Iron incorporated heterogeneous catalyst from rice husk ash

Journal of Colloid and Interface Science 304 (2006) 137–143 www.elsevier.com/locate/jcis

Iron incorporated heterogeneous catalyst from rice husk ash Farook Adam ∗ , Kalaivani Kandasamy, Saraswathy Balakrishnan School of Chemical Sciences, Universiti Sains Malaysia, 11800 Penang, Malaysia Received 5 April 2006; accepted 27 August 2006 Available online 1 September 2006

Abstract Silica supported iron catalyst was prepared from rice husk ash (RHA) via the sol–gel technique using an aqueous solution of iron(III) salt in 3.0 M HNO3 . The sample was dried at 110 ◦ C and labeled as RHA-Fe. A sample of RHA-Fe was calcined at 700 ◦ C for 5 h and labeled as RHAFe700. X-ray diffraction spectrogram showed that both RHA-Fe and RHA-Fe700 were amorphous. The SEM/EDX results showed that the metal was present as agglomerates and the Fe ions were not homogeneously distributed in RHA-Fe but RHA-Fe700 was shown to be homogeneous. The specific surface areas for RHA-Fe and RHA-Fe700 were determined by BET nitrogen adsorption studies and found to be 87.4 and 55.8 m2 g−1 , respectively. Both catalysts showed high activity in the reaction between toluene and benzyl chloride. The mono-substituted benzyltoluene was the major product and both catalysts yielded more than 92% of the product. The GC showed that both the ortho- and para-substituted monoisomers were present in about equal quantities. The minor products consisting of 16 di-substituted isomers were also observed in the GC–MS spectra of both catalytic products. The catalyst was found to be reusable without loss of activity and with no leaching of the metal. © 2006 Elsevier Inc. All rights reserved. Keywords: Rice husk ash; Heterogeneous catalyst; Iron supported catalyst; Sol–gel; Friedel–Craft benzylation

1. Introduction Rice husk is a major waste product of the rice milling industry. The Malaysian Ministry of Agriculture’s statistic [1] shows that approximately one million ton of rice husk was generated in 1994. Rice husk is composed of 20% ash, 38% cellulose, 22% lignin, 18% pentose, and 2% of other organic components [2]. The silica, SiO2 , content [3] of the ash is more than 94%. Silica is known to be the main precursor in the ceramic industry. Guha and Roy [4] have prepared silica gel from rice husk ash (RHA) for gas chromatography. The silica gel from rice husk ash was comparable with commercial silica gel. Lewis acid, metal cation-exchange clays, metal salt-impregnated clays and pillared clays have been used for Friedel–Craft reactions [5]. Very recently a newer type of metal salt–metal ion exchanged montmorillonite have been prepared by Phukan et al. [6], and evaluated as solid acid catalysts for Friedel–Craft reactions particularly for benzylation of benzene. In their work, they have reported that only certain supported * Corresponding author. Fax: +60 4 6574 854.

E-mail addresses: [email protected], [email protected] (F. Adam). 0021-9797/$ – see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.jcis.2006.08.051

metal ion or metal salt on montmorillonite clay may play a key role as solid acid catalysts. Recently Shrigadi et al. [5] have reported the use of iron oxyhydroxides and oxides in the Friedel–Craft benzylation reaction using benzyl chloride and benzyl alcohol. These catalysts showed very good catalytic activity. The activity of such a catalyst depends on various factors, notably the nature of the metal cation and the activation temperature. Besides this, the catalytic activity and selectivity of a catalyst are mainly governed by the size of the metal particles and its surface properties. Many researchers are currently engaged in the design and synthesis of environmental friendly catalyst. In this regard heterogeneous catalyst needs special mention and is being developed at a tremendous pace judging by the number of papers appearing in the literature. Selection of environmentally friendly catalyst should also take into account its reusability, ease of product recovery, low temperature and leaching effect of the metal if they are present, which may cause problems during the purification of the products. As we know, over the years, silica powders and gels have been widely used in industry as fillers, adsorbents, chromatographic agents, catalyst and as catalytic supports. Adam et al. [7,8] have reported a series of metal sup-

138

F. Adam et al. / Journal of Colloid and Interface Science 304 (2006) 137–143

ported on rice husk ash, which could find application in many catalytic reactions. In previous work the author has studied the use of rice husk ash as an adsorbent for the removal of free fatty acids in palm oil manufacturing and palm oil products [9–11]. In this paper we report a systematic study on the synthesis and characterization of a catalyst from agricultural waste specifically rice husk ash and its use in the benzylation reaction of toluene with benzyl chloride. The incorporation of the iron into the rice husk silica matrix was carried out using the sol–gel technique. Chang et al. [12,13] had published the use of rice husk ash as a matrix for metal supported catalyst. In both cases the incipient-wet method was used to incorporate the metal into the rice husk ash silica matrix. Prasetyoko [14] described the preparation and catalytic activity of silica–alumina from rice husk ash. This work is an attempt to further advance the use of rice husk ash silica from the agricultural waste. Herein we report the synthesis of an environmental friendly catalyst and its catalytic activity that can be used to replace existing processes for the manufacture of fine chemicals.

