Energy & Fuels 2009, 23, 2537–2543
2537
Removal of Refractory Sulfur Compounds in Diesel Using Activated Carbon with Controlled Porosity Hong-Joo Jeon,†,‡ Chang Hyun Ko,† Sung Hyun Kim,*,‡ and Jong-Nam Kim*,† Korea Institute of Energy Research, 71-2 Jang-Dong, Yuseong-Gu, Daejeon 305-343, Korea, and Department of Chemical and Biological Engineering, Korea UniVersity, Anam-Dong, Sungbuk-Gu, Seoul 136-701, Korea ReceiVed December 2, 2008. ReVised Manuscript ReceiVed February 18, 2009
Carbon-based adsorbents with controlled porosity were prepared to investigate the adsorption capacity of refractory sulfur compounds. The porosity of carbon-based adsorbents was controlled by thermal treatment at 1173 K in CO2 environment. The equilibrium sulfur adsorption capacities of each adsorbent were measured for both model and commercial diesel. Pore widening of carbon-based adsorbents by CO2 activation increased BET surface area, total pore volume, micropore volume, and sulfur adsorption capacity. However, the sulfur adsorption capacity did not increase linearly according to the increase of BET surface area, total pore volume, and BJH pore volume, which included the properties of meso- and macropores. Specific micropore volume, whose pore diameters range from 0.63 to 1.2 nm, showed good linear relationship with sulfur adsorption capacity for commercial diesel. Model diesel adsorption tests supported that adsorbents with proper pore size for target sulfur molecules should be prepared for the enhancement of sulfur adsorption capacity.
1. Introduction When a sulfur compound in transportation fuels is converted to SOx during combustion of fuel in an internal-combustion engine, it not only makes a contribution to acid rain but also acts as a poison to the catalytic converter for exhaust emission treatment. Governments worldwide regulate the sulfur contents in transportation fuels to protect the environment. The regulations proposed that the upper limit of sulfur concentration in diesel be reduced from 500 ppmw to 15 ppmw by 2010 in the U.S, from 50 ppmw to 10 ppmw by 2010 in the EU, and from 30 ppmw to 10 ppmw by 2010 in Korea, respectively.1-5 Considering these regulations, development of desulfurization technology for transportation fuels seems to be one of the most urgent research subjects to be accomplished to meet those regulations. The conventional hydrodesulfurization (HDS) process is less effective for refractory sulfur compounds, such as 4,6-dimethyldibenzothiophene (4,6-DMDBT), because of steric hindrance toward active sites in HDS catalysts. The catalysts of conventional HDS are Ni-Mo/Al2O3 and Co-Mo/Al2O3, with various modifications by variation in additives, promoters, and improved preparation methods. The HDS process using those catalysts requires severe operating conditions and a large reactor volume to comply with the new regulations.2,3 Recently, although the unsupported transition metal sulfide catalysts with higher activity * To whom correspondence should be addressed. Tel: +82-42-860-3112 (J.-N.K.). E-mail: (J.-N.K.)
[email protected]; (S.H.K.)
[email protected]. † Korea Institute of Energy Research. ‡ Korea University. (1) Clean Air Conservation Act, Table 30, Preparation Standard for Transportation Fuel and Its Additives, Ministry of Environment, Korea, 2004. (2) Song, C. Catal. Today 2003, 86, 211–263. (3) Song, C.; Ma, X. Appl. Catal., B 2003, 41, 207–238. (4) Bhandari, V. M.; Ko, C. H.; Park, J. G.; Han, S. S.; Cho, S. H.; Kim, J. N. Chem. Eng. Sci. 2006, 61, 2599–2608. (5) Kim, J. H.; Ma, X.; Zhou, A.; Song, C. S. Catal. Today 2006, 111, 74–83.
