Enhancement of the Direct Sulfation of Limestone ... - ACS Publications

Ind. Eng. Chem. Res. 2007, 46, 5295-5303

5295

Enhancement of the Direct Sulfation of Limestone by Alkali Metal Salts, Calcium Chloride, and Hydrogen Chloride Guilin Hu,* Kim Dam-Johansen, and Stig Wedel CHEC, Department of Chemical Engineering, Technical UniVersity of Denmark, 2800 Lyngby, Denmark

Jens P. Hansen FLSmidth A/S, 2500 Valby, Denmark

The enhancement of the direct sulfation of limestone by various Li+-, Na+-, and K+-containing inorganic salts, calcium chloride, and hydrogen chloride has been studied in a bench-scale quartz reactor. Experimental results indicate that Li+- and Na+-containing salts enhance the sulfation process by increasing ionic mobility in both the solid reactant (calcite) and the solid product (anhydrite), whereas K+-containing salts and CaCl2 promote the direct sulfation of limestone only by increasing ionic mobility in the solid reactant. The significant increase in ionic mobility in limestone causes serious sintering of the limestone particles and formation of significantly fewer but larger product nuclei/grains. The significant increase in ionic mobility in the product phase causes serious deformation and easy coalescence of the product grains. The enhancement by HCl is related to a eutectic which is formed by simultaneous formation of calcium chloride by the chlorination of limestone and calcium sulfate by the sulfation of limestone. Introduction Limestone is a widely used sorbent for high-temperature desulfurization of flue gases because of its wide availability and low cost. The key reaction for the desulfurization of flue gas by limestone at high temperatures is the sulfation reaction between SO2 and limestone particles. This reaction can proceed via two different routes depending on whether calcination of the limestone takes place under given reaction conditions. If the CO2 partial pressure in the system is lower than the equilibrium decomposition pressure of limestone, the limestone first decomposes to form CaO, which then reacts with SO2 and O2. This process is often called indirect sulfation of limestone and is expressed by the following overall reactions:

CaCO3(s) f CaO(s) + CO2(g)

(1)

CaO(s) + SO2(g) + 0.5O2(g) f CaSO4(s)

(2)

If the CO2 partial pressure in the system is higher than the equilibrium decomposition pressure of limestone, the limestone may react directly with SO2 and O2. This process is often called direct sulfation of limestone and is expressed by the following overall reaction:

CaCO3(s) + SO2(g) + 0.5O2(g) f CaSO4(s) + CO2(g) (3) The direct sulfation of limestone takes place for example in the desulfurization of flue gas by direct dry limestone injection during pressurized fluid-bed combustion (PFBC) and in the cyclone preheater used in cement production. A number of investigations in the past have shown that both direct and indirect sulfation of limestone can be enhanced by addition of various additives, which is a potential method to increase the desulfurization effectiveness and probably the efficiency of limestone usage as well. In the paper by Hu et * Corresponding author. telephone: +45 45252525. fax: +45 45882258. E-mail: [email protected].

al.1 the published results related to the enhancement of both indirect and direct sulfation of limestone by additives are reviewed and discussed. The majority of the published results in the literature are related to the indirect sulfation of limestone. Various alkali metal salts, CaCl2, HCl, and ferric oxide (Fe2O3) are among those additives that are found to have enhancing effects on the indirect sulfation of limestone. The available results related to the direct sulfation of limestone are very limited. Fuertes and Fernandez2 performed a rather comprehensive study on the enhancement of the direct sulfation of limestone by various metallic salts in the temperature interval from 973 to 1148 K. The direct sulfation of limestone was shown to be enhanced by Li+-, Na+-, and K+-containing inorganic salts. Fuertes and Fernandez2 concluded that the enhancement of the direct sulfation of limestone by these alkali metal salts is due to reduced resistance of solid-state diffusion, based on the observation that the enhancing effects were more pronounced at higher conversions at which the sulfation process is assumed to be controlled by solid-state diffusion through the product layer. Partanen et al.3 observed that the direct sulfation of limestone was enhanced by the presence of HCl in the gas phase. It was suggested that formation of a eutectic between CaCO3, CaSO4, and CaCl2 was the reason for the observed enhancement. In general, the current understanding of the enhancement mechanisms of the different additives is still limited to very general concepts such as “increased solid-state diffusivity”, first suggested by Borgwardt et al.4 Few details about how and where solid-state diffusivity is increased by the additives are known. The work presented in this paper is part of a project to explore methods to reduce SO2 emission from the cyclone preheater used in cement production. Hu et al.5 in an earlier study have shown that the direct sulfation of limestone involves oriented nucleation and growth of anhydrite (the solid product) and that the sulfation process is restricted by solid-state diffusion. This study is a continuation of the earlier study. Its purpose is to explore the effects of different additives on the direct sulfation of limestone and to assess the related mechanisms.

