864
Energy & Fuels 2005, 19, 864-868
Effect of Pore Size Distribution of Calcium Oxide High-Temperature Desulfurization Sorbent on Its Sulfurization and Consecutive Oxidative Decomposition Shengji Wu,† Md. Azhar Uddin,‡ and Eiji Sasaoka*,† Faculty of Environmental Science and Technology, Okayama University, 3-1-1 Tsusima-naka, Okayama, 700-8530, Japan, and School of Chemical Engineering and Industrial Chemistry, The University of New South Wales, Sydney, NSW 2052, Australia Received July 28, 2004. Revised Manuscript Received February 27, 2005
CaS formed from the CaO sorbent during desulfurization in coal gasifiers must be converted to CaSO4 before disposal. CaS is mainly decomposed to CaO and SO2, and then the CaO is converted to CaSO4 by SO2 in the presence of H2O and O2. However, the inner portion of the CaS particles cannot be converted to CaO and CaSO4, because the pores are plugged and oxidative gases (H2O and O2) cannot come into contact with the interior of the CaS particles. In this study, the effect of the pore-size distribution of the CaO sorbent on the sulfurization and consecutive oxidative decomposition of the formed CaS in the presence of H2O at high temperature have been investigated, using three samples with different pore structures but similar surface areas. The following results have been obtained: (i) the macroporous CaS can be easily converted, because the oxidative decomposition of the macroporous CaS occurs without pore plugging; (ii) pores with a size of >100 nm in the sorbent can have an important role during the sulfurization and consecutive decomposition of the formed CaS; and (iii) the molar CaO/CaSO4 ratio in the decomposed CaS sample is affected by the pore structure.
Introduction
The following reaction has also been reported:
Limestone is a very important high-temperature desulfurization sorbent that is used in coal gasifiers for the in-bed removal of H2S. In coal gasifiers, the decomposition of limestone to CaO and CO2 is dependent on the concentration and/or pressure of CO2, as well as the temperature.1 Therefore, two sulfurization reactions must be considered:
CaO + H2S f CaS +H2O
(1)
CaCO3 + H2S f CaS + H2O + CO2
(2)
CaS + 2O2 f CaSO4
(5)
Whether the reaction described in reactions 3 and 4 or reaction 5 occurs is dependent on the reaction conditions.3,4 If a layer of CaSO4 covers the outside of the CaS particle, then CaS is not converted to CaSO4 until the CaSO4 decomposes to CaO, SO2, and O2 at high temperature (>1200 °C).4-6
1 CaSO4 f CaO + SO2 + O2 2
(6)
The CaS must be converted to CaSO4 before disposal, because H2S is released from the reaction between CaS and water (H2O).2 The conversion of CaS to CaSO4 is expressed as follows:
The following reaction between CaS and CaSO4 has been reported, in addition to reaction 6:7
3 CaS + O2 f CaO + SO2 2
(3)
Recently, the following reaction was also reported for the CaS decomposition:8,9
1 CaO + SO2 + O2 f CaSO4 2
(4)
CaS + 3CO2 f CaO + 3CO +SO2
* Author to whom correspondence should be addressed. Telephone: +81-86-251-8900. Fax: +81-86-251-8900. E-mail address: sasaokae@ cc.okayama-u.ac.jp. † Okayama University. ‡ The University of New South Wales. (1) Yrjas, K. P.; Zevenhoven, C. A. P.; Hupa, M. M. Ind. Eng. Chem. Res. 1996, 35, 176-183. (2) Ninomiya, Y.; Sato, A.; Watikinson, A. P. Oxidation of Calcium Sulfide in Fluidized Bed Combustion/Regeneration Condition. Proc. Int. Conf. Fluid. Bed Combust. 1995, 13th, 1027-1033.
