SiMCM-41 cyclohexanone oxime to ε-caprolactam

SiMCM-41 cyclohexanone oxime to ε-caprolactam

Applied Catalysis A: General 248 (2003) 291–301 Beckmann rearrangement over phosphotungstic acid/SiMCM-41 cyclohexanone oxime to ε-caprolactam R. Mah...

201KB Sizes 0 Downloads 11 Views

Applied Catalysis A: General 248 (2003) 291–301

Beckmann rearrangement over phosphotungstic acid/SiMCM-41 cyclohexanone oxime to ε-caprolactam R. Maheswari a , K. Shanthi a , T. Sivakumar a , Sankarasubbier Narayanan b,∗ b

a Department of Chemistry, Anna University, Chennai 600025, India Inorganic and Physical Chemistry Division, Indian Institute of Chemical Technology, Hyderabad 500007, India

Received 14 September 2002; received in revised form 6 January 2003; accepted 20 February 2003

Abstract Vapor-phase Beckmann rearrangement of cyclohexanone oxime to ε-caprolactam is reported for the first time on phosphotungstic acid (PWA)-supported SiMCM-41 catalyst. The catalyst was prepared in the laboratory following standard procedures and was characterised by XRD and FT-IR. By N2 adsorption, BET surface area, pore size and pore diameter were measured. The conversion of cyclohexanone oxime over the catalyst was studied in the temperature region of 250–400 ◦ C using mainly acetonitrile as a solvent at different contact times. Increase in the reaction temperature to 325 ◦ C enhances oxime conversion and the selectivity of ε-caprolactam. An impressive catalytic performance of cyclohexanone oxime conversion >99% and ε-caprolactam selectivity of 75% was observed on 30 wt.% PWA/SiMCM-41 at 325 ◦ C using a feed containing10 wt.% oxime in acetonitrile at WHSV = 3.24 h−1 . The catalytic efficiency for this rearrangement reaction was studied, varying the PWA content from 10 to 50 wt.% on SiMCM-41 and also using benzene and ethanol as alternate solvents in place of acetonitrile. The rearrangement reaction and ε-caprolactam selectivity with respect to catalyst physical characteristics, PWA content and type of solvent are discussed. © 2003 Elsevier Science B.V. All rights reserved. Keywords: Beckmann rearrangement; Cyclohexanone oxime; ε-Caprolactam; SiMCM-41; Phosphotungstic acid

1. Introduction ε-Caprolactam is an important precursor for nylon-6 and plastics. The conventional method of preparation of ε-caprolactam from cyclohexanone oxime via Beckmann rearrangement results in the formation of a large amount of by-products, especially ammonium sulfate and approximately 4–5 tonnes of (NH4 )2 SO4 ∗ Corresponding author. Tel.: +91-40-27-19-32-00; fax: +91-40-2716-09-21. E-mail address: [email protected] (S. Narayanan).

per tonne of ε-caprolactam is inevitably obtained [1,2]. Further problems encountered include handling of a large amount of oleum and corrosion of the apparatus. To overcome the above-mentioned problems, solid acid catalysts were investigated for vapor-phase Beckmann rearrangement of cyclohexanone oxime. There are a number of reports on the use of heterogeneous catalysts such as alumina, heteropolyacids, boronphosphate, phosphoric acid, silica–alumina and boric acid on alumina and silica-supported tantalum oxide catalysts [3] for the rearrangement of a variety of ketoximes to amides [2,4–13]. There are sev-

0926-860X/$ – see front matter © 2003 Elsevier Science B.V. All rights reserved. doi:10.1016/S0926-860X(03)00182-0

