- Email: [email protected]
Adsorptive removal of sulfur and nitrogen species from a straight run gas oil over activated carbons for its deep hydrodesulfurization
Applied Catalysis B: Environmental 49 (2004) 219–225
Adsorptive removal of sulfur and nitrogen species from a straight run gas oil over activated carbons for its deep hydrodesulfurization Yosuke Sano, Ki-Hyouk Choi, Yozo Korai, Isao Mochida∗ Institute for Advanced Materials Chemistry and Engineering, Kyushu University, Kasugakouen 6-1, Fukuoka 816-8580, Japan Received 18 September 2003; received in revised form 11 December 2003; accepted 12 December 2003
Abstract As a pre-treatment step for ultra deep hydrodesulfurization (UD-HDS), sulfur and nitrogen species were adsorptively removed from the straight run gas oil (SRGO) over activated carbon materials. An activated carbon having the largest surface area and highest surface polarity showed the highest total capacity for adsorption of sulfur and nitrogen species as 0.098 g sulfur and 0.039 g nitrogen per 1 g of the activated carbon, respectively, among the carbon materials examined. Removal of both nitrogen and sulfur species was found very effective to achieve UD-HDS under conventional HDS conditions. SRGO treated over activated carbon contained only 11 ppm of sulfur, while non-treated SRGO contained still 193 ppm of sulfur in its HDS product. © 2004 Elsevier B.V. All rights reserved. Keywords: Hydrodesulfurization; Inhibition; Activated carbon; Adsorption
1. Introduction World refining industry is facing new and stricter regulations on the heteroatom contents in their major products such as diesel and gasoline. The sulfur content of diesel fuel must be lowered to less than 15 ppm S from current 500 ppm S and even further regulations may be implemented with accelerated concerns on the atmospheric pollution [1]. Several new approaches have been proposed and some of them were practically examined to achieve the ultra deep hydrodesulfurization (UD-HDS) [2]. However, more efficient and more economical process is still required in spite of various efforts. Critical barriers for achieving ultra deep hydrodesulfurization are very low reactivity of refractory sulfur species under conventional conditions and strong inhibition by H2 S, NH3 , nitrogen, and even aromatic species against UD-HDS. Such inhibitors retard the hydrodesulfurization, very markedly in the region of 0–100 ppm S [3–7]. There are some proposals of the pre-treatment for gas oil to remove nitrogen species by silica or silica–alumina prior to ∗
Corresponding author. Tel.: +81-92-5837797; fax: +81-92-837798. E-mail address: [email protected] (I. Mochida).
0926-3373/$ – see front matter © 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.apcatb.2003.12.007
hydrodesulfurization [8–11]. However, there are still some windows to improve the cost/performance in terms of the adsorbent. Removal level of nitrogen species, efficiency and effects of sulfur removal, regeneration of adsorbent and oil recovery must be delicately balanced. Song and Ma [12] proposed selective adsorption for removing sulfur at ambient temperature (SARS) process to achieve ultra clean diesel and gasoline. They used 5 wt.% metal loaded on silica gel as an adsorbent. Feed oil was not real diesel, but model species which did not contain the nitrogen species. Although detail information on the adsorbent was not disclosed, their adsorbent showed rather fast breakthrough over 10 ppm after eluting about 5 ml of model diesel. IRVAD and S-Zorb process also utilize adsorption to remove the heteroatoms in gasoline and diesel. But, operating temperature of them are reported to be over 240 ◦ C (IRVAD) and 340–410 ◦ C (S-Zorb) [13–15]. Yang and coworkers [16,17] reported that Cu and Ag-exchanged Y-type zeolites were very effective to selectively adsorb the thiophene and the used adsorbents could be regenerated by heating them at 350 ◦ C under flowing air. Activated carbon and activated carbon fiber have been recognized to be versatile absorbents for gaseous and liquid
220
Y. Sano et al. / Applied Catalysis B: Environmental 49 (2004) 219–225
substrates [18,19]. Their surface structure and properties can be controllable to propose better adsorbent [20–22]. In this study, sulfur and nitrogen species were removed from straight run gas oil (SRGO) through the dynamic adsorption over some activated carbon materials at room temperature. A particular activated carbon of very large surface area was examined in detail. The treated oil was hydrodesulfurized under conventional conditions to evaluate the effects of pre-treatment over activated carbons on the HDS activity to establish the facile UD-HDS process. It must be emphasized in the present study that carbon adsorbent removes sulfur as well as nitrogen species in the gas oil. The reduction of both species must be favorable for the deep hydrodesulfurization. Furthermore, the regeneration of the adsorbent was successfully performed with a conventional solvent, toluene, at low temperature.