3CaSO4 + CaS f 4CaO + 4SO2
(7)
(8)
(3) Torres-Ordonez, R. J.; Wall, T. F.; Longwell, J. P.; Sarofim, A. F. Energy Fuels 1989, 3, 506-515. (4) Torres-Ordonez, R. J.; Wall, T. F.; Longwell, J. P.; Sarofim, A. F. Energy Fuels 1989, 3, 595-603. (5) Lynch, D. C.; Elliott, J. F. Met. Trans. B 1980, 11B, 415-424. (6) Fenouil, L. A.; Lynn, S. Ind. Eng. Chem. Res. 1996, 35, 10241043. (7) Davies, N. H.; Laughlin, K. M.; Hayhurst, A. N. The Oxidation of Calcium Sulphide at the Temperatures of Fluidised-Bed Combustors. In Proceedings of the 25th Symposium (International) on Combustion; Combustion Institute: Pittsburgh, PA, 1994; pp 211-218.
10.1021/ef049818m CCC: $30.25 © 2005 American Chemical Society Published on Web 03/30/2005
Effect of Pore-Size Distribution of CaO Sorbent
Energy & Fuels, Vol. 19, No. 3, 2005 865
The conversion of CaS to CaCO3 and H2S in liquid phase has also been reported.10 We have found a preparation method that uses acetic acid to produce a highly active macroporous lime. The macroporous CaS formed from the macroporous lime was easily converted to CaSO4 and CaO. However, the CaS formed from natural lime was partially converted to CaSO4 and CaO, because only a small number of pores were present in the CaS sample.11,12 In our previous work,13 the decomposition of CaS was studied in the presence of H2O using a CaS reagent, because hot combustion gas that contains moisture is usually used for the decomposition of CaS. The role of H2O in the oxidative decomposition of CaS was clarified. That is, the direct reaction between CaS and H2O occurred and this reaction was accelerated by the presence of O2. The consecutive reactions are as follows:14
CaS + 3H2O f CaO + SO2 +3H2
(9)
1 H2 + O2 f H2O 2
(10)
1 CaO + SO2 + O2 f CaSO4 2
(4)
In this work, the effects of pore structure of the CaO sample on the sulfurization and the oxidative decomposition of the sulfurized sample were investigated using three CaO samples with different pore structures, in the presence of H2O at high temperature. Experimental Section Preparation of Macroporous Limes. The raw limestone used in this study was composed of 55.6 wt % CaO, 0.29 wt % SiO2, 0.18 wt % Al2O3, 0.08 wt % Fe2O3, and 0.04 wt % MgO. The limestone was obtained from the Yabashi Company. Raw lime was prepared from the limestone by heating it at 800 °C for 1 h in an air atmosphere. For the water-acetic acid swelling method, crushed lime particles with a diameter of 100 nm in size) was dependent on the preparation method. That is, the order of the development was as follows: Water-AC > Water > Lime. Apparatus and Procedure. Sulfurization of CaO. The reactivity of modified lime and raw lime with H2S were measured at 850 °C for 2 h, using a flow-type thermogravimetric apparatus (Shimadzu, model TGA50S) that was equipped with a quartz tubular reactor (inside diameter of 1.5 cm). Approximately 0.1 cm3 of the sample was placed in a platinum wire net sample holder (1.3 cm diameter). In the sulfurization experiments, a mixture of H2S (1500 ppm), H2 (10%), H2O (13.2%), and N2 (balance gas) was fed into the reactor at a rate of 500 cm3/min at standard temperature and pressure (STP). Oxidative Decomposition of CaS. The consecutive sulfurization and oxidative decomposition of the samples was performed using a flow-type thermogravimetric apparatus (Shimadzu, model TGA50S) under atmospheric pressure. The reaction
866
Energy & Fuels, Vol. 19, No. 3, 2005
Wu et al
Figure 2. Reactivity of the three samples with H2S. (Temperature, 850 °C; particle diameter, 2.0 mm.)
temperature examined was 850 °C. The reaction commenced when a mixture of 1% O2, 13.2% H2O, and N2 (balance gas) was fed into the reactor at 500 cm3/min at STP. Weight gain of the sample during the reaction was measured by the thermobalance. The amount of SO2 evolved during the oxidation was measured via a wet absorption method (Arsenazo III). The composition of the oxidized sample was calculated from the weight gain and the amount of SO2 evolved.12 The remaining CaS and products (CaSO4 and CaO) and CaS were analyzed via X-ray diffractometry (XRD) (Shimadzu, model XRD-6100).