or selectivity than that of the traditional γ-Al2O3-supported Ni/ Co-Mo/W catalysts were developed to apply in existing HDS units,6 it is an established fact that they consume high energy and require high partial H2 pressure because of the technology based on hydrogenation. As an alternative or complementary desulfurization technology to produce ultra-low-sulfur diesel (ULSD), extraction using ionic liquid,7 oxidative desulfurization,8,9 and adsorption4,5,10-21 have been proposed. Among them, adsorptive desulfurization has attracted much attention due to some advantages, such as low-energy consumption, because it can be performed at ambient temperature and pressure without hydrogen or oxygen consumption. Also, adsorbents can be reused by the proper regeneration process. (6) Eijsbouts, S.; Mayo, S. W.; Fujita, K. Appl. Catal., A 2007, 322, 58–66. (7) Nie, Y.; Li, C.; Sun, A.; Meng, H.; Wang, Z. Energy Fuels 2006, 20, 2083–2087. (8) Etemadi, O.; Yen, T. F. Energy Fuels 2007, 21, 2250–2257. (9) Etemadi, O.; Yen, T. F. Energy Fuels 2007, 21, 1622–1627. (10) Herma´ndez-Maldonado, A. J.; Yang, R. T. Catal. ReV. 2004, 46, 111–150. (11) Herma´ndez-Maldonado, A. J.; Yang, R. T. Ind. Eng. Chem. Res. 2003, 42, 123–129. (12) Herna´ndez-Maldonado, A. J.; Yang, R. T. Ind. Eng. Chem. Res. 2003, 42, 3103–3110. (13) Yang, R. T.; Herna´ndez-Maldonado, A. J.; Yang, F. H. Science. 2003, 301, 79–81. (14) Velu, S.; Ma, X.; Song, C. S. Ind. Eng. Chem. Res. 2003, 42, 5293– 5304. (15) Ko, C. H.; Park, J. G.; Park, J. C.; Song, H. J.; Han, S. S.; Kim, J. S. Appl. Surf. Sci. 2007, 253, 5864–5867. (16) Zhou, A.; Ma, X.; Song, C. S. J. Phys. Chem. B 2006, 110, 4699– 4707. (17) Salem, S. H. Ind. Eng. Chem. Res. 1994, 33, 336–340. (18) Salem, S. H.; Hamid, H. S. Chem. Eng. Technol. 1997, 20, 342– 347. (19) Ma, L.; Yang, R. T. Ind. Eng. Chem. Res. 2007, 46, 2760–2768. (20) Sano, Y.; Choi, K. H.; Korai, Y.; Mochida, I. Energy Fuels. 2004, 18, 644–651. (21) Sano, Y.; Choi, K. H.; Korai, Y.; Mochida, I. Appl. Catal., B 2004, 49, 219–225.
10.1021/ef801050k CCC: $40.75 2009 American Chemical Society Published on Web 03/19/2009
2538
Energy & Fuels, Vol. 23, 2009
There are ongoing studies to develop adsorbents for the desulfurization of transportation fuel based on zeolites,4,10-14 mesoporous materials,15 and activated carbon materials.16,20,21 Also, many comparative studies were investigated to seek effective adsorbents for desulfurization. Salem carried out the study of naphatha desulfurization using activated carbons, zeolites 5A and 13X. The study indicated that zeolite 13X has a high capacity for sulfur in low concentration ranges and should be used when the sulfur content in the organic fraction is less than 25 ppm. When the concentration of sulfur compound increases, the capacity of activated carbon can be 3 times greater than that on a zeolite.17,18 Recently, Yang and co-workers made a comparative study of ion-exchanged zeolite and metal halideimpregnated carbon for selective adsorption of desulfurization in liquids fuels. The results indicated that metal halideimpregnated carbon has a more selective adsorption property than that of ion-exchanged zeolite for refractory sulfur compounds, which are resistant to hydrodesulfurization.19 Song and co-workers made a comparative study of alumino-silica gelimpregnated Ni, activated alumina, and activated carbon for selective adsorption in the desulfurization of model fuels containing sulfur compounds, aromatic compounds, and nitrogen compounds. The study reported each adsorbent made a difference in selective adsorption properties, and activated carbon showed higher adsorptive capacity and selectivity for both sulfur and nitrogen compounds, especially for the sulfur compounds with methyl groups.5 Mochida and co-workers studied activated carbon for desulfurization and denitrogenation of straight-run gas oil (SRGO). These studies showed that adsorptive capacity for sulfur compounds increased in proportion to the oxygen content, surface area, and total pore volume.20,21 Previous studies showed that a series of activated carbons are the most suitable for desulfurization of fuels that have a high sulfur concentration. The object of this article is to investigate which physical property (surface area, pore volume, pore width, etc) plays an important role in the adsorption of organic sulfur species on activated carbon. Although surface area is an important factor in adsorption to the active site, many cases showed poor linear correlation between surface area of activated carbon and adsorption capacity as shown in the study of Song and co-workers using commercial activated carbon.16 These phenomena mainly come from the different physical factors and chemical functionality of activated carbon, depending on the preparation method selected by manufacturing companies. To determine the predominant physical property of activated carbon that plays an important role in sulfur adsorption, a series of carbon-based adsorbents were prepared, which have the same functionality but different surface area, pore volume, etc. For this purpose, a CO2 activation process, which treated carbon samples at 1173 K in a CO2 environment, was selected. Owing to the CO2 activation, the main surface functional group of activated carbons, used in this study, was expected to be the carboxylic group.22 On this experimental basis, various physical properties, such as BET surface area, total pore volume, and micropore volume, were correlated with sulfur adsorption capacity. Micropore volume, the diameter of which is similar to the critical dimension of sulfur compounds, plays a crucial role in sulfur compound adsorption. 2. Experimental Section 2.1. Preparation of Adsorbents. To exclude effects of chemical properties for sulfur adsorption, we made carbon adsorbents with ¨ zyuguran, A. Fuel 2008, 87, (22) Karatepe, N.; Orbak, I.; Yavuz, R.; O 3207–3215.
Jeon et al. Table 1. Model Diesel Compositiona MD-0 composition
MD-20 composition
MD-40 composition
name
purity, %
BT DBT 4,6-DMDBT etcc total
99 98 95 -
toluene naphthalene
99.5 99
Aromatic Compounds 0 wt% 10 wt% 0 wt% 10 wt%
25 wt% 15 wt%
n-tetradecane
98
Aliphatic Compound balance balance
balance
Sulfur Compounds 68 ppmb 139 ppm 139 ppm 32 ppm 378 ppm
68 ppm 138 ppm 138 ppm 32 ppm 376 ppm
68 ppm 139 ppm 139 ppm 32 ppm 378 ppm
a
T: thiophene, BT: benzothiophene, DBT: dibenzothiophene, 4,6-DMDBT: 4,6-dimethyldibenzothiophene. b ppm: ppm w/v (weight per volume, mg sulfur per 1 L of solvent). c Etc: Sulfur concentration contributed from solvents (toluene, naphthalene, n-tetradecane).
different physical properties but with similar chemical properties using the CO2 activation method as function of treatment time. Two commercially available carbon materials were selected to activate with CO2 gas. One is carbon molecular sieve (CMS) manufactured by Shirasaki and the other is activated carbon manufactured by Calgon. Coal is the raw material for these carbon materials. The starting carbon materials were further activated at 1173 K with a CO2 flow rate of 200 sccm in a custom-made vertical reactor. For clarity, the prepared carbon adsorbents were individually identified by source material name and activation time. For example, CMS5h means carbon adsorbent made from CMS supplied by Shirasaki and CO2 activation time was 5 h. 2.2. Characterization of CO2-Activated Carbon Samples. The morphologies of CO2-activated carbon adsorbents were examined by scanning electron microscopy (SEM, S-4700 Hitachi Co., Japan). The contents of carbon, hydrogen, and nitrogen were measured with TruSpec Elemental Analyzer (LECO Co.). The analysis was based on thermal conductivity detection for measuring carbon, hydrogen, and nitrogen, after combustion and reduction. The textural structure of activated carbons was characterized by adsorption/desorption of nitrogen at 77 K using a Micromeritics ASAP2010 surface area and porosimetry analyzer. The BET surface area (SBET) was calculated from the N2 adsorption isotherms at 77 K by applying the Brunauer, Emmett, and Teller (BET) equation.23 Total pore volume (Vtotal) was calculated from the N2 adsorbed amount at P/P0 ) 0.98. The total micropore volume (Vmicro, widths smaller than 2 