Zirconium Containing Al-MCM-41 – Synthesis, Characterisation and Catalytic performance in 1-Hexene Isomerisation I.Eswaramoorthia* , V.Sundaramurthya and N.Lingappanb a

Department of Chemistry, Anna University, Chennai-600025, India.

b

Science and Humanities Division, Madras Institute of Technology, Chennai. India

Zirconium containing Al-MCM-41 samples with different Si/Zr ratio were synthesised and characterised by XRD, DRS, BET surface area, FT-IR and pyridine adsorption studies. The higher value of d-spaceing of Zr-Al-MCM-41 samples compared with Al-MCM-41 indicates the incorporation of Zr in the framework. Further, presence of Zr (IV) in the tetrahedral co-ordination is confirmed by absorption band around 210 nm in diffuse reflectance spectra. The FT-IR studies of pyridine adsorbed Zr-Al-MCM-41 show that they have higher acidity than Al-MCM-41 sample used for comparison. 1-Hexene isomerisation over Zr-Al-MCM-41 samples, Al-MCM-41 and zirconium impregnated Al-MCM-41 was carried out at temperatures 250, 300 and 350ºC. At all temperatures higher conversion of 1-hexene and higher selectivity of skeletal isomerised products were observed when higher is the Zr content in the framework of Al-MCM-41 catalyst.

1. INTRODUCTION Heteroatom substituted mesoporous molecular sieves have attracted much interest due to their potential use as hosts, adsorbents and catalysts (1). Like microporous materials, substitution of heteroatoms such as V, Cr, Ti in MCM-41 has been reported (2,3). Ratnasamy et al (4) made an attempt to incorporate zirconium into the silicalite framework. They used Zr-silicalite for hydroxylation of phenol with H2 O2 and obtained poor yield. They accounted for the poor yield in terms of strains in the crystalline framework due to the incorporation of larger size Zr cation (larger compared to aluminium). Tuel and co-workers (5) synthesized Zr containing mesoporous silicas with Si/Zr ratios in the range of 15-1000. They hydrolysed a alcoholic solution containing TEOS and Zr-isopropoxide in the presence of hexadecylamine. The Zr mesoporous silicas obtained had surface area, pore volume, and average pore diameter in the ranges 865-1049 m2 /g, 0.6-0.72 cm3 /g, and 3.3-3.8 nm respectively. Further they reported that the Zr content does not influence the crystal properties. However they used the catalyst for the oxidation of large substrates such as aniline cyclohexane norborylene and 2,6 DTBP using both H2 O2 and tert-butyl hydroperoxide separately and found that increase of Zr content increases the activity and selectivity of the catalysts. Jones et al (6) reported the synthesis of mesoporous zirconium silicates from gel of Si/Zr ratio of 50, 25 and 5 and the samples had surface area of 1302, 1386, 973cm3 /g and an average pore diameter of 2.8, 2.5, 2.8 nm respectively. Further they reported that the IR spectra of samples calcined and -----------------------------* Corresponding author, Email: [email protected]