Figure 3. X-ray diffraction (XRD) patterns of the sulfurized samples: (0) CaO and (4) CaS.
Results and Discussion Effect of the Pore-Size Distribution of CaO Sorbent on Its Sulfurization. The reactivity of the samples as a high-temperature desulfurization sorbent was examined at 850 °C. The reactivity of each sample is shown in Figure 2. It was confirmed that the reactivity of the lime was drastically improved by the modifications. The conversion of the modifiedWater-AC and Water samples after the 2 h of sulfurization was 90 and 61 mol%, respectively. These values were considerably larger than that of the Lime sample (43 mol%). The specific surface areas of the three samples were almost the same, as shown in Table 1; therefore, the activation by the modifications may be explained by the development of the macropores in the lime, because the order of the reactivity was consistent with the order of the development of large macropores (>100 nm in size) shown in Figure 1. When the macropores are formed in the sample by swelling, the diffusion resistance in the pore becomes small and the macropores also prevent pore plugging by the sulfurization. Thus, the reactivity of the sample was drastically improved. Figure 3 shows XRD patterns of the Water-AC and Lime samples sulfurized at 850 °C. CaS and CaO were the main components in the samples. The XRD measurement also suggests that, in the case of the WaterAC sample, almost all of the CaO was converted to CaS. This result is consistent with the high conversion shown in Figure 2. When CaO sample is converted to CaS, the pore volume of the sample decreases, because the CaS/CaO molar volume ratio is 1.7, and it can also be supposed that the pore structure changes. Figure 4 shows the pore-size distribution before and after sulfurization. This observation confirmed that the volume of the pores in the size range of 50-200 nm was drastically decreased by the sulfurization of the sample. Particularly, macropores >200 nm in size were scarcely observed in the Lime sample after the sulfurization.
Figure 4. Pore-size distributions of the samples before and after sulfurization: (a) water-acetic acid (Water-AC) sample, (b) Water sample, and (c) Lime sample.
Figure 5 shows scanning electron microscopy (SEM) images before and after the sulfurization. In the case of the Water-AC sample, the large macropores were observed in the sulfurized samples but not in the sulfurized Lime sample. Effect of the Pore-Size Distribution of the CaS Formed on Its Oxidative Decomposition. After sulfurization at 850 °C for 2 h, the system was substituted with N2 for ∼30 min and then a mixture of 1% O2, 13.2% H2O, and N2 (balance gas) was fed into the reactor for the oxidative decomposition of the formed CaS. Figure 6 shows weight gains of the samples during the consecutive sulfurization and the oxidative decomposition of the three samples. The weight gain of the sulfurized Water-AC sample during the decomposition was largest and most quickly attained a constant value among the three samples. The conversion of CaS to CaO was accompanied by a weight loss; however, the conversion of CaS to CaSO4 was accompanied by a weight gain. The weight change ceased when the reaction was completed. The order of the finish time of the sample was as follows: Water-AC < Water < Lime.
Effect of Pore-Size Distribution of CaO Sorbent
Energy & Fuels, Vol. 19, No. 3, 2005 867
Figure 5. Scanning electron microscopy (SEM) images of the Water-AC sample (top) and Lime (bottom) sample (a) before sulfurization and (b) after sulfurization.
Figure 6. Consecutive sulfurization and oxidation.
This order was opposite to the order of the number of macropores >100 nm in size (see Figure 4a-c). From this result, it was supposed that, if the pores 100 nm in size in the sample. The remaining CaS and the products (CaSO4 and CaO) were confirmed using XRD. The compositions of the samples after the oxidative decomposition for 2 h was calculated from the weight gain and the amount of SO2 evolved during the decomposition. Figure 7 shows the compositions of the three samples. The composition was calculated based on the CaS formed from the CaO sample. Therefore, the bases for calculation of the Water-AC, Water, and Lime samples were 90, 61, and 43 mol %, respectively. As shown in Figure 7, a large fraction of CaS in the modified limes was mainly converted to CaSO4 and partially converted to CaO, but only ca. 40% of CaS in the Lime sample was converted
Figure 7. Compositions of the three samples (Water-AC, Water, and Lime) after the oxidative decomposition.