2. Experimental 2.1. Adsorption SRGO (11,780 ppm S and 260 ppm N) used in this study was provided by a Japanese commercial refinery. Fig. 1 shows its carbon-, sulfur-, nitrogen-chromatograms. Most of sulfur species in SRGO were benzothiophenes (BT), dibenzothiophenes (DBTs). Quinolines, indoles, and carbazoles were found in nitrogen-specific chromatogram of the feed gas oil. Activated carbon materials, which were dried at 110 ◦ C under vacuum oven prior to adsorption, were packed into the stainless steel tube of 50 mm length and 6 mm diameter. Key properties of carbon materials examined in the present study were listed in Table 1. SRGO was fed into the tube
2000
Carbon Chromatogram 1500
(B)
1000 500
(A) AED Response(Arb.Unit)
0 5
10
300
15
20
25
30
15
20
25
30
Sulfur Chromatogram
200
(B)
100
(A)
0 5
10
Nitrogen Chromatogram
20
(B) 10
(A) 0 15
20
25
30
Retention Time(min) Fig. 1. Carbon, sulfur, and nitrogen chromatograms of (A) SRGO and (B) adsorptively treated SRGO. Table 1 The characteristics of the carbon materials examined in the present study Carbon material
MAXSORB-II OG-20A MGC-B
Elemental analysis (wt.%) C
H
N
O
85.00 93.83 80.11
0.49 0.69 0.60
0.14 0.25 3.23
14.16 5.15 7.27
Surface area (cm2 /g)
Total pore volume (cm3 /g)
2972 2000 683
1.75 1.10 0.59
Y. Sano et al. / Applied Catalysis B: Environmental 49 (2004) 219–225
by an HPLC pump (Shimadzu, LC-10AD) at the rate of 0.1 ml/min under the pressure of 0.14 MPa. The temperature of the adsorption tube was kept at 10, 30, and 50 ◦ C by water bath. The eluted oil was sampled for 30 s (0.05 ml) at every 60 min and analyzed by GC-AED (atomic emission detector, HP6890P and G2350A). Activated carbon used in the adsorption experiment was regenerated by toluene extraction at 80 ◦ C and then dried at 135 ◦ C under vacuum. 2.2. HDS reaction The eluted oil was hydrodesulfurized over a current commercial catalyst (CoMo/SiO2 -Al2 O3 ) in an autoclave-type reactor (100 ml internal volume). The 10 g oil, 1 g catalyst, which was pre-sulfided by 5% H2 S/H2 at 360 ◦ C for 2 h, and 50 kg/cm2 hydrogen gas were charged into the reactor at room temperature. The temperature of the reactor was raised to 340 ◦ C by 50 min and maintained at that temperature for 2 h. The HDS product was sampled through filter fitted in the reactor and analyzed by GC-AED.
3. Results 3.1. Removal of nitrogen and sulfur over MAXSORB-II, a particular activated carbon of the largest surface area The carbon-, sulfur-, nitrogen-chromatograms of treated SRGO (120 ml oil/1 g MAXSORB-II) are also illustrated in Fig. 1. There was no difference between the carbonchromatograms of SRGO and treated-SRGO. Treated-SRGO contained certainly less amounts of larger sulfur species, such as dimethydibenzothiophens (DMDBTs) and trimethy-
221
dibenzothiophens (TMDBTs), than the original SRGO. The selective removal of such sulfur species is definitely observed over the activated carbon. Unlike sulfur- and carbon-chromatograms, the distributions of nitrogen species of treated SRGO were very different from that of the feed, although the nitrogen contents were much smaller in the treated SRGOs than that of the feed SRGO. Fig. 2 shows the breakthrough profiles during adsorptive removal of the sulfur and nitrogen species over the activated carbon at various temperatures (10, 30, and 50 ◦ C). The removal extents of nitrogen species were basically independent on the adsorption temperature as shown in Fig. 2. In contrast, adsorption at the higher temperature reduced the adsorption of total sulfur species as indicated in Fig. 2. However, refractory sulfur species such as 4,6-DMDBTs were found to be more selectively removed at the higher temperatures. Fig. 3 illustrates sulfur and nitrogen contents in treated SRGOs sampled at various elution times. The 37% of total sulfur species were left in the first 0.05 ml of the treated SRGO. The content increased rapidly to 90% by 20 ml elution and then slowly to 100% by 45 ml elution. The slow increase of sulfur species from 20 to 40 ml elution corresponded to the removal of refractory sulfur species such as 4,6-DMDBT and 4,6,X-TMDBT as described below. Only 10% of refractory sulfur species were left in the first 0.05 ml of the eluted SRGO. The content increased slowly to reach 100% by 55 ml elution, showing longer breakthrough profile than the total sulfur content. Preferential adsorption of refractory sulfur species was definite over MAXSORB-II. Adsorption of nitrogen species was much more effective than that of sulfur species. The nitrogen content remained at much lower level than the sulfur content in SRGO. First 250
12000
Nitrogen
Sulfur 200
Sulfur Content(ppm)
6000
o
10 C o 30 C o 50 C
3000
Nitrogen Content(ppm)
o
9000
10 C o 30 C o 50 C
150
100
50
0 0 0
10
20
30
40
Treated Amount(ml-oil/g-absorbent)
0
50
100
150
200
250
Treated Amount(ml-oil/g-absorbent)
Fig. 2. Sulfur and nitrogen breakthrough profiles of SRGO on MAXSORB-II at various adsorption temperatures.
222
Y. Sano et al. / Applied Catalysis B: Environmental 49 (2004) 219–225
Content in product/Content in feed(%)
100
Total Sulfur
80
Refractory Sulfur
60
40
Total Nitrogen
20
0 0
50
100
150
Treated Amount(ml-oil/g-absorbent) Fig. 3. Sulfur and nitrogen breakthrough profiles of SRGO on MAXSORB-II at 30 ◦ C.
0.05 ml of eluted SRGO contained no nitrogen species at all. The nitrogen content increased gradually to reach 100% by 150 ml elution. The removed amounts of total sulfur, refractory sulfur, and nitrogen species until their 100% breakthrough were calculated; 0.098 g total sulfur, 0.042 g refractory sulfur, and 0.039 g total nitrogen over 1 g MAXSORB-II, respectively, at 30 ◦ C.
adsorption capacity while their surface area may not be the most important factor because MGC-B of low surface area showed larger removal extent than OG-20A of large surface area. The present materials can remove sulfur species by 12–28%, surface area of carbon materials is appeared to be the most important factor.
3.2. Recovery of treated oils and their nitrogen removal
Fig. 5 compares sulfur-chromatograms of HDS products from the original SRGO and adsorptively treated SRGOs.
80
60
40 20
Removal Ratio(%)
About 0.5 ml of oil remained on 0.3 g of MAXSORB-II after the adsorptive treating of 120 ml SRGO due to the void volume in the reactor. Since this amount of the oil stayed in the reactor regardless of eluted amount of oil, increasing the amount of treated oil improved remarkably the recovery of treated oil. The treated oil, however, contained more sulfur and nitrogen species according to the increasing amount of the treated oil. When 12 ml oil was treated over 0.3 g of MAXSORB-II, about 97% of oil was recovered where 95% of nitrogen species being removed. When 36 ml of oil was treated over MAXSORB-II, about 99% of oil was recovered while only 77% of nitrogen species were removed.