292

R. Maheswari et al. / Applied Catalysis A: General 248 (2003) 291–301

eral studies on the rearrangement of cyclohexanone oxime to ε-caprolactam over zeolites Y [4,14], ZSM-5 [15], Beta [16] and mordenite [17]. The principal by-product was cyanopentene together with traces of cyclohexanone and cyclohexanol. Aucejo et al. [14] reported that the by-product cyanopentene was formed mainly on the Na+ ions of the zeolite. The Beckmann rearrangement is believed to take place even at 100 ◦ C, but the desorption of the product takes place only at temperatures higher than 300 ◦ C [18]. Above 360 ◦ C, there is a decrease in selectivity due to the decomposition of ε-caprolactam on the catalyst surface [19]. Medium polar solvents are more favorable than high polar or non-polar solvents [20]. Chung et al. [21] proposed that the high polarity is preferred for promoting the migration of OH2 + group in the rearrangement step. However, the efficiency of the solid acid catalysts is low towards ε-caprolactam formation, because of the rapid deactivation during the reaction [22–24]. Sato et al. [15,25] have reported that the Beckmann rearrangement occurs on the external surface of ZSM-5 type zeolites. Röseler et al. [26] and Dahlhoff and co-workers [27,28] claim that only extremely weak acidic sites on zeolitic catalysts are needed for the Beckmann rearrangement. The silanol nests of SiMCM-41 are the most suitable for the reaction; the vicinal silanol groups are more favorable than the terminal silanols. Yashima et al. [29] concluded by studying the vapor-phase Beckmann rearrangement of cyclohexanone oxime on zeolites with different pore windows that the selective rearrangement reaction occurred on the external surface of zeolite crystals. Takahashi et al. [22] studied the kinetics of Beckmann rearrangement on HZSM-5 and the effect of acid strength on the reaction. They concluded that, if the reaction had proceeded on the outer surface of the zeolite, the rate constant should be directly proportional to the outer surface area. However, the rate constant was not proportional to the outer surface area, so the result indicated that the acid sites on the outer surface area were not necessarily effective for the reaction of cyclohexanone oxime. Other reports [19,30] on AlMCM-41 and MCM-22 also reveal the importance of weak acidic sites, larger outer surface and pore structure for the above transformation. Silanol groups on SiMCM-41 are not sufficiently acidic [31,32] to catalyse the rearrangement and the neutral silanol

groups cannot be the active centres for Beckmann rearrangement [33]. However, there are reports claiming that the medium and weak Bronsted acid sites or even neutral silanol groups are most favourable for Beckmann rearrangement [7,14,34–36]. Even though silanol groups provide considerable conversion and selectivity towards ε-caprolactam, they are also responsible for the formation of by-products. Chaudhari et al. [19] reported that, while acidic sites appear to catalyse the rearrangement, SiOH groups catalyse the formation of hydrolysis product. It has also been reported that siliceous materials are not good enough due to high hydrophobicity and that the main by-product is cyanopentene [37]. There is a considerable influence of moderate acidity and silanol groups on the Beckmann rearrangement; most of the reports support the assertion that the silanol groups on the external surface are responsible [38–40]. The production of ε-caprolactam via cyclohexanone oxime intermediate involves three different steps: (i) the synthesis of cyclohexanone; (ii) amoximation of cyclohexanone to its oxime; (iii) the Beckmann rearrangement of the oxime to ε-caprolactam. The first step involves the conversion of benzene to cyclohexene/cyclohexane and to cyclohexanol/ cyclohexanone. Cyclohexanone can also be obtained from phenol. Narayanan and Krishna [41–44] have reported a highly selective synthesis of cyclohexanone from phenol over Pd/hydrotalcite. The present work concerns the Beckmann transformation of the cyclohexanone oxime to ε-caprolactam. There is a considerable scope for understanding the Beckmann rearrangement of oxime from the point of view of the mechanism, the dependence on physical factors and acidity of solid acids. Here, the focus of attention is on the preparation of SiMCM-41 and phosphotungstic acid (PWA)-loaded (10–50 wt.%) SiMCM-41 and evaluation of them as catalysts for the first time in the Beckmann rearrangement reaction of cyclohexanone oxime to ε-caprolactam. The effects of various reaction parameters, such as temperature, time on stream, use of different solvents, contact time and concentration of oxime in the feed are investigated. The influences of physical properties of the PWA/SiMCM-41 are cor-

R. Maheswari et al. / Applied Catalysis A: General 248 (2003) 291–301

293

Table 1 Physical characteristics of SiMCM-41 and PWA/SiMCM-41 Sample

PWA (wt.%)

d1 0 0 spacing (Å)

SBET (m2 /g)

Pore volume (ml/g)

Pore diameter (Å)

SiMCM-41 SiMCM-41/PWA (10) SiMCM-41/PWA (30) SiMCM-41/PWA (50)

– 10 30 50

41.63 37.02 34.92 30.56

1120 963 789 606

0.80 0.63 0.47 0.23

31 28 28 26

related with oxime transformation and ε-caprolactam selectivity.