nm) was calculated from the statistical thickness of adsorbed layers (t-method).24 Meso-macro pore surface area (SBJH) and volume (VBJH) were calculated according to the Barrett-Joyner-Halenda (BJH) model.25 Especially, the micropore size distribution of some samples (CMS-4h, CMS-5h, Calgon-1h, Calgon-2h, Calgon-3h, Calgon-5h, Calgon-6h) were calculated by adsorption/desorption of Ar at 87 K according to Horvath-Kawazoe (HK) analysis.26 The specific micropore volume was defined as follows. The micropore volume corresponding to pore diameter between 0.6 and 1.2 nm was designated as “specific pore volume (0.6-1.2 nm)”. The specific volumes of 0.62-1.2 nm and 0.62-2.89 nm were calculated by subtracting the HK cumulative pore volume (V0.62-1.2nm ) V1.2nm - V0.62nm, V0.62-2.89nm ) V2.89nm - V0.62nm, V0.44-2.89nm ) V2.89nm - V0.44nm). 2.3. Preparation of Model Diesel Fuels and Analysis of Commercial Diesel. The compositions of three model diesel fuels (MD0, MD20, MD40) used in the present work are listed in Table 1. A model diesel fuel was prepared by adding an aromatic compound and sulfur compounds to liquid alkanes (tetradecane). (23) Brunauer, S.; Emmett, P. H.; Teller, E. J. Am. Chem. Soc. 1938, 60, 309–319. (24) Harkins, W. D.; Jura, G. J. Am. Chem. Soc. 1944, 66, 919–927. (25) Barrett, E. P.; Joyner, L. G.; Halenda, P. H. J. Am. Chem. Soc. 1951, 73, 373–380. (26) Horvath, G.; Kawazoe, K. J. Chem. Eng. Jpn. 1983, 16, 470–475.
RemoVal of Refractory Sulfur Compounds in Diesel
Energy & Fuels, Vol. 23, 2009 2539
C(s) + CO2(g) f 2CO(g)
Figure 1. GC-PFPD chromatogram of commercial diesel.
For clarity, the prepared model diesel will be individually identified with corresponding “MD (model diesel)-weight fraction of aromatic compound”. To investigate the influence of increasing contents of aromatic compound on sulfur adsorption capacity of carbon material, the sulfur compounds of the same composition were added to MD-20 and MD-40. The composition was that benzothiophene (BT) approximately is 68 ppm w/v, dibenzothiophene (DBT) is 138 ppm w/v, and 4,6-dimethyldibenzothiophene (4,6-DMDBT) is 138 ppm w/v. Solvents such as toluene, naphthalene, and ntetradecane also have a trace amount of sulfur compounds. The sulfur concentration contributed from solvents was about 32 ppm w/v. Therefore, the sulfur concentration of model diesel was approximately 378 ppm w/v. A commercial diesel was purchased from the gas station near Daejeon. A GC-PFPD chromatogram of the diesel with identification of major peaks is shown in Figure 1. The total sulfur concentration in the diesel was 320 ppm w/v according to our analysis using an Antek 9000s total sulfur analyzer. 2.4. Adsorption Experiments. All carbon adsorbents were dried in 573 K with N2 for 6 h just before sulfur adsorption. Static adsorption experiments were conducted by allowing 10 mL of model or commercial diesel to come in contact with 1.0 g of adsorbents in a glass vial with stirring. Adsorption experiments were held at room temperature. After the desired adsorption time (6 h) was reached, the sample was filtered, and the sulfur concentration in the sample was analyzed to determine the adsorption capacity of the activated carbon adsorbents. 2.5. Analysis and Measurement of Sulfur Compounds. A pulsed flame photometric detector (PFPD) was used for identification and qualification of the treated commercial diesel samples and quantitative analysis of the treated model diesel. A Varian CP3800 gas chromatograph with a fused silica capillary column (SPB-1 sulfur, Supelco, 30 m × 0.32 mm × 4.0 µm film thickness) and a split mode injector (75:1) was used with ultrahigh purity helium as a carrier gas. The total sulfur concentration in the commercial diesel and desulfurized commercial diesel were determined using an Antek 9000LLS total sulfur analyzer. The instrument was calibrated to at least three difference sulfur concentration ranges using standard samples, and linear calibration curves were obtained for each calibration range. The sulfur detection limit of the total sulfur analyzer in the normal working range is 0.5 ppm w/v sulfur.
