Friedel Crafts acylation and related reactions catalysed by heteropoly acids

Transcription

1 Applied Catalysis A: General 256 (2003) 3 18 Friedel Crafts acylation and related reactions catalysed by heteropoly acids I.V. Kozhevnikov The Leverhulme Centre for Innovative Catalysis, Department of Chemistry, University of Liverpool, Liverpool L69 7ZD, UK Received 30 September 2002; received in revised form 10 January 2003; accepted 14 January 2003 Abstract Recent studies on catalysis by heteropoly acids (HPA) for the Friedel Crafts acylation of arenes and related Fries rearrangement of aryl esters are reviewed. It is demonstrated that HPA-based solid acids are efficient and environmentally friendly catalysts for these reactions, usually superior in activity to the conventional acid catalysts such as H 2 SO 4 or zeolites Elsevier B.V. All rights reserved. Keywords: Heteropoly acid; Heterogeneous catalysis; Friedel Crafts acylation; Fries rearrangement 1. Introduction The Friedel Crafts aromatic acylation (Eq. (1)) and related Fries rearrangement of aryl esters, e.g. phenyl acetate (Eq. (2); Ac = acetyl), catalysed by strong acids are the most important routes for the synthesis of aromatic ketones that are intermediates in manufacturing fine and speciality chemicals as well as pharmaceuticals [1,2]. Tel.: ; fax: address: kozhev@liverpool.ac.uk (I.V. Kozhevnikov). (1) (2) These reactions involve acylium ion intermediates that are generated from an acylating agent or aryl ester by interaction with an acid catalyst. For the Friedel Crafts chemistry, present industrial practice uses acyl chlorides or acid anhydrides as acylating agents and requires a stoichiometric amount of soluble Lewis acids (e.g. AlCl 3 ) or strong mineral acids (e.g. HF or H 2 SO 4 ) as catalysts, which results in substantial amount of waste and corrosion problems [2]. The overuse of catalyst is caused by-product inhibition the formation of strong complexes between the aromatic ketone and the catalyst. In view of the increasingly strict environmental legislation, the application of heterogeneous catalysis has become attractive. In the last couple of decades, considerable effort has been put into developing heterogeneously catalysed Friedel Crafts chemistry using solid-acid catalysts such as zeolites, clays, Nafion-H, heteropoly acids (HPA), etc. [2], zeolites being the most studied catalysts ([2 6] and references therein). Likewise, the environmentally benign aromatic acylation with carboxylic acids (Eq. (3)) instead of the anhydrides and acyl chlorides, resulting in the formation of water X/$ see front matter 2003 Elsevier B.V. All rights reserved. doi: /s x(03)00406-x

2 4 I.V. Kozhevnikov / Applied Catalysis A: General 256 (2003) 3 18 as the only by-product, has been attempted, mostly with zeolites as catalysts [7 10]. Despite the numerous studies, solid-acid catalysts have hitherto been proved efficient only for the acylation of activated arenes (e.g. anisole), whereas non-activated arenes remain beyond the reach of heterogeneous catalysis. The acylation of anisole with acetic anhydride using a zeolite catalyst has been commercialised by Rhodia [2]. (3) Heteropoly acids are promising solid-acid catalysts for the Friedel Crafts reactions [11 15,17 20]. They are stronger than many conventional solid acids such as mixed-oxides, zeolites, etc. The Keggin-type HPAs typically represented by the formula H 8 x [XM 12 O 40 ], where X is the heteroatom (most frequently P 5+ or Si 4+ ), x is its oxidation state, and M is the addenda atom (usually W 6+ or Mo 6+ ), are the most important for catalysis [11 16]. They have been widely used as acid and oxidation catalysts for organic synthesis and found several industrial applications (for a recent comprehensive review, see the monograph [15]). The aim of this paper is to review recent studies on catalysis by HPA for the Friedel Crafts acylation and related Fries rearrangement of aryl esters. Use of HPA catalysts for Friedel Crafts alkylation of aromatic compounds has been reviewed elsewhere [14,15]. 2. Heteropoly acid catalysts 2.1. Acid sites In acid-catalysed reactions, bulk and supported heteropoly acids as well as heteropoly salts are used as catalysts. Several types of acid sites are present in these catalysts [12 15]. 1. Proton sites in heteropoly acids (e.g. H 3 [PW 12 O 40 ]). 2. Proton sites in acidic salts (e.g. Cs 2.5 H 0.5 [PW 12 O 40 ]). 3. Lewis acid sites in salts (metal countercations, e.g. in La III [PMo 12 O 40 ]). 4. Proton sites generated by dissociation of coordinated water: Ln(H 2 O) n 3+ Ln(H 2 O) n (OH) 2+ + H + 5. Proton sites generated by reduction of salts: Pd 2 [SiW 12 O 40 ] + 4{H} 2Pd 0 + H 4 [SiW 12 O 40 ] 6. Protons generated by partial hydrolysis of polyanions: [PW 12 O 40 ] 3 + 2H 2 O [PW 11 O 39 ] Bulk heteropoly acids +{WO 3 }+4H + Solid heteropoly acids possess purely Brønsted acidity and are stronger than such conventional solid acids as SiO 2 -Al 2 O 3, H 3 PO 4 /SiO 2, and HX and HY zeolites [11 16]. The acids H 3 [PW 12 O 40 ], H 4 [SiW 12 O 40 ], H 3 [PMo 12 O 40 ], and H 4 [SiMo 12 O 40 ] are readily available and most frequently used as acid catalysts, the first two usually being preferred. These acids have fairly high thermal stabilities, decomposing at 465, 445, 375, and 350 C, respectively [14,15]. Decomposed molybdenum acids may be reconstructed under exposure to water vapour [12]. For much less labile tungsten acids such reconstruction is unlikely. The acid strength of crystalline heteropoly acids decreases in the series [11 15]: H 3 [PW 12 O 40 ] > H 4 [SiW 12 O 40 ] > H 3 [PMo 12 O 40 ] > H 4 [SiMo 12 O 40 ] Usually relative catalytic activities of heteropoly acids are consistent with this order both in homogeneous and heterogeneous systems [12 15]. The drawback to the bulk acids is their low surface area (typically, 1 10 m 2 g 1 ) and low porosity (<0.1 cm 3 g 1 ). Because of this, for Friedel Crafts reactions, supported heteropoly acids are usually preferred Supported heteropoly acids Supported heteropoly acid catalysts have much greater number of surface acid sites than the bulk