2. Experimental 2.1. Catalyst preparation SiMCM-41 material was synthesised by an isothermal method [45]. Sodium meta silicate (E-Merck) was used as silica source. Sodium meta silicate (10.6 g) was dissolved in 60 g of water and the mixture was thoroughly stirred until a clear solution was obtained. Cetyltrimethylammonium bromide (3.36 g; CTAB, E-Merck, India) was dissolved in 20 g of ethanol. To this, sodium meta silicate solution was added dropwise. The resultant mixture was stirred for 1 h; then the pH of the resulting gel was adjusted to 11 using 4N sulfuric acid, followed by 3 h stirring. The resulting homogenous solution was transferred to an autoclave and heated to 140 ◦ C in static conditions for 12 h. The molar composition of the gel was SiO2 : 9.0 EtOH : 0.20 CTAB : 160 H2 O The resulting precipitate was recovered by filtration, washed with deionised water, dried in air at ambient temperature and finally calcined at 550 ◦ C for 1 h in a nitrogen flow and then in air for 12 h. Catalyst samples with different loadings of PWA (E-Merck, AR Grade, India), were prepared by stirring 1.0 g of SiMCM-41 with 10 ml of an aqueous solution containing 0.1, 0.3 or 0.5 g PWA at room temperature for 24 h. The resultant mixture was filtered, dried at 80 ◦ C and calcined at 200 ◦ C for 2 h.

2.2. Characterisation The mesoporous materials were characterised by XRD (Rigaku, D-Max/III–VC model) using nickel filtered Cu K␣ radiation, λ = 1.5406 Å; surface area measurements were done by a NOVA 1200, Quantachrome (USA) instrument. FT-IR spectra were taken on the instrument Bio-RAD FTS 175c FT-IR using catalyst samples supported on KBr wafers. Surface area, pore volume and pore size description details of the catalysts are given in Table 1. 2.3. Reaction procedure Cyclohexanone oxime (Aldrich), acetonitrile, benzene and ethanol (Analytical Grade, E-Merck, India) were used as solvents for dissolving oxime without further purification. The vapor-phase Beckmann rearrangement of cyclohexanone oxime to ε-caprolactam was carried out using a fixed bed vertical down flow reactor (length, 45 cm; diameter, 19 mm) with a catalyst charge of 1.5 g (10–20 mesh). Prior to the start of the experiment, the catalyst was activated at 330 ◦ C for 3 h in flowing air. The temperature of activation is kept below 350 ◦ C, because above this temperature the heteropoly acid would decompose. The catalyst was brought to the reaction temperature (250–400 ◦ C) in a flow of dry nitrogen. The reaction feed containing a solution of 10 wt.% cyclohexanone oxime in acetonitrile was introduced into the reactor from the top using a calibrated motorised syringe, along with nitrogen carrier gas (20 ml/min). The products along with solvent in liquid state was collected in a trap at room temperature and are analysed by gas chromatography (HP 6890 series) with HP-1 (100% dimethylpolysiloxane, 30 m × 0.25 mm) capillary column and FID detector. The analysis was

294

R. Maheswari et al. / Applied Catalysis A: General 248 (2003) 291–301

confirmed by GC-MS (QP 2000A, Shimadzu SE-52 column). 3. Results and discussion 3.1. Catalyst characterization Table 1 gives the physical characteristics of SiMCM-41 and PWA-supported catalysts. SiMCM-41 has a large surface area of 1120 m2 /g with pore volume 0.80 ml/g and diameter 31 Å. Impregnation of this SiMCM-41 with 10, 30 and 50 wt.% of PWA decreases surface area, pore volume and pore diameter. Addition of 10 wt.% PWA brings down the surface area of SiMCM-41 by approximately 15% from 1120 to 963 m2 /g. The surface area of SiMCM-41 decreases to nearly half with the addition of 30 and 50 wt.% PWA. The decrease of surface area may be mainly attributed to the corresponding decrease of pore volume as well as pore diameter. A significant decrease in the pore volume suggests that PWA occupies volume inside the pore as well as around the pore mouth. A reasonably large surface area of PWA-loaded SiMCM-41 suggests that the mesoporous structure is rather maintained and this is supported by XRD (Fig. 1). XRD patterns of SiMCM-41 and PWA/SiMCM-41 are shown in Fig. 1. SiMCM-41 has a typical XRD pattern exhibiting an intense peak at d = 41.63 Å and a much less intensive peak at d = 17.42 Å, which is in accordance with the earlier observations reported in [45,46]. Addition of 10 wt.% of PWA did not affect the XRD pattern very much. However, addition of 30 wt.% PWA alters the intensity of the peaks at d = 41.63 Å as well as at d = 17.42 Å, and new peaks start to appear. In the case of 50 wt.% PWA-loaded SiMCM-41, the intensity of the peaks around d = 17.42 Å increases along with the decrease of the peak at d = 41.63 Å. A comparison of the XRD patterns of pristine SiMCM-41 and PWA-loaded SiMCM-41 shows that, with the PWA loading, the mesoporous structure is rather intact. This is corroborated with the reasonable surface area of PWA/SiMCM-41, which is one of the characteristics of mesoporous structure. On higher loading, PWA spreads on the surface especially in and around the pore mouth of mesoporous SiMCM-41, contributing to the crystallinity observed

Fig. 1. XRD pattern of (a) SiMCM-41; (b) 10 wt.% PWA/SiMCM41; (c) 30 wt.% PWA/SiMCM-41; (d) 50 wt.% PWA/SiMCM-41.

in the XRD pattern as well as to the decreases in the pore volume and surface area. FT-IR spectra of bulk PWA as well as SiMCM-41supported PWA are shown in Fig. 2. The IR frequencies of bulk PWA are in good agreement with the reported values (cm−1 ): 1081 (P–O), 985 (W=O) and 897, 803 (W–O–W) [47]. The IR spectra of supported SiMCM-41 are similar to that of bulk PWA, even though the peaks are less intensive. It appears that the PWA keggin structure is retained in the supported sample. One sees that the bands for outer groups (W=O, corner- and edge-bridging W–O–W) show a small shift of 4–12 cm−1 compared to the spectrum of bulk PWA. This may be due to a chemical interaction between the PWA anion and the SiMCM-41 inner surface. 3.2. Catalysis 3.2.1. Effect of temperature The transformation of cyclohexanone oxime was studied in the temperature region 250–400 ◦ C on

R. Maheswari et al. / Applied Catalysis A: General 248 (2003) 291–301

295

Fig. 2. FT-IR spectra of (a) PWA; (b) 10 wt.% PWA/SiMCM-41; (c) 30 wt.% PWA/SiMCM-41; (d) 50 wt.% PWA/SiMCM-41.

SiMCM-41 as well as on PWA-supported SiMCM-41 catalysts. Fig. 3 shows the variation of oxime conversion as well as the selectivity of prime product ε-caprolactam and others. Cyclohexanone oxime conversion increases from around 70% at 250 ◦ C and reaches a steady state of nearly 100% around 300 ◦ C. With increase of temperature from 250 ◦ C, ε-caprolactam selectivity generally decreases and the decrease is significant beyond 325 ◦ C. Of the catalysts studied, 30 wt.% PWA shows a conversion of ∼100% and ε-caprolactam selectivity of ∼75% at 300 ◦ C. The pristine SiMCM-41 and 10 wt.% PWA/SiMCM-41 show ε-caprolactam selectivity of 60% at the same temperature. Fifty percent by weight of PWA/SiMCM-41 shows the least selectivity to ε-caprolactam at same temperature, even though the

oxime conversion is more or less similar to that of other catalysts. The other products such as cyanopentene and cyclohexanone follow a selectivity pattern opposite to that of ε-caprolactam with respect to temperature. Of the side products formed, cyanopentene is always more than cyclohexanone at higher temperatures. From conversion and selectivity patterns, it may be said that the 30 wt.% PWA/SiMCM-41 is selective for ε-caprolactam at 325 ◦ C. 3.2.2. Effect of space velocity The influence of feed rate of the reactant oxime on the surface of 30 wt.% PWA/SiMCM-41 was studied by using 10 wt.% cyclohexanone oxime in acetonitrile at 325 ◦ C (Fig. 4). While the catalyst weight was kept at 1.5 g, the contact time was varied by chang-

296

R. Maheswari et al. / Applied Catalysis A: General 248 (2003) 291–301

Fig. 3. Effect of reaction temperature on cyclohexanone oxime transformation: (a) SiMCM-41; (b) 10 wt.% PWA/SiMCM-41; (c) 30 wt.% PWA/SiMCM-41; (d) 50 wt.% PWA/SiMCM-41. Catalyst weight = 1.5 g; feed = 10 wt.% cyclohexanone oxime in acetonitrile; WHSV = 3.24 h−1 ; N2 carrier = 20 ml/h; time on stream = 2 h.

ing the feed rate, for example, 3, 5, 10 and 15 ml/h (i.e. 1.5–9.6 WHSV h−1 ). With increase of feed rate, the oxime conversion decreases from nearly 100 to 70%, as expected, and the ε-caprolactam selectivity increases from 25 to ∼80%. The other products, cyanopentene and cyclohexanone, follow the opposite trend to that of ε-caprolactam. Similar studies on the effect of contact time on conversion and selectivity were carried out on 30 wt.% PWA/SiMCM-41 at 325 ◦ C, by changing the concentration of oxime in the feed and keeping the feed rate at 5 ml/h (Table 2). With increase in oxime concentration from 5 to 20 wt.% in the feed, the oxime conversion decreases from nearly 100 to 90%. The ε-caprolactam selectivity under the

same condition decreases from 84 to 62%, while the other products, cyanopentene and cyclohexanone increase from 11 to 23% and 5 to 16%, respectively. The contact time studies also suggest an inverse relationship in the selectivity of ε-caprolactam and other products, as seen in the case of the temperature effect. 3.2.3. Effect of solvent and time on stream Several polar as well as non-polar solvents have been used for dissolving the cyclohexanone oxime. It has been reported that these solvents, depending on the polarity, contribute to 1,2-H shift or migration of proton during this transformation [21]. The mechanism

R. Maheswari et al. / Applied Catalysis A: General 248 (2003) 291–301

297

Fig. 4. Effect of WHSV (h−1 ) on cyclohexanone oxime transformation over 30 wt.% PWA/SiMCM-41. Catalyst weight = 1.5 g; feed = 10 wt.% cyclohexanone oxime in acetonitrile; temperature = 325 ◦ C; N2 carrier = ml/h; time on stream = 2 h.

of proton shift is visualised to take place as follows:

We have chosen benzene (non-polar), ethanol (medium polar) and acetonitrile (highly polar) as solvents for cyclohexanone oxime to study the polarity of solvents on conversion and the selectivity of the products for this reaction. Each reaction has been carried out for a period of 7 h, collecting the product

samples at every hour and analysing them. The solvent effect and time on stream on cyclohexanone oxime transformation are shown in Fig. 5. Ten percent by weight of cyclohexanone oxime in the selected solvent has been chosen and the feed rate has been maintained at WHSV 3.24 h−1 and reaction

298

R. Maheswari et al. / Applied Catalysis A: General 248 (2003) 291–301

Table 2 Effect of oxime concentration in the feed on cyclohexanone oxime transformation over 30 wt.% PWA/SiMCM-41 Oxime concentration (wt.%)

5 10 20

WHSV (h−1 )

2.92 3.24 3.88

Oxime conversion (%)

∼100 99 90

Product selectivity (%) A

B

C

84 74 62

11 18 23

5 9 16

ε-Caprolactam yield (%) 84 73 56

(A) ε-caprolactam; (B) cyanopentene + cyanopentane; (C) cyclohexanone + cyclohexenone. Catalyst weight = 1.5 g; solvent = acetonitrile; feed = 5 ml/h; N2 carrier = 20 ml/h; temperature = 325 ◦ C.

Fig. 5. Effect of solvent and time on stream on cyclohexanone oxime transformation using (a) benzene; (b) ethanol; (c) acetonitrile as solvent over 30 wt.% PWA/SiMCM-41. Catalyst weight = 1.5 g; feed = 10 wt.% cyclohexanone oxime in solvent; WHSV = 3.24 h−1 ; temperature = 325 ◦ C; N2 carrier = 20 ml/h.

temperature at 325 ◦ C in all the cases. Irrespective of the solvent used, the conversion is nearly 100% during the experimental period up to 7 h. The solvents seem to affect only the selectivity patterns. For example, when benzene and acetonitrile are used as solvents, the initial ε-caprolactam selectivity is around 70% and it decreases with time on stream; the decrease however is sharper beyond 4 h in the case of benzene than in the case of acetonitrile. The selectivity is maintained around 55% up to 7 h. When ethanol is used as a solvent, the ε-caprolactam selectivity is only between

20 and 40% during the reaction period studied. Generally, cyanopentene increases with time on stream and increases significantly in the acetonitrile solvent from 20 to 45%. Correspondingly, the cyclohexanone selectivity is low in the case of acetonitrile solvent. The solvent and time on stream studies bring out the interdependence of ε-caprolactam and cyanopentene selectivity and their independence with respect to conversion. This is due to the competitive nature of formation of caprolactam and cyanopentene. The reaction mechanism of formation of the by-products

R. Maheswari et al. / Applied Catalysis A: General 248 (2003) 291–301

299

involving silanol group is given below:

3.3. Effect of PWA loading Cyclohexanone oxime transformation was studied on SiMCM-41 as well as on 10, 30 and 50 wt.% PWA-loaded SiMCM-41. The detailed studies on the effect of temperature on such catalysts have been discussed already in the previous section (Fig. 3). Here, we compare the oxime conversion and ε-caprolactam

selectivity on these catalysts at a chosen temperature of 325 ◦ C for the reactions carried out using 10 wt.% cyclohexanone oxime in acetonitrile as feed, WHSV = 3.24 h−1 (Fig. 6). Pristine SiMCM-41 and PWA-supported SiMCM-41 (irrespective of the PWA loading) show nearly the same oxime conversion of >99%. The ε-caprolactam selectivity of SiMCM-41 and 10 wt.% PWA/SiMCM-41 is the same, around

300

R. Maheswari et al. / Applied Catalysis A: General 248 (2003) 291–301

Fig. 6. Effect of PWA loading of SiMCM-41on cyclohexanone oxime transformation. Catalyst weight = 1.5 g; feed = 10 wt.% cyclohexanone oxime in acetonitrile; WHSV = 3.24 h−1 ; temperature = 325 ◦ C; N2 carrier = 20 ml/h.

50% whereas 30 wt.% PWA catalyst shows a higher selectivity of 75%. Increasing the PWA loading to 50 wt.% brings down the ε-caprolactam selectivity to 20%, even though the conversion is maintained at >99%. The fact that there is a good conversion comparable to that of the PWA-loaded SiMCM-41 shows that the SiMCM-41 itself is active for this transformation reaction. Many reports claim that the SiMCM-41 is nearly neutral or weakly acidic. It appears that the inherent acidity in SiMCM-41 is sufficient for the oxime conversion. The observation that 10 wt.% PWA shows the same conversion and ε-caprolactam selectivity as that of pristine SiMCM-41 indicates that the added PWA is not available for the reaction. This is confirmed by the identity of the XRD patterns of SiMCM-41 and 10 wt.% PWA/SiMCM-41 being the same (Fig. 1). On 30 wt.% PWA loading of SiMCM-41, XRD patterns show crystalline PWA on the surface. These crystalline PWA, which may be well dispersed, contribute to the ε-caprolactam selectivity. However, further loading of SiMCM-41 with PWA (50 wt.%) brings down the ε-caprolactam selectivity, though the oxime conversion is the same as in the case of other catalysts. In 50 wt.% PWA/SiMCM-41, the PWA may be large and bulky agglomerates and these may not contribute to the selectivity. These observations lead to the conclusion that the cyclohexanone oxime transformation can take place even on