10.1021/ie070208u CCC: $37.00 © 2007 American Chemical Society Published on Web 06/27/2007

5296

Ind. Eng. Chem. Res., Vol. 46, No. 16, 2007

Table 1. Tested Additives and Their Dosages additive name

dosage, mol %

additive name

NaCl Na2CO3 Na2SO4 Li2CO3

2 1 1 1

KCl K2CO3 CaCl2 HCl (gas)

dosage, mol % 2 1 1 1000 ppm (in the gas phase)

Experimental Section The experiments were performed in a bench-scale, electrically heated, quartz fixed-bed reactor, which was originally developed by Dam-Johansen et al.6-9 Details about the reactor system and the experimental procedure have been described elsewhere in the paper by Hu et al.5 The reaction conditions were kept within the following ranges: temperature, 723-923 K; SO2 concentration, 900-1800 ppm; O2 concentration, 1-3 vol %; CO2 concentration, 15-30 vol %. These are typical conditions in the cyclone preheater used in cement production. Conversion of the limestone was calculated on the basis of the difference between the inlet SO2 concentration and the outlet SO2 concentration. In the conversion rate data presented in the following sections, values in the first minute are not included because of the significant influence of the residence time distribution (RTD) in the system in the first minute of each experiment. After this short period, the change of the outlet SO2 concentration was relatively slow; the RTD influence became insignificant. The standard deviation of the conversion rate is about (2.5%, which is calculated by using data from two sets of experiments repeated under identical reaction conditions (900 data pairs). This deviation is expected to hold for all conversion rates presented in this paper. Preparation of Samples. The limestone used for the experiments is a soft and porous bryozoan limestone from Faxe Kalk in Denmark (it will be called Faxe Bryozo later on in the text). Details of this limestone have been described elsewhere in the paper by Hu et al.5 Faxe Bryozo was in powder form when purchased. The limestone was sieved, and the fraction with particle sizes between 0.18 and 0.25 mm was used for the experiments. Calculations indicate that intraparticle diffusion resistance under the applied reaction conditions was insignificant with such particle sizes. The additives were introduced into the limestone samples by impregnation. A molar ratio of 0.02 between the relevant additive ions and Ca2+ in the limestone samples was used for all the tested additives. For the impregnation, the relevant additive was first dissolved in water and then mixed with the limestone particles to form a slurry. The slurry was then heated to evaporate most of the water. The still-wet sample was further dried in an electrically heated oven at 393 K for approximately 12 h. The dried sample was gently ground and sieved again. The fraction between 0.18 and 0.25 mm was used for the experiments. Table 1 shows the tested additives and the dosages. The dosages are calculated values. Preliminary experiments showed that the conversion rate of a doped limestone sample was strongly influenced by its thermal history before the sulfation reaction. To investigate this effect, some of the doped limestone samples were thermally treated at different temperatures in a gas consisting of 30% CO2, 3% O2, and 67% N2 before they were used for the sulfation experiments. Results and Discussion Apparent Effect of the Additives. The effects of the different additives listed in Table 1 are demonstrated in Figures 1 and 2.

Figure 1. Conversion rate vs time curves of Faxe Bryozo doped with different additives (other conditions: T, 823 K; P, 0.11 MPa; inlet SO2, 1800 ppm; O2, 3%; CO2, 30%; N2, balance).

Figure 2. Conversion rate vs time curves of doped and thermally pretreated Faxe Bryozo (other conditions: thermal pretreatment before sulfation, 923 K for 1 h; T, 823 K; P, 0.11 MPa; inlet SO2, 1800 ppm; O2, 3%; CO2, 30%; N2, balance).

Figure 1 shows the conversion rate vs time curves obtained with the samples without a thermal pretreatment, while Figure 2 shows the conversion rate vs time curves obtained with the samples with a thermal pretreatment at 923 K for 1 h. As shown in these two figures, the sulfation process was enhanced by all the tested additives. The conversion rates were increased about 6-8 times with the most effective ones which are Li2CO3 and Na2CO3 in the case without the thermal pretreatment and NaCl, KCl, Na2CO3, and Na2SO4 in the case with the thermal pretreatment. The observed significant enhancing effects of the tested alkali metal salts and HCl on the direct sulfation of limestone are in good agreement with the experimental results obtained by Fuertes and Fernandez2 and Partanen et al.3 These curves show also that the behaviors of these additives differed. Without the thermal pretreatment, the conversion rates of the samples doped with Li2CO3, Na2CO3 K2CO3, Na2SO4, and CaCl2 were relatively high at the start but decreased relatively fast with reaction time. In contrast, conversion rates with the samples doped with NaCl and KCl were not as high as with the above five additives at the start but were kept almost constant for a long time. Except for CaCl2 and Li2CO3, the situation is the same with the samples thermally pretreated. The enhancing effects of these additives were strongly affected by the thermal pretreatment. As is seen in Figures 1 and 2, after the thermal pretreatment, NaCl and KCl became the most effective additives, whereas Li2CO3 and CaCl2, the most ineffective ones. Another interesting phenomenon is the appearance of upcurved conversion rate vs time curves with