to CaSO4 and CaO. The order of the amount of remaining CaS in the sample was as follows: Water-AC (1.3) < Water (1.8) , Lime (18.6). From a comparison of this result and the pore distributions of the samples shown in Figure 4, it was strongly suggested again that the presence of macropores >100 nm in size has an important role in the oxidative decomposition of CaS to CaSO4 and CaO. It may be supposed that the big macropores were not plugged by the formation of CaSO4 and presented a channel for the diffusion of water vapor and oxygen. In the case of the Lime sample, it was also supposed that the remaining small macropores might be shrunk or plugged by the formation of CaSO4, because the conversion of CaS to CaSO4 is accompanied with a volume increase (CaSO4/ CaS molar volume ratio of 1.6). The change of the poresize distribution before and after the oxidative decomposition will be discussed hereafter. The molar ratios of the formed CaO to the formed CaSO4 in the decomposed Water-AC, Water, and Lime samples were 23%,
868
Energy & Fuels, Vol. 19, No. 3, 2005
Wu et al
in size were decreased by the sulfurization and also the oxidative decomposition. However, no change in the number of pores >1000 nm in size was observed. These results suggest that larger pores (>1000 nm) do not have an important role during the two reactions. In the case of the Lime sample, the number of pores >100 nm in size were decreased by the sulfurization and also by the oxidative decomposition. After the oxidative decomposition, a considerable amount of pores ∼100 nm in size were observed; however, not much more decomposition occurred, as shown in Figure 6. If the pores present routes for the transfer of water and oxygen from the outside into inside of the particle, the oxidative decomposition should continue. Therefore, it was supposed that the inside wall of the pores produced during the oxidative decomposition were composed of CaSO4 and water/oxygen did not go inside the wall. The reason the pore size of the pores produced during the oxidative decomposition was larger than those from before the decomposition is unknown. Conclusion Figure 8. Change of the pore-size distribution of the WaterAC sample (top panel) and Lime sample (bottom panel) during the consecutive sulfurization and oxidative decomposition.
19%, and 13%, respectively. Usually, CaSO4 formed from the following reactions:
3 CaS + O2 f CaO + SO2 2
(3)
1 CaO + SO2 + O2 f CaSO4 2
(4)
The reason the CaO/CaSO4 ratio of the Water-AC sample was largest among the three samples may be explained by the difference of the pore-size distribution among the three samples. If the reaction described in reaction 4 occurs quickly, the CaO/CaSO4 molar ratio becomes small. Furthermore, SO2 formed from CaS in the small pore of the sorbent particle reacts with CaO before exiting the particle. However, if, in the sorbent particle, the big macropores are well-developed, some SO2 produced from CaS can exit the particle without reaction with CaO, because the mass transfer occur readily in the big macropores. The change in the pore-size distribution of the WaterAC and Lime samples before use, after sulfurization, and after the oxidative decomposition is shown in Figure 8. It was confirmed that, in the case of the Water-AC sample, the number of pores that measured ∼100 nm
In this work, the effect of pore-size distribution on sulfurization of CaO and the oxidative decomposition of the CaS formed from the CaO at high temperature were investigated, using three CaO samples that were prepared from the same raw lime. The specific surface area of the three CaO samples were almost the same within analytical error range; however, their pore structures were quite different from each other. The following results were obtained: (1) The macroporous CaO could be easily converted to CaS and the CaS formed was also easily converted to CaSO4 and CaO, because the sulfurization of the CaO and the consecutive oxidative decomposition of the formed CaS occurred without pore plugging by the product. (2) Pores >100 nm in size in the sorbent could have an important role during the sulfurization and the consecutive decomposition of the formed CaS. (3) The CaO/CaSO4 molar ratio in the decomposed CaS sample was affected by the pore structure: In the case of the Water-AC sample, which had well-developed larger macropores, the value of the CaO/CaSO4 ratio became larger than that of the Lime sample. Acknowledgment. This work was supported by the Ministry of Education, Science, Sport and Culture of Japan through the Grant-in-Aid on Priority-Area Research (Region No. 737, Grant No. 11218207). EF049818M
