Synthesis of Mesoporous TS-1 for Cat

1778 Bull.KoreanChem. Soc. 2009, Vol. 30, No.8 Chui- Woo Parket al.

Synthesis of Mesoporous TS-1 for Cat

ytic Oxidative Desulfurization

Chul-Woo Park, Tae-Kyung Kim, andWha-Seung Ahn*

DepartmentofChemical Engineering, Inha University, Incheon 402-751,Korea

*E-mail: whasahn@inha.ac.kr

ReceivedMay 18, 2009, AcceptedJune 25, 2009

Mesoporous TS-1 catalysts were prepared via a nanocasting route using two different carbon template sources of CMK-3 and commercial carbon black. Products were characterized by XRD, UV-Vis spectroscopy, SEM, TEM, and N2 adsorp­

tion-desorption measurement. The catalytic performances of the samples for allylchloride epoxidation and oxidative desulfurization of the representative refractory sulfur compounds, dibenzothiophene and 4,6-dimethyldibenzothiophene, were compared against those of conventional TS-1. Whilst the allychloride epoxidation activity for the mesoporus TS-1 samples were similar, mesoporous TS-1 exhibited significantly higher catalytic activities than conventional TS-1 in oxidative desulfurization.

Key Words: Mesoporous TS-1, CMK-3, Carbonblack, Oxidativedesulfurization

Introduction

Sulfurremoval in transportation fuelsisnecessary in order toreduce the pollution level inthe atmosphere, andcurrently ultra-deep desulfurization of fuel oil is receiving a worldwide attention ever since the announcement of the stringentU. S.

environmental regulationsimposed,limitingthe sulfurlevel in dieselto< 15 ppmw.1,2 The mostcommonindustrialprocess fordesulfurization isthetreatment of thefuel withhydrogen at hightemperature and high pressureconditions.3Whilst this conventional hydrodesulfurization (HDS) is highly efficient in removingthiols,sulfides, and disulfides,4 the HDSprocess toachieve ultra-deepdesulfurization of diesel requires increas­ ingly severe operating conditions with high consumption of hydrogen, thus incurring a penalty of significant capital expen- diture.5 Obviously,development of analternativedesulfuriza­ tion process is desirable, and nowoxidative desulfurization (ODS)under mild conditions(atmospheric pressure and tem­

perature lower than 80 oC) is being consideredasone of the mostpromisingoptions for theultra-deepdesulfurization.6,7

ForODS, Ti-containing zeolites have been extensively inves­ tigated ascatalystbecauseoftheir outstanding oxidation pro­ perty andutilization of peroxides having an environmentally- benign natureas oxidant.8-10 It isknownthat some smallsul­ fides such as mercaptan,thioether,eventhiophene, ahighly stableorganicsulfide,can beeffectivelyoxidizedbyTS-1 ca­

talyst.11 However,conventional microporous TS-1 showspoor catalytic activity in the oxidation of bulkiersulfurcompounds such as dibenzothiophene (DBT)andits derivatives, due to intracrystalline diffusion limitationsimposed by small pores.

Therefore, efforts havebeenmade to resolve this diffusion limitation problem in TS-1 byintroducingmesopores, and va­

rioustemplate synthesistechniques that enable apreparation of zeolites havingmesopores havenewlyemerged. Inparti­ cular, much attention hasbeengivento thesynthesis of meso­

pore-containingzeolitesvia nanocasting byemploying various sacrificial carbon templates in the hydrothermalsynthesisof zeolites.12-17

Inthiswork,weconducted a systematic comparison of the synthesis routes for makingmesoporousTS-1, and detailed characterization of themesoporous TS-1zeolites prepared by employingtwo different carbon templates (carbonblackand CMK-3) was reported. We carriedout a probereaction of ally- chloride epoxidation using H2O2 as oxidant to confirmthe nature of catalyticTisitesinalltheTS-1 samples prepared.

Finally, it was demonstrated that themesoporous TS-1 sam­ ples exhibit superior catalytic performances in oxidative desul­ furization of selected bulkysulfurcompoundstoconventional TS-1, which werealso influencedbytheproperty ofthe car­

bon template used.

Experiment

^시

Synthesis ofmesoporous TS-1. For the synthesis of meso- porous TS-1, CMK-3 and carbon black(N762, Degussa) were usedas solidtemplate. SBA-15 was prepared followingthe literature procedure reported earlier,18 andthecalcined SBA-15 was then used astemplate for making CMK-3. CMK-3 was preparedaccording to theliterature procedure.19 The CMK-3 produced had the BET area of 1070 m2g-1 and the total pore volume of 1.1 cm3g-1. N762 has, onthe otherhand, had the BETarea of 124m2g-1 andthetotalporevolume of 0.22cm3g-1.

The carbon blackconsists of nearlysphericalprimary particles with typical diametersranging from 20 to 150nm.

Mesoporous TS-1 wasprepared following thedrygelsyn­ thesisprocedurereportedby Jacobsen et a" CMK-3 (1.0g) orcarbon black(4.0 g) wasimpregnated with a clearsolution containingtetrapropylammoniumhydroxide(TPAOH,40wt%, TCI, 3.3 g)and ethanol(3.3 g). After evaporating ethanol at 40 oC, a mixture of tetraethylorthosilicate (TEOS, 98 wt%, Aldrich) and tetrabutylorthotitanate (TBOT, 97 wt%, Aldrich) was added with incipient wetness. The molar composition of theresulting synthesis gelwaskept to20 TPA2：TiO2：100 SiO2： 200H2O.The gelwas transferred into aTeflon linercontaining sufficient water to produce saturated steam. The linerwas placedinside a stainless steelautoclave and heatedat 170oC

MesoporousTS-1 for Catalytic OxidativeDesulfurization Bull. Korean Chem. Soc. 2009, Vol. 30, No. 8 1779

for 72 h.After the hydrothermalreaction, the autoclavewas quenched in cold waterand the powder was recovered byfil­ trationand washed with de-ionized water. Thefinal powder was calcined in air at 550 oC for 8 h to removethe organic struc­ ture-direction agent. For comparison, a conventional micro- porous TS-1 with similar titanium content synthesized accord­ ingto Ref. 20.

Characterization.Thecrystallinity of thesamples prepared was measuredby X-ray diffraction usingCuKa radiation (Phillips, PW-1700).Nitrogenadsorption-desorption isotherms weremea­ sured using aMicromeriticsASAP-2020instrumentat77 K.

UV-Visible spectra were obtained underambient conditionsusing dehydrated MgO asa reference in therange of 190- 500 nm on a Varian CARY-3E doublebeam spectrometer. Scanning elec­ tron microscopy(SEM)was performed using HitachiS-4300 electron microscope, and transmission electron microscopy (TEM) measurements were taken on a Philips CM200.

Cat

^겐

(n._쯔m_su--

Figure 1. XRD patterns of (a) TS-1 synthesized with CMK-3, (b) TS-1 synthesized with carbon black, and (c) conventional TS-1.

200 250 300 350 400 450 500

Wavelength (nm)

Figure 2. UV-Vis spectra of (a) TS-1 synthesized with CMK-3, (b) TS-1 synthesized with carbon black, and (c) conventional TS-1.

PPS-2510 fitted with a condenser (Eyela). The reaction was carried out at 318K for 2 h using 32mmolsubstrate, 17 mmol H2O2, 20 mLmethanol,and 0.15 g catalystunder vigorous stirring.

Theproductswereanalyzedbyusingagas chromatography (HP5890 seriesII GC). The concentrationof H2O2 after are­ actionwas determinedby 0.1 M cerium sulfate titration using ferroin indicator.

The ODSreactivity of model sulfurcompounds wasstudied in a 100mLglass batch reactor equipped with a temperature controller, a condenser,andmechanical stirrer. Typically, 48 g of feed(n-heptane 80 wt% andtoluene 20wt%) containing modelsulfur compounds of200 ppmw-S witha molar ratio of tert-butyl hydroperoxide(TBHP)/Sof 2.5 was heated at 80 oC andstirredat 400 rpm followed by addition of 0.12 gof catalyst.

The reaction products wereanalyzed bygaschromatography (YoungLin ACME 6000 GC, Korea) equippedwitha PFPD (Pulsed FlamePhotometric) detector.

Results andDiscussion

X-ray powder diffractionpatterns of thecalcined TS-1 sam­ ples prepared areshown inFig. 1.All theTS-1sampleswere highly crystalline, showingthe identicalcharacteristicpeaks corresponding tothezeoliteMFIstructure.

The UV-Vis spectra of the samples (Ti/Si = 0.01) are shown in Fig. 2. The spectraof the conventional TS-1 sample (c) showan absorption bandaround210nm,which corresponds to tetrahedral Ti(IV)coordination species in the framework.