2.3. Reaction

2. Experimental

Fig. 1 shows the FTIR spectra of the RHA, RHA-Fe, and RHA-Fe700. The broad band in the range 3430–3480 cm−1

Toluene (80 mL, Analar, 99.5%) and benzyl chloride (5.2 mL, Merck, 99%) was mixed and used as the stock solution. Toluene was present in large excess. Each reaction was carried out in a 50 mL round bottom flask equipped with a reflux condenser and magnetic stirrer. The catalyst (0.1 g) was weighed directly into the flask. The stock solution (20.0 mL) was pipetted into this flask. The content was stirred for 3 h in normal atmosphere at various temperatures up to the boiling point of toluene. The reaction mixture was filtered to remove the catalyst and the resultant filtrate was analyzed by using GC–MS (Trace MS, Italy, with Supelcowax GC column, 30 m). The temperature program for the GC analysis was as follows: initial temperature: 90 ◦ C; initial time: 0 min; Ramp 1: 4 ◦ C/min; Temp 2: 175 ◦ C; Ramp 2: 15 ◦ C/min; final temperature: 240 ◦ C; detection: mass spectroscopy. The reaction can be illustrated as shown in Scheme 1. 3. Results and discussion

2.1. Sample preparation Rice husk from a local rice mill, Leong Guan Sdn. Bhd., Seberang Perai Utara, Penang, Malaysia, was washed with copious amount of water and rinsed with distilled water to wash away all the dirt and mud. It was then dried in air for at least 48 h. The rice husk was then pyrolyzed in a muffle furnace at 700 ◦ C for 5 h to produce white silica powder. The collected silica powder was treated with 1.0 M HNO3 for 24 h. It was then filtered and washed thoroughly with deionized water until a constant pH value was obtained. This was labeled as RHA. The RHA was subsequently dried in an oven at 110 ◦ C overnight. 5.0 g of RHA was added into 250 mL of 6.0 M NaOH, stirred for 12 h and the sodium silicate solution was filtered to remove undissolved material. The filtrant was titrated with 3.0 M HNO3 which contained 10 wt% Fe3+ [3.62 g Fe(NO3 )3 ·9H2 O was dissolved in 200 mL of 3.0 M HNO3 ]. The acid was added via a burette at a fast flow rate until pH 8 was reached. The titration rate was reduced to drop by drop until a final pH 5 was obtained. A soft gel was formed which was aged for 4 days. The gel was filtered through suction filtration and washed with distilled water. It was dried in an oven at 110 ◦ C for 24 h. The product was labeled as RHA-Fe. One half of the RHA-Fe was then calcined at 700 ◦ C for 5 h to produce RHA-Fe700.

Scheme 1. The expected products from the Friedel–Craft reaction between toluene and benzyl chloride.

2.2. Sample characterization The prepared catalysts were characterized by FTIR (Perkin– Elmer System 2000) spectroscopy, BET surface area analysis (Micromeritics Instrument Corporation Model ASAP 2000, Norcross), powder X-ray diffractometry (Siemens diffractometer D5000, Kristalloflex), scanning electron microscopy (SEM) (Leica Cambridge S360) and energy dispersive spectrometry (EDX) (Edax Falcon System).

Fig. 1. The FTIR spectra of RHA, RHA-Fe, and RHA-Fe700.

F. Adam et al. / Journal of Colloid and Interface Science 304 (2006) 137–143

is due to the stretching vibration of O–H bond. This band is due to the silanol OH groups and the adsorbed water bound to the silica surface. The band at 1636 cm−1 is due to the bending vibration of the water molecules, which are trapped in the matrix of the silica. The strong peak at 1095 cm−1 is due to the structural siloxane bond, Si–O–Si. This peak is observed in both RHA and metal incorporated RHA. The band at 800–803 and 466–470 cm−1 in all spectra are due to the deformation of Si–O bond. The RHA-Fe showed an extra band at 1384 cm−1 which is characteristic for the NO− 3 group [15,16]. In RHA-Fe the shoulder at 960 cm−1 is due to the Fe–O–Si vibration [17]. The intensity of this shoulder is decreased on calcination, as evident in RHA-Fe700. Measurements of specific surface area were based on the Brunauer–Emmett–Teller (BET) method. RHA-Fe700 shows lower specific surface area, pore volume and a higher average pore diameter than RHA-Fe. This is because calcination at 700 ◦ C results in the inter-particle condensation of free hydroxyl group, which is combined with the rearrangement of

Fig. 2. The nitrogen adsorption isotherm for RHA-Fe700. (F) is the adsorption branch, and (P) is the desorption branch.

139

Table 1 BET data of RHA-Fe and RHA-Fe700 Sample

BET surface area (m2 /g)

Micropore volume (cc/g)

Average diameter of pore (Å)

RHA-Fe RHA-Fe700

87.40 55.83

0.0081 0.0017

54.89 110.61

silica globules to produce a more stable configuration. This phenomenon is highly favored by smaller pores, resulting in faster collapse of pore structure [18]. Calcination closes the narrower pores, effectively decreasing the specific surface area. The BET analysis data for the RHA-Fe and RHA-Fe700 are summarized in Table 1. The specific surface area of RHA-Fe is greater than RHA-Fe700. The condensation of adjacent silanol groups at the pore openings will form siloxane bonds that essentially close the narrower pores [18,19]. These narrow pores are thus not accessible to adsorption by nitrogen. This results in the lowering of the specific surface area and also results in the average diameter of the pores being larger. This is reflected in Table 1, where the average pore diameter of RHA-Fe is smaller (larger surface area) than that of RHA-Fe700 (lower surface area). Both samples show mesoporous characteristic [20]. Fig. 2 shows the nitrogen adsorption isotherm for RHAFe700. A similar isotherm was obtained for RHA-Fe. The hysteresis found in Fig. 2 corresponds to type H3 under the IUPAC classification [20]. It has been generally agreed that type H3 hysteresis are due to plate like pores that has uniform pore size distribution. Fig. 3 shows the pore size distribution of RHA-Fe700. Two minor pore ranges are distinctly visible. The first is observed between 20 and 50 Å. The second narrow range occurs between 50 and 90 Å. The major pore range is observed to be between 250 and 600 Å. These pore ranges are within the mesoporous range. The SEM micrographs of RHA-Fe and RHA-Fe700 are shown in Fig. 4. Fig. 4a reveals that RHA-Fe is more porous than RHA-Fe700 (in Fig. 4b). The higher porosity results in the solid having a higher specific surface area. This corrob-

Fig. 3. The graph of desorption pore size distribution for RHA-Fe700.

140

F. Adam et al. / Journal of Colloid and Interface Science 304 (2006) 137–143

(a)

(b)

(c) Fig. 4. The SEM micrographs of RHA-Fe and RHA-Fe700. (a) RHA-Fe (×1.50 K), (b) RHA-Fe700 (×1.50 K), and (c) RHA-Fe700 (×3.00 K).

orates very well with the BET results. The higher magnification (×3.0 K) of RHA-Fe700 reveals that the surface was still porous after the sample had been calcined at 700 ◦ C for 5 h. Fig. 5 shows the X-ray diffraction spectrograph of RHAFe700. The XRD analysis showed that the RHA-Fe was present in an amorphous state. Upon calcinations, RHA-Fe700, retained its amorphous nature although its specific surface area decreased. The amorphous nature of RHA-Fe700 can be seen in Fig. 4c. The quantitative analysis of the elements on the surface of RHA-Fe and RHA-Fe700 determined by EDX are shown in Table 2. Three spots from the SEM micrographs were randomly chosen at which the EDX analysis was done. From the EDX results, it was found that the Si and Fe atoms in RHA-Fe were not homogeneously dispersed in the silica matrix. However, the Si and Fe in RHA-Fe700 were more homogeneous and the Si:Fe ratio was found to be close to 12:1. The most likely reason for this homogeneity was that the calcination process removed the adsorbed water molecules and NO− 3 ions yielding a much more homogeneous catalyst. The loss of NO− 3 species can be deduced from the FTIR spectrum in Fig. 1, −1 was greatly diwhere the band due to NO− 3 , i.e., 1384 cm minished on calcination. There is an apparent increase of Fe after first use, which could not be accounted for at the present time. However it could just be local in-homogeneity that is resulting in the apparent increase of Fe. However, the Si:Fe ratio remained virtually unchanged during several reuse cycles. It can thus be concluded that the leaching of the metal ion into the reaction solution does not take place with this metal supported catalyst. The most probable reason for this stability to leaching is due to the fact that the Fe is chemically bonded to the silica matrix in the form of Si–O–Fe bonded phase. Benzylation of toluene using benzyl chloride was carried out without catalyst and also in the presence of RHA-Fe and RHAFe700 as catalyst under refluxing conditions. Fig. 6 shows the graph of percentage yield versus time of reaction under refluxing conditions. The minimum time required for complete conversion to products was determined to be 60 min under the refluxing condition. In subsequent tests the reaction was carried out for 60 min. Fig. 7 shows the graph of percentage yield versus temperature when the reactions were carried out for 60 min. Maximum yield was obtained at a reaction temperature of 100 ◦ C, which was close to the refluxing temperature of toluene used as the solvent and as one of the reagents. No product was obtained when the reaction was carried out with the catalyst at room temperature. Formation of products was detected only above 80 ◦ C. Reaction carried out in the presence of RHA-Fe or RHA-Fe700 showed 100% conversion of the benzyl chloride to products under the optimum conditions determined above. Fig. 8 shows the gas chromatogram of the product mixture under refluxing conditions. The sharp peak with retention time of 1.09 min was identified as that of toluene by its mass spectrum. The benzyl chloride, which appeared at the retention time of 2.33, has completely disappeared. Two major peaks at 14.94 and 15.19 min were found in the GC. Both gave the same mass spectrum that was determined to be monobenzyl toluene.

F. Adam et al. / Journal of Colloid and Interface Science 304 (2006) 137–143

141

Table 2 The EDX results of RHA-Fe and RHA-Fe700 Sample RHA-Fe RHA-Fe700 RHA-Fe700 R1 RHA-Fe700 R2 RHA-Fe700 R3

Silicon, Si (atom%) Spot 1

Spot 2

Spot 3

Average

Spot 1

Iron, Fe (atom%) Spot 2

Spot 3

Average

Ratio of Si:Fe

10.21 17.31 9.08

2.40 21.10 8.12

7.15 18.13 6.22

– 18.8 7.80

0.90 1.57 1.10

0.23 1.34 1.15

0.62 1.62 1.04

– 1.51 1.10

12 : 1 7:1

6.81

6.56

7.20

6.85

1.00

1.00

0.93

0.98

7:1

8.27

7.80

8.28

8.12

0.67

1.20

1.08

0.98

8:1

Note. R1 = after 1st use, R2 = after 2nd use, R3 = after 3rd use.

Fig. 5. The X-ray diffraction spectrogram of RHA-Fe700. (RHA-Fe gave a similar X-ray diffraction spectrogram.)

Fig. 7. Graph of percentage yield versus reaction temperature. Reactions were carried out for 60 min. Fig. 6. Graph of percentage yield versus time under refluxing condition.

The ortho-substituted monobenzyl toluene is expected to be present in a larger quantity due to the presence of 2 ortho-positions for substitution on the toluene molecule. Based

on this assumption, the peak at 15.19 min was attributed to the ortho-product while the peak at 14.95 was attributed to the para-product. Sixteen (16) minor peaks were observed in the gas chromatogram with retention times above 20 min (not

142

F. Adam et al. / Journal of Colloid and Interface Science 304 (2006) 137–143

Fig. 8. The GC chromatogram of products with RHA-Fe as catalyst. (RHA-Fe700 gave a similar chromatogram.)

shown). All these peaks gave identical mass spectra. Analysis of the mass spectra showed that these minor peaks correspond to various isomers and other products having similar molecular weight to the di-substituted product. The major product distribution from the reaction is shown in Table 3. From Table 3 it can be seen that RHA-Fe700 gave a slightly higher percentage of the mono-substituted product compared to RHA-Fe. Both catalysts gave similar distribution of the oand p-mono-substituted products. The RHA-Fe700 gave a significantly lower yield of the di-substituted benzylation product. Further enhancement of the mono-substituted product was observed when the catalyst was reused. The mono-substituted product increased to 98.24% when the catalyst was used for the third time (second reuse, T2, in Table 3). However there was not much difference in the distribution of the ortho- and para-derivatives.

Table 3 The product distribution of benzylation of toluene with RHA-Fe and RHAFe700 as catalyst Catalyst

Product distribution Mono

Total di-substituted product

Without catalyst (reflux)

nil

nil

RHA-Fe room temperature RHA-Fe700 room temperature RHA-Fe (reflux)

nil

nil

nil

nil

92.96

44.84 (p-) 48.12 (o-)

7.04

RHA-Fe700 (reflux)

95.54

46.01 (p-) 49.53 (o-)

4.46

RHA-Fe700 (reflux) T1

97.18

48.56 (p-) 48.62 (o-)

2.82

4. Conclusion

RHA-Fe700 (reflux) T2

97.13

48.50 (p-) 48.63 (o-)

2.87

Metal incorporated silica from rice husk ash had been prepared by the sol–gel technique and found to be amorphous with a large surface area. The pores in the RHA-Fe700 was in the mesoporous range. The iron-incorporated catalysts were found to be active for the Friedel–Craft benzylation of toluene using benzyl chloride. The optimum condition for the conversion to products was determined to be 100 ◦ C and 60 min of reaction time. Other metal supported catalysts similarly prepared are being investigated in our laboratory.

Note. T1 = 1st reuse and T2 = 2nd reuse.

Acknowledgments The authors wish to thank the Universiti Sains Malaysia for the facilities and the Malaysian Government for financial support in the form of IRPA grant (09-02-05-2148 EA 004) and the Fundamental Research (FRGS) grant (Account No. 304.PKIMIA.670005).

F. Adam et al. / Journal of Colloid and Interface Science 304 (2006) 137–143

References [1] Malaysian Agricultural Directory & Index 1995/96, sixth ed., Ministry of Agriculture, Kuala Lumpur, Malaysia, 1995, p. 107. [2] J. James, M.S. Rao, Am. Ceram. Soc. Bull. 65 (1986) 1177. [3] F. Adam, The Preparation and Adsorption Studies on Silica from Rice Husk Ash, M.Sc. thesis, Universiti Sains Malaysia, Penang, Malaysia, 1991. [4] O.K. Guha, J. Roy, J. Chromatogr. 437 (1988) 244. [5] N.B. Shrigadi, A.B. Shinde, S.D. Samant, Appl. Catal. A 252 (2003) 23. [6] A. Phukan, J.N. Ganguli, D.K. Dutta, J. Mol. Catal. A Chem. 202 (2002) 279. [7] F. Adam, B. Saad, B. Saraswathy, M. Fazliawati, Silica Supported Metal Oxide Catalysts from Rice Husk Ash, paper presented at the Regional Conference of Young Chemist (RCYC 2004), 13–14th of April 2004, Universiti Sains Malaysia, Pulau Pinang, Malaysia. [8] F. Adam, C.J. Hann, J. Colloid Interface Sci. 280 (2004) 55–61. [9] F. Adam, M.I. Saleh, J. Am. Oil Chem. Soc. 71 (1994) 1363. [10] F. Adam, S. Ravendran, J. Am. Oil Chem. Soc. 77 (2000) 437. [11] F. Adam, M.I. Saleh, in: M.R. Nordin, K.Y. Liew, A.A. Zainal (Eds.), Surface Science and Heterogeneous Catalysis, Universiti Sains Malaysia, Penang, 1993, p. 99.

143

[12] F.W. Chang, M.S. Kuo, M.T. Tsay, M.C. Hsieh, Appl. Catal. A Gen. 247 (2003) 309–320. [13] F.W. Chang, W.Y. Kuo, K.C. Lee, Appl. Catal. A Gen. 246 (2003) 253– 264. [14] D. Prasetyoko, Direct Synthesis of β-Zeolite from Rice Husk Ash and Its Use as a Heterogeneous Catalyst in the Friedel–Craft Reaction, M.Sc. thesis, Universiti Teknologi Malaysia, 2001. [15] J.B. Lambert, H.F. Shurvell, D.A. Lightner, R.G. Cooks, Organic Structural Spectroscopy, Prentice–Hall, NJ, USA, 1998, p. 191. [16] E.A.V. Ebsworth, D.W.H. Rankin, S. Cradock, Structural Methods in Inorganic Chemistry, second ed., Blackwell Scientific Publication, Oxford, United Kingdom, 1991, p. 222. [17] W.J. Stark, R. Strobel, D. Gunther, S.E. Pratnis, A. Baiker, J. Mater. Chem. 12 (2002) 3620–3625. [18] E.F. Vansant, P.V.D. Voort, K.C. Vrancken, Characterization and Chemical Modification of the Silica Surface, vol. 93, Elsevier, Amsterdam, The Netherlands, 1995, pp. 22–23, 31–47. [19] F. Adam, I.Ab. Rahman, M.I. Saleh, in: R. Othman (Ed.), Seramik Nusantara, Universiti Sains Malaysia, 1990, pp. 261–273. [20] S.J. Gregg, K.S.W. Sing, Adsorption, Surface Area and Porosity, Academic Press, London, 1967.