subsequently adsorbed with pyridine showed the band characteristic of Lewis acid sites only, with a linear correlation between the Zr content and acid site density. Ramasamy et al (7) incorporated Zr into Al- β framework and reported that Zr incorporation generates additional acidity due to the strong polarization of possible Si-Oδ-----Zrδ+ linkages. They used the Zr-Alβ catalyst for m-xylene isomerisation and found that the activity is higher than that of Al- β and Zr impregnated Al- β catalysts. Further they explained that higher activity of Zr-Al- β is due to strengthening of Bronsted acid sites (due to Al3+ ion) by Zr4+ Lewis acid sites. No report is available in literature on the incorporation of zirconium in mesoporous Al-MCM-41 framework till date. Hence in the present study we report the synthesis, characterisation of Zr containing Al-MCM-41 and its catalytic behaviour in 1-hexene isomerisation. 2. EXPERIMENTAL 2.1. Catalyst Preparation The Zr-Al-MCM-41 samples using initial gel of Si/Zr ratio of 50, 100, and 200 were synthesized under hydrothermal conditions. The gels had the following composition. SiO 2 : 0.12 CTAB: x ZrO 2 : 0.01Al2 O3 : 0.19Na2 O: 35H2 O, where x=0.02-0.005 and CTAB stands for Cetyl Trimethyl Ammonium Bromide, a structure directing agent. Ethyl silicate-40 (ES-40), zirconyl chloride and aluminium sulphate were used as sources of Si, Zr and Al respectively. For comparison purpose Al-MCM-41 (Si/Al=50) without Zr and 0.1 wt. % Zr impregnated Al-MCM-41 were prepared. The as-synthesized samples were calcined in the flowing N2 for 1 h and in air for 6 h at 550ºC. The calcined materials were converted into their ammonium form by repeated ion exchange with 1M NH4 Cl solution. H-form of the each catalyst was obtained by calcination at 450ºC for 5-6 h. The catalyst Al-MCM-41 free from Zr, and three Zr- Al-MCM-41 obtained from gel of Si/Zr ratio 200,100 and 50 are designated as catalyst 1, 2, 3, and 4 respectively. 2.2. Characterisation The powder X-ray diffraction patterns of calcined Zr-Al-MCM-41 samples were recorded on a Rigaku X-ray diffractometer in the scan range of 2θ between 1° and 10° using Cu/Ni 40 KV/30 mA radiation. The IR spectra of the samples were recorded on a Nicolet AVATAR 360 FT-IR Spectrophotometer by KBr pellet technique. The co-ordination of zirconium in the framework was studied by diffuse reflectance spectroscopy (DRS). The surface area measurements of the all Zr-Al-MCM-41 and Al-MCM-41 samples were carried out in Sorptomatic 1990 CE instrument following BET procedure using N2 as adsorbent at liquid nitrogen temperature. The acidity of H-form of the catalysts was studied by pyridine adsorption. The pyridine adsorption was carried out in a quartz tube, which is packed with 0.5 g of each catalyst samples. The samples were first activated at 500°C for 4 h and were subjected to high vaccum (1.5 ×10-3 Torr) conditions. Then the system was cooled under vacuum to 150°C. The pyridine was passed through to permeate the samples. Then the FT-IR spectra of pyridine adsorbed samples were recorded on a Nicolet AVATAR FT-IR Spectrophotometer using KBr pellet technique. 2.3. Catalytic Studies 1-Hexene isomerisation was carried out over the H-form of Zr-Al-MCM-41, Al-MCM-41 and Zr impregnated Al-MCM-41 catalysts in a fixed bed continuous down flow quartz reactor at 250, 300 and 350°C with WHSV of 3 h-1.The products were collected at the time interval of 1 h. The products were analysed by Gas Chromatography equipped with FID and further confirmed by GC-Mass Spectrometry.

3. RESULTS AND DISCUSSION 3.1. Characterisation The powder XRD patterns of catalysts 1-4 exhibit (Figure 1) a major peak at 2θ~2º along with three small peaks due to 110, 200, and 210 plane reflection lines. These four reflection lines are generally indexed for the hexagonal unit cell. In the case of catalysts 2-4 the d-spaceings are found to increase with increasing Zr content (catalyst 2, 3 and 4 have d-spaceings 45.3, 46.0, 47.2 respectively), which is in agreement with the general trend that insertion of a metal ion larger than Si brings about an increase in the unit cell parameters because of the larger M-O bond distance. The higher values of d-spaceing on comparision with that of catalyst 1 (d-spaceing of Al-MCM-41 is 44) indicates the presence of Zr in the Zr-Al-MCM-41 framework. The diffuse reflectance spectra of catalysts 2-4 are shown in Figure.2. All the samples exhibit single narrow band around 210nm confirming the presence of Zr (IV) in the framework of Al-MCM-41 (8). This type of absorption in UV region may be attributed to the ligand to metal charge transfer involving isolated Zr (IV) atoms in the tetrahedral coordination. Catalysts 1-4 have BET surface area 958, 951, 955, and 948 m2 /g respectively and exhibit type IV isotherm with characteristic step at P/P 0 ~0.38 confirming the mesoporous nature of the samples with narrow pore size distribution. The FT-IR spectra of catalysts 1-4 are shown in Figure 3 and they are similar to that of amorphous silica (9). The FT-IR spectra of pyridine adsorbed samples are shown in Figure 4. All the samples gave bands similar to that reported by Mokaya et al (10) for both Bronsted and Lewis acid sites. The absorptions at 1547 cm-1 and 1633 cm-1 are associated with pyridine adsorbed at Bronsted acid sites of framework Al and absorptions at 1445 cm-1 and 1625 cm-1 are attributed to pyridine adsorbed on Lewis acid sites. And absorption at 1490 cm-1 is supposed to be due to the superposition of bands of pyridine adsorbed at Lewis and Bronsted acid sites.This observation is looking similar to that of Miller et al (11). The intensity of the all peaks are found to increase with increasing Zr content indicating that the acidity of the catalysts increases proportionate to Zr content. Thus it offers itself as a proof to the report of Ramasamy et al (7), which states that the increase in acidity of the catalyst is due to the increasing polarization of possible Si-Oδ----Zrδ+ linkages.