Ind. Eng. Chem. Res., Vol. 46, No. 16, 2007 5297

Figure 3. SEM images of Faxe Bryozo particles doped with 2% NaCl and sulfated at 873 K to different conversions (other conditions: thermal pretreatment before sulfation, 923 K for 1 h; P, 0.11 MPa; inlet SO2, 1800 ppm; O2, 3%; CO2, 30%; N2, balance): (a) unsulfated particles; (b) sulfated for 5 min (x ) ca. 1.9%; white arrows indicate the crystal grains); (c) sulfated for 10 min (x ) ca. 4.4%); (d) sulfated for 20 min (x ) ca. 9.0%); (e) sulfated for 60 min (x ) ca. 14.8%).

maximum (it will just be mentioned as “upcurved” later on) obtained with the samples doped with NaCl, KCl, Na2CO3, Na2SO4, and HCl. The upcurved form means that the sulfation reaction was somehow accelerated in the period before the maximum point. This upcurved form is more prominent with the samples thermally pretreated. To see the morphological changes, the doped and sulfated samples were examined by scanning electron microscope (SEM; Zeiss 1530). The SEM examinations revealed formation of product crystals at the surface of the limestone particles doped with the tested alkali metal salts and CaCl2 and a meltlike layer at the surface of limestone particles sulfated in a gas containing HCl. The crystals and the meltlike layer were confirmed by elemental analysis with EDS (energy dispersive spectrometry) X-ray microanalysis, and powder patterns of X-ray powder diffraction to be anhydrite (CaSO4). This is also the only solid product of the sulfation reaction found, the same as without

addition of the additives.5,10-12 In the following sections the enhancement mechanisms of these different additives are discussed and assessed along with the presentation of the relevant experimental results. Enhancement by NaCl. (a) Mechanism of the Enhancement. Hu et al.5 demonstrated that the direct sulfation of limestone involves oriented nucleation and growth of crystal grains of the solid product (anhydrite). SEM examinations of the sulfated samples demonstrate that in the presence of NaCl the direct sulfation of limestone involves nucleation and crystal grain growth of the solid product as well. Figure 3 shows a series of SEM images that demonstrate the nucleation and subsequent growth of product crystals with samples doped with NaCl and thermally pretreated. The product crystals are seriously deformed when compared to the normal orthorhombic form of anhydrite. With increasing conversion, the calcite grain surface is gradually covered by the product

5298


crystals. The product crystals coalesce when in touch with each other and form a poreless product layer. The limestone particles were also sintered to a certain degree which is reflected by the generally rounded particle shape as shown in Figure 3a. The situation is similar to particles sulfated at 823 K. Compared to the situation without additives, the nucleation and growth process in the presence of NaCl differs in various aspects: (i) The number of nuclei formed in the presence of NaCl is significantly fewer than without the additive; the distance between the formed nuclei is relatively long. (ii) The product crystal grains formed in the presence of the additive are seriously deformed and have almost totally lost the characteristic orthorhombic form of anhydrite crystals. (iii) The product crystals grow laterally and coalesce when touching. (iv) The nucleation and growth of the product crystals are not oriented any more. These changes caused by NaCl indicate a significantly increased solid-state diffusivity/mobility in both the solid reactant (calcite) and the solid product (anhydrite), most probably by the mechanism of formation of more extrinsic point defects13 in the crystal lattices of calcite and anhydrite by the incorporation of Na+ and/or Cl- ions in their lattice structures. The increased solid-state mobility in calcite makes it possible for product ions to migrate longer distances to reach the nucleation and growth sites, while the increased ionic mobility in the product crystal grains is the main reason for the deformation of the product crystal grains and their easy coalescence. The increased ionic mobility in the product crystals and the probably less stable lattice structure at the surface of calcite because of the significantly increased ionic mobility are most likely the main reasons for the nonoriented nucleation and growth. Considering the significant enhancing effect of CaCl2 (as shown in Figure 1), the effect of NaCl should contain the contributions of both Na+ and Cl- ions. As shown in Figures 1 and 2, the enhancing effects of Na2CO3 and Na2SO4 decreased a little (about 10-20%) after a thermal pretreatment at 923 K for 1 h, whereas the enhancing effect of NaCl increased about 100% after the same thermal pretreatment. This is most likely due to the enhanced effect of Cl- by the thermal pretreatment. The increase in ionic mobility in the product crystals is most likely caused by the diffusion of sodium ions (Na+) from the calcite phase into the solid product phase during the nucleation and growth process, as analysis of the sulfated particles by EDS X-ray microanalysis revealed the presence of Na+ ions in the product crystals, even with a concentration significantly higher than in the calcite phase. The sulfation reaction may take place on both the surface of calcite and the surface of product crystals. The observed conversion rate is the sum of the contributions from the sulfation reaction on both surface types. The increase in ionic mobility in both the calcite phase and the solid product phase may increase carbonate concentrations at both surfaces and thus also the reactivity of these two surfaces. The enhancement by NaCl thus most likely comes from three major contributions: the first from the increased reactivity on the calcite surface, the second from the increased reactivity on the product surface, and the third from the slower shielding of the calcite surface because of formation of fewer but larger product nuclei/grains. (b) Kinetic Controlling Mechanism. It has been concluded by Hu et al.5 that the direct sulfation of limestone consists of five general steps, including gas film diffusion, pore diffusion in product layer, chemical reaction, solid-state diffusion, and

