expansion/contraction, which was one of the most important causes of the change of the physical properties of sorbents during sulfidation and regeneration. Under middle-temperature conditions, the phase separation of the ZTC sorbent including the cobalt oxide intermediate in addition to the ZnCoTiO4 spinel structure was observed. Therefore, the changes in the physical properties of sorbents in the initial stages of sulfidation/regeneration were thought to be due to phase separation. Also, the cobalt additive increased the regeneration capacity of Zn-based sorbents because it played an important role as the catalyst for oxidation. Acknowledgment We appreciate the financial support from RaCER (the R&D Management Center for Energy & Resources, Korea) and also in part from IAE (Institute for Advanced Engineering, Korea) for this work. Literature Cited

Figure 13. XRD patterns of cobalt oxide to various regeneration temperatures: (a) fresh; (b) regenerated at 480 °C.

and most of the sulfur was desorbed in the temperature range 600-630 °C. Two peaks at 480 and 780 °C were observed for cobalt oxides (Co3O4) in Figure 12. A peak observed at low temperatures was due to Co9S8, which is formed during sulfidation, while another peak was observed at high temperatures due to CoSO4, which formed during the regeneration of Co9S8 in the oxygen stream. To confirm these results, the sorbents regenerated at 480 °C after sulfidation at middle-temperature conditions were collected and analyzed by XRD. As shown in Figure 13, a sulfide peak of Co9S8 was not found, but the peaks for Co3O4 and CoSO4 were observed instead. A SO2 peak from the thermal decomposition of CoSO4 was observed at a temperature higher than 750 °C. In the case of the ZTC sorbent, the sulfides were ZnS and Co9S8. Most of them were desorbed below 600 °C, and only small amounts of SO2 peaks were shown at 780 °C in the TPD experiment. As was shown previously, this peak resulted from the thermal decomposition of unregenerated CoSO4 because of an insufficient regeneration temperature. The reason most of the sulfide was desorbed below 600 °C could be explained by the catalytic role of a cobalt additive in the oxidation or the exothermic reaction of cobalt sulfide which could supply the necessary heat to initiate the reaction of zinc sulfide.

(1) Woudstra, T.; Woudstra, N. Exergy analysis of hot-gas cleanup in IGCC systems. J. Inst. Energy 1995, 68, 157. (2) Gibson, J. B.; Harrison, D. P. The Reaction between Hydrogen Sulfide and Spherical Pellets of Zinc Oxide. Ind. Eng. Chem. Prod. Res. Dev. 1980, 19, 231. (3) Gangwal, S. K.; Stogner, J. M.; Harkins, S. M. Testing of Novel Sorbents for H2S Removal from Coal Gas. Environ. Prog. 1989, 8, 26. (4) Lew, S.; Jothimurugesan, K.; Flytzani-Stephanopoulos, M. High-Temperature H2S Removal from Fuel Gases by Regenerable Zinc Oxide-Titanium Dioxide Sorbents. Ind. Eng. Chem. Res. 1989, 28, 535. (5) Siriwardane, R. V.; Poston, J. A.; Evans, G., Jr. Spectroscopic Characterization of Molybdenum-Containing Zinc Titanate Desulfurization Sorbents. Ind. Eng. Chem. Res. 1994, 33, 2810. (6) Tamhankar, S. S.; Bagajewicz, M.; Gavalas, G. R.; Sharma, P. K.; Flytzani-Stephanopoulos, M. Mixed-Oxide Sorbents for High-Temperature Removal of Hydrogen Sulfide. Ind. Eng. Chem. Prod. Res. Dev. 1986, 25, 429. (7) Patrick, V.; Gavalas, G. R.; Flytzani-Stephanopoulos, M.; Jothimurugesan, K. High-Temperature Sulfidation-Regeneration of CuO-Al2O3 Sorbents. Ind. Eng. Chem. Res. 1989, 28, 931. (8) Li, Z.; Flytzani-Stephanopoulos, M. Cu-Cr-O and CuCe-O Regenerable Oxide Sorbents for Hot Gas Desulfurization. Ind. Eng. Chem. Res. 1997, 36, 187. (9) Tamhankar, S. S.; Hasatani, M.; Wen, C. Y. Kinetic studies on the reactions involved in the hot gas desulfurization using a regenerable iron oxide sorbents1. Chem. Eng. Sci. 1981, 36, 1181. (10) Fuda, K.; Palmer, A. D.; Sears, P.; Blais, D.; Furimsky, E. Chemical changes occuring during sulphidation and regeneration of iron-containing sorbents. Fuel 1991, 70, 100. (11) Sasaoka, E.; Sakamoto, M.; Ichio, T.; Kasaoka, S.; Sakata, Y. Reactivity and Durability of Iron Oxide High-Temperature Desulfurization Sorbents. Energy Fuels 1993, 7, 632. (12) Atakul, H.; Wakker, J. P.; Gerritsen, A. W.; Van den berg, P. J. Removal of H2S from fuel gases at high-temperatures using MnO/γ-Al2O3. Fuel 1995, 74, 187. (13) Atakul, H.; Wakker, J. P.; Gerritsen, A. W.; Van den berg, P. J. Regeneration of MnO/γ-Al2O3 used for high-temperature desulfurization of fuel gases. Fuel 1996, 75, 373.