3.4. HDS of feed and treated SRGOs
0
3.3. Adsorption performance of active carbon materials Fig. 4 illustrates the removal extents of the sulfur and nitrogen species over some activated carbon materials, where their removal ratios of total sulfur and nitrogen in 36 ml recovered oil over 0.3 g of carbon materials are illustrated. The nitrogen removal over MAXSORB-II was very remarkable, removing 77% of nitrogen while MGC-B and OG-20A did 41 and 43%, respectively under the same conditions. The oxygen content of carbon materials appears to reflect the
C
Nitrogen
OG-20A
B Sulfur
MGC-B MAXSORB-II
Fig. 4. Capacities of activated carbon materials for removing nitrogen and sulfur species in gas oil. Treated oil: 36 ml; adsorbent: 3 g; removal ratio = removed amount of sulfur and nitrogen species/total sulfur and nitrogen content in feed oil.
Y. Sano et al. / Applied Catalysis B: Environmental 49 (2004) 219–225
223
15 Nitrogen Chromatogram
Sulfur Chromatogram
30
AED Response(Arb. Unit)
(A) 10 20
(A) (B) 5
(B)
10
(C)
(C) 0
0 5
10
15
20
25
Retention Time(min)
30
5
10
15
20
25
30
Retention Time(min)
Fig. 5. Sulfur and nitrogen chromatograms of HDS products from (A) SRGO, (B) adsorptively treated SRGO (total nitrogen content: 60 ppm), and (C) adsorptively treated SRGO (total nitrogen content: 20 ppm).
The sulfur removal levels, by adsorptive treatment, of treated SRGOs were 22 and 26%, respectively. Nitrogen contents of the treated SRGO were estimated to be 60 and 20 ppm N. The adsorptive treatment was found very effective to enhance the HDS, reducing the sulfur levels of HDS products from 193 ppm of feed SRGO to 11 and 8 ppm of treated SRGOs, respectively, after HDS as indicated in Table 2. By these levels of nitrogen removal, sulfur level less than 15 ppm can be achieved by the hydrodesulfurization under the conventional conditions. All hydrodesulfurized products contained the refractory sulfur species exclusively. Although the contents of reactive sulfur species as well as total sulfur species were different between the feed and treated oils, enhancement for HDS of 4,6-DMDBT and reduction of some heavier sulfur species by the adsorption appear to contribute significantly to the achievement of deep HDS. HDN was also performed along with HDS. Under the present HDS conditions, almost all nitrogen species are removed through the hydrodenitrogenation. Table 2 Total sulfur and nitrogen contents of SRGO, adsorptively treated SRGO and the total sulfur contents of their HDS products Before HDS (ppm)
After HDS (ppm)
Total nitrogen content
Total sulfur content
Refractory sulfur content
Total sulfur content
250a 60b 20b
11340a 9200b 8700b
1680a 980b 860b
193 11 8
a b
SRGO. Adsorptively treated SRGO.
3.5. Regeneration of active carbon for the extractive removal of adsorbed nitrogen and sulfur species MAXSORB-II, which was used in the adsorptive treatment of 120 ml SRGO and then washed by toluene to remove adsorbed species, was evaluated in the same adsorption conditions. Fig. 6 compares the removal of nitrogen and sulfur species in SRGO over virgin and regenerated MAXSORB-II. Regenerated MAXSORB-II showed almost the same breakthrough profiles of nitrogen and sulfur species. Toluene extraction at 80 ◦ C was confirmed to regenerate the adsorptive activity of MAXSORB-II for repeated removal of sulfur and nitrogen species in SRGO.
4. Discussion The present paper emphasizes three points: 1. Adsorptive treatment of SRGO over the activated carbon at ambient temperature was very effective to achieve very low sulfur level less than 15 ppm by the conventional HDS of the treated SRGO. 2. Adsorptive treatment removed nitrogen and refractory sulfur species in gas oil simultaneously to contribute to the achievement of ultra deep hydrodesulfurization. 3. A particular activated carbon, MAXSORB-II, showed the excellent performance in the adsorptive treatment. Its adsorption activity was successfully regenerated by extractive recovery of the adsorbed species with toluene. The nitrogen species of both basic and neutral natures are certainly inhibitors for deep hydrodesulfurization of
224
Y. Sano et al. / Applied Catalysis B: Environmental 49 (2004) 219–225 10000 250 8000
Virgin 2nd 3rd
6000
Virgin 2nd 3rd 4000
Nitrogen Content(ppm)
Sulfur content (ppm)
200
150
100
50
2000
0 0 0
5
10
15
20
Treated Amount(ml-oil/g-absorbent)
0
50
100
150
200
250
Treated Amount(ml-oil/g-absorbent)
Fig. 6. Sulfur and nitrogen breakthrough profiles of SRGO on virgin and regenerated MAXSORB-II at 30 ◦ C. Second and third runs were performed after extractive regeneration of the virgin adsorbent.