Fig.1. XRD patterns of Catalysts 1-4

Fig.2. Diffuse reflectance spectra of Catalysts 2-4

Fig.3. FT-IR Spectra of calcined catalysts 1-4

Fig.4. FT-IR Spectra of pyridine adsorbed catalysts 1-4

3.2. Catalytic studies The catalysts 1-4 in their H-form were subjected to 1-hexene isomerisation at the temperatures 250, 300 and 350°C with WHSV of 3 h-1. The conversion of 1-hexene over the catalysts 1-4 are presented in Figure 5. 1-hexene conversion was found to increase linearly with increase of reaction temperature as well as the Zr content of the catalysts. Further it is seen that the conversion of 1-hexene remains higher over the Zr containing catalysts (catalysts 2-4) compared to that over Zr free Al-MCM-41 (catalyst 1) at all the temperatures. For example at 250ºC, 65.0, 68.3, 73.0 and 77.0 wt.% conversions are observed over catalysts 1-4 respectively. The increase of 1-hexene conversion with increasing Zr content is attributed to the enhancement of acidity of Zr-Al-MCM-41 by incorporation of Zr4+ ion in the framework lattice. It is presumed that Zr4+ Lewis acid sites strengthen the Bronsted acid sites (due to Al3+ ions) and hence influence the isomerisation of 1-hexene. This is similar to the observation reported by Ramasamy et al (7) in the Zr incorporated β-zeolite study where they found

100

Conversion (wt%)

90

80

70

60

Catalyst-1

Catalyst-2

Catalyst-3

Catalyst-4

50 250

300

350

Temperature (ºC )

Fig.5. Conversion of 1-hexene over catalysts 1-4 at different temperatures increase in isomerisation of m-xylene with increase in Zr content of β-zeolite. The observation that increase in Lewis acid sites with increasing Zr content is supported by FT-IR studies of pyridine adsorbed samples.

Table 1 shows the product selectivity in 1-hexene isomerisation at the reaction temperatures 250, 300 and 350ºC. The isomerised products such as trans-2-hexene (t-2H), cis-2-hexene (c-2H) and (cis+trans)-3-hexene (c+t)-3H are formed by 1-2 double bond shift and 2-methyl-2-pentene (2M2P), 3-methyl-2-pentene (3M2P), 4-methyl-2-pentene (4M2P), 2-methyl-1-pentene (2M1P), 3-methyl-1-pentene (3M1P) are formed by skeletal isomerisation. The products distribution is similar to that in the report of 1-hexene isomerisation over B-ZSM-5 (12). The selectivity of double bond shift products over catalysts 2-4 are found to be decreasing with increase of Zr content as well as increase of temperature and remains lower always with respect to catalyst 1. For example at 300°C, 95.9, 92.7, 86.3 and 77.4 % are the total double bond shift products formed over catalysts 1-4 respectively. Due to increasing Zr content in the framework, which in turn increases the strength of Bronsted acid sites of Al3+, more of skeletal isomerised products are forming. This is again similar to the observation of Ramasamy et al (7) in their m-xylene isomerisation to o- and pproducts over Zr containing Al- β catalyst. As a consequence of formation of more skeletal products from 1-hexene, double bond shift products show a fall in their yield. Among the double bond shift products, the formation of 2-hexenes remains always higher compared to 3-hexenes because formation of 3-hexene is kinetically controlled. The kinetic control is

supposed to be due to the shift of carbenium ion to the adjacent position followed by deprotonation. More of t-2H with respect to c-2H is due to the more stability of the trans species than cis isomer though both form by deprotonation of C6 carbenium ion. Table 1 Product selectivity in 1- Hexene isomerisation on Al-MCM-41 (catalyst-1) and Zr-Al-MCM-41 (catalysts 2-4) WHSV=3 h-1

Product t-2H c-2H (c+t)-3H Total 2M2P 3M2P 4M2P 2M1P 3M1P Total Cracked products

Time =1h

Product selectivity on catalysts at 250º C 1 2 3 4 51.0 50.5 49.3 48.0 29.1 29.0 28.5 27.3 18.7 18.2 17.5 16.4 98.8 97.7 95.3 91.7 1.2 1.8 2.7 4.2 0.5 1.4 2.6 0.6 1.5 1.2 2.3 4.7 8.3 -

-

-

-

Weight of catalyst=1.0 g

Product selectivity on catalysts at 300ºC 1 2 3 4 48.0 46.1 45.0 41.3 28.1 26.3 24.0 20.4 19.8 20.3 17.3 15.7 95.9 92.7 86.3 77.4 2.8 4.2 7.1 10.0 1.3 1.9 2.4 4.0 1.2 1.1 2.2 1.1 3.3 2.2 4.1 7.3 11.7 21.6 -

-

-

1.0

Product selectivity on catalysts at 350ºC 1 2 3 4 37.3 33.1 31.7 29.0 20.6 15.5 13.1 10.2 19.9 11.6 9.9 8.0 77.8 60.2 54.7 47.2 10.6 13.0 12.4 12.9 3.1 9.9 9.7 10.0 2.2 3.9 3.9 3.7 2.7 5.5 6.8 7.1 2.2 2.2 3.5 4.3 20.8 34.5 36.3 38.0 2.0