3. Results and Discussion 3.1. Characterization of the CO2 Activated Carbon Materials. Figure 2 showed SEM images of CMS and Calgon (activated carbon) samples before and after CO2 activation. The flat surface of untreated CMS turned into a rugged surface after CO2 activation. The partially uneven surface of Calgon became much rougher by CO2 activation. As expected from eq 1, some portion of carbon is removed by reaction with CO2.
(1)
The morphological change of carbon materials seems to be closely related to pore generation and pore widening. Table 2 presents the elemental composition of original carbon materials and activated samples. The content of carbon decreased and that of oxygen increased along with the increase of activation time. As the surface carbons were eroded and gasified through CO2 activation, existing original functional groups on the surface of adsorbents would be removed. At the same time, CO2 activation might generate surface carboxylic groups (COOH) and phenolic groups (COH).22 Thus, all samples that were prepared by CO2 activation had similar chemical functional groups on their surface. Such an increase of oxygen content is also useful for sulfur compound adsorption.16 The N2 adsorption-desorption isotherms for CO2-activated samples are shown in Figure 3. In Figure 3A, it is observed that the adsorption isotherms of CMS samples are close to type I of the IUPAC classification typical of microporous carbon material. As activating time increases, most of the nitrogen adsorption took place in micropores at low pressure range, and long activation time samples have a larger amount of nitrogen adsorbed than short activation time samples at low pressure. So, we can clearly observe a gradual micropore evolution in these carbon materials from the nitrogen adsorption-desorption isotherms. In the CO2 activation with CMS series samples, micropore generation predominates. Also, some samples activated for a long time (CMS-5h, -6h) showed increased N2 adsorption capacities at higher pressures, which are caused by the adsorption in macropores formed during the activation. In Figure 3B, the isotherms of Calgon samples showed steady N2 adsorption capacity in micropores at low pressure range. As activating time passes, it is observed that nitrogen is adsorbed on macropores at higher partial pressure (Calgon-4h, -5h, and -6h). This means that CO2 activation generated macropores on these samples. Also, there are hysteresis loops in the mesopore range corresponding to a relative pressure of 0.4-0.6. The hysteresis loop is a H4 type, and it is a characteristic of the slit-shaped pores where the adsorption and desorption branches are parallel.27 In the CO2 activation with Calgon series samples, micropore generation predominates and meso- and macropore generation also occurred. It is consistent with the analytical data in Table 3. Table 3 showed the BET surface areas and the total pore volumes obtained from the N2 adsorption isotherms. The surface areas increased gradually from 8.4 m2 g-1 for pristine CMS to 1976 m2 g-1 for CMS-5h and from 974 m2 g-1 for pristine Calgon to 1561 m2 g-1 for Calgon-5h. Correspondingly, the total pore volumes increased from 0.0 to 0.86 cm3 g-1 for the series of CMS samples and from 0.45 to 0.84 cm3 g-1 for the series of Calgon samples, respectively. CO2-activated CMS samples showed that micropore volume obtained by the t-plot method increased more than twice from 0.31 cm3 g-1 for CMS1h to 0.76 cm3 g-1 for CMS-5h. Also, Calgon samples showed that micropore volume increased from 0.38 cm3 g-1 for Calgon1h to 0.61 cm3 g-1 for Calgon-5h. These increases of pore volume are probably due to opening of closed pores, generation of new micropores, and widening of pre-existent pores. CO2 activation turned CMS and Calgon samples into high surface area adsorbents with similar surface properties. CO2activated CMS samples mainly had micropores, but CO2activated Calgon samples not only had micropores but also had (27) Marsh, H.; Francisco, R. R. ActiVated carbons; Elsevier Science & Technology Books: New York, 2006; pp 230-232.
2540
Energy & Fuels, Vol. 23, 2009
Jeon et al.