refractory sulfur species especially at their low level in the range of 10–200 ppm [6]. Basic nitrogen species tend to show strong inhibition, however many of them were denitrogenated at rather initial stage of hydrodesulfurization. The neutral nitrogen species such as carbazoles which are of low reactivity tend to stay through the whole stage of deep hydrodesulfurization where the last portion of refractory sulfur species were hydrodesulfurized to achieve sulfur level less than 15 ppm [23]. Hence, their removal prior to the hydrotreatment is very effective for deep HDS as previously reported [8–11], even if they were denitrogenated at the last minute of HDS. It must be recognized that removal of refractory sulfur species is also very effective to achieve the deep hydrodesulfurization, although the reduction of total sulfur species by adsorption treatment is very limited. Because the heavier refractory sulfur species in the feed is of very low reactivity and highly inhibited [24], their removal can be suggested to accelerate the hydrodesulfurization in the region of ultra deep hydrodesulfurization. MAXSORB-II was found particularly active to remove both nitrogen and sulfur species in SRGO although three carbon materials examined in the present study showed much higher removal capacity than silica gel and alumina, which had been reported as effective adsorbents for nitrogen species in the gas oil. Its adsorption capacity (0.098 g total sulfur and 0.039 g total nitrogen) was much larger than Cu-exchanged Y-zeolite (1.40 wt.%) [16]. Its very large surface area and high oxygen content appear to be reasons for its high performance. MAXSORB-II has been disclosed to be prepared through KOH activation
of green coke [25,26]. Hence it is very expensive at present. It must be clarified for exploring the effective activated carbon of low cost whether its very large surface area over 2000 m2 /g is indispensable or not. The polar surface of active carbon may be responsible to adsorb nitrogen species of high polarity. In contrast, the removal of refractory sulfur species looks dependent on the surface area of active carbons. Their lower solubility in the gas oil may be a major driving force for their preferential adsorption. The adequate surface of active carbons can be designed for the respective removal of basic, polar nitrogen species or heavier sulfur species. The modification of activated carbon and carbon fiber to improve the adsorption capacity for sulfur and nitrogen species will be attempted in a following paper. Furthermore, detail mechanism and driving force in selective adsorption of sulfur and nitrogen species in gas oil will be investigated to improve the adsorption capacity of activated carbon. Finally, adsorption capacity of MAXSORB-II to achieve the prescribed sulfur level in the following HDS is 120 ml oil/1 g carbon. Such a capacity is fairly high compared with that of silica gel. In addition, rather easy regeneration of activated carbon by toluene extraction was confirmed. In contrast, silica or other oxide absorbents have been known not to be easily regenerated. In fact, alcohol or polar solvent is required to remove strongly adsorbed species from the adsorbent [11]. However, activated carbon was successfully regenerated by conventional hydrocarbon solvent, toluene, which is often found in refinery plant. Thus, the active carbon appears to be better adsorbent than silica gel for practical application of adsorptive removal to establish the pre-treatment for the facile deep HDS of gas oil in the usual refinery.
Y. Sano et al. / Applied Catalysis B: Environmental 49 (2004) 219–225
5. Conclusion Carbon material could be applied as an absorbent for the removal of both sulfur and nitrogen species of gas oil which was found effective for acceleration of HDS. Adsorption capacity of carbon materials depended on their surface properties as well as surface area. The gas oil free from nitrogen species with reduced content of heavier refractory sulfur species showed significant enhancement of hydrodesulfurization to achieve 15 ppm level of sulfur.
References [1] T.R. Halbert, G.B. Brignac, R.A. Demmin, E.M. Roundtree, Hydrocarbon Eng., (2000) 2. [2] I.V. Babich, J.A. Moulijn, Fuel 82 (2003) 607. [3] D.D. Whitehurst, T. Isoda, I. Mochida, Adv. Catal. 42 (1998) 345. [4] D.D. Whitehurst, H. Farag, T. Nagamatsu, K. Sakanishi, I. Mochida, Catal. Today 45 (1998) 299. [5] I. Mochida, K.-H. Choi, J. Jpn. Petrol. Insts. (2004), in press. [6] S. Shin, H. Yang, K. Sakanishi, I. Mochida, D.A. Groudoski, J.H. Shinn, Appl. Catal. A 243 (2003) 207. [7] P. Zeuthen, K.G. Knudsen, D.D. Whitehurst, Catal. Today 65 (2001) 307. [8] S.D. Sumbogo, M.H. Yang, K.H. Choi, T. Korai, I. Mochida, Appl. Catal. A 252 (2003) 331.
225
[9] B.C. Gates, H. Topsoe, Polyhedron 16 (1997) 18. [10] K.-H Choi, T. Korai, I. Mochida, J.W. Ryoo, W. Min, Appl. Catal. B (2004), in press. [11] W.-S. Min, K.-I. Choi, S.-Y. Khang, D.-S. Min, J.-W. Ryu, K.-S. Yoo, J.-H. Kim, US Patent 6 248 230, to SK Corporation (2001). [12] C. Song, X. Ma, Appl. Catal. B 41 (2003) 207. [13] G.P. Khare, US Patent 6 346 190, to Phillips Petroleum Company (2002). [14] A.S.H. Salem, H.S. Hamid, Chem. Eng. Technol. 20 (1997) 342. [15] R.L. Irvine US Patent 5 730 860, to the Pritchard Corporation (1998). [16] A.J. Hernandez-Maldonado, R.T. Yang, Ind. Eng. Chem. Res. 42 (2003) 123. [17] A. Takahashi, F.H. Yang, R.T. Yang, Ind. Eng. Chem. Res. 41 (2003) 2487. [18] N. Shirahama, S.H. Moon, K.-H. Choi, T. Enjoji, S. Kawano, Y. Korai, M. Tanoura, I. Mochida, Carbon 40 (2002) 2605. [19] R. Leboda, J. Skubiszewska-Ziba, W. Tomaszewski, V.M. Gunko, J. Colloid Interface Sci. 263 (2003) 533. [20] J. Hayashi, T. Horikawa, I. Takeda, K. Muroyama, F.N. Ani, Carbon 40 (2002) 2381. [21] A.M. Puziy, O.I. Poddubnaya, A. Mart´ınez-Alonso, F. Suárez-Garc´ıa, J.M.D. Tascón, Appl. Surf. Sci. 200 (2002) 1493. [22] A. Takahashi, F.H. Yang, R.T. Yang, Ind. Eng. Chem. Res. 41 (2002) 2487. [23] A. Szymanska, M. Lewandowski, C. Sayag, G.D. Mariadassou, J. Catal. 218 (2003) 24. [24] M. Macaud, A. Milenkovic, E. Schulz, M. Lemaire, M. Vrinat, J. Catal. 193 (2000) 255. [25] T. Otowa, R. Tanibata, M. Itoh, Gas Sep. Purif. 4 (1993) 241. [26] T. Otowa, Y. Nojima, T. Miyazaki, Carbon 35 (1997) 1315.