5.6

9.0

14.7

The selectivity of skeletal isomerised products increases with increase of reaction temperature as well as Zr content of the catalysts. For example, at 300°C, 4.1, 7.3, 11.7 and 21.5 % total skeletal isomerised products are obtained over the catalysts 1-4 respectively. When temperature increases from 250°C to 300°C and then to 350°C, 8.3, 21.6 and 38 % are the total skeletal isomerised products formed over catalyst 4 and similar trend is observed over the other catalysts. The increase in skeletal isomerised products over all the catalysts with temperature is attributed to the activation of Al Bronsted acid sites (with or without Zr nearby) with temperature. The increase of skeletal isomerisation with increasing Zr content in the framework at any fixed temperature is attributed to the strengthening effect of Zr on Al Bronsted acidity. Thus more is the Zr content in the framework more is the number of Al sites strengthened with respect to its acidity leading to more skeletal isomerised products. The skeletal isomerised products are formed by a sequence of hydrogen and methyl shifts. These are energy demanding steps and need strong acidity of the catalysts. It is further observed from Table 1, that the amount of cracked products, though very less, (may be due to the subsequent decomposition of skeletal isomerisation products) increases with increasing Zr content and temperature. For example, at 350°C, 2.0, 5.6, 9.0 and 14.7 % are the total cracked products observed over the catalysts 1-4 respectively. Similarly 0, 1, and 14.7 % are the cracked products observed, as a typical case, at 250, 300 and 350°C respectively over the catalyst 4. This may be due to increase in cracking of skeletal isomerisation products that form in considerable amount with increase of temperature and acidity of catalyst.

4. CONCLUSION Using the molar gel composition of SiO 2 : 0.12 CTAB: x ZrO 2 : 0.01Al2 O3 : 0.19Na2 O: 35H2 O, where x = 0.02-0.005, zirconium containing Al-MCM-41 samples were synthesised and characterised by XRD, DRS, BET surface area, FT-IR and pyridine absorption studies. The increase of d-spaceing with increase of Zr content indicates the incorporation of Zr in the framework. In diffuse reflectance spectra, an absorption band around 210nm confirms the presence of Zr (IV) in the tetrahedral co-ordination. The pyridine adsorbed FT-IR studies shows that the increasing incorporation of Zr in Al-MCM-41 increases the acidity of Zr-AlMCM-41. Further the acidity enhancement is supported by the increase in selectivity of skeletal isomerised products in 1-hexene isomerisation. Zirconium free Al-MCM-41 and zirconium impregnated Al-MCM-41 show lower 1-hexene conversion and lower selectivity of skeletal isomerised products than that of Zr-Al-MCM-41 catalysts. This confirms the presence of Zr in the framework of MCM-41 and accounts for the increase in acidity. Hence the incorporation of Zr in Al-MCM-41 framework facilitates the formation of skeletal isomerised products there by decreasing considerably double bond shift products. ACKNOWLEDGEMENT The authors gratefully acknowledge the financial support from the Department of Science and Technology, New Delhi. REFERENCES 1. 2.

X.S. Zhao, G. Q. Lu and G.J. Millar, Ind. Eng. Chem. Res., 35 (1996) 2075. K..M. Reddy, I. Moudrakovski and A.Sayari, J. Chem. Soc. Chem. Commun., (1994) 1059. 3. A. Corma, M.T. .Navarro and J. Perez Pariente, J. Chem. Soc. Chem. Commun., (1994) 147. 4. M. K. Dongare, P. Singh, P.P. Moghe and P. Ratanasamy, Zeolites, 11 (1991) 690. 5. A. Tuel,S. Gontier and R.Teissier, J. Chem. Soc. Chem. Commun., (1996) 651. 6. R. Mokaya and W. Jones, Chem. Commun., (1996) 983. 7. B. Rakshe, V. Ramasamy and A.V. Ramasamy, J. Catal., 188 (1999) 252. 8. G. R. Wang, X. Q.Wang, X.S. Wang and S.X. Yu, Stud. Surf. Sci. Catal., 83 (1994) 67. 9. Sophie Biz and Mario. L.Occelli, Catal. Rev- Sci. Eng., 40 (3) (1998) 329. 10. R. Mokaya, W. Zones, Z. Luan, M. D. Alba and Klinowski, Catal. Lett., 37 (1996) 113. 11. T. M. Millar, and V.H . Grassian., Catal. Lett., 46 (1997) 213. 12. V. Sundaramurthy and N. Lingappan, J. Mol. Catal., 160 (2000) 367.