Figure 2. SEM images of (A) untreated CMS, (B) CMS after CO2 activation for 5 h, (C) untreated Calgon, and (D) Calgon after CO2 activation for 5 h. Table 2. Elemental Analysis of CO2-Activated Carbon-Based Adsorbents samples
C (wt%)
H (wt%)
N (wt%)
O (wt%)
CMS CMS-1h CMS-4h CMS-5h Calgon Calgon-2h Calgon-4h Calgon-5h Calgon-6h
96.05 94.53 93.87 93.30 94.20 93.15 87.45 87.05 89.20
0.56 0.67 0.63 0.47 0.54 0.60 0.93 0.84 0.68
0.46 0.47 0.36 0.43 0.36 0.63 0.52 0.42 0.40
2.93 4.33 5.14 5.8 4.9 5.62 11.1 11.69 9.72
meso- and marcopores. Such porosity control with similar surface functionality induced by CO2 activation will be useful to determine predominant physical properties of carbon-based adsorbents for sulfur adsorption. 3.2. Effect of Physical Properties of Activated Carbon on Sulfur Adsorption Capacity in Commercial Diesel. Figure 4 shows the relationship between sulfur adsorption capacity for commercial diesel and BET surface area. The sulfur adsorption capacities of Calgon and CMS samples were improved according to the activation time. Optimum activation time (5 h) maximized both BET surface area and sulfur adsorption capacity regardless of sample series. It means that large surface area is favorable for the improvement of sulfur adsorption capacity. Although each series of samples has its own linearity, there is a loose linear correlation between CMS and Calgon series samples. Especially, CMS series samples with shorter activation time showed a large deviation from linear correlation. Figure 5 displays the correlation between sulfur adsorption capacity for commercial diesel and total pore volume. As total pore volumes of adsorbents were increased by CO2 activation, sulfur adsorption capacity increased gradually. The linear relationship trend in Figure 5 is almost the same as that in Figure 4. To find out more related factors of carbon-based adsorbents
with sulfur adsorption capacity, not only total pore volume but also micropore volume of Calgon-6h and Calgon-5h were compared. The two samples had almost the same total pore volume: 0.83 cm3 g-1 for Calgon-5h and 0.84 cm3 g-1 for Calgon-6 h, but the sulfur adsorption capacity of Calgon-5h (2.08 mg-S/g-ads.) was larger than that of Calgon-6h (1.86 mgS/g-ads.). This means that total pore volume, which includes micro-, meso-, and marcropore volumes, is not a suitable parameter for the linear fitting of sulfur adsorption capacity. Figure 6 shows the relationship between the sulfur adsorption capacity and BJH cumulative pore volume. The BJH pore volume is a cumulative volume between meso- and macropores except micropores less than 1.7 nm. The sulfur adsorption capacity of samples increased gradually as the BJH volume increased until the BJH volume was smaller than 0.1 cm3 g-1. When the BJH pore volume was larger than 0.1 cm3 g-1, the increase of BJH pore volume does not have much contribution to sulfur adsorption capacity. In general, meso- and macropores are regarded as diffusion passages toward adsorption sites in micropores. For the proper access of sulfur compounds toward adsorption sites in carbon-based adsorbents, a minimum portion of diffusion passage should be secured by pore opening and widening. As Calgon series samples had well-developed mesoand marcropores in addition to micropores, CMS series samples, which originally had micropores only, must have meso- and macropores by CO2 activation because they should be compared under equal adsorption conditions. A BJH pore volume of 0.1 cm3 g-1 seems to be a criterion for the appropriate generation of diffusion passages. Thus, CMS samples with smaller BJH pore volumes due to short activation times (CMS-1h, CMS-2h, and CMS-3h) should be excluded in further correlation studies. Figure 7 displays the correlation between micropore volume obtained by t-plot method and sulfur adsorption capacity. CMS1h, -2h, and -3h were eliminated in correlation fitting, because
RemoVal of Refractory Sulfur Compounds in Diesel
Energy & Fuels, Vol. 23, 2009 2541
Figure 4. Relationship between sulfur adsorption capacities in commercial diesel and surface areas of CO2-activated carbon-based adsorbents.
Figure 3. N2 adsorption-desorption isotherms at 77 K of (A) CO2activated CMS (carbon molecular sieve) samples, (B) CO2-activated Calgon samples. Table 3. Physical Properties of CO2-Activated Carbon-Based Adsorbents samples CMS CMS-1h CMS-2h CMS-3h CMS-4h CMS-5h Calgon Calgon-1h Calgon-2h Calgon-3h Calgon-4h Calgon-5h Calgon-6h
Figure 5. Change of sulfur adsorption capacities in commercial diesel according to total pore volume of CO2-activated carbon-based adsorbents.
V total V micro V micro/ S BJH V BJH S BET (m2g -1) (cm3g-1) (cm3g-1) V total (m2g-1) (cm3g-1) 8.4 821 1107 1339 1610 1976 974 1170 1322 1457 1513 1561 1510
0.018 0.33 0.45 0.56 0.68 0.86 0.45 0.55 0.63 0.72 0.75 0.83 0.84
0.002 0.31 0.42 0.52 0.62 0.76 0.38 0.45 0.51 0.56 0.59 0.61 0.59
0.11 0.94 0.94 0.92 0.89 0.85 0.84 0.83 0.81 0.78 0.78 0.73 0.70
7.42 31 44 64 128 239 107 126 147 225 212 338 375
0.016 0.03 0.04 0.06 0.10 0.18 0.09 0.12 0.15 0.23 0.24 0.34 0.37
the degree of pore opening and pore widening for these samples is not enough for sulfur compounds to access adsorption sites in micropores, as mentioned in Figure 6. Pristine CMS and Calgon samples without CO2 activation were also excluded to eliminate the influence of different surface functional groups on sulfur adsorption capacity. The criterion of linear relationship for micropore volume with sulfur adsorption capacity (R2 ) 0.8824) is not so much improved compared with that for BET surface area (R2 ) 0.8904) or total pore volume (R2 ) 0.8254) with sulfur adsorption capacity. Although many points are
Figure 6. Change of sulfur adsorption capacities in commercial diesel according to meso- and macropore volume obtained by the BJH method.
considered, micropore volume obtained by the t-plot method is not enough for the linear correlation with sulfur adsorption capacity.
2542
Energy & Fuels, Vol. 23, 2009
Figure 7. Relationship between sulfur adsorption capacities in commercial diesel and micropore volume obtained by the t-plot method.
Figure 8. Micropore volume distribution of carbon-based adsorbents obtained by the H-K method from Ar adsorption-desorption isotherms.
According to results from Figures 4-7, BET surface area, total pore volume, micropore volume, and BJH pore volume, which were obtained from N2 adsorption, did not provide good linear correlation with sulfur adsorption capacity. The major sulfur compound contained in commercial diesel is 4,6-DMDBT as mentioned previously in Figure 1. The critical diameter of this molecule is known to be 0.63 nm. In order to investigate the major factor for sulfur compound adsorption on carbonbased adsorbents considering the diameter of micropores and critical diameter of the sulfur molecule, micropore size distributions using the H-K method from the Ar adsorption isotherm were additionally obtained for these samples. We notice that CMS-4h and Calgon-3h have similar sulfur adsorption capacities. The sulfur adsorption capacity of CMS-4h is 1.91 mg/gads and that of Calgon-3h is 1.88 mg/g-ads. To find out similar values in various physical properties, micropore size distributions, obtained by the H-K method, of those samples were compared, as shown in Figure 8. Although their pore size distribution was different, an overlapping micropore range could be found around 0.68-1.2 nm. Such an overlapping pore range might be related to sulfur adsorption capacity. In other words, sulfur adsorption seemed to be related to the diameter of the micropore. On the basis of the result in Figure 8, the specific micropore volumes were correlated with sulfur adsorption capacity. Linear relationships between sulfur adsorption capacity and specific
Jeon et al.
Figure 9. Correlations between specific micropore volumes and sulfur adsorption capacity in commercial diesel: specific micropore volumes of carbon-based adsorbents were obtained by the H-K method in Ar adsorption-desorption isotherms.
micropore volumes were examined to obtain the optimum result by changing the lower limit (0.44 nm) and higher limit (2.89 nm) of micropore size. Figure 9 shows that the best correlation coefficient (R2 ) 0.9883) between sulfur adsorption capacity and specific micropore volume was obtained when the micropore diameter ranges from 0.62 to 1.2 nm. This limited range of pore diameter seems to be closely related with physical dimensions of carbon adsorbents and sulfur compounds. As the major sulfur compound in commercial diesel is 4,6-DMDBT, our discussion for sulfur adsorption is focused on 4,6-DMDBT. Considering that the critical diameter of 4,6-DMDBT is 0.63 nm, the lower limit of 0.62 nm means that sulfur compounds could not be adsorbed in a certain range of micropores whose diameter is smaller than that of sulfur compounds. The physical meaning for the higher limit of 1.2 nm seems to be related with monolayer adsorption, because 1.2 nm is less than two layers of 4,6DMDBT. It means that multilayers adsorption of 4,6-DMDBT cannot occurr in each specific micropore at once. Therefore, much more sulfur adsorption capacity could be expected in the case of controlling adsorbent pore structure by decreasing invalid micropores with pore diameters under 0.62 nm and increasing valid specific micropore volume, which corresponds to sulfur molecule size. 3.3. Effect of Aromatic Compound Ratio on Sulfur Adsorption Capacity. Figure 10 shows the change of sulfur adsorption capacities of CMS samples depending on their BET surface areas and the concentration of aromatic compounds. As the concentration of aromatic compounds increased in model diesel, sulfur adsorption capacity decreased. Competitive adsorption between sulfur and aromatic compounds might reduce the sulfur adsorption capacity of CMS samples. At the same aromatic concentration, BET surface area did not affect the sulfur adsorption capacity in model diesel. However, sulfur adsorption in commercial diesel increased as the BET surface area increased. Such a different adsorption behavior mainly comes from the size difference of sulfur compounds in model diesel and commercial diesel. In commercial diesel, the main sulfur compound is 4,6-DMDBT and most of the other sulfur compounds are larger than 4,6-DMDBT, as shown in Figure 1. On the other hand, three kinds of sulfur compounds (BT, DBT, 4,6-DMDBT) with equal amounts were dissolved in model diesel. Considering molecular structures, critical diameters of sulfur compounds in model diesel are smaller than those in commercial diesel. As shown in Figure 11, CMS samples had
RemoVal of Refractory Sulfur Compounds in Diesel
Energy & Fuels, Vol. 23, 2009 2543
micropores with similar molecular size. These smaller micropores ranging from 0.4 to 0.6 nm are useless for the adsorption of larger sulfur molecules in commercial diesel. Thus, sulfur adsorption capacities in model diesel are higher than those in commercial diesel. Sulfur adsorption capacities in model diesel with an aromatic concentration of 40 wt% and those of commercial diesel converged at high BET surface area, because CO2 activation increased not only BET surface area but also the volume of larger micropores with a size between 0.6 and 1.2 nm. These experimental result strongly support that adsorbents with proper pore-sizes for target sulfur compounds should be prepared for the enhancement of sulfur adsorption capacity. 4. Conclusion
Figure 10. Sulfur adsorption capacities on carbon-based adsorbents at different aromatic concentrations in model diesel and commercial diesel.
Figure 11. Micropore volume distribution of selected carbon-based adsorbents obtained by the H-K method from Ar adsorption-desorption isotherms.
well-developed micropores ranging from 0.4 to 0.6 nm. Small size sulfur compounds in model diesel (0.4-0.63 nm) are easily adsorbed on these samples because CMS samples had a lot of
Two types of commercial carbon material were activated with CO2 at 1173 K. One is carbon molecular sieve with micropores around 0.3 nm and the other is general activated carbon material with both micro- and mesopores. Along with the increase of activation time, the physical properties of samples (surface area, pore volume) were increased. The activated carbon samples were prepared to investigate the sulfur adsorption capacities in commercial and model diesel. Although sulfur adsorption capacities of CO2-activated adsorbents for commercial diesel were increased according to the increase of BET surface area, total pore volume, and BJH pore volume, their linear relationship is very poor. However, specific micropore volume, the pore diameters of which ranged from 0.62 to 1.2 nm obtained by the H-K method, showed a very good linear relationship with sulfur adsorption capacity. The lowest limit of the specific micropore volume reflected the critical diameter of the main sulfur compound (4,6-DMDBT) in commercial diesel. Model diesel with smaller sulfur compounds compared to those in commercial diesel were prepared to investigate more evidence for the role of micropore in sulfur adsorption. The model diesel adsorption test also supported that adsorbents with the proper pore size for target sulfur molecules should be prepared for the enhancement of sulfur adsorption capacity. Further work should focus on the preparation of carbon material with more specific micropore volumes and the modification of functional groups of the carbon materials to reduce competitive adsorption with aromatic compounds. EF801050K