Adsorptive removal of sulfur and nitrogen species from a straight run gas oil over activated carbons for its deep hydrodesulfurization Yosuke Sano, Ki-Hyouk Choi, Yozo Korai, Isao Mochida∗ Institute for Advanced Materials Chemistry and Engineering, Kyushu University, Kasugakouen 6-1, Fukuoka 816-8580, Japan Received 18 September 2003; received in revised form 11 December 2003; accepted 12 December 2003
Abstract As a pre-treatment step for ultra deep hydrodesulfurization (UD-HDS), sulfur and nitrogen species were adsorptively removed from the straight run gas oil (SRGO) over activated carbon materials. An activated carbon having the largest surface area and highest surface polarity showed the highest total capacity for adsorption of sulfur and nitrogen species as 0.098 g sulfur and 0.039 g nitrogen per 1 g of the activated carbon, respectively, among the carbon materials examined. Removal of both nitrogen and sulfur species was found very effective to achieve UD-HDS under conventional HDS conditions. SRGO treated over activated carbon contained only 11 ppm of sulfur, while non-treated SRGO contained still 193 ppm of sulfur in its HDS product. © 2004 Elsevier B.V. All rights reserved. Keywords: Hydrodesulfurization; Inhibition; Activated carbon; Adsorption
1. Introduction World refining industry is facing new and stricter regulations on the heteroatom contents in their major products such as diesel and gasoline. The sulfur content of diesel fuel must be lowered to less than 15 ppm S from current 500 ppm S and even further regulations may be implemented with accelerated concerns on the atmospheric pollution [1]. Several new approaches have been proposed and some of them were practically examined to achieve the ultra deep hydrodesulfurization (UD-HDS) [2]. However, more efficient and more economical process is still required in spite of various efforts. Critical barriers for achieving ultra deep hydrodesulfurization are very low reactivity of refractory sulfur species under conventional conditions and strong inhibition by H2 S, NH3 , nitrogen, and even aromatic species against UD-HDS. Such inhibitors retard the hydrodesulfurization, very markedly in the region of 0–100 ppm S [3–7]. There are some proposals of the pre-treatment for gas oil to remove nitrogen species by silica or silica–alumina prior to ∗
Corresponding author. Tel.: +81-92-5837797; fax: +81-92-837798. E-mail address: [email protected] (I. Mochida).
0926-3373/$ – see front matter © 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.apcatb.2003.12.007
hydrodesulfurization [8–11]. However, there are still some windows to improve the cost/performance in terms of the adsorbent. Removal level of nitrogen species, efficiency and effects of sulfur removal, regeneration of adsorbent and oil recovery must be delicately balanced. Song and Ma [12] proposed selective adsorption for removing sulfur at ambient temperature (SARS) process to achieve ultra clean diesel and gasoline. They used 5 wt.% metal loaded on silica gel as an adsorbent. Feed oil was not real diesel, but model species which did not contain the nitrogen species. Although detail information on the adsorbent was not disclosed, their adsorbent showed rather fast breakthrough over 10 ppm after eluting about 5 ml of model diesel. IRVAD and S-Zorb process also utilize adsorption to remove the heteroatoms in gasoline and diesel. But, operating temperature of them are reported to be over 240 ◦ C (IRVAD) and 340–410 ◦ C (S-Zorb) [13–15]. Yang and coworkers [16,17] reported that Cu and Ag-exchanged Y-type zeolites were very effective to selectively adsorb the thiophene and the used adsorbents could be regenerated by heating them at 350 ◦ C under flowing air. Activated carbon and activated carbon fiber have been recognized to be versatile absorbents for gaseous and liquid
220
Y. Sano et al. / Applied Catalysis B: Environmental 49 (2004) 219–225
substrates [18,19]. Their surface structure and properties can be controllable to propose better adsorbent [20–22]. In this study, sulfur and nitrogen species were removed from straight run gas oil (SRGO) through the dynamic adsorption over some activated carbon materials at room temperature. A particular activated carbon of very large surface area was examined in detail. The treated oil was hydrodesulfurized under conventional conditions to evaluate the effects of pre-treatment over activated carbons on the HDS activity to establish the facile UD-HDS process. It must be emphasized in the present study that carbon adsorbent removes sulfur as well as nitrogen species in the gas oil. The reduction of both species must be favorable for the deep hydrodesulfurization. Furthermore, the regeneration of the adsorbent was successfully performed with a conventional solvent, toluene, at low temperature.
2. Experimental 2.1. Adsorption SRGO (11,780 ppm S and 260 ppm N) used in this study was provided by a Japanese commercial refinery. Fig. 1 shows its carbon-, sulfur-, nitrogen-chromatograms. Most of sulfur species in SRGO were benzothiophenes (BT), dibenzothiophenes (DBTs). Quinolines, indoles, and carbazoles were found in nitrogen-specific chromatogram of the feed gas oil. Activated carbon materials, which were dried at 110 ◦ C under vacuum oven prior to adsorption, were packed into the stainless steel tube of 50 mm length and 6 mm diameter. Key properties of carbon materials examined in the present study were listed in Table 1. SRGO was fed into the tube
2000
Carbon Chromatogram 1500
(B)
1000 500
(A) AED Response(Arb.Unit)
0 5
10
300
15
20
25
30
15
20
25
30
Sulfur Chromatogram
200
(B)
100
(A)
0 5
10
Nitrogen Chromatogram
20
(B) 10
(A) 0 15
20
25
30
Retention Time(min) Fig. 1. Carbon, sulfur, and nitrogen chromatograms of (A) SRGO and (B) adsorptively treated SRGO. Table 1 The characteristics of the carbon materials examined in the present study Carbon material
MAXSORB-II OG-20A MGC-B
Elemental analysis (wt.%) C
H
N
O
85.00 93.83 80.11
0.49 0.69 0.60
0.14 0.25 3.23
14.16 5.15 7.27
Surface area (cm2 /g)
Total pore volume (cm3 /g)
2972 2000 683
1.75 1.10 0.59
Y. Sano et al. / Applied Catalysis B: Environmental 49 (2004) 219–225
by an HPLC pump (Shimadzu, LC-10AD) at the rate of 0.1 ml/min under the pressure of 0.14 MPa. The temperature of the adsorption tube was kept at 10, 30, and 50 ◦ C by water bath. The eluted oil was sampled for 30 s (0.05 ml) at every 60 min and analyzed by GC-AED (atomic emission detector, HP6890P and G2350A). Activated carbon used in the adsorption experiment was regenerated by toluene extraction at 80 ◦ C and then dried at 135 ◦ C under vacuum. 2.2. HDS reaction The eluted oil was hydrodesulfurized over a current commercial catalyst (CoMo/SiO2 -Al2 O3 ) in an autoclave-type reactor (100 ml internal volume). The 10 g oil, 1 g catalyst, which was pre-sulfided by 5% H2 S/H2 at 360 ◦ C for 2 h, and 50 kg/cm2 hydrogen gas were charged into the reactor at room temperature. The temperature of the reactor was raised to 340 ◦ C by 50 min and maintained at that temperature for 2 h. The HDS product was sampled through filter fitted in the reactor and analyzed by GC-AED.