The absence of ashoulder at 260- 270nm correspondingto oligomericTispecies and330 - 350 nm peak correspondingto anatasephase in the UV-Vis spectrasuggeststhat the sample is essentiallyfree from extra-frameworkTispecies.The TS-1 samples prepared using carbontemplates(a) and(b), on the other hand, exhibited a trace of oligomericTispecies near 260 nm.The UV-Vis spectral quality for the TS-1samplesprepared with carbontemplates deteriorated further with an increasingly visibleshoulderat 260 nm when the samples were prepared withthehigheramount of titaniumin Ti/Si mol ratio of 2(not shown). It seemsthe presence of the heterogeneous carbon

400

-

-

-

-

­

0 0 0 0 0 5

0

5

0

5 3

3

2

2

1

(d is

으

% 으

pw」잉

즒

。>

0.2 0.4 0.6 0.8 1.0

100 0.0

Relative Pressure (P/P0)

Figure 3. Nitrogen adsorption/ desorption isotherms and pore size distribution curves obtained from the desorption branch by the BJH method (inset) of (a) TS-1 synthesized with CMK-3, (b) TS-1 synthesized with carbon black, and (c) conventional TS-1.

1780 Bull. KoreanChem. Soc. 2009, Vol. 30, No. 8 Table 1. Textural properties of the TS-1 samples prepared

Sample Template source

Vmicro Vmeso

(cm3/g)a (cm3/g)^“

Dmeso

(nm)“

Sext

(m2/g)a Sbet

(m2/g)c

meso-TS-1 CMK-3 0.12 0.31 14.7 110 388

meso-TS-1 Carbon

black 0.15 0.14 6.8 106 440

conv-TS-1 - 0.16 0.07 - 121 428

"Calculated

by the

method.

by the

tion).

Calculated by the

method.

Figure 4. SEM images of (a) TS-1 synthesized with CMK-3, (b) TS-1 synthesized with carbon black, and (c) conventional TS-1.

phase in the substratemixtureinhibits the uniformmixingof titaniumandsiliconalkoxideprecursors during the synthesis of TS-1.

Fig. 3shows the nitrogenadsorption-desorptionisotherms of the TS-1 samples afterremoving the carbontemplates by calcination. Theisotherms of TS-1 synthesizedwith CMK-3 (a) and carbon black (b) exhibited ahysteresisloopatrelative pressures higher than P/P^o= 0.4, clearly indicating the existence of mesopores.21 TheBETsurfaceareas,t-plotexternalsurface

Chul- Woo Parket al.

areas, porevolumes andthe BJH average mesoporoussizes are given in Table1. Whilst the microporevolumesestimated by the t-plot methodfor the three TS-1 samples are reasonably closeto one another, TS-1 synthesizedwith CMK-3 has the largestmesoporevolume (0.31 cm3/g) over the conventional TS-1 (0.07 cm3/g) andhasasignificantlylarge averagepore size. TS-1 synthesized with carbon black also has shown a mesoporevolume (0.14 cm3/g) accompaniedbythe average poresize in the mesoporerangeassociated with texturalmeso­ porosity. The pore size distribution curves of the TS-1 catalysts prepared were shown in the inset ofFig. 3. TheTS-1 synthesized with CMK-3 shows a narrow peak centered at 3.7 nm and broadrange of poresdistributedbetween 10 to 80 nm. Onthe otherhand, the TS-1 synthesizedwithcarbonblackexhibited abimodalpore size distribution; one in the range of 1.8 to 3 nmand the other 40 to60 nm. As expected, the conventional TS-1 shows negligible mesopore volume. This analysis of pore size distributionshowsthatmesopores are successfully createdbyremoving the carbontemplatesbycalcination, but asmallportion of macropores were also detected.

Fig.4 shows thescanning electron microscopy (SEM) images ofTS-1 samples after removalof the carbon templates. All TS-1 samplesappear to behighly crystalline with no amorphous phasesdetectedontheirexternalsurfaces. The crystal sizes of TS-1 catalystsasdeterminedfromSEM were ca. 0.2 - 0.3 gm formesoporousTS-1with CMK-3,0.2 -1.2 gm for mesoporous TS-1with carbon black, and 0.3 - 0.4gm for conventional TS-1, respectively. It is surprising that we could have TS-1 with smaller particle size than the conventional sample usingCMK-3 templates,which willprovebeneficialinliquid phase catalytic reactionsinvolving large organic molecules dueto theirshorter diffusion path. Weare speculating that the CMK-3 particles could have acted asa seed for TS-1 nucleation.TS-1with carbon black, on the otherhand,producedabimodaldistribution of particle size, which maybe aconsequenceof the relativelywide particlesizedistributionin the carbon black, N762(Degussa) asmentioned earlier.

Transmission electron microscopy (TEM) was attempted to visualize the actualnanoporous structure ofthe TS-1 sam­ plesprepared, and Fig. 5shows the correspondingTEM images.