3 I.V. Kozhevnikov / Applied Catalysis A: General 256 (2003) acids, hence they are more important for applications [11 16]. The acidity and catalytic activity of supported heteropoly acids depend on the type of carrier, the HPA loading, conditions of pretreatment, etc. Acidic or neutral substances such as SiO 2, active carbon, acidic ion-exchange resin, etc. are all suitable as supports, the most often used being SiO Tungstophosphoric acid supported on titania [21] and zirconia [22,23] has been characterised. Basic solids like MgO tend to decompose heteropoly acids [24]. Catalysts comprising H 3 [PW 12 O 40 ] supported on various porous silicas have been characterised [12,14,15,25,26]. Silica is relatively inert towards heteropoly acids, at least above a certain loading level, although some chemical interaction takes place between heteropoly acids and SiO 2, as shown by 1 H and 31 P magic-angle spinning (MAS) NMR [27 30]. The thermal stability of heteropoly acids on SiO 2 seems to be comparable to or slightly lower than that of the parent HPA [12,31]. At low loadings, H 3 [PW 12 O 40 ] and H 4 [SiW 12 O 40 ] form finely dispersed species on the SiO 2 surface; HPA crystal phase on silica ( m 2 g 1 ) is developed at an HPA loading above 20 wt.% [32,33]. Various HPA forms have been observed on the silica surface by transmission electron microscopy (TEM): discrete molecules, clusters 50 Å in size and large crystallites of 500 Å. Their relative amounts depend on the HPA loading [33]. As shown by microcalorimetry [34], when loading H 3 [PW 12 O 40 ] (20 wt.%) on SiO 2, the proton sites become weaker and less uniform than those in the bulk H 3 [PW 12 O 40 ]. The acid strength of H 3 [PW 12 O 40 ]/SiO 2 increases with HPA loading [35]. According to the ammonia thermal desorption data [34], the acid strength of supported H 3 [PW 12 O 40 ] decreases in the series of carriers: SiO 2 > -Al 2 O 3 > activated carbon. Heteropoly acids supported on certain activated carbons have been considered to be promising fixed-bed acid catalysts for liquid-phase reactions, e.g. esterification, because of their high stability towards HPA leaching from the carrier [36,37]. However, as shown by microcalorimetry [38,39], the acid strength of HPA is greatly reduced in such catalysts. As evidenced by IR and 31 P MAS NMR, H 3 [PW 12 O 40 ] and H 4 [SiW 12 O 40 ] supported on a chemically (H 3 PO 4 ) activated carbon retain the Keggin structure at the HPA loading >5 wt.% but decompose at lower loadings [40] Intrazeolite heteropoly acids Incorporation of heteropoly acids into zeolite pores to obtain shape-selective catalysts has long been a challenge. However, conventional zeolites are not suitable for this because their pores are too small to adsorb large (12 Å) HPA molecules. H 3 [PW 12 O 40 ] encapsulated into a mesoporous puresilica molecular sieve MCM-41 (BET surface area of 1200 m 2 g 1, uniform pores 32 Å in size) has been prepared [41,42] and characterised by nitrogen physisorption, XRD, Fourier-transform infrared spectroscopy (FTIR), TEM, and 31 P MAS NMR [30,41]. The H 3 [PW 12 O 40 ]/MCM-41 compositions with HPA loadings from 10 to 50 wt.% have 30 Å uniformly sized mesopores. Heteropoly acid forms finely dispersed species on the MCM-41 surface. No HPA crystal phase is observed at HPA loadings as high as 50 wt.%. As shown by TEM [30], the H 3 [PW 12 O 40 ] species are mainly located inside the MCM-41 pores rather than on the outer surface. H 4 [SiW 12 O 40 ]/MCM-41 has been characterised [43]; it is very similar to H 3 [PW 12 O 40 ]/MCM-41. Cesium salt of H 3 [PW 12 O 40 ] supported on MCM-41 mesoporous silica has been studied [44]. The synthesis of 12-molybdophosphoric acid in the supercages of Y-type zeolite has been reported [45]. This ship-in-the-bottle catalyst has been used for liquid-phase reactions [46]. The effect of the Si/Al ratio and the countercation in zeolite Y on the encapsulation of Keggin heteropoly acids has been studied [47] Heteropoly salts The nature of countercation in heteropoly salts is critical to their acidity, solubility, porosity, and thermal stability. Salts with rather small cations resemble the parent heteropoly acids; they are readily soluble in water, nonporous, and possess surface areas under 10 m 2 g 1. In contrast, salts with large monovalent cations, such as NH 4 +, K +, Cs +, etc. are water-insoluble, have a rigid microporous/mesoporous structure and can be prepared with surface areas over 100 m 2 g 1 [16,48]. The stoichiometric salts prepared by precipitation from aqueous solutions contain residual quantities of protons, which are apparently responsible for the catalytic activity of these salts [16,49]. As demonstrated by Misono and others [12,50,51],

4 6 I.V. Kozhevnikov / Applied Catalysis A: General 256 (2003) 3 18 the acidic Cs + salt, Cs 2.5 H 0.5 [PW 12 O 40 ], has strong acid sites, high surface area ( m 2 g 1 ) and hydrophobicity and is a very efficient solid-acid catalyst for a variety of organic reactions, especially in liquid-phase. The Cs + content controls the pore size of the Cs x H 3 x [PW 12 O 40 ] salts [52]. Various methods of the preparation of bulk and supported heteropoly salts have been described [53 56]. The preparation of silica-supported Cs + salts of H 3 [PW 12 O 40 ] with the egg-white morphology has been reported [57]. The proton sites in H 3 [PW 12 O 40 ] and its cesium salts have been characterised by 1 H, 2 H, 31 P MAS NMR and inelastic neutron scattering [58] and in situ FTIR [59]. The 129 Xe NMR technique has been applied to characterise the pore structure of NH 4 +,K +, and Cs + salts of Keggin heteropoly acids and confirmed the presence of microporosity in these salts [60]. Certain neutral heteropoly salts can also gain proton sites upon interaction with reaction medium. Two mechanisms of the proton generation in heteropoly salts are distinguished: the dissociation of coordinated water (for salts with the cations like Al 3+,Zn 2+, etc.) and the reduction of the metal cation (e.g. Ag +,Cu 2+, and Pd 2+ ) [61,62]. Al 3+ + H 2 O Al(OH) 2+ + H + Ag H 2 Ag 0 + H Sol gel catalysts H 3 [PW 12 O 40 ] and its Cs + salt, Cs 2.5 H 0.5 [PW 12 O 40 ], have been included in the silica matrix by means of a sol gel technique using the hydrolysis of tetraethyl orthosilicate to become water-insoluble and easily separable microporous solid-acid catalysts [63,64]. The catalysts thus obtained have large surface areas ( m 2 g 1 ) and are thermally more stable than Amberlyst-15. They catalyse the hydrolysis of ethyl acetate in aqueous phase, showing higher turnover frequencies than Amberlyst-15 and HZSM-5. The Keggin-type heteropoly acids included into a silica matrix by sol gel technique have been tested in Friedel Crafts alkylations [65]. However, the sol gel HPA catalysts, because of stronger interactions of their protons with the silica matrix, appear to have a weaker acid strength compared to silica-supported HPA [66,67]. 3. General features of HPA catalysis Heteropoly acids catalyse a wide variety of reactions in homogeneous or heterogeneous (liquid solid, gas solid or liquid liquid biphasic) systems, offering strong options for more efficient and cleaner processing compared to conventional mineral acids [11 16]. Being stronger acids, heteropoly acids will have significantly higher catalytic activity than conventional catalysts such as mineral acids, mixed-oxides, zeolites, etc. In particular, in organic media, the molar catalytic activity of heteropoly acid is often times higher than that of H 2 SO 4 [12,14,15]. This makes it possible to carry out the catalytic process at a lower catalyst concentration and/or at a lower temperature. Further, heteropoly acid catalysis lacks side reactions such as sulfonation, chlorination, nitration, etc. which occur with mineral acids [14,15]. As stable, relatively nontoxic crystalline substances, heteropoly acids are also preferable regarding safety and ease of handling. The relative activity of Keggin heteropoly acids primarily depends on their acid strength. Other properties, such as the oxidation potential as well as the thermal and hydrolytic stability are also important. These properties for the most common heteropoly acids are summarised as follows [14,15]. Acid strength Oxidation potential Thermal stability Hydrolytic stability PW > SiW PMo > SiMo PMo > SiMo PW > SiW PW > SiW > PMo > SiMo SiW > PW > SiMo > PMo Usually, tungsten heteropoly acids are the catalysts of choice because of their stronger acidity, higher thermal stability and lower oxidation potential compared to molybdenum acids [14,15]. Generally, if the reaction rate is controlled by the catalyst acid strength, H 3 [PW 12 O 40 ] shows the highest catalytic activity in the Keggin series. However, in the case of less demanding reactions as well as in reactions at higher temperatures in the presence of water, H 4 [SiW 12 O 40 ], having lower oxidation potential and higher hydrolytic stability, could be superior to H 3 [PW 12 O 40 ]. The major problem, limiting the utility of homogeneously catalysed processes, is the well-known difficulty in catalyst recovery and recycling. As the cost of heteropoly acids is higher than that of mineral acids,

5 I.V. Kozhevnikov / Applied Catalysis A: General 256 (2003) the recycling of HPA catalysts is the key issue to their application. Only a few homogeneous reactions, such as the hydration of lower olefins, allow for easy recycling. In some cases, heteropoly acid can be recovered from polar organic solution without neutralisation by precipitating with a hydrocarbon solvent. Heteropoly acid can also be extracted from an acidified aqueous solution of its salt with a polar organic solvent [14,15]. Separation of products and recovery and recycling of catalysts often becomes much easier when homogeneously catalysed reactions are performed in biphasic systems consisting of two immiscible liquid-phases a catalyst phase and a product/reactant phase with intense mass transfer between them. Heteropoly acids due to their special solubility properties, i.e. high solubility in a variety of polar solvents and insolubility in nonpolar solvents, are promising catalysts for operating under phase-transfer conditions [14,15]. The amount of heteropoly acid in the product phase must be negligible to allow easy catalyst separation. The important advantage of heterogeneous systems including solid HPA catalysts over homogeneous ones is the easy separation of catalyst from reaction products. Furthermore, since heteropoly acids are soluble only in wet nucleophilic solvents levelling the acid strength (dehydrated heteropoly acids are scarcely soluble in dry polar media), their intrinsic strong acidity cannot be fully utilised in homogeneous systems. Hence, for catalysing highly demanding reactions, such as the Friedel Crafts reaction, heteropoly acid must be used as a solid-acid catalyst in a dry non-nucleophilic medium. Supported heteropoly acids (usually on silica) are generally preferred because of their large surface area. HPA loadings vary from about 10 to 50 wt.% or even higher. At lower loadings, the acidity of heteropoly acid decreases because of interaction with support; such catalysts are also quite sensitive to poisoning by impurities that may be present in the support or feed. For these reasons, the catalysts containing less than 10 wt.% HPA usually show poor activities and are rarely used. Control of water content in heteropoly acid catalysts is essential for their efficient performance. This can be achieved by thermal pretreatment of the catalysts, typically at C. The effect of water may be attributed to the HPA acid strength and the number of proton sites as well as to catalyst deactivation. Excess water causes a decrease in the HPA acid strength, and thus in its activity. Dehydration of the catalyst increases the acid strength but decreases the number of acid sites, which may reduce the overall catalytic activity. The strong acid sites thus created tend to deactivate (coke) faster [14,15]. Like other solid-acid catalysts, the solid heteropoly acids tend to deactivate during organic reactions because of the formation of carbonaceous deposit (coke) on the catalyst surface. Subsequent regeneration of heteropoly acid catalyst is quite difficult. This problem remains to be solved to put heterogeneous heteropoly acid catalysis in practice. 4. Mechanistic principles Generally, reactions catalysed by heteropoly acids may be represented by the conventional mechanisms of Brønsted acid catalysis. In a simple case of single proton transfer, the mechanism may include the protonation of the substrate followed by the conversion of the ionic intermediate to yield the reaction product [14,15]: S 1 + H + S 1 H +(S 2) P + H + (4) In this equation, S 1 and S 2 are the substrates and P is the product. In accordance with this mechanism, the catalytic activity of heteropoly acids, both in homogeneous and heterogeneous systems, usually parallels their acid strength, i.e. H 3 [PW 12 O 40 ]>H 4 [SiW 12 O 40 ] >H 3 [PMo 12 O 40 ]>H 4 [SiMo 12 O 40 ] [14,15]. Being stronger acids and therefore more efficient proton donors, heteropoly acids usually exhibit higher catalytic activities than the conventional acid catalysts. Relatively strong oxidants, molybdenum heteropoly acids are frequently deactivated due to their reduction by the organic reaction medium; it is not uncommon for them to show lower activities than those expected from their acid strengths [14,15]. For heterogeneous acid catalysis by solid heteropoly compounds, Misono and others [12,13] advanced a mechanistic classification which distinguishes two types of catalysis, namely (i) surface type and (ii) bulk type (or pseudoliquid ). The surface type is a conventional acid catalysis on the gas solid or liquid solid interface. This type applies to reactions occurring on the surface of bulk or supported heteropoly compounds.

6 8 I.V. Kozhevnikov / Applied Catalysis A: General 256 (2003) 3 18 In this case, the catalytic activity usually depends on the surface acidity of heteropoly acid, i.e. the reaction rate is parallel to the number and the strength of the accessible surface acid sites. The bulk type of mechanism is largely relevant to reactions of polar substrates (e.g. alcohol, ether, ketone, amine, etc.) on bulk heteropoly compounds. These substrates are capable of absorbing into the catalyst bulk, forming HPA solvates [12]. In this case, solid heteropoly acids behave like highly concentrated solutions, and all protons, both in the bulk and on the surface of heteropoly acid, are suggested to participate in the catalytic reaction. Unlike polar molecules, nonpolar reactants (e.g. hydrocarbons) are incapable of absorbing in the bulk of heteropoly acids. They interact only with the surface of the catalyst. Water-insoluble heteropoly salts (e.g. Cs 2.5 H 0.5 [PW 12 O 40 ]) scarcely absorb polar molecules into the bulk and hence exhibit surface-type catalysis towards both polar and nonpolar molecules [12]. The pseudoliquid behaviour appears to be important for reactions of polar molecules at relatively low temperatures, i.e. when the sorption of substrate in the catalyst bulk is significant. Surface and bulk catalysis may operate with strongly differing selectivities [12,13]. 5. Friedel Crafts acylation and Fries reaction catalysed by heteropoly acids 5.1. Acylation The Friedel Crafts acylation of aromatic compounds is the most important route to aromatic ketones that are intermediates in various organic syntheses [1]. The reaction occurs by interacting the aromatic compound with a carboxylic acid derivative (e.g. acid anhydride, acyl chloride, or the acid itself) in the presence of an acid catalyst. Reaction mechanism involves acylium ion intermediates that are generated from the acylating agent by interaction with the acid catalyst [1]. In the last couple of decades, solid-acid catalysts such as zeolites, clays, Nafion-H, heteropoly acids, etc. have been explored for aromatic acylation to replace the conventional soluble acids (e.g. AlCl 3, HF, etc.), zeolites being the most studied catalysts ([2 10] and references therein). Although relatively active catalysts, the zeolites (e.g. H-beta) are deactivated during the acylation [6 8]. The main deactivation is deemed to be reversible; this is attributed to the strong adsorption of the acylation product on the catalyst, blocking access to the active sites. Another type of deactivation, which is irreversible, is caused by tar deposition on the catalyst surface (coking). Heteropoly acids are promising solid-acid catalysts for aromatic acylation. Below is described their use for these reactions Acylation by acyl chlorides and acid anhydrides Izumi et al. [11] pioneered the use of heteropoly acids as catalysts for aromatic acylation. Silicasupported acids H 4 [SiW 12 O 40 ] and H 3 [PW 12 O 40 ] both effectively catalyse the acylation of p-xylene with benzoyl chloride (Eq. (5); Bz: benzoyl) and remain unchanged on the SiO 2 surface after the reaction. (5) But in a more polar reaction medium such as chlorobenzene even H 4 [SiW 12 O 40 ] leaches from the silica support and decomposes in the course of reaction. Benzoic anhydride can be used as an acylating agent with the H 4 [SiW 12 O 40 ]/SiO 2 catalyst, but benzoic acid cannot. A weaker solid acid, H 4 [SiW 12 O 40 ]/ carbon, can also catalyse the acylation, but less efficiently than H 4 [SiW 12 O 40 ]/SiO 2 [11]. In contrast, H 3 [PMo 12 O 40 ] in the HPA/SiO 2 catalyst was found to decompose during the acylation of p-xylene with benzoyl chloride [11]. Probably, the real active species is not the supported H 3 [PMo 12 O 40 ], but some soluble species which might be formed by the interaction between H 3 [PMo 12 O 40 ] and benzoyl chloride. Cs 2.5 H 0.5 [PW 12 O 40 ] shows high efficiency in the acylation of activated arenes, such as p-xylene, anisole, mesitylene, etc. by acetic and benzoic anhydrides and acyl chlorides (Table 1). This catalyst provides higher yields of acylated arenes than the parent acid H 3 [PW 12 O 40 ], the latter being partly soluble in the reaction mixture [68]. The recent study [20] has shown that bulk and silica-supported H 3 [PW 12 O 40 ] exhibit a very high activity in the acylation of anisole (Eq. (6)) with

7 I.V. Kozhevnikov / Applied Catalysis A: General 256 (2003) Table 1 Friedel Crafts acylation [68] Substrates Product yield a Acylating agent Arene Cs 2.5 H 0.5 [PW 12 O 40 ] (PhCO) 2 O p-xylene 57 3 (PhCO) 2 O Anisole Ac 2 O Anisole n-c 7 H 15 COCl Mesitylene H 3 [PW 12 O 40 ] a Acylating agent/arene/catalyst = 5/100/0.01 mmol, reflux, 2 h. Table 2 Acylation of anisole with acetic anhydride (2 h) [20] Catalyst (amount, wt.%) a AN/AA (mol mol 1 ) T ( C) Yield b (%) p- MOAP o- MOAP H 3 [PW 12 O 40 ] (0.83) c H 3 [PW 12 O 40 ] (0.83) % PW/SiO 2 (0.83) % PW/SiO 2 (0.83) % PW/SiO 2 (0.88) d % PW/SiO 2 (0.83) % PW/SiO 2 (0.83) % PW/SiO 2 (0.83) H 4 [SiW 12 O 40 ] (0.83) c 40% SiW/SiO 2 (0.83) c H 3 [PMo 12 O 40 ] (0.83) % PMo/SiO 2 (0.83) Cs 2.5 H 0.5 [PW 12 O 40 ] (0.83) a The amount of catalysts per total reaction mixture: PW = H 3 [PW 12 O 40 ]; SiW = H 4 [SiW 12 O 40 ]; and PMo = H 3 [PMo 12 O 40 ]. b Yield based on acetic anhydride. c The yield of o-moap ca. 2 3%. d Yield in 10 min. acetic anhydride in liquid-phase, yielding up to 98% para- and 2 4% ortho-isomer of methoxyacetophenone (MOAP) at C and an anisole to acetic anhydride molar ratio AN/AA = (Table 2). (6) Catalyst pretreatment is essential. Fig. 1 shows the effect of catalyst pretreatment on the yield of p-moap and on the initial rate of anisole acylation with bulk H 3 [PW 12 O 40 ] and 50% H 3 [PW 12 O 40 ]/SiO 2. The activity passes a maximum at a pretreatment temperature of 150 C. From TGA, the amount of water remaining in the catalyst after pretreatment at 150 C is about 3 4H 2 O molecules per Keggin unit. Excess water causes a decrease in the HPA acid strength, and thus in its catalytic activity. Dehydration of the catalyst increases the acid strength but decreases the number of acid sites, which will reduce the catalytic activity unless the reaction is highly demanding for the catalyst acid strength. In addition, the very strong acid sites thus created tend to deactivate (coke) faster. The acylation of anisole appears to be heterogeneously catalysed; no contribution of homogeneous catalysis by HPA is observed. The H 3 [PW 12 O 40 ] catalyst is reusable, although gradual decline of activity was observed due to the coking of the catalyst. It should be noted that H 3 [PW 12 O 40 ] is almost a factor of 100 more active than the zeolite H-beta (Table 3), which is in agreement with the higher acid strength of HPA [20]. Anisole acylation is first-order in acetic anhydride, the order in catalyst is 0.66, and the apparent activation energy is 41 kj mol 1 in the temperature range of C [20]. The reaction is inhibited by-product because of adsorption of p-moap on the catalyst surface (Fig. 2). Applying the Langmuir Hinshelwood kinetic model, the ratio of adsorption coefficients of p-moap and anisole has been found to be 37 at 90 C. The effect of H 3 [PW 12 O 40 ] loading on silica upon the initial rate of anisole acylation is shown in Fig. 3. In these experiments, the total amount of H 3 [PW 12 O 40 ] Table 3 Acylation of anisole with acetic anhydride: HPA vs. zeolite (90 C, 2h) [20] Reaction conditions Catalyst 10% H 3 [PW 12 O 40 ]/ SiO 2 10% H 3 [PW 12 O 40 ]/ SiO 2 H-beta a [3] Catalyst amount (wt.%) AN/AA (mol mol 1 ) Yield of p-moap (%) b TON TOF (min 1 ) a Si/Al = b Yield based on acetic anhydride.

8 10 I.V. Kozhevnikov / Applied Catalysis A: General 256 (2003) 3 18 Fig. 1. Effect of HPA catalyst pretreatment (specified temperature/0.1 Torr, 1.5 h) in the acylation of anisole at AN/AA = 10 mol mol 1 : (a) yield of p-moap (50% H 3 [PW 12 O 40 ]/SiO 2 (0.83 wt.%), 50 C, 2 h); (b) initial rate (bulk H 3 [PW 12 O 40 ] (0.83 wt.%), 70 C) [20]. is kept constant. The activity of H 3 [PW 12 O 40 ] increases with the loading, passing a maximum at about 50% loading. Such behaviour may be explained as a result of increasing the HPA acid strength, on the one hand, and decreasing the HPA surface area, on the other, as the loading increases. It should be noted that the specific catalytic activity (per Keggin unit) of supported HPA is greater than that of bulk HPA. This demonstrates that the reaction occurs via the surface-type catalysis in terms of Misono and others classification ( bulk versus surface type ) [12,13]. In contrast to anisole, the acylation of toluene with HPA is far less efficient than that with H-beta [20]. These results have been explained by the well-known strong affinity of bulk HPA towards polar oxygenates, which would lead to the preferential adsorption of acetic anhydride on HPA, blocking access for toluene to the catalyst surface. It appears that the hydrophobic Fig. 2. Inhibition by-product in the acylation of anisole (40% H 3 [PW 12 O 40 ]/SiO 2 (0.83 wt.%), AN/AA = 10 mol mol 1,70 C, 2 h): (1) no p-moap added; (2) p-moap added initially, AA/p-MOAP = 2 mol mol 1 ; (3) p-moap added initially, AA/p-MOAP = 1 mol mol 1. Yields are based on acetic anhydride [20].

9 I.V. Kozhevnikov / Applied Catalysis A: General 256 (2003) Fig. 3. Effect of H 3 [PW 12 O 40 ] loading in the H 3 [PW 12 O 40 ]/SiO 2 catalyst on the initial rate of anisole acylation at AN/AA = 20 mol mol 1 and a constant total amount of H 3 [PW 12 O 40 ] (0.010 mmol): (a) 70 C; (b) 90 C [20]. zeolites with high Si/Al ratios less strongly differentiate the adsorption than the hydrophilic HPA and, therefore, are more suitable catalysts for the acylation of nonpolar aromatics like toluene Acylation by acids The aromatic acylation with carboxylic acids (Eq. (3)) instead of acid anhydrides and acyl chlorides has attracted interest, because it is an environmentally benign reaction, resulting in the formation of water as the only by-product. It has been attempted with zeolites and clays as catalysts [7 10]. Heteropoly acids have proved to be more active catalysts for this reaction [17,18,69]. Silica-supported 12-tungstophosphoric acid and its Cs + salts catalyse the acylation of toluene, p-xylene and m-xylene with crotonic acid (Eq. (7)) [17,18]. Some alkylation of aromatic compounds with crotonic acid also takes place. Heteropoly acid is more active than zeolites HY and H-beta in the acylation. Recently, the acylation of toluene and anisole with C 2 C 12 aliphatic carboxylic acids in liquid-phase catalysed by Cs 2.5 H 0.5 [PW 12 O 40 ] has been reported [69]. The acylation of toluene is carried out at a molar ratio PhMe/RCOOH = 50 and 110 C (reflux) in the presence of ca. 10 wt.% Cs 2.5 H 0.5 [PW 12 O 40 ] in the reaction mixture. The reaction is clearly heterogeneous, as Cs 2.5 H 0.5 [PW 12 O 40 ] is insoluble; it stopped when the catalyst was filtered off the reacting mixture. With acetic, propionic and butyric acids, the yield of acylated products is very low, though increasing in this series, similar to that observed for zeolites [7,8]. This may be due to the preferential adsorption of the lower acids on Cs 2.5 H 0.5 [PW 12 O 40 ], blocking access for toluene to the catalyst surface. The higher acids C 6 C 12 are more reactive in acylation, yielding 31 51% aromatic ketones (Table 4). All three possible isomers, ortho, meta and para, are formed, the para-isomers being the major products (7)

10 12 I.V. Kozhevnikov / Applied Catalysis A: General 256 (2003) 3 18 Table 4 Acylation of toluene (100 mmol) with carboxylic acids (2.0 mmol) at 110 C (reflux), 48 h [69] Catalyst (g) Acid Yield (%) a Product distribution (%) para ortho meta Cs 2.5 H 0.5 [PW 12 O 40 ] (1.0) Hexanoic Cs 2.5 H 0.5 [PW 12 O 40 ] (1.0) Octanoic Cs 2.5 H 0.5 [PW 12 O 40 ] (1.0) Dodecanoic Cs 2.5 H 0.5 [PW 12 O 40 ] (1.0) b Dodecanoic H 3 [PW 12 O 40 ] (1.0) Dodecanoic % H 3 [PW 12 O 40 ]/SiO 2 (2.5) Dodecanoic a The yield of aromatic ketones based on carboxylic acid; the yield based on toluene is 50 times lower. b A reuse of the above run. The catalyst was filtered off, washed with CH 2 Cl 2, dried and rerun. (55 73%), as expected. The reaction selectivity is virtually 100%, no other products are formed. The yield increases in the series of acids from hexanoic to dodecanoic acid like in the reaction with CeY zeolite [7,8] and cation-exchanged montmorillonite [9]. Cs 2.5 H 0.5 [PW 12 O 40 ](S BET, 112 m 2 g 1 ) is a much more efficient catalyst than the bulk H 3 [PW 12 O 40 ] (S BET, 7m 2 g 1 ) (Table 4), which may be explained by a greater number of H + surface sites in Cs 2.5 H 0.5 [PW 12 O 40 ]. The Cs + salt is also more active than the silica-supported HPA, 40% H 3 [PW 12 O 40 ]/ SiO 2, which may be the result of the higher hydrophobicity of Cs 2.5 H 0.5 [PW 12 O 40 ] [12,13], favouring the adsorption of nonpolar reactants on the catalyst surface and making Cs 2.5 H 0.5 [PW 12 O 40 ] more resistant towards deactivation by co-product water compared to the more hydrophilic H 3 [PW 12 O 40 ]. After the reaction, Cs 2.5 H 0.5 [PW 12 O 40 ] can be easily separated by filtration and reused. Some catalyst deactivation was observed, though, which was probably caused by coking. The most important advantage of Cs 2.5 H 0.5 [PW 12 O 40 ] catalyst is that it gives much higher productivity in aromatic ketones than the zeolite and clay catalysts reported so far, which may be attributed to the stronger acidity of Cs 2.5 H 0.5 [PW 12 O 40 ]. Thus, for the acylation of toluene with dodecanoic acid, Cs 2.5 H 0.5 [PW 12 O 40 ] gives a 1.0% yield of ketone based on toluene (Table 4) which is three times that reported for CeY (0.31%) [7] and for Al 3+ -montmorillonite (0.32%) [9]. (For CeY, a 96% yield based on dodecanoic acid at PhMe/acid = 313 has been obtained (150 C, 48 h) [7]. For Al 3+ -montmorillonite, a 60% yield at PhMe/acid = 187 (110 C, 24 h) has been reported [9].) It should be noted, however, that CeY gives a higher selectivity to the para-acylation (94%) [7] than Cs 2.5 H 0.5 [PW 12 O 40 ], which may be the result of shape-selective catalysis by the zeolite. The acylation of anisole with C 2 C 12 acids has been carried out under the same conditions as that of toluene, except a shorter reaction time (5 h) [69]. The acylated anisole forms as the major product (para/ortho = 59:1 96:1 and no meta-isomers) together with esterification products methyl esters of carboxylic acids and phenol (Table 5). No phenyl esters form. The selectivity to esters increases from acetic to dodecanoic acid, reaching 40% for the latter. The acylation of anisole, in contrast to that of toluene, is most efficient with C 2 C 6 acids, giving a 62 65% yield of acylated products and only 2 6% of methyl esters. When the acylation of anisole by acetic, propanoic or butyric acid is carried out in air Table 5 Acylation of anisole (100 mmol) with carboxylic acids (2.0 mmol) catalysed by Cs 2.5 H 0.5 [PW 12 O 40 ] (1.0 g) at 110 C, 5 h [69] Acid Conversion (%) Product distribution (%) a para ortho Methyl ester Acetic Propionic Butyric Hexanoic Octanoic Dodecanoic a Based on carboxylic acid converted; para and ortho are the corresponding acylated anisoles. Phenol in a molar ratio of 1:4 1:5 to acylated anisole is also formed.

11 I.V. Kozhevnikov / Applied Catalysis A: General 256 (2003) instead of nitrogen, 4-methoxy-4 -acylderivatives of diphenyl ether in a molar ratio to acylated anisole of 1:3 1:5 are obtained. Only traces of such products form in the case of C 6 C 12 acids. Apparently, these products are formed by the C O oxidative coupling of phenol and anisole, followed by acylation. The acylation of anisole with HZSM-5 zeolite (Si/Al = 30) as a catalyst proceeds differently [10]. With C 2 C 3 acids, at 120 C, PhOMe/acid = 4 and 20% HZSM-5, the phenyl esters are the main products; no methyl esters have been found. At 150 C and otherwise the same conditions, a 2:1 5:1 mixture of acylated anisole and phenyl ester forms at an % acid conversion. The conversion drops sharply for the acids higher than C 3, down to 0.6% for C 12, probably because of restricted access into zeolite pores. Thus Cs 2.5 H 0.5 [PW 12 O 40 ] is a more active as well as more selective catalyst than HZSM-5 for the anisole acylation Fries rearrangement The Fries rearrangement of aryl esters is an important route to aromatic hydroxyketones that are intermediates in manufacturing fine and speciality chemicals as well as pharmaceuticals [1,2]. The rearrangement of esters, e.g. phenyl acetate (Eq. (2)) to yield 2- and 4-hydroxyacetophenones (2HAP and 4HAP) and 4-acetoxyacetophenone (4AAP) together with phenol, involves acylium ion intermediates that are generated from the ester by interaction with an acid catalyst, typically a soluble Lewis acid (e.g. AlCl 3 ) or mineral acid (e.g. HF or H 2 SO 4 ). Heterogeneous catalysis, using solid acids such as zeolites, clays, heteropoly acids, etc. has attracted considerable interest to replace the conventional soluble acids in these reactions ([2,4 6] and references therein). Mechanisms for the formation of products have been discussed ([4] and references therein). 2HAP, 4AAP and phenol are considered to be the primary products, 2HAP being formed by the intramolecular rearrangement of PhOAc (Eq. (8)) and 4AAP and PhOH by the self-acylation (Eq. (9)). (8) (9) (10) In contrast, 4HAP appears to be the secondary product formed by the intermolecular acylation of phenol with PhOAc (Eq. (10)). Usually, the yield of phenol is greater than that of 4AAP, as part of PhOH results from the decomposition and/or hydrolysis of PhOAc that also produce ketene, acetic acid and acetic anhydride. Solvent plays a significant role in Fries reaction, polar solvents favouring the formation of the para-acylation products (4AAP and 4HAP). Heteropoly acids, especially H 3 [PW 12 O 40 ], have been demonstrated to be promising catalysts for Fries rearrangement of aryl esters [67,70]. The HPAcatalysed rearrangement of phenyl acetate (Eq. (2)) occurs in liquid-phase at C(Table 6). One of important advantages of HPA, as compared to zeolites or mineral acids, is that the reaction can be carried out both homogeneously and heterogeneously. The homogeneous process occurs in polar media, for example, in neat PhOAc or polar organic solvents like nitrobenzene (PhNO 2 )oro-dichlorobenzene that are commonly used for Fries reaction. All these media will dissolve H 3 [PW 12 O 40 ] at elevated temperatures (ca. 100 C). On the other hand, when using nonpolar solvents, such as higher alkanes (e.g. dodecane) that will not dissolve HPA, the reaction proceeds heterogeneously over solid HPA catalysts. In the latter case, supported HPA, preferably on silica, is the catalyst of choice, as bulk HPA possesses a low surface area. The HPA catalysts are easily separated from the heterogeneous system by filtration and can be reused, though with reduced activity. From the homogeneous systems, HPA can be effectively separated without its neutralisation by extraction with water and reused or utilised otherwise. The heterogeneous catalysis in the PhOAc dodecane media was clearly proved by filtering off the catalyst from the reacting system,

12 14 I.V. Kozhevnikov / Applied Catalysis A: General 256 (2003) 3 18 Table 6 Fries rearrangement of phenyl acetate (2 h) a [67,70] Catalyst (wt.%) Solvent (PhOAc, wt.%) T ( C) Conversion (%) Selectivity (%) PhOH 2HAP 4HAP 4AAP H 3 [PW 12 O 40 ] (0.60) PhOAc (100) H 3 [PW 12 O 40 ] (3.0) PhOAc (100) H 3 [PW 12 O 40 ] (3.0) PhNO 2 (25) H 3 [PW 12 O 40 ] (0.60) PhNO 2 (25) H 3 [PW 12 O 40 ] (0.60) PhNO 2 (50) H 2 SO 4 (1.4) PhNO 2 (25) H 3 [PW 12 O 40 ] (0.60) Dodecane (25) % PW/SiO 2 (1.5) Dodecane (25) % PW/SiO 2 (1.5) b Dodecane (25) % PW/SiO 2 (6.0) Dodecane (25) % PW/SiO 2 (3.3) c Dodecane (36) H-beta (1.3) c,d Dodecane (36) Cs 2.5 H 0.5 [PW 12 O 40 ] (0.67) PhNO 2 (25) a The reaction with H 3 [PW 12 O 40 ] is homogeneous in PhOAc and PhNO 2 and heterogeneous in dodecane. b Reuse of the above run. c 5h. d Si/Al = 11 [4]. which completely terminated the reaction. In contrast, filtration did not affect the reaction course in homogeneous systems, e.g. PhOAc PhNO 2 [67,70]. Strong inhibition of HPA-catalysed process with reaction products takes place both in homogeneous and heterogeneous systems like in anisole acylation. Addition of more HPA catalyst allows reaching a higher PhOAc conversion. Some irreversible catalyst deactivation is also observed [67,70]. The total selectivity towards the sum of PhOH, 2HAP, 4HAP and 4AAP is over 98%. Some acetic acid and acetic anhydride are also formed. The homogeneous reaction is more efficient than the heterogeneous one because it makes less phenol and more acetophenones, the selectivity to the more valuable para-acetophenones, 4AAP and 4HAP, being also higher. In terms of turnover frequencies, HPA is almost 200 times more active than H 2 SO 4 in homogeneous reaction, as well as more selective to acetophenones. In heterogeneous systems, HPA is also two orders of magnitude more active than H-beta zeolite, which is one of the best zeolite catalysts for this reaction [6]. However, H-beta shows a higher total selectivity to acetophenones than HPA (Table 6). It should be pointed out that HPA in homogeneous systems gives a higher selectivity to para-acetophenones 4AAP and 4HAP than H-beta. The efficiency of solid HPA (at constant loading) increases in the order H 3 [PW 12 O 40 ] < 40% H 3 [PW 12 O 40 ]/SiO 2 < 10% H 3 [PW 12 O 40 ]/SiO 2 in which the number of accessible proton sites increases. The insoluble salt Cs 2.5 H 0.5 [PW 12 O 40 ]isanefficient solid catalyst for the reaction in polar media such as PhNO 2 (Table 6). Although less active per unit weight than the homogeneous H 3 [PW 12 O 40 ]orthe solid catalyst H 3 [PW 12 O 40 ]/SiO 2, it is more selective to acetophenones than the parent HPA. The explanation of this may be that the less hydrophilic Cs + salt possesses stronger proton sites than the partially hydrated solid H 3 [PW 12 O 40 ]orh 3 [PW 12 O 40 ]/SiO 2 that contained 4 6H 2 O molecules per Keggin unit [70]. In contrast to silica-supported H 3 [PW 12 O 40 ], the sol gel H 3 [PW 12 O 40 ] catalysts prepared by the hydrolysis of tetraethyl orthosilicate show only a negligible activity in the Fries reaction of phenyl acetate, yielding mainly phenol with % selectivity. This may be explained by a weaker acid strength of the sol gel catalysts due to strong interaction of the HPA protons with the silica matrix and the presence of relatively high amount of water in sol gel catalysts [67]. This is in agreement with the fact that the gas-phase isomerisation of butene occurs much slower over the sol gel H 3 [PW 12 O 40 ] than over silica-impregnated H 3 [PW 12 O 40 ] [66].

13 I.V. Kozhevnikov / Applied Catalysis A: General 256 (2003) A continuous gas-phase catalytic acetylation of phenol with acetic anhydride to phenyl acetate followed by simultaneous Fries rearrangement to yield ortho- and para-hydroxyacetophenones over a silica-supported heteropoly acid has been described [71]. A complete conversion of phenol to phenyl acetate is achieved at 140 C. Upon increasing the temperature to 200 C, the phenyl acetate formed rearranges into hydroxyacetophenones with 10% yield and 90% selectivity to the para-isomer. The rearrangement of phenyl benzoate (PhOBz) occurs similarly to that of PhOAc (Eq. (11)), yielding 2- and 4-hydroxybenzophenones (2HBP and 4HBP), 4-benzoxybenzophenone (4BBP) and phenol together with benzoic acid. Table 7 Fries rearrangement of phenyl benzoate in 75:25 (wt.%) PhNO 2 PhOBz mixture (130 C, 2 h) [67] Catalyst (wt.%) Conversion (%) Selectivity (%) PhOH 2HBP 4HBP 4BBP H 3 [PW 12 O 40 ] (0.60) a Cs 2.5 H 0.5 [PW 12 O 40 ] (0.67) b a Homogeneous reaction. b Heterogeneous reaction. p-cresol (Eq. (12)) together with acetic acid and acetic anhydride. A very small amount of the meta-acylation product 3-hydroxy-6-methylacetophenone (3H6MAP) may also be formed. (11) Table 7 shows examples of homogeneous (with H 3 [PW 12 O 40 ]) and heterogeneous (with Cs + salt) rearrangement of PhOBz in PhNO 2 solution [67]. The product selectivities and catalyst activities are quite similar to those observed for PhOAc. The difference is that the amount of phenol formed is nearly equal to that of 4BBP, indicating that the hydrolysis of PhOBz is less significant in this case. In the case of rearrangement of p-tolyl acetate (p-toloac), the acylation in the para-position is no longer possible. Hence the major products are 2-hydroxy-5-methylacetophenone (2H5MAP) and (12) The homogeneous reaction with H 3 [PW 12 O 40 ] in PhNO 2 or o-dichlorobenzene gives almost equal amounts of 2H5MAP and p-cresol, no 3H6MAP being formed (Table 8) [67]. The heterogeneous reaction with the Cs + salt gives 2H5MAP with a remarkably high selectivity of 82% and only 17% of p-cresol. A little of 3H6MAP (1.4%) is also formed. This indicates that the hydrolysis p-toloac is less significant with the Cs + salt than with H 3 [PW 12 O 40 ], which is in agreement with the higher hydrophobicity of the Cs + salt. Table 8 Fries rearrangement of p-tolyl acetate (130 C, 2 h) [67] Catalyst (wt.%) Solvent (TolOAc, wt.%) Conversion (%) Selectivity (%) p-cresol 2H5MAP 3H6MAP H 3 [PW 12 O 40 ] (0.60) a PhNO 2 (25) H 3 [PW 12 O 40 ] (0.60) a o-cl 2 C 6 H 4 (25) Cs 2.5 H 0.5 [PW 12 O 40 ] (1.4) b o-cl 2 C 6 H 4 (25) a Homogeneous reaction. b Heterogeneous reaction.

14 16 I.V. Kozhevnikov / Applied Catalysis A: General 256 (2003) Deactivation and regeneration of solid heteropoly acid catalysts A serious problem with the solid heteropoly acid catalysts is their deactivation during organic reactions because of the formation of carbonaceous deposit (coke) on the catalyst surface. Conventional regeneration by burning coke at C, which is routinely used in the case of aluminosilicates and zeolites, is not applicable to heteropoly acids because their thermal stability is not high enough [15]. Therefore, for heteropoly acids to be more widely used for heterogeneous acid catalysis an efficient and reliable methodology of their regeneration would be beneficial. The development of a technique leading to a reduction in the temperature of coke removal would be of importance for the regeneration of deactivated solid heteropoly acid catalysts. The coking and regeneration of silica-supported H 3 [PW 12 O 40 ] during propene oligomerisation has been studied [72,73]. Coke formation causes rapid deactivation of the catalyst. The coked versus fresh catalysts have been characterised by 31 P and 13 C MAS NMR, XRD, XPS and TGA/TPO to reveal that the Keggin structure of the catalysts remains unaffected by coke deposition. The Pd doping has been shown to affect the nature of coke formed, inhibiting the formation of polynuclear aromatics. Co-feeding water to the propene flow greatly inhibits coke formation. The removal of coke from HPA catalysts has been attempted using solvent extraction, ozone treat- ment and aerobic oxidation. The aerobic burning of coke on the undoped H 3 [PW 12 O 40 ]/SiO 2 proceeds to completion at the temperature of C, which is higher than the temperature of H 3 [PW 12 O 40 ] decomposition. Doping the catalyst with Pd significantly decreases this temperature (Fig. 4) to allow catalyst regeneration at temperatures as low as 350 C without loss of catalytic activity. The catalyst comprising H 3 [PW 12 O 40 ] supported on sulphated zirconia doped with iron(iii) for the alkylation of benzene with propene ( C, 4 MPa) has been regenerated under an air flow at 350 C for 2h [74]. It has been claimed that a heteropoly acid catalyst, e.g. a molybdophosphoric acid catalyst, whose activity had been lowered could be regenerated by dissolving and/or suspending it in an aqueous medium and then treating with an inorganic ion-exchange material, e.g. crystalline antimonic acid [75]. Generally, prevention of catalyst deactivation is preferable to catalyst regeneration because the latter is usually difficult and expensive. In the case of gas-phase oligomerisation of propene catalysed by H 4 [SiW 12 O 40 ], the catalyst cokes less rapidly when supported on silica than in the pure form [76]. Diluting bulk heteropoly acid catalyst with acid-washed sand (1 part catalyst to 10 parts sand) dramatically increases the product yield and catalyst lifetime in the gas-phase oligomerisation of propene (230 C, 5 MPa) [77]. Treatment of the deactivated heteropoly acid catalyst for the gas-phase 1-butene and n-pentane Fig. 4. TGA/TPO for 20% H 3 [PW 12 O 40 ]/SiO 2 coked with propene [72,73].

15 I.V. Kozhevnikov / Applied Catalysis A: General 256 (2003) isomerisation with water recovers activity, whereas treatment in air or N 2 O is ineffective. It is suggested that water treatment regenerates acid sites by rehydrating partially decomposed Keggin units [35]. Co-feeding water has been reported to be crucial to the stable activity of the silica-supported H 4 [SiW 12 O 40 ] catalyst for the industrial synthesis of ethyl acetate from ethylene and acetic acid in vapour phase by the new BP Avada TM process [78]. The process produces 220,000 tonnes of ethyl acetate per year in BPs chemical complex near Hull, UK, since 2001 [79]. Some of the above recommendations, such as catalyst doping with transition metals and controlled addition of water to the HPA catalyst, might prove useful to prolong the lifetime of HPA catalysts in Friedel Crafts and related reactions. 7. Conclusion The recent studies reviewed here demonstrate that HPA-based solid acids, including bulk and supported heteropoly acids (preferably H 3 [PW 12 O 40 ]) as well as acidic heteropoly salts (e.g. Cs 2.5 H 0.5 [PW 12 O 40 ]), are active and environmentally friendly catalysts for the Friedel Crafts acylation of aromatic compounds and related Fries rearrangement of aryl esters. These solid acids are superior in activity to the conventional acid catalysts such as H 2 SO 4 or zeolites, which is in line with the stronger acidity of HPA. The HPA catalysts can be reused after a simple work-up, albeit with reduced activity. Similarly to zeolite catalysts, the HPA-catalysed acylations are inhibited by-products because of strong adsorption of the products on the catalyst surface. Consequently, to achieve higher conversions a larger amount of the catalyst is needed or a flow technique should be applied. Adsorption of aromatic substrate and acylating agent on the catalyst, especially preferential adsorption of one of them (e.g. the acylating agent), can affect (inhibit) the activity of HPA catalyst in acylation. The irreversible deactivation (coking) of HPA catalysts in Friedel Crafts reactions is an issue that needs to be addressed. References [1] G.A. Olah, in: Friedel Crafts and Related Reactions, vols. I IV, Wiley/Interscience, New York, ; G.A. Olah, in: Friedel Crafts and Related Reactions, Wiley/ Interscience, New York, [2] P. Metivier, in: R.A. Sheldon, H. van Bekkum (Eds.), Fine Chemicals Through Heterogeneous Catalysis, Wiley-VCH, Weinheim, 2001, p [3] E.G. Derouane, G. Crehan, C.J. Dillon, D. Bethell, H. He, S.B. Abd Hamid, J. Catal. 194 (2000) 410. [4] F. Jayat, M.J. Sabater Picot, M. Guisnet, Catal. Lett. 41 (1996) 181. [5] A. Vogt, H.W. Kouwenhoven, R. Prins, Appl. Catal. A 123 (1995) 37. [6] M. Guisnet, G. Perot, in: R.A. Sheldon, H. van Bekkum (Eds.), Fine Chemicals Through Heterogeneous Catalysis, Wiley-VCH, Weinheim, 2001, p [7] B. Chiche, A. Finiels, C. Gauthier, P. Geneste, J. Org. Chem. 51 (1986) [8] C. Gauthier, B. Chiche, A. Finiels, P. Geneste, J. Mol. Catal. 50 (1989) 219. [9] B. Chiche, A. Finiels, C. Gauthier, P. Geneste, J. Mol. Catal. 42 (1987) 229. [10] Q.L. Wang, Y. Ma, X. Ji, H. Yan, Q. Qiu, Chem. Commun. (1995) [11] Y. Izumi, K. Urabe, M. Onaka, Zeolite, Clay and Heteropoly Acid in Organic Reactions, Kodansha/VCH, Tokyo, [12] T. Okuhara, N. Mizuno, M. Misono, Adv. Catal. 41 (1996) 113. [13] M. Misono, Chem. Commun. (2001) [14] I.V. Kozhevnikov, Chem. Rev. 98 (1998) 171. [15] I.V. Kozhevnikov, Catalysts for fine chemicals, in: Catalysis by Polyoxometalates, vol. 2, Wiley, Chichester, [16] J.B. Moffat, Metal-Oxygen Clusters. The Surface and Catalytic Properties of Heteropoly Oxometalates, Kluwer Academic Publishers, New York, [17] C. Castro, J. Primo, A. Corma, J. Mol. Catal. A: Chem. 134 (1998) 215. [18] C. Castro, A. Corma, J. Primo, J. Mol. Catal. A: Chem. 177 (2002) 273. [19] B.M. Devassy, S.B. Halligudi, C.G. Hedge, A.B. Halgeri, F. Lefebvre, Chem. Commun. (2002) [20] J. Kaur, K. Griffin, B. Harrison, I.V. Kozhevnikov, J. Catal. 208 (2002) 448. [21] J.C. Edwards, C.Y. Thiel, B. Benac, J.F. Knifton, Catal. Lett. 51 (1998) 77. [22] E. Lopez-Salinas, J.G. Hernandez-Cortez, M.A. Cortes- Jacome, J. Navarrete, M.A. Llanos, A. Vazquez, H. Armendariz, T. Lopez, Appl. Catal. A 175 (1998) 43. [23] E. Lopez-Salinas, J.G. Hernandez-Cortez, I. Schifter, E. Torres-Garcia, J. Navarrete, A. Gutierrez-Carrillo, T. Lopez, P.P. Lottici, D. Bersani, Appl. Catal. A 193 (2000) 215. [24] T. Matsuda, A. Igarashi, Y. Ogino, J. Jpn. Petrol. Inst. 23 (1980) 30. [25] F. Marme, G. Coudurier, J.C. Vedrine, Microporous Mesoporous Mater. 122 (1998) 151. [26] C. Trolliet, G. Coudurier, J.C. Vedrine, Top. Catal. 15 (2001) 73. [27] V.M. Mastikhin, S.M. Kulikov, A.V. Nosov, I.V. Kozhevnikov, I.L. Mudrakovsky, M.N. Timofeeva, J. Mol. Catal. 60 (1990) 65.