pristine SiMCM-41. Addition of PWA up to 30 wt.% contributes to the selectivity of ε-caprolactam, because of the contribution to acidity by well-dispersed PWA. However, bulk PWA that may be too strong in acidity is not helpful for ε-caprolactam selectivity. 4. Conclusions The cyclohexanone oxime transformation reactions over SiMCM-41 as well as PWA-loaded SiMCM-41 are studied at different reaction conditions of temperature, contact time and time on stream. SiMCM-41 itself is an active catalyst for this reaction. Addition of PWA improves mainly the selectivity of ε-caprolactam. Between 300 and 325 ◦ C, a conversion of >99% with ε-caprolactam selectivity of 75% is achieved on 30 wt.% PWA/SiMCM-41 using 10 wt.% cyclohexanone oxime in acetonitrile feed. Well-dispersed PWA helps ε-caprolactam selectivity, even though oxime conversion is insensitive to PWA addition. Conversion being the same, the selectivity of ε-caprolactam and of other products, cyanopentene and cyclohexanone, depend on reaction variables; the selectivity of ε-caprolactam is inversely related to that of cyanopentene because of the competitive nature of formation of these products from the oxime.

R. Maheswari et al. / Applied Catalysis A: General 248 (2003) 291–301

Acknowledgements SN thanks the CSIR, New Delhi for the Emeritus Scientist Position. The authors wish to thank Department of Science and Technology (DST), New Delhi, for providing financial support. We thank the Director of IICT for extending the facilities to carry out the characterization work. References [1] Kirk-Othmer. Encyclopedia of Chemical Technology, vol. 4, fourth ed., Wiley, New York, 1995, p. 827. [2] O. Immel, H.H. Schwarz, G. Starke, W. Swodenk, Chem. Ing. Technol. 56 (1984) 612. [3] T. Ushikubo, K. Wada, J. Catal. 148 (1994) 138. [4] P.B. Venuto, P.S. Landis, J. Catal. 6 (1966) 245. [5] D.C. England, Process for Manufacturing Caprolactam. US Patent 2,634,269 (1953). [6] I.S. Fisher, Improvements in or Relating to the Production of Amides and Lactams, UK Patent 881,276 (1961). [7] A. Costa, P.M. Deya, J.V. Sinisterra, J.M. Marinas, Can. J. Chem. 58 (1980) 1266. [8] P. Vitarelli, S. Cavallaro, R. Maggiore, G. Cimino, C. Caristi, S. Galvagno, Gazz. Chim. Ital. 112 (1982) 493. [9] T. Curtin, J.B. McMonagle, B.K. Hodnett, Appl. Catal. A 93 (1992) 91. [10] T. Curtin, J.B. McMonagle, B.K. Hodnett, Appl. Catal. A 93 (1992) 75. [11] T. Curtin, J.B. McMonagle, M. Ruwet, B.K. Hodnett, J. Catal. 142 (1993) 172. [12] S. Sato, K. Hirose, M. Kitamura, H. Tojima, N. Ishii, European Patent 236,092 (1987). [13] S. Sato, H. Sakurai, K. Urabe, Y. Izumi, Chem. Lett. 3 (1985) 277. [14] A. Aucejo, M.C. Burguet, A. Corma, V. Fornes, Appl. Catal. 22 (1986) 187. [15] H. Sato, K. Hirose, M. Kitamura, Y. Nakamura, Stud. Surf. Sci. Catal. 49 (1989) 1213. [16] L.-X. Dai, R. Hayasaka, Y. Iwaki, K.A. Koyano, T. Tatsumi, Chem. Commun. 9 (1996) 1071. [17] H. Sato, K. Hirose, N. Ishii, Y. Umada, European Patent 234,088 (1986). [18] H. Sato, K. Hirose, Y. Nakamura, Chem. Lett. (1993) 1987. [19] K. Chaudhari, R. Bal, A.J. Chandwadkar, S. Sivasanker, J. Mol. Catal. A 177 (2002) 247. [20] T. Komatsu, T. Maeda, T. Yashima, Microporous Mesoporous Mater. 35–36 (2000) 173.

301

[21] Y.M. Chung, H.K. Rhee, J. Mol. Catal. A: Chem. 159 (2000) 389. [22] T. Takahashi, M. Nishi, Y. Tagawa, T. Kai, Microporous Mater. 3 (1995) 467. [23] J.S. Reddy, R. Rvaisankar, S. Sivasanker, P. Ratnasamy, Catal. Lett. 17 (1993) 139. [24] P. O’Sullivan, L. Forni, B.K. Hodnett, Ind. Eng. Chem. Res. 40 (2001) 1471. [25] H. Sato, N. Ishii, K. Hirose, S. Nakamura, Stud. Surf. Sci. Catal. 28 (1986) 755. [26] J. Röseler, G. Heitmann, W.F. Hölderich, Appl. Catal. A 144 (1996) 319. [27] G.P. Heitmann, G. Dahlhoff, W.F. Hölderich, J. Catal. 186 (1999) 12. [28] G. Dahlhoff, G.P. Heitmann, J.P.M. Niederer, W.F. Hölderich, J. Catal. 194 (2000) 122. [29] T. Yashima, K. Miura, T. Komatsu, Stud. Surf. Sci. Catal. 84 (1994) 1897. [30] G. Dahlhoff, U. Barsnick, W.F. Hölderich, Appl. Catal. A 210 (2001) 83. [31] M. Busio, J. Janchen, J.H.C. Van Hoff, Microporous Mater. 5 (1995) 211. [32] A. Corma, V. Fornes, M.T. Navarro, J. Perez-Pariarte, J. Catal. 148 (1994) 569. [33] M. Kitamura, H. Ichihashi, Stud. Surf. Sci. Catal. 90 (1994) 67. [34] N. Katada, T. Tsubouchi, M. Niwa, Y. Murakami, Appl. Catal. A 124 (1995) 1. [35] E.R. Herrero, O.A. Anunziata, L.B. Pierella, O.A. Oria, Latin Am. Appl. Res. 24 (1994) 195. [36] T. Curtin, J.B. McMonagle, B.K. Hodnett, Catal. Lett. 17 (1993) 145. [37] A. Corma, H. Garcia, J. Primo, Zeolites 11 (1991) 593. [38] H. Sato, K. Hirose, Chem. Lett. (1993) 1765. [39] G.P. Heitmann, G. Dalhlhoff, W.F. Holderich, J. Catal. 186 (1999) 12. [40] T. Yashima, K. Miura, T. Komatsu, Stud. Surf. Sci. Catal. 84 (1994) 1897. [41] S. Narayanan, K. Krishna, Chem. Commun. (1997) 1991. [42] S. Narayanan, K. Krishna, Appl. Catal. A 174 (1998) 221. [43] S. Narayanan, K. Krishna, Catal. Today 49 (1999) 57. [44] S. Narayanan, K. Krishna, Appl. Catal. A 198 (2000) 13. [45] J.S. Beck, J.C. Vartuli, W.J. Roth, M.E. Leonowicz, C.T. Kresge, K.D. Schmitt, C.T.-W. Chu, D.H. Olson, E.W. Sheppard, S.B. McCullen, J.B. Higgins, J.L. Schlenker, J. Am. Chem. Soc. 114 (1992) 10834. [46] C.-Y. Chen, H.-X. Li, M.E. Davis, Microporous Mater. 2 (1993) 17. [47] C. Rocchiccioli-Deltcheff, M. Fournier, R. Frank, R. Thouvenot, Inorg. Chem. 22 (1983) 207.