Whilst it was difficult to make a clear observation of the mesopores created by carbon template as was reported by others,16,17wecould locate several grey spotsof irregular meso­ pores in the TS-1 samplessynthesizedwithcarbontemplates in (a) and (b), which were absent in the conventional TS-1 sampleshown in (c).

Epoxidation of allylchloride (ALC)was chosenas aprobe reaction to comparethe catalytic propertiesof the TS-1samples prepared; ALC is a sufficiently small organicmolecule that can be accommodated inside the 10-memberedringpores of the conventional microporous TS-1. As shown in Table 2, conventionalTS-1 showed thehighestALC conversion (21.7%) and mesoporous TS-1 catalysts exhibited a little lower ca­ talytic activitiesthanconventional TS-1; TS-1 prepared with CMK-3 (18.7% ALC conversion) was somewhat better in catalyticperformancethan theTS-1 prepared with carbonblack (17.5%). We believe the differences in catalyticactivitiesare causedby theminordifferences inTi states mentioned inFig. 2

Mesoporous TS-1 for Catalytic Oxidative Desulfurization Bull. Korean Chem. Soc. 2009, Vol. 30, No.8 1781

Figure 5. TEM images of (a) TS-1 synthesized with CMK-3, (b) TS-1 synthesized with carbon black, and (c) conventional TS-1 (The circles in figures (a) and (b) indicate the regions of meso/macro pores).

Table 2. Catalytic activities of the TS-1 samples prepared in allylchlo­

ride epoxidation using H²O²

Cl

+ HO-OH $ A Cl^/1\

Sample Template ^C ^ECH

source ALC ECH H2O2 H2O2

Conversion Selectivity Conversion Selectivity

(%) (%) (%) (%)

meso-TS-1 CMK-3 18.7 > 99% 35.8 98.3

meso-TS-1 Carbon

black 17.5 > 99% 34.7 95.0

conv-TS-1 - 21.7 > 99% 41.2 99.9

(汶

」으 )_

은 으

\ |

」

|-

뜨 ( _

Figure 6. Catalytic oxidation of DBT using TBHP as an oxidant by (a) TS-1 synthesized with CMK-3, (b) TS-1 synthesized with carbon black, and (c) conventional TS-1.

earlier, since itisknown that tetrahedral Ti sitesbeingmore catalyticallyactive than the oligomericonesin TS-1. AllTS-1 catalysts showed satisfactory H2O2 selectivities; 98.3% for TS-1 synthesizedwith CMK-3, 95.0% for TS-1 synthesized withcarbon black, and 99.9%for conventionalTS-1. Product selectivityto epichlorohydrin (ECH) was closeto 100% with only trace amounts of side productssuch as dichlorohydrin or 3-chloro-1,2-propandioldetected.

Finally, oxidativedesulfurization of DBT and 4,6-DMDBT with tert-butyl hydroperoxide as oxidizing agent over the TS-1 catalysts were conductedandthe results are shownin Fig. 6and Fig.7, respectively.As shown inthefigures,substan­ tialenhancement in catalytic activitywas achieved in meso- porous TS-1 catalysts comparedwithconventional TS-1.More specifically,catalyticODS activities of thecatalystsvaried in the following order: conventional TS-1 <<mesoporous TS-1 with carbonblack< mesoporousTS-1 with CMK-3 for both DBT and 4,6-DMDBT.These resultscanbe understood when we realize that the molecular dimensions of DBT (8.0 x 12.2 A22)and 4,6-DMDBT (8.7 x 12.2 A23)aresignificantlylarger than the pores of conventional TS-1 zeolite (5.6x 5.3 A). There­ fore, it is reasonableto assume that molecules likeDBTand 4,6-DMDBT cannot have an accessto theredox-activesites located inside the pores of conventional TS-1,and theimproved catalytic activities of the mesoporousTS-1 catalystsmustbe attributedto their improved diffusion propertiescompared with conventionalTS-1. In addition, mesoporous TS-1 prepared

(汶

)

으 은

(D>u oo

Figure 7. Catalytic oxidation of 4,6-DMDBT using TBHP as an oxi­

dant by (a) TS-1 synthesized with CMK-3, (b) TS-1 synthesized with carbon black, and (c) conventional TS-1.

using CMK-3 produced better catalytic performances than the one prepared using carbon black, since the former has the larger mesopore volumes as well as the largeraverage pore diameters in TS-1. In addition, smaller particle size of the

1782 Bull.KoreanChem. Soc. 2009, Vol. 30, No.8 Chui- Woo Parket al.

TS-1prepared with CMK-3 willalsobe usefulfor the reactions duetothe shorter diffusion pathof thereactants.

As shown in Fig. 6 and Fig. 7, conventional TS-1 still pro­ ducedsignificantconversions of 29.6% for DBT and 22.7%

for 4,6-DMDBT after 3 h reaction. Apparently,ODSreaction can also take placeat the catalyst externalsurfacesas well as insidethepores of TS-1 catalysts.

Conclusions

Mesoporous TS-1 catalysts were prepared using CMK-3 and carbon black as templates. Thesynthesizedmesoporous TS-1 catalysts were foundtoexhibit substantially enhanced catalyticactivitiesthan a conventionalmicroporous TS-1 ca­

talyst in theoxidative desulfurization (ODS) of bulky sulfur compoundsby providing easy access of the reactantstothe active redox sites.CMK-3 was provenmoreeffective togene­ ratemesopores in TS-1 and alsoresulted insmallerparticle size useful in liquidphaseODS reactions.

Acknowledgments. This work wassupportedby Inha Uni­ versity (2009). Experimentalsupportby Dr. K. E. Jeong from KRICT is greatlyappreciated.

References

1. US EPA, regulatory announcement: Heavy-Duty Engine and Vehicle Standards and Highway Diesel Fuel Sulfur Control Requirements, December 2000.

2. Liotta, F. J.; Han, Y. Z. Production of Ultra-Low Sulfur Fuels by Selective Hydroperoxide Oxidation； NPRA AM-03-23, Washing­

ton, DC, March 23-25, 2003.

3. Rothlishberger, A.; Prins, R. J. Catal. 2005, 235, 229.

4. Lu, H.; Gao, J.; Jiang, Z.; Jing, F.; Yang, Y; Wang, G.; Li, C. J.

Catal. 2006, 239, 369.

5. Palomeque, J.; Clacens, J. M.; Figueras, F. J. Catal. 2002, 211, 103.

6. Otsuki, S.; Nonaka, T.; Takashima, N.; Qian, W. Energy Fuels 2000, 14,1232.

7. Jeong, K. E.; Chae, H. J.; Kim, U. C.; Jeong, S. Y.; Ahn, W. S.

Solid State Phenom. 2008, 135, 89.

8. Hulea, V; Fajula, F.; Bousquet, J. J. Catal. 2001, 198,179.

9. Kong, L.; Li, G.; Wang, X. Catal Today 2004, 93, 341.

10. Jin, C.; Li, G.; Wang, X.; Wang, Y; Zhao, L.; Sun, D.

MicroporousMesoporousMater. 2008, 111, 236.

11. Kong, L.; Li, G.; Wang, X. Catal. Lett. 2004, 92, 163.

12. Jacobsen, C. J. H.; Madsen, C.; Houzvicka, J.; Schmidt, I.;

Carlsson, A. J. Am. Chem. Soc. 2000, 122, 7116.

13. Kustova, M. Y.; Hasselriis, P.; Christensen, C. H. Catal. Lett.

2004, 96,205.

14. Wei, X.; Smirniotis, P. G. Microporous Mesoporous Mater.

2006, 89, 170.

15. Egeblad, K.; Kustova, M.; Klitgaard, S. K.; Zhu, K.; Christensen, C. H. MicroporousMesoporousMater. 2007, 101, 214.

16. Schmidt, I.; Krogh, A.; Wienberg, K.; Carlsson, A.; Brorson, M.;

Jacobsen, C. J. H. Chem. Commun. 2000, 2157.

17. Fang, Y.; Hu, H. Catal. Commun. 2007, 8, 817.

18. Zhao, D.; Huo, Q.; Feng, J.; Chmelka, B. F.; Stucky, G. D. J. Am.

Chem. Soc. 1998, 120, 6024.

19. Jun, S.; Joo, S. H.; Ryoo, R.; Kruk, M.; Jaroniec, M.; Liu, Z.;

Ohsuna, T.; Terasaki, O. J. Am. Chem. Soc. 2000, 122, 10712.

20. Serrano, D. P; Uguina, M. A.; Ovejero, G.; Grieken, R. V.;

Camacho, M. MicroporousMater. 1996, 7, 309.

21. On, D. T.; Lutic, D.; Kaliaguine, S. Microporous Mesoporous Mater. 2001, 44, 435.

22. Daage, M.; Chianelli R. R. J. Catal. 1994, 149, 414.

23. Meng, X.; Wu, Y; Li, Y. J. Porous Mater. 2006, 13, 365.