3. Results 3.1. Removal of nitrogen and sulfur over MAXSORB-II, a particular activated carbon of the largest surface area The carbon-, sulfur-, nitrogen-chromatograms of treated SRGO (120 ml oil/1 g MAXSORB-II) are also illustrated in Fig. 1. There was no difference between the carbonchromatograms of SRGO and treated-SRGO. Treated-SRGO contained certainly less amounts of larger sulfur species, such as dimethydibenzothiophens (DMDBTs) and trimethy-
221
dibenzothiophens (TMDBTs), than the original SRGO. The selective removal of such sulfur species is definitely observed over the activated carbon. Unlike sulfur- and carbon-chromatograms, the distributions of nitrogen species of treated SRGO were very different from that of the feed, although the nitrogen contents were much smaller in the treated SRGOs than that of the feed SRGO. Fig. 2 shows the breakthrough profiles during adsorptive removal of the sulfur and nitrogen species over the activated carbon at various temperatures (10, 30, and 50 ◦ C). The removal extents of nitrogen species were basically independent on the adsorption temperature as shown in Fig. 2. In contrast, adsorption at the higher temperature reduced the adsorption of total sulfur species as indicated in Fig. 2. However, refractory sulfur species such as 4,6-DMDBTs were found to be more selectively removed at the higher temperatures. Fig. 3 illustrates sulfur and nitrogen contents in treated SRGOs sampled at various elution times. The 37% of total sulfur species were left in the first 0.05 ml of the treated SRGO. The content increased rapidly to 90% by 20 ml elution and then slowly to 100% by 45 ml elution. The slow increase of sulfur species from 20 to 40 ml elution corresponded to the removal of refractory sulfur species such as 4,6-DMDBT and 4,6,X-TMDBT as described below. Only 10% of refractory sulfur species were left in the first 0.05 ml of the eluted SRGO. The content increased slowly to reach 100% by 55 ml elution, showing longer breakthrough profile than the total sulfur content. Preferential adsorption of refractory sulfur species was definite over MAXSORB-II. Adsorption of nitrogen species was much more effective than that of sulfur species. The nitrogen content remained at much lower level than the sulfur content in SRGO. First 250
12000
Nitrogen
Sulfur 200
Sulfur Content(ppm)
6000
o
10 C o 30 C o 50 C
3000
Nitrogen Content(ppm)
o
9000
10 C o 30 C o 50 C
150
100
50
0 0 0
10
20
30
40
Treated Amount(ml-oil/g-absorbent)
0
50
100
150
200
250
Treated Amount(ml-oil/g-absorbent)
Fig. 2. Sulfur and nitrogen breakthrough profiles of SRGO on MAXSORB-II at various adsorption temperatures.
222
Y. Sano et al. / Applied Catalysis B: Environmental 49 (2004) 219–225
Content in product/Content in feed(%)
100
Total Sulfur
80
Refractory Sulfur
60
40
Total Nitrogen
20
0 0
50
100
150
Treated Amount(ml-oil/g-absorbent) Fig. 3. Sulfur and nitrogen breakthrough profiles of SRGO on MAXSORB-II at 30 ◦ C.
0.05 ml of eluted SRGO contained no nitrogen species at all. The nitrogen content increased gradually to reach 100% by 150 ml elution. The removed amounts of total sulfur, refractory sulfur, and nitrogen species until their 100% breakthrough were calculated; 0.098 g total sulfur, 0.042 g refractory sulfur, and 0.039 g total nitrogen over 1 g MAXSORB-II, respectively, at 30 ◦ C.
adsorption capacity while their surface area may not be the most important factor because MGC-B of low surface area showed larger removal extent than OG-20A of large surface area. The present materials can remove sulfur species by 12–28%, surface area of carbon materials is appeared to be the most important factor.
3.2. Recovery of treated oils and their nitrogen removal
Fig. 5 compares sulfur-chromatograms of HDS products from the original SRGO and adsorptively treated SRGOs.
80
60
40 20
Removal Ratio(%)
About 0.5 ml of oil remained on 0.3 g of MAXSORB-II after the adsorptive treating of 120 ml SRGO due to the void volume in the reactor. Since this amount of the oil stayed in the reactor regardless of eluted amount of oil, increasing the amount of treated oil improved remarkably the recovery of treated oil. The treated oil, however, contained more sulfur and nitrogen species according to the increasing amount of the treated oil. When 12 ml oil was treated over 0.3 g of MAXSORB-II, about 97% of oil was recovered where 95% of nitrogen species being removed. When 36 ml of oil was treated over MAXSORB-II, about 99% of oil was recovered while only 77% of nitrogen species were removed.