Figure 4. Variation of the apparent reaction order of SO2 with reaction temperature with NaCl-doped and presulfated Faxe Bryozo (2, with the sample doped with 2 mol % NaCl, thermally pretreated at 923 K for 1 h, and presulfated to 4.4% conversion; 0, with the sample doped with 2 mol % NaCl, thermally pretreated at 923 K for 1 h and presulfated to 14.8% conversion; the “I” bars are the estimated maximum deviations.) (Other conditions: P, 0.11 MPa; inlet SO2, 900-1800 ppm; O2, 3%; CO2, 30%; N2, balance.)

nucleation/growth of the solid product, and that at temperatures lower than about 973 K the sulfation process is under mixed control by both chemical reaction and solid-state diffusion. It seems that addition of NaCl does not change this situation under the reaction conditions used in this study. This is well-illustrated by the variation of the apparent reaction order of SO2 and the apparent activation energy with the reaction conditions. Figure 4 shows the apparent reaction order of SO2 at different conversions and temperatures. To ensure close reaction conditions (such as same limestone conversion and same product layer structure), presulfated samples (the same samples for SEM images c and e in Figure 3) were used for the evaluation. The reaction order was evaluated by a step decrease of SO2 concentration from 1800 to 900 ppm. As shown in Figure 4, the apparent reaction order of SO2 was significantly influenced by the temperature. With the sample presulfated to a conversion of about 4.4%, the apparent reaction order of SO2 increased from about 0.2 at 823 K to about 0.5 at 923 K. This may be explained by the reduced resistance of solid-state diffusion at higher temperatures. With the sample presulfated to a conversion of about 14.8%, the increase in the reaction order with increasing temperature was significantly depressed, most likely related to the fact that the sulfation reaction takes place only at the surface of the product layer due to the total shielding of the calcite surface at such a conversion level. With the same presulfated samples as used for the evaluation of the apparent reaction order of SO2, the apparent activation energy was evaluated in the temperature interval from 823 to 923 K. The apparent activation energy was observed to be significantly influenced by SO2 concentration in the gas. With the sample presulfated to a conversion of about 4.4%, the apparent activation energy was determined to be about 98 kJ/ mol in a gas containing 1800 ppm SO2. With the same sample and in a gas containing 900 ppm SO2, the apparent activation energy was determined to be only about 77 kJ/mol. Similarly, with the sample presulfated to a conversion of about 14.8%, the apparent activation energy decreased from about 100 kJ/ mol to about 90 kJ/mol when the SO2 concentration was decreased from 1800 to 900 ppm. This phenomenon may be explained by a decreased resistance of solid-state diffusion at a lower SO2 concentration. The relatively less decrease in the apparent activation energy with the sample presulfated to a conversion of 14.8% may be explained as well by a generally increased resistance of solid-state diffusion at higher conversions.


Figure 5. Conversion rate vs time curve of Faxe Bryozo doped with 2% NaCl and sulfated at 873 K represents 1.0 × 10-4 (other conditions: thermal pretreatment before sulfation, 923 K for 1 h; P, 0.11 MPa; inlet SO2, 1800 ppm; O2, 3%; CO2, 30%; N2, balance). (Points a-d are indications for sampling positions for Figure 3).

Figure 6. Influence of temperature on the conversion rate vs time curve of Faxe Bryozo doped with 2% NaCl (other conditions: thermal pretreatment before sulfation, 923 K for 1 h; P, 0.11 MPa; inlet SO2, 1800 ppm; O2, 3%; CO2, 30%; N2, balance).

(c) Reasons for the Upcurved Form of the Conversion Rate vs Time Curve. Figure 5 shows the conversion rate vs time curve of NaCl-doped sample at 873 K. Points a-d correspond to sampling positions for SEM images b-e in Figure 3, respectively. The upcurved form of the conversion rate vs time curve could be caused by a combination of the progressive shielding of the more reactive calcite surface by the solid product crystals and the presence of significant influence of solid-state diffusion. In the sulfation process, sulfate ions formed at the uncovered calcite surface diffuse into the product crystals. The diffusion rate toward a product crystal may be roughly assumed to be proportional to the length of the boundary of the crystal and the sulfate concentration gradient around the crystal. The sulfation rate at the uncovered calcite surface is proportional to the uncovered calcite surface area and the concentration of carbonate ions at the uncovered calcite surface. A balance between the formation of sulfate ions at the uncovered calcite surface and the diffusion of the formed sulfate ions to the product crystal is supposed to exist. With increasing product crystal size, the uncovered calcite surface area shrinks. Initially, the conversion is relatively low. The crystals are small and cover only a small fraction of the calcite surface. The increasing conversion rate with increasing conversion or reaction time is most likely caused by two major reasons. One is the higher percentage of increase in the size of the crystals than the percentage of reduction in the uncovered calcite surface area. The other is the significant resistance of solid-state diffusion. The rate by which the formed sulfate ions diffuse into a product crystal increases with the increase in the size of the product crystal, which in turn results in a decrease of the concentration of sulfate ions and correspondingly an increase in the concentration of carbonate ions at the uncovered calcite surface. This increased concentration of carbonate ions results in a higher sulfation rate at the uncovered calcite surface which keeps in balance with the rate by which the sulfate ions diffuse into the crystal. The apparent result is thus an increasing conversion rate. After the conversion reaches a certain level, the product crystals become large and cover a large percentage of the calcite surface. The conversion rate now decreases with increasing conversion because the percentage of reduction in the uncovered calcite surface area is now significantly larger than the percentage of increase in the size of the product crystal. The increase in the reaction rate caused by the increased concentration of carbonate ions is now not sufficient to compensate for the reduction in the uncovered surface area because the concentration of carbonate ions is limited to that in pure calcite.

The maximum rate may be the transition point between these two situations. As shown in Figures 3 and 5, the maximum in the conversion rate vs time curve appeared when the calcite surface was only half-covered by product crystals. This indicates that the reactivity of the calcite surface was higher than the reactivity of the surface of the product crystals, as expected. The increasing reactivity of the uncovered calcite surface and the gradual shielding of the surface by the solid product crystals with increasing conversion are therefore most likely the two key factors for the upcurved form of the conversion rate vs time curves. However, the shape of a conversion rate vs time curve may depend on a number of factors such as the relative dominance of chemical reaction and solid-state diffusion, the morphological properties of the limestone particles, the ionic mobility in the product phase, the number of nuclei, and the sulfation rate at the product surface. Figure 6 shows conversion rate vs time curves of NaCl-doped samples at different temperatures. This figure demonstrates the movement of the maximum toward the left side with increasing temperature and the significantly flattened and almost invisible peak at temperatures lower than about 823 K. The explanation could be that at sufficiently high temperatures (greater than about 923 K), chemical reaction begins to become the dominant control mechanism. At temperatures lower than about 823 K, solid-state diffusion becomes the dominant control mechanism. Enhancement by Other Alkali Metal Salts. Other alkali metal salts tested in this study were not investigated as extensively as NaCl. However, SEM examinations showed that the morphological changes of the sulfated samples doped with Na2CO3 and Na2SO4 are similar to the samples doped with NaCl. On the basis of the similar morphological changes and the upcurved conversion rate vs time curves with samples doped with Na2CO3, Na2SO4, and NaCl, the enhancement mechanisms of Na2CO3 and Na2SO4 are most likely similar to that of NaCl. Li2CO3 caused a severe sintering of the limestone particles and the formation of highly deformed product crystals as demonstrated in Figure 7 (the part with a smooth look at the left side of the figure is product crystals). The conversion rate vs time curve of the sample doped with Li2CO3 was also observed to be upcurved at 873 K. Except the much stronger sintering effect of Li2CO3 on limestone particles, it seems to enhance the sulfation process in the same way as those Na+containing salts. KCl affects the sulfation process differently than Li2CO3 and the Na+-containing salts, reflected by the difference in the shape of the product crystals as shown in Figure 8. In contrast to the significantly deformed crystals with NaCl-doped samples, KCl caused the formation of well-shaped product crystals. Analysis

5300


Figure 7. SEM image of Faxe Bryozo doped with 1% Li2CO3 and sulfated to a conversion of about 1.5% at 823 K (other conditions: thermal pretreatment before sulfation, 923 K for 1 h; P, 0.11 MPa; inlet SO2, 1800 ppm; O2, 3%; CO2, 30%; N2, balance).

Figure 8. SEM image of Faxe Bryozo doped with 2% KCl and sulfated at 823 K for 30 min (x ) ca. 6%). (Other reaction conditions: thermal pretreatment before sulfation, 923 K for 1 h; P, 0.11 MPa; inlet SO2, 1800 ppm; O2, 3%; CO2, 30%; N2, balance.)

of the sulfated particles by EDS X-ray microanalysis showed no significant amount of K+ or any increase in K+ concentration in the product crystals with the KCl-doped sample. This indicates that diffusion of K+ into the product crystals during the nucleation and growth process was limited, most likely due to the fact that the size of K+ is significantly larger than the size of Ca2+. Thus, KCl most likely affects the sulfation reaction by increasing solid-state diffusivity only in the solid reactant. Enhancement by CaCl2. The mechanism of the enhancement by CaCl2 is different from that of the tested alkali metal salts judged by SEM images of the reacted limestone particles (Figure 9). Figure 9 demonstrates that CaCl2 caused a severe sintering of the limestone particles just as Li2CO3. However, in contrast to the totally deformed solid product crystals with the samples doped with Na+-containing salts and Li2CO3, the solid product crystals are well-shaped and in slice form. The situation is the same with CaCl2-doped particles reacted at 873 K and with or without the thermal pretreatment. The enhancing effect of CaCl2 appears to be solely due to the increased ionic mobility in the limestone, indicated by the significantly sintered appearance of the limestone particle and the relatively long distance between the product crystals. The ionic mobility in the limestone may be increased due to the formation of more cation vacancies by the incorporation of the single valent chloride ions (Cl-). The well-shaped form of the product crystals shown in Figure 9 indicates that ionic mobility in product crystals was not increased noticeably by the additive, most likely because of the

difficulties of the incorporation of chloride ions into the crystal lattice of the solid product (anhydrite). Analysis of the sulfated particles by EDS X-ray microanalysis showed that Cl- was present in significant amount in the limestone but not in the product crystals. A probable reason for this phenomenon could be difficulties for Cl- ions to diffuse into product crystals because of too-large differences in both the size and the structure of Cl- and SO42-. With the same sample for Figure 9, an upcurved conversion rate vs time curve was observed at 873 K. The upcurved form may be explained as for samples doped with NaCl. Partanen et al.3 suggested that a eutectic between CaCO3, CaSO4, and CaCl2 might be formed at a temperature above 853 K on the basis of thermodynamic model calculations. However, SEM examinations of the CaCl2-doped Faxe Bryozo particles that were sulfated at 873 K showed that such a eutectic apparently was not formed under the reaction conditions used in this study, possibly because of too-low CaCl2 concentrations at the limestone surface. Enhancement by HCl. The sulfation of Faxe Bryozo was enhanced by the presence of HCl in the gas phase, as shown in Figure 1. SEM examinations revealed that the addition of HCl in the gas phase resulted in the formation of a meltlike product layer as demonstrated in Figure 10. As is seen in Figure 1, the conversion rate vs time curve with HCl addition is upcurved with the maximum rate appearing relatively early. This could mean a relatively fast and progressive covering of the more reactive calcite surface by the meltlike product phase. Partanen et al.3 observed the formation of CaCl2 by the chlorination reaction between limestone and HCl in the gas. In this study, it was observed that the conversion rate jumped instantly to a significantly higher level after the addition of HCl was stopped and maintained at this new level for a relatively long time. Analysis of the meltlike product layer by EDS X-ray microanalysis showed also a relatively high content of chloride. These findings are clear evidence for the existence of chlorination of the limestonesa competing reaction to the sulfation of the limestone. A comparison between the results with CaCl2 addition and the results with HCl addition indicates that the formation of the meltlike product layer may be related to the simultaneous formation of CaCl2 and CaSO4, which probably creates conditions for formation of the eutectic suggested by Partanen et al.3 However, formation of a eutectic may not always mean a faster conversion rate. Compared to samples doped with alkali metal salts, the conversion rate of the limestone in the presence of HCl was significantly lower. Blocking of the internal surface area for the sulfation reaction by the eutectic and the presence of the competing reaction, the chlorination of the limestone, may be part of the reason for a lower conversion rate. Thermal Pretreatment. The results shown earlier demonstrate that a thermal pretreatment of the doped limestone samples had large influence on the enhancing effects of the additives. Three major effects of the thermal pretreatment have been observed. The first is the increase in reactivity of the limestone, most probably due to the formation of more point defects because of increased incorporation of the additive in the crystal lattice of calcite. The second is the decrease in the total surface area of the thermally pretreated sample. As shown in Table 2, the total surface areas of the doped samples are all reduced in different degrees after a thermal pretreatment at 923 K for 1 h. The reduction in the total surface area is most significant with samples doped with Li2CO3 and Na2CO3.


Figure 9. SEM image of sulfated Faxe Bryozo particles doped with 1% CaCl2 (other conditions: thermal pretreatment before sulfation, 923 K for 1 h; T, 823 K; P, 0.11 MPa; inlet SO2, 1800 ppm; O2, 3%; CO2, 30%; N2, balance): (a) unreacted particles; (b) sulfated at 823 K (x ) ca. 1.3%).

Figure 10. SEM image of the surface of Faxe Bryozo particles sulfated in the presence of 1000 ppm HCl in the gas at 823 K (x ) ca. 2.7%). (Other conditions: P, 0.11 MPa; inlet SO2, 1800 ppm; O2, 3%; CO2, 30%; N2, balance.) Table 2. Total Surface Areas of the Undoped and Some of the Doped Samples before and after Thermal Pretreatment at 923 K for 1 h (Only 15 min for the Undoped Sample) measured total surface areaa (m2/g) sample

before the thermal pretreatment

after the thermal pretreatment

blank Faxe Bryozo Faxe Bryozo + NaCl Faxe Bryozo + KCl Faxe Bryozo + Li2CO3 Faxe Bryozo + Na2CO3 Faxe Bryozo + K2CO3 Faxe Bryozo + Na2SO4

0.79 0.62 0.55 0.62 0.56 0.47 0.6

0.62 (heated for 15 min) 0.51 0.50 0.22 0.23 0.41 0.46

a

Determined by BET (Micrometrics ASAP 2000).

The third is the influence on the ionic mobility in the solid product, which can be illustrated by SEM image of the sulfated Faxe Bryozo particles doped with Li2CO3. As shown in Figure 11, without the thermal pretreatment, relatively well shaped product crystal grains of around 1 µm diameter were formed on the particle surface, which is very different from the sample thermally pretreated as shown in Figure 7. Samples doped with NaCl, Na2CO3, and Na2SO4 showed enhanced sintering of the product phase as well when the samples were thermally pretreated before sulfation. However, the same effect was not observed with samples doped with CaCl2. These phenomena indicate that the thermal pretreatment has great influence on the diffusion of Li+ and Na+ ions into the product phase but not K+ and Cl- ions during the nucleation and growth process. The increased diffusion of Li+ and Na+ ions after the thermal pretreatment could be caused by an increase in the concentra-

Figure 11. SEM image of Faxe Bryozo doped with 1% Li2CO3 and sulfated to a conversion of about 4.6% at 823 K (other conditions: thermal pretreatment, none; P, 0.11 MPa; inlet SO2, 1800 ppm; O2, 3%; CO2, 30%; N2, balance).

tions of Li+ and Na+ ions in the calcite lattice and/or the influence of the thermal pretreatment on the lattice site types in calcite which are occupied by these ions. The apparent conversion rate of a doped sample depends on all the above three aspects. For example, as shown in Figures 1 and 2, the significantly reduced conversion rate of the sample doped with Li2CO3 and thermally pretreated before the sulfation may be explained by the significantly reduced total surface area and the significantly increased ionic mobility in the solid product by the thermal pretreatment, whereas the significantly increased conversion rate with the sample doped with NaCl and thermally pretreated before the sulfation may be explained by the significantly increased ionic mobility and a limited decrease in the total surface area by the thermal pretreatment. The temperatures used for the thermal pretreatment have great influence on the enhancing effect of an additive. Figures 12 and 13 demonstrate that the enhancing effect of NaCl on the sulfation process was significantly increased with increasing temperature of the thermal pretreatment. This may be explained by the formation of more point defects by an increased incorporation of the additive in the crystal lattice of calcite at a higher temperature. No extra benefit was obtained with durations of the thermal pretreatment longer than 1 h. The insignificant effect of the longer durations for the thermal pretreatment may be due to the limitation on the incorporation of the additive in the crystal lattice of calcite determined by its solid-state solubility in calcite.

5302


Figure 12. Influence of the temperature of the thermal pretreatment on the conversion rate of Faxe Bryozo doped with 2% NaCl (other conditions: T, 823 K; P, 0.11 MPa; inlet SO2, 1800 ppm; O2, 3%; CO2, 30%; N2, balance).

Figure 13. Influence of the duration of the thermal pretreatment on the conversion rate of Faxe Bryozo doped with 2% NaCl (other conditions: T, 823 K; P, 0.11 MPa; inlet SO2, 1800 ppm; O2, 3%; CO2, 30%; N2, balance).

Conclusions The direct sulfation of limestone has been observed to be significantly enhanced by various Li+-, Na+-, and K+-containing inorganic salts (Li2CO3, NaCl, Na2CO3, Na2SO4, KCl, and K2CO3), CaCl2, and HCl. The additives cause also the conversion rate vs time curves to become upcurved with a maximum and the solid product (anhydrite) to be formed in different physical forms depending on the additive types and reaction conditions. These phenomena are explained by increased ionic mobility in the solid phases by the additives. Depending on the additive types, ionic mobility may be increased in both the solid reactant (limestone) and the solid product (anhydrite) or just in the solid reactant. The sulfation process in the presence of the alkali metal salts involves nucleation and growth of the solid product crystals similar to the case without additives. Li+- and Na+-containing salts may enhance the sulfation process by increasing ionic mobility in both the solid reactant (calcite) and the solid product (anhydrite). K+-containing salts enhance the sulfation process mainly by increasing ionic mobility in the solid reactant. An increase in ionic mobility in the solid reactant results in the formation of fewer but larger nuclei/crystals of anhydrite, while a significant increase in ionic mobility in the anhydrite crystals causes them to lose their normal orthorhombic form and coalesce easily. The increase in ionic mobility in the solid product is caused by diffusion of the relevant alkali metal ions into the product phase during the nucleation and growth process. A progressive covering of the surface of the limestone particles/ grains by the (coalesced) product crystals generally takes place during the sulfation process. In the presence of CaCl2, the sulfation process involves nucleation and crystal growth of the solid product as well. CaCl2 enhances the sulfation process by increasing only ionic mobility

in the solid reactant (calcite). The ionic mobility in the solid product is not increased significantly by CaCl2 mainly because of the difficulties for chloride ions to diffuse into the product phase. The significant increase in ionic mobility solely in the solid reactant results in formation of relatively large, individual, and well-shaped anhydrite crystals. The enhancement by HCl in the gas phase is related to a eutectic which is formed by the simultaneous formation of CaCl2 by the chlorination reaction of limestone and CaSO4 by the sulfation of limestone. The upcurved conversion rate vs time curve of the doped limestone is a combined result of an increase in the reactivity of the uncovered limestone surface with increasing conversion and a gradual shielding of the limestone surface by the product crystals or eutectic. The thermal experience of the doped limestone particles before the sulfation reaction has significant influence on the sulfation kinetics mainly because of its influence on the physical/ chemical properties of the doped limestone (such as total surface area and solid-state diffusivity/mobility) and the later diffusion of the relevant additive ions into the product phase during the sulfation process. The results obtained in this study show that the rate of the direct sulfation of limestone at temperatures around 823 K can be enhanced about 6-8 times by the addition of alkali metal salts in a relatively low dosage. This means that with proper engineering the application of additives may make the direct sulfation of limestone practical for the desulfurization of flue gases at a temperature significantly lower than 973 K. This may be especially valuable for those industrial processes that generate sulfur-containing flue gases at a relatively low temperature such as in cement production. Nomenclature P ) total pressure, MPa T ) temperature, K t ) reaction time, s x ) conversion of limestone, dimensionless Acknowledgment This work is part of the research program of the CHEC (Combustion and Harmful Emission Control) Research Center funded a.o. by the Technical University of Denmark, the Danish Technical Research Council, the European Union, the Nordic Energy Research, Dong Energy A/S, Vattenfall A.B., F L Smidth A/S, and Public Service Obligation funds from Energinet.dk and the Danish Energy Research program. This particular work is financially supported by the Technical University of Denmark and FLSmidth A/S. Literature Cited (1) Hu, G.; Dam-Johansen, K.; Wedel, S.; Hansen, J. P. Review of the direct sulfation reaction of limestone. Prog. Energy Combust. Sci. 2006, 32, 386. (2) Fuertes, A. B.; Fernandez, M. J. The effect of metallic salt additives on direct sulfation of calcium carbonate and on decomposition of sulfated samples. Thermochim. Acta 1996, 276, 257. (3) Partanen, J.; Backman, P.; Backman, R.; Hupa, M. Absorption of HCl by limestone in hot flue gases. Part III: Simultaneous absorption with SO2. Fuel 2005, 84, 1685. (4) Borgwardt, R. H.; Bruce, K. R.; Blake, J. An investigation of productlayer diffusivity for CaO sulfation. Ind. Eng. Chem. Res. 1987, 26, 1993. (5) Hu, G.; Dam-Johansen, K.; Wedel, S.; Hansen, J. P. Direct sulfation of limestone. AIChE J. 2007, 53 (4), 945.

Ind. Eng. Chem. Res., Vol. 46, No. 16, 2007 5303 (6) Dam-Johansen, K.; Østergaard, K. High-temperature reaction between sulfur dioxide and limestonesI. Comparison of limestone in two laboratory reactors and a pilot plant. Chem. Eng. Sci. 1991, 46 (3), 827. (7) Dam-Johansen, K.; Østergaard, K. High-temperature reaction between sulfur dioxide and limestonesII. An improved experimental basis for a mathematical model. Chem. Eng. Sci. 1991, 46 (3), 839. (8) Dam-Johansen, K.; Hansen, P. F. B.; Østergaard, K. High-temperature reaction between sulfur dioxide and limestonesIII. A grainmicrograin model and its verification. Chem. Eng. Sci. 1991, 46 (3), 847. (9) Dam-Johansen, K.; Østergaard, K. High-temperature reaction between sulfur dioxide and limestonesIV. Chem. Eng. Sci. 1991, 46 (3), 855. (10) Murthy, K. S.; Howes, J. E.; Nack, H. Emissions from pressurized fluidized-bed combustion processes. EnViron. Sci. Technol. 1979, 13 (2), 197.

(11) Ljungstro¨m, E.; Lindqvist, O. Measurement of in-bed gas and solid compositions in a combustor operating at pressures up to 20 bar. Int. Conf. Fluid. Bed Combust. (7th Conf., Philadelphia, PA) 1982, 465. (12) Tullin, C.; Nyman, G.; Ghardashkhani, S. Direct Sulfation of CaCO3: The influence of CO2 partial pressure. Energy Fuels 1993, 7, 512. (13) West, A. R. Basic Solid State Chemistry; John Wiley & Sons: Chichester, U.K., 1999.

ReceiVed for reView February 5, 2007 ReVised manuscript receiVed May 24, 2007 Accepted May 24, 2007 IE070208U

Enhancement of the Direct Sulfation of Limestone ... - ACS Publications

Recommend Documents