WO2000013792A1

WO2000013792A1 - Supported bf3-complex solid acid catalyst, preparation and use

Info

Publication number: WO2000013792A1
Application number: PCT/EP1999/006528
Authority: WO
Inventors: James H. Clark; Karen Wilson
Original assignee: The University Of York
Priority date: 1998-09-07
Filing date: 1999-09-06
Publication date: 2000-03-16
Also published as: EP1115487A1; GB9819456D0; WO2000013792B1

Abstract

A novel method to prepare solid acid catalysts by the reaction of liquid BF3 precursors with inorganic oxide supports is reported. The resulting solid acid catalyst exhibits strong Brønsted acidity that can be tuned by appropriate selection of cocatalyst, solvent or calcination temperