Environ. Sci. Technol. 2007, 41, 4375-4381
A New Generation of Sludge-Based Adsorbents for H2S Abatement at Room Temperature
(5). However, it was only recently that results obtained at the City University of New York (CUNY) have shown the real potential of the sludge-derived materials for H2S abatement (1). Since then, several papers have been published reporting different attempts to enhance the H2S retention ability of sewage-sludge-based adsorbents/catalysts (6-9).
ANNA ROS,† M. ANGELES LILLO-RO Ä DENAS,‡ CARLA CANALS-BATLLE,† ENRIQUE FUENTE,§ M I G U E L A . M O N T E S - M O R AÄ N , § M A R ´ı A J . M A R T I N , * , † A N D ANGEL LINARES-SOLANO‡ Laboratori d’Enginyeria Quı´mica i Ambiental (LEQUIA), Institut de Medi Ambient, Facultat de Cie`ncies, Universitat de Girona, E-17071 Girona, Spain, Grupo de Materiales Carbonosos y Medioambiente (MCMA), Departamento de Quı´mica Inorga´nica, Universidad de Alicante, Apartado 99, E-03080 Alicante, Spain, and Instituto Nacional del Carbo´n, CSIC, Apartado 73, E-33080 Oviedo, Spain
Studies focusing on the mechanism of hydrogen sulfide removal on these particular materials have been also reported (10-12). In spite of their heterogeneous nature, the key role played by the surface properties of the adsorbents/catalysts tested so far seems clear. Specific surface area is thus an essential parameter, and several activation procedures have been carried out to increase the adsorptive capacity of the sludge-based materials (3, 6, 7, 13). An overall conclusion that can be drawn from those previous studies is that sewage sludge should be considered a modest precursor for obtaining adsorbents with relatively high surface areas. The porosity values obtained were rather poor, and apparent BET surface areas higher than 600 m2 g-1 have never been reported by direct activation of the precursor. However, we have quite recently succeeded in obtaining materials with high surface areas from sewage sludge by alkaline hydroxides activation (14). Based on these results, the aim of the present work is to test the ability of various adsorbents/catalysts prepared by chemical activation with alkaline hydroxides in H2S removal experiments at room temperature. The selected materials comprise a wide range of porosities with the purpose of understanding the role of such porosity. Finally, the performance of the sludge-derived materials is compared to that of commercial sorbents (including some activated carbons).
The present paper discusses H2S removal by a new generation of sewage-sludge-derived materials which are characterized by their outstanding textural properties when compared to previous materials obtained by pyrolysis and/or activation of similar precursors. Alkaline hydroxide activation was used to prepare adsorbents/catalysts covering a wide range of porosities (SBET values from 10 to 1300 m2 g-1). Our results outline that textural properties are important for H2S abatement. However, not only highly porous sorbents, but also a high metallic content and a basic pH of these materials are required to achieve good performances. Proper combinations of textural properties and alkalinity render superior performances with retention values (x/M) as high as 456 mg of H2S removed per g of material. These retention capacities outperform previously published data for sewage-sludge derived materials and those achieved with commercial materials (including some activated carbons). Sulfur titration shows that most H2S is removed in the form of elemental sulfur, especially in the sewage/NaOH materials.
Introduction Sewage sludge has been postulated as a precursor of materials that can be successfully employed in several environmental applications (1-5). In all these studies, raw sludges were submitted to different treatments in order to improve their adsorptive capacities for removing contaminants in liquid or gas effluents. In addition, traces of metals still present in the sludge-based adsorbents might eventually act as catalysts in certain processes. Desulfurization of gases at room temperature constitutes one of the examples where materials prepared from sewage sludge behave as both adsorbents and catalysts. This particular application was explored more than 10 years ago * Corresponding author phone: + 34 972 418 161; fax: + 34 972 418 150; e-mail:
[email protected]. † Universitat de Girona. ‡ Universidad de Alicante. § Instituto Nacional del Carbo ´ n, CSIC. 10.1021/es062358m CCC: $37.00 Published on Web 05/16/2007
2007 American Chemical Society
Experimental Section Materials. The sewage sludge samples used in this study were obtained from two Spanish wastewater treatment plants (WWTPs) located in Girona, an area in the northeast of Spain: Banyoles-Terri (B) and Lloret de Mar (L). The influent of these selected facilities is mainly of domestic origin. Details of their particular sludge lines are reported elsewhere (10). About 10 kg of sludge was collected over a month. Special care was taken to ensure that the sampling period excluded atypical operating/weather conditions. The collected samples were stored at 4 °C and dried at 105 °C for 48 h. Following the drying of the sludge samples, the solids (referred to as raw or dried sludges) were ground and sieved to obtain a particle size range of 0.4-2 mm. Dried sluges will be referred to as SB and SL. Pyrolysis of the dried sludge samples (SB or SL) was carried out in a tubular furnace (Hobersal ST15 PADP). A sample of the dried material (SB or SL), weighing about 100-150 g, was put into a crucible, heated to 700 °C in 5 L min-1 of flowing nitrogen, using a heating rate of 15 °C min-1. The maximum temperature was held for 30 min. Then, the samples were allowed to cool down to room temperature in an inert atmosphere. Pyrolyzed samples are referred to as *-P. Acid-washed pyrolyzed samples (*-P-AW) were obtained by thoroughly washing the thermal treated samples with 5 M HCl, and then with distilled water until absence of chloride ions in the washing water (pH ) 6) was achieved. Activation by NaOH was carried out following the impregnation (I) and the ground physical mixing (G) methods, as described elsewhere (14). After appropriate mixing of the precursor and activating agent in both methods, the dried mixture was heated to 700 °C, at a 5 °C min-1 heating rate in a horizontal furnace with a cylindrical quartz tube (65 mm i.d.). Holding time at the maximum temperature was 1 h. VOL. 41, NO. 12, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
4375
TABLE 1. Elemental Composition and Ash Contents of the Sludges and Sludge-Based Adsorbents/catalysts selected elements (mg g-1)
elemental analysis (wt %) sample
yielda (wt
%)
ash (wt %)
C
H
N
S
O
Si
Fe
Ca
Al
SB SB-P SB-P-AW SB-P-AI SB-P-AI-AW SB-P-AGN-AW SB-P-AGK-AW
44.6 53.0 81.3 57.7 39.6 34.5
31.2 66.1 45.5 68.0 44.9 46.2 48.4
31.4 24.2 37.2 24.6 38.1 35.7 32.5
4.8 1.0 1.6 1.0 1.6 1.7 1.9
4.4 2.7 4.1 2.4 4.0 2.7 1.7
1.1 0.4 0.7 0.3 0.6 0.5 0.3
21.5 14.9 11.8 12.8 10.1 13.8 30.4
56.3 94.6 149 73.1 128 151 109
10.5 25.4 6.9 26.4 8.1 8.5 6.9
38.3 89.2 3.0 91.0 10.9 2.0 0.8
11.5 27.5 14.0 24.9 29.1 5.2 4.6
SL SL-P SL-P-AW SL-P-AI SL-P-AI-AW SL-P-AGN-AW SL-P-AGK-AW
61.5 22.7 80.8 17.2 11.5 5.5
53.4 64.9 35.8 72.1 40.4 48.2 47.6
27.9 17.7 45.6 12.4 38.2 30.1 30.8
4.2 0.5 1.9 0.8 2.2 1.9 1.6
1.5 0.5 2.3 0.4 1.3 1.3 1.1
0.2 0.2 1.4 0.3 0.7 0.6 0.7
24.0 20.7 12.6 20.8 19.2 8.4 16.8
28.3 44.5 173 102 245 173 119
40.2 55.0 54.0 55.9 2.8 0.8 11.4
206 394 244 220.1 7.3 3.4 2.7
6.5 11.6 11.5 2.7 18.5 3.2 13.8
a
Yields correspond to each successive step, according to samples nomenclature.
The heat-treatment stage was carried out in nitrogen, at a flow rate of 500 mL min-1. Once cooled to room temperature, all samples were first thoroughly washed with 5 M HCl, and then with distilled water until absence of chloride ions in the washing water (pH ) 6) was achieved. The so-called “activated samples” will be referred to as *-P-AI-AW or *-P-AGN-AW depending on the activation method employed (impregnation or physical mixing, respectively). Additionally, activated samples labeled *-P-AGK-AW were obtained, under similar conditions, using ground KOH instead of NaOH. Results obtained are compared with those of commercial products: (i) two commercial sorbents (Sorbalit) from Ma¨rker Umwelttechnik GmbH, Germany, and (ii) the WV-A1100 activated carbon from MeadWestvaco, as well as an activated carbon (AA) obtained from the NaOH activation of an anthracite, using the method described elsewhere (15). Methods. Characterization of the Sewage-Sludge-Based Materials. Chemical composition analyses, and pH and XRD measurements of these materials were performed. Details of the characterization methods can be found in a previous publication (10). The porosity characterization of the different precursors and adsorbents was carried out using physical adsorption of N2 and CO2 at 77 and 273 K, respectively. Experimental details and calculation methods are detailed elsewhere (14, 16, 17). H2S Adsorption/Oxidation Experiments. The H2S removal tests were performed in a quartz fixed-bed reactor (9 mm i.d.) (Figure SI1 in the Supporting Information). The adsorption bed was composed of 250 mg of adsorbent and 1 g of sand to provide sufficient bed length (15 mm) to ensure uniform flow throughout and to avoid axial dispersion. Two types of materials were used in the fixed-bed reactor as adsorbents in H2S removal experiments: (i) sewage-sludgederived adsorbents/catalysts, and (ii) a 1/1 physical mixture of sludge-based adsorbent/catalyst plus ground NaOH. All samples were pre-humidified prior to testing. An influent concentration of 1000 ppmv of H2S in moist air (flow rate 150 mL min-1, relative humidity 30%) was used. To obtain the breakthrough curves, inlet and outlet H2S concentrations were monitored using a photoionization detector (Phockeck 5000, Ion Science Ltd.) until the outlet concentration equaled the inlet concentration. Adsorption/removal capacities, x/M (mg H2S/g of material), were then calculated by integrating the corresponding breakthrough curves and by applying eq 1
Q × MW x (c × ts ) M w × VM 0 4376
9
∫
ts
0
c(t)dt)
(1)
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 12, 2007
where Q is the total inlet flow rate (m3 s-1), w is the weight of sludge-based material introduced into the column (g), MW is the molecular weight of H2S, VM is the molar volume, co is the inlet gas H2S concentration (ppmv), c(t) is the gas outlet concentration (ppmv), and ts is the bed saturation/ exhaustion time (s). To achieve statistical soundness, at least three experiments were carried out with each of the samples under consideration. Average x/M values were therefore calculated, with typical errors being less than 10% of the reported averages. In the experiments carried out using the sewage-sludge-based adsorbent/NaOH physical mixture, retention capacities results (i.e., x/M values, eq 1) are expressed as mg of H2S per g of the adsorbent/NaOH admixture. Furthermore, elemental sulfur, sulfate, and sulfide contents of the raw and exhausted materials were determined as detailed elsewhere (10).
Results and Discussion H2S Retention in Sewage-Sludge-Based Adsorbents/Catalysts. The two selected sludges differ strongly in terms of their chemical and physical properties (Table 1, Figure SI2) (10, 14). Pyrolysis significantly increases the basic character of the materials in concomitance with the enrichment in metals such as Ca and Fe (Tables 1 and 2), whereas, as expected, the porosity development at this stage is still limited (Table 2). The main mineralogical changes observed after the pyrolysis of the raw sludges are phase transformations due to the loss of crystallization water of hydrated compounds present in the original solids and, eventually, mineral decomposition (Figure SI2) (10, 14). The acid washing of the pyrolyzed samples is effective in dissolving a significant amount of the inorganic fraction present in the chars. Thus, ash contents are reduced in the *-P-AW solids, with this effect being accompanied by the subsequent enrichment of carbonaceous matter in the materials (Table 1). A direct consequence of this carbon enhancement is the wide bump centered at approximately 22° 2θ that can be observed in the XRD patterns of both acid washed samples (Figure SI2). On the other hand, it is worth noticing that, in spite of the initial differences in the mineralogy of SB-P and SL-P, the washing treatment with HCl renders solids (*-P-AW) with very similar XRD profiles (Figure SI2). One of the most relevant results, from the point of view of the material’s final application proposed in the present paper, is the outstanding increase of the apparent surface area and other textural parameters of the *-P-AW solids when
TABLE 2. Textural Properties and pH Values of the Adsorbents/Catalysts under Studya x/M (mg H2S g-1)b ads/cat ads/cat + NaOH
sample
BET (m2 g-1)
VDR-N2 (cm3 g-1)
VDR-CO2 (cm3 g-1)
Vt (cm3 g-1)
pH
SB-P SB-P-AW SB-P-AI SB-P-AI-AW SB-P-AGN-AW SB-P-AGK-AW
13 188 118 294 725 1058
0.01 0.09 0.05 0.12 0.30 0.46
0.03 0.07 0.04 0.10 0.20 0.34
0.04 0.22 0.14 0.32 0.71 0.84
8.2 3.4 8.7 3.6 3.4 3.2
73 48 280 93 95 119
113 149 242 266 333 298
SL-P SL-P-AW SL-P-AI SL-P-AI-AW SL-P-AGN-AW SL-P-AGK-AW
49 428 52 505 943 1301
0.02 0.16 0.02 0.21 0.38 0.60
0.01 0.11 0.01 0.13 0.27 0.34
0.08 0.55 0.16 0.59 0.91 0.99
10.2 2.7 10.7 4.1 3.8 3.2
183 153 288 179 -
201 386 262 456 350 446
AAc WV-A1100d
1494 1767
0.64 0.67
0.48 0.36
0.74 1.19
2.7 6.6
16 45
315 330
a The last two columns correspond to the H S retention capacities calculated in experiments carried out with pure materials (ads/cat) and 1/1 2 material/NaOH physical mixtures. b -: Not determined c Activated anthracite (15) (see Experimental Section). d Activated carbon from Westvaco (see Experimental Section).
compared to the pyrolyzed sludges (Table 2). In contrast to previous findings (1), such a high enhancement of the apparent BET surface areas is well above the expected values if only the yield of the washing process is taken into account. It seems that the inorganic matter acts as a template in both pyrolyzed samples SB-P and SL-P, the removal of which leads to a relatively highly porous carbonaceous fraction in SBP-AW and SL-P-AW (14). The H2S retention capacities of these materials (i.e., pyrolyzed and pyrolyzed plus acid washed sludges) were calculated from the breakthrough curves (Figure SI3). Results are compiled in Figure 1 (together with the sulfur species titration) and in Table 2. The x/M values of the *-P and *-PAW samples are fairly high, particularly if the environmental issues linked to sludge management are taken into consideration (1, 18). As reported previously (10), differences in performance between SB- and SL-based adsorbents/catalysts depend on the particular characteristics of the raw sludges. Thus, the SL family of adsorbents renders higher retention capacities than the corresponding SB sludge-based counterparts, most likely due to the relatively high Fe and Ca concentrations of the former (Table 1) (10), as well as to the higher porosity development observed for the SL-based adsorbents/catalysts (Table 2). Nevertheless, the retention capacities reported for this first set of samples compares reasonably well with those observed, under similar test conditions, for two commercial adsorbents based on activated carbon and lime (Sorbalit2, x/M ) 257 mg g-1; Sorbalit4, x/M ) 130 mg g-1) which are specifically designed for gaseous emissions control. The H2S removal ability of the pyrolyzed samples is not significantly improved after activation (either by impregnation or by the physical mixing method) (Table 2). This was, at first glance, surprising, considering the substantial enhancement of the textural properties of the activated materials. This lack of performance could be linked to the washing off of the inorganic matter, a similar effect just abovementioned when discussing the ash content of the *-P-AW samples (Table 1). The amount of elements that could contribute to eliminate H2S (i.e., Ca, Fe) drops considerably in all activated samples under study. The typical profile recorded for any *-P-AG*-AW is shown in Figure 2, which is dominated by the amorphous carbon wide band at approx 22° 2θ. The relatively high ash contents of the *-P-AG*-AW samples (Table 1) correspond mainly to amorphous silica. The X-ray diffractograms of the *-P-AI-AW (Figure 2) are very comparable to the *-P-AW counterparts (Figure SI2), i.e.,
mainly peaks corresponding to quartz, micas, feldspars, and eventually cristobalite could be identified. However, differences in composition should not be the main factor undermining the H2S retention ability of the activated samples. Indeed, a closer inspection of the results obtained for the *-P-AW (pyrolyzed plus acid washed) samples suggests that the mineral content is not always determinant to predict the performance of the materials under study. The x/M values of the *-P-AW samples are lower than those of the *-P, despite the significant increase of the apparent surface areas of the acid washed samples (Table 2). The elemental analysis of SB-P-AW and SL-P-AW reveals important differences regarding the behavior of Fe and Ca compounds during the acid treatment (Table 1). As a result, most of the Ca and Fe are removed in SB-P-AW whereas SL-P-AW only shows a minor reduction of the Ca content, with the Fe content remaining unaltered. In other words, Fe and most of the Ca are present in SL-P-AW forming part of compounds which are stable during the acid washing of SL-P. Hence, better retention ability should be expected for SL-P-AW in comparison to SL-P, which is not the case (Table 2). The reason behind the worse performance of the *-P-AW adsorbents/catalysts when compared to *-P is almost certainly the drop of the material’s pH values after the acid washing, an effect that should also affect the functioning of the activated samples. Surface pH has been reported to be a critical parameter in experiments using activated carbons for H2S abatement at room temperature (19, 20). In these studies, it has been shown that below a given threshold value (pH ) 5) the H2S dissociation in aqueous solution is impaired, thus restricting the ability of the solid to remove hydrogen sulfide from the gas phase. In spite of their surface acidity, the retention ability of the *-P-AW materials is still reasonably high due to the remarkable enhancement of their textural properties (Table 2). Considering the influence of the solid’s pH, H2S removal tests were performed under similar conditions (see Experimental Section) using two intermediate materials obtained during the preparation of the activated adsorbents/catalysts by impregnation (*-P-AI-AW). These two samples, labeled *-P-AI, are the solids resulting after the thermal treatment of the impregnated pyrolyzed sludges, i.e., prior to the acid washing treatment carried out to remove the remaining activating agent byproducts. Accordingly, the porosity of these materials is much lower than that of the activated samples, VOL. 41, NO. 12, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
4377
FIGURE 1. Sulfur balances and sulfur speciation (expressed as mg of S per gram of adsorbent) after exhaustion of *-P, *-P-AI, *-P-AW, and *-P-AI-AW adsorbents/catalysts. Top: pure solids. Bottom: 1/1 solid/NaOH physical admixtures. whereas their pH values are well above 7 units (Table 2). As shown in Table 1, the elemental composition of both *-P and *-P-AI sets of samples is very similar. Figure 2 shows the XRD diffraction patterns of the pyrolyzed sludges, *-P-AI solids, and the activated materials. Minimal changes are observed between the mineralogical species present in *-P and *-P-AI. The x/M values of the *-P-AI samples clearly outperform the results discussed so far (Table 2 and Figure 1). The sulfur species titration carried out in both *-P and *-P-AI sets of exhausted samples reveals a very high selectivity of the H2S oxidation to elemental sulfur. The autocatalytic effect of elemental sulfur during H2S removal experiments is responsible for the local maxima observed in the breakthrough curves recorded for these samples, especially *-P-AI (Figure SI3) (21). On the other hand, selectivity is not distinctive of the removal process when the acidic sets of samples (*-P-AW and *-P-AI-AW) are tested, except in the case of the SL-P-AW adsorbent/catalyst. For these samples, approximately similar quantities of sulfates and elemental sulfur are detected. The shift toward sulfur species with higher oxidation states (S4+ or even S6+) has been reported to be mechanistically favored under surface acidic conditions in H2S removal by activated carbons experiments (22). The high selectivity observed for the SL-P-AW adsorbent/catalyst notwithstanding its low pH value is most likely due to the high contents of Ca and Fe still present after the acid washing of the pyrolyzed SL sludge (Table 1 and discussion above). Thus, these elements promote charge mediation mechanisms 4378
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 12, 2007
(i.e., redox mechanisms involving Mx+ and S2- species) rather than the oxygen activation reaction path required for the formation of oxygenated sulfur species (23, 10). H2S Retention in Sewage-Sludge-Based Adsorbents/ NaOH Admixtures. This section is devoted to discussing the results obtained after attempting an approach to take advantage of the potential of the high-surface-area materials prepared from SL and SB. To overcome the acidic character of most of the sludge-based adsorbents/catalysts under study, they were physically mixed with ground NaOH, and H2S removal tests were carried out in a similar way. It is recalled here that a similar approach is used commercially for manufacturing Sorbalit. First, removal tests were performed using different proportions of NaOH in the solid mixture in order to establish the influence of the adsorbent/catalyst bed dilution on the H2S conversion (24, 25). Figure 3 shows the results of such a set of experiments for the SL-P-AW adsorbent/catalysts. The x/M values exhibit a maximum value at 10% NaOH/material concentration. From then on, the retention ability of the admixtures seems to be independent of the amount of NaOH added. This behavior matches the conclusions of previous studies carried out with activated carbons (26). The x/M plateau value of Figure 3 suggests that the NaOH is essential at the first stages of the H2S retention, easing the dissociation of the H2S required to promote its eventual oxidation in the aqueous surface. Thus, it was decided to prepare 1/1 adsorbent/NaOH admixtures (see Experimental Section) to ensure a good mixing.
FIGURE 4. H2S retention capacities vs total pore volume plot for the adsorbents/catalysts under study. Full symbols: experiments carried out with pure solids; void symbols: experiments carried out with 1/1 material/NaOH physical admixtures.
FIGURE 2. XRD diffraction patterns of the *-P, *-P-AI, *-P-AW, and *-P-AI-AW adsorbents/catalysts: (top) SB series; (bottom) SL series. Band labeling: B, brushite; Ca, calcite; Cn, iron ferrocyanide; Cr, cristobalite; D, dicalcium ferrite; F, Na-Ca feldspars (albite, anorthite); I, illite; P, portlandite; Q, quartz; W, whitlockite.
FIGURE 3. Variation of the H2S retention capacities of SL-P-AW/ NaOH physical admixtures using different mixing ratios. Retention capacities of these admixtures are collected also in Table 2 and shown in Figure 1. Two different behaviors can be observed. On one hand, the material/NaOH mixture does not significantly improve the performance of adsorbents/catalysts exhibiting a basic character (i.e., *-P and *-PAI). However, a stunning rise of the x/M values occurs when the acidic materials (*-AW) are mixed with NaOH. Particularly, two SL-based adsorbents/catalysts reach almost 50 wt % sulfur stored in the exhausted materials. The sulfur species titration accomplished on these exhausted admixtures
indicates a clear shift toward the H2S oxidation to elemental sulfur, i.e., the specificity of the removal process always increases after addition of NaOH to the adsorbents/catalysts. Regarding the sulfur speciation of the exhausted admixtures, residual amounts of sulfides are detected only in one of the materials/NaOH combinations (SB-P + NaOH, Figure 1), in spite of the NaOH ability to neutralize significant quantities of H2S. This result agrees well with the behavior observed in Figure 3. The oxidation of the dissociated sulfydric acid seems to be preferred to the plain neutralization (H2S(aq) + 2NaOH f Na2S + 2H2O), due to the presence of the sludgebased adsorbent/catalysts in the vicinity of NaOH particles. To further corroborate this, we performed a detailed SEM/ EDX analysis of different exhausted samples. Figure SI4 shows the micrographs of pure ground NaOH and a SL-P-AW + NaOH admixture, both after complete exhaustion of the fixed bed. Despite a very different morphology, EDX revealed decisive differences between the two samples. In the case of exhausted NaOH, the EDX spectrum corresponds to Na2S (Figure SI5a), while in the exhausted admixture a prevalence of the signal coming from sulfur is observed to the detriment of that corresponding to Na (Figure SI5b). Furthermore, several spots of unreacted NaOH could be found in this latter micrograph, although a relatively strong S signal was normally present in the spectra (Figure SI5c). In spite of the spatial resolution of the EDX technique, that S signal could be interpreted as elemental sulfur passivating the surface of NaOH particles. Role of the Textural Properties in H2S Removal. In view of the results just discussed, it seems clear that the differences observed in the x/M values of the adsorbents/catalysts/NaOH admixtures should depend on the particular characteristics of the sludge-based materials. The best performing materials appear to be those most texturally developed. Moreover, a previous section has emphasized the role of the composition and pH. This implies that a combination of porosity, chemical composition, and pH of the sorbents governs the H2S removal. Some authors have claimed that mesopores strongly contribute to the storage of elemental S (11, 21). In the present section, the influence of different textural parameters was investigated by comparing similar samples. Therefore, the contribution of other factors rather than textural properties to the differences in H2S removal behavior is removed. When such a comparison is performed (Figure 4), the most important conclusion is that acceptable correlations between the retention capacities of the materials and their total pore volumes, Vt (Table 2), can be drawn if the samples are organized in groups depending on their inorganic content VOL. 41, NO. 12, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
4379
and/or pH. This correlation, however, is not observed if all the analyzed materials are considered without taking into account their differences in metallic content and pH. When H2S is totally or mostly removed in the form of elemental sulfur, pores are desired to be larger than narrow micropores (0.7 nm) (21). For other samples in which some H2S is also removed in the form of SO42- (samples *-AW, Figure 3), this point is not so crucial and all pore sizes, even narrow micropores, should contribute to H2S removal. In principle, the experimental points of Figure 4 come out grouped in three different areas, according to the Vt values of the adsorbents/catalysts. The first set of samples corresponds to the *-P and *-P-AI materials, i.e., solids with relatively low Vt values (