3556


(14) Ben-Slimane, R.; Hepworth, M. T. Desulfurization of hot coal-gas desulfurized fuel gases with manganese-based regenerable sorbents. 2. Regeneration and multicycle tests. Energy Fuels 1994, 8, 1184. (15) Gangwal, S. K.; Portzer, J. W.; Gupta, R.; Turk, B. S. Sulfur Recovery in Advanced IGCC systems. Proceedings of the 12th Korea-U.S. Joint Workshop on Energy and Environment, Korea Institute of Energy Research and Federal Energy Technology Center, Taejon, Korea, Oct 6-11, 1997; p 186. (16) Gupta, R.; Turk, B. S.; Gangwal, S. K. Bench-Scale Development of Fluid-bed Spray Dried Sorbents. Proceedings of Advanced Coal Fired Power Systems ’96 Contractor’s Review Meeting, Federal Energy Technology Center, Morgantown, WV, July 1996 (available as CD-Rom). (17) Lew, S.; Sarofim, A. F.; Flytzani-Stephanopoulos, M. Sulfidation of Zinc Titanate and Zinc Oxide Solids. Ind. Eng. Chem. Res. 1992, 31, 1890. (18) Woods, M. C.; Gangwal, S. K.; Jothimurugensan, K.; Harrison, D. P. Reaction Between H2S and Zinc Oxide-Titanium Oxide Sorbents. 1. Single-Pellet Kinetic Studies. Ind. Eng. Chem. Res. 1990, 29, 1160. (19) Gupta, R. P.; Gangwal, S. K.; Jain, S. C. Development of Zinc Ferrite Sorbents for Desulfurization of Hot Coal Gas in a Fluid-Bed Reactor. Energy Fuels 1992, 6, 21.

(20) Jothimurugesan, K.; Gangwal, S. K. Regeneration of Zinc Titanate H2S Sorbents. Ind. Eng. Chem. Res. 1998, 37, 1929. (21) Sasaoka, E.; Sada, N.; Manabe, A.; Uddin, M. A.; Sakata, Y. Modification of ZnO-TiO2 High-Temperature Desulfurization Sorbent by ZrO2 addition. Ind. Eng. Chem. Res. 1999, 38, 958. (22) Schrodt, J. T.; Hilton, G. B.; Rogge, C. A. High-temperature Desulfurization of Low-CV Fuel Gas. Fuel 1975, 54, 269. (23) Ibarra, J. V.; Cilleruelo, C.; Garcia, E.; Pineda, M.; Palacios, J. M.; Vibrational spectroscopy study of zinc-containing mixed oxides as regenerable sulphur sorbents at high temperature. Vib. Spectrosc. 1998, 16, 1. (24) Nyquist, R. N.; Kagel, R. O. Infrared Spectra of Inorganic Compounds (3800-45 cm-1); Academic Press: London, 1971. (25) Siriwardane, R. V.; Poston, J. A.; Evans, G., Jr. Spectroscopic Characterization of Molybdeum-Containing Zinc Titanate Desulfurization Sorbents. Ind. Eng. Chem. Res. 1994, 33, 2810.

Received for review December 19, 2000 Revised manuscript received April 23, 2001 Accepted May 8, 2001 IE0011167

A Study of ZnâTi-Based H2S Removal Sorbents Promoted with Cobalt

Recommend Documents

A Study of ZnâTi-Based H2S Removal Sorbents Promoted with Cobalt