3.4. HDS of feed and treated SRGOs
0
3.3. Adsorption performance of active carbon materials Fig. 4 illustrates the removal extents of the sulfur and nitrogen species over some activated carbon materials, where their removal ratios of total sulfur and nitrogen in 36 ml recovered oil over 0.3 g of carbon materials are illustrated. The nitrogen removal over MAXSORB-II was very remarkable, removing 77% of nitrogen while MGC-B and OG-20A did 41 and 43%, respectively under the same conditions. The oxygen content of carbon materials appears to reflect the
C
Nitrogen
OG-20A
B Sulfur
MGC-B MAXSORB-II
Fig. 4. Capacities of activated carbon materials for removing nitrogen and sulfur species in gas oil. Treated oil: 36 ml; adsorbent: 3 g; removal ratio = removed amount of sulfur and nitrogen species/total sulfur and nitrogen content in feed oil.
Y. Sano et al. / Applied Catalysis B: Environmental 49 (2004) 219–225
223
15 Nitrogen Chromatogram
Sulfur Chromatogram
30
AED Response(Arb. Unit)
(A) 10 20
(A) (B) 5
(B)
10
(C)
(C) 0
0 5
10
15
20
25
Retention Time(min)
30
5
10
15
20
25
30
Retention Time(min)
Fig. 5. Sulfur and nitrogen chromatograms of HDS products from (A) SRGO, (B) adsorptively treated SRGO (total nitrogen content: 60 ppm), and (C) adsorptively treated SRGO (total nitrogen content: 20 ppm).
The sulfur removal levels, by adsorptive treatment, of treated SRGOs were 22 and 26%, respectively. Nitrogen contents of the treated SRGO were estimated to be 60 and 20 ppm N. The adsorptive treatment was found very effective to enhance the HDS, reducing the sulfur levels of HDS products from 193 ppm of feed SRGO to 11 and 8 ppm of treated SRGOs, respectively, after HDS as indicated in Table 2. By these levels of nitrogen removal, sulfur level less than 15 ppm can be achieved by the hydrodesulfurization under the conventional conditions. All hydrodesulfurized products contained the refractory sulfur species exclusively. Although the contents of reactive sulfur species as well as total sulfur species were different between the feed and treated oils, enhancement for HDS of 4,6-DMDBT and reduction of some heavier sulfur species by the adsorption appear to contribute significantly to the achievement of deep HDS. HDN was also performed along with HDS. Under the present HDS conditions, almost all nitrogen species are removed through the hydrodenitrogenation. Table 2 Total sulfur and nitrogen contents of SRGO, adsorptively treated SRGO and the total sulfur contents of their HDS products Before HDS (ppm)
After HDS (ppm)
Total nitrogen content
Total sulfur content
Refractory sulfur content
Total sulfur content
250a 60b 20b
11340a 9200b 8700b
1680a 980b 860b
193 11 8
a b
SRGO. Adsorptively treated SRGO.
3.5. Regeneration of active carbon for the extractive removal of adsorbed nitrogen and sulfur species MAXSORB-II, which was used in the adsorptive treatment of 120 ml SRGO and then washed by toluene to remove adsorbed species, was evaluated in the same adsorption conditions. Fig. 6 compares the removal of nitrogen and sulfur species in SRGO over virgin and regenerated MAXSORB-II. Regenerated MAXSORB-II showed almost the same breakthrough profiles of nitrogen and sulfur species. Toluene extraction at 80 ◦ C was confirmed to regenerate the adsorptive activity of MAXSORB-II for repeated removal of sulfur and nitrogen species in SRGO.
4. Discussion The present paper emphasizes three points: 1. Adsorptive treatment of SRGO over the activated carbon at ambient temperature was very effective to achieve very low sulfur level less than 15 ppm by the conventional HDS of the treated SRGO. 2. Adsorptive treatment removed nitrogen and refractory sulfur species in gas oil simultaneously to contribute to the achievement of ultra deep hydrodesulfurization. 3. A particular activated carbon, MAXSORB-II, showed the excellent performance in the adsorptive treatment. Its adsorption activity was successfully regenerated by extractive recovery of the adsorbed species with toluene. The nitrogen species of both basic and neutral natures are certainly inhibitors for deep hydrodesulfurization of
224
Y. Sano et al. / Applied Catalysis B: Environmental 49 (2004) 219–225 10000 250 8000
Virgin 2nd 3rd
6000
Virgin 2nd 3rd 4000
Nitrogen Content(ppm)
Sulfur content (ppm)
200
150
100
50
2000
0 0 0
5
10
15
20
Treated Amount(ml-oil/g-absorbent)
0
50
100
150
200
250
Treated Amount(ml-oil/g-absorbent)
Fig. 6. Sulfur and nitrogen breakthrough profiles of SRGO on virgin and regenerated MAXSORB-II at 30 ◦ C. Second and third runs were performed after extractive regeneration of the virgin adsorbent.
refractory sulfur species especially at their low level in the range of 10–200 ppm [6]. Basic nitrogen species tend to show strong inhibition, however many of them were denitrogenated at rather initial stage of hydrodesulfurization. The neutral nitrogen species such as carbazoles which are of low reactivity tend to stay through the whole stage of deep hydrodesulfurization where the last portion of refractory sulfur species were hydrodesulfurized to achieve sulfur level less than 15 ppm [23]. Hence, their removal prior to the hydrotreatment is very effective for deep HDS as previously reported [8–11], even if they were denitrogenated at the last minute of HDS. It must be recognized that removal of refractory sulfur species is also very effective to achieve the deep hydrodesulfurization, although the reduction of total sulfur species by adsorption treatment is very limited. Because the heavier refractory sulfur species in the feed is of very low reactivity and highly inhibited [24], their removal can be suggested to accelerate the hydrodesulfurization in the region of ultra deep hydrodesulfurization. MAXSORB-II was found particularly active to remove both nitrogen and sulfur species in SRGO although three carbon materials examined in the present study showed much higher removal capacity than silica gel and alumina, which had been reported as effective adsorbents for nitrogen species in the gas oil. Its adsorption capacity (0.098 g total sulfur and 0.039 g total nitrogen) was much larger than Cu-exchanged Y-zeolite (1.40 wt.%) [16]. Its very large surface area and high oxygen content appear to be reasons for its high performance. MAXSORB-II has been disclosed to be prepared through KOH activation
of green coke [25,26]. Hence it is very expensive at present. It must be clarified for exploring the effective activated carbon of low cost whether its very large surface area over 2000 m2 /g is indispensable or not. The polar surface of active carbon may be responsible to adsorb nitrogen species of high polarity. In contrast, the removal of refractory sulfur species looks dependent on the surface area of active carbons. Their lower solubility in the gas oil may be a major driving force for their preferential adsorption. The adequate surface of active carbons can be designed for the respective removal of basic, polar nitrogen species or heavier sulfur species. The modification of activated carbon and carbon fiber to improve the adsorption capacity for sulfur and nitrogen species will be attempted in a following paper. Furthermore, detail mechanism and driving force in selective adsorption of sulfur and nitrogen species in gas oil will be investigated to improve the adsorption capacity of activated carbon. Finally, adsorption capacity of MAXSORB-II to achieve the prescribed sulfur level in the following HDS is 120 ml oil/1 g carbon. Such a capacity is fairly high compared with that of silica gel. In addition, rather easy regeneration of activated carbon by toluene extraction was confirmed. In contrast, silica or other oxide absorbents have been known not to be easily regenerated. In fact, alcohol or polar solvent is required to remove strongly adsorbed species from the adsorbent [11]. However, activated carbon was successfully regenerated by conventional hydrocarbon solvent, toluene, which is often found in refinery plant. Thus, the active carbon appears to be better adsorbent than silica gel for practical application of adsorptive removal to establish the pre-treatment for the facile deep HDS of gas oil in the usual refinery.
Y. Sano et al. / Applied Catalysis B: Environmental 49 (2004) 219–225
5. Conclusion Carbon material could be applied as an absorbent for the removal of both sulfur and nitrogen species of gas oil which was found effective for acceleration of HDS. Adsorption capacity of carbon materials depended on their surface properties as well as surface area. The gas oil free from nitrogen species with reduced content of heavier refractory sulfur species showed significant enhancement of hydrodesulfurization to achieve 15 ppm level of sulfur.
References [1] T.R. Halbert, G.B. Brignac, R.A. Demmin, E.M. Roundtree, Hydrocarbon Eng., (2000) 2. [2] I.V. Babich, J.A. Moulijn, Fuel 82 (2003) 607. [3] D.D. Whitehurst, T. Isoda, I. Mochida, Adv. Catal. 42 (1998) 345. [4] D.D. Whitehurst, H. Farag, T. Nagamatsu, K. Sakanishi, I. Mochida, Catal. Today 45 (1998) 299. [5] I. Mochida, K.-H. Choi, J. Jpn. Petrol. Insts. (2004), in press. [6] S. Shin, H. Yang, K. Sakanishi, I. Mochida, D.A. Groudoski, J.H. Shinn, Appl. Catal. A 243 (2003) 207. [7] P. Zeuthen, K.G. Knudsen, D.D. Whitehurst, Catal. Today 65 (2001) 307. [8] S.D. Sumbogo, M.H. Yang, K.H. Choi, T. Korai, I. Mochida, Appl. Catal. A 252 (2003) 331.
225
[9] B.C. Gates, H. Topsoe, Polyhedron 16 (1997) 18. [10] K.-H Choi, T. Korai, I. Mochida, J.W. Ryoo, W. Min, Appl. Catal. B (2004), in press. [11] W.-S. Min, K.-I. Choi, S.-Y. Khang, D.-S. Min, J.-W. Ryu, K.-S. Yoo, J.-H. Kim, US Patent 6 248 230, to SK Corporation (2001). [12] C. Song, X. Ma, Appl. Catal. B 41 (2003) 207. [13] G.P. Khare, US Patent 6 346 190, to Phillips Petroleum Company (2002). [14] A.S.H. Salem, H.S. Hamid, Chem. Eng. Technol. 20 (1997) 342. [15] R.L. Irvine US Patent 5 730 860, to the Pritchard Corporation (1998). [16] A.J. Hernandez-Maldonado, R.T. Yang, Ind. Eng. Chem. Res. 42 (2003) 123. [17] A. Takahashi, F.H. Yang, R.T. Yang, Ind. Eng. Chem. Res. 41 (2003) 2487. [18] N. Shirahama, S.H. Moon, K.-H. Choi, T. Enjoji, S. Kawano, Y. Korai, M. Tanoura, I. Mochida, Carbon 40 (2002) 2605. [19] R. Leboda, J. Skubiszewska-Ziba, W. Tomaszewski, V.M. Gunko, J. Colloid Interface Sci. 263 (2003) 533. [20] J. Hayashi, T. Horikawa, I. Takeda, K. Muroyama, F.N. Ani, Carbon 40 (2002) 2381. [21] A.M. Puziy, O.I. Poddubnaya, A. Mart´ınez-Alonso, F. Suárez-Garc´ıa, J.M.D. Tascón, Appl. Surf. Sci. 200 (2002) 1493. [22] A. Takahashi, F.H. Yang, R.T. Yang, Ind. Eng. Chem. Res. 41 (2002) 2487. [23] A. Szymanska, M. Lewandowski, C. Sayag, G.D. Mariadassou, J. Catal. 218 (2003) 24. [24] M. Macaud, A. Milenkovic, E. Schulz, M. Lemaire, M. Vrinat, J. Catal. 193 (2000) 255. [25] T. Otowa, R. Tanibata, M. Itoh, Gas Sep. Purif. 4 (1993) 241. [26] T. Otowa, Y. Nojima, T. Miyazaki, Carbon 35 (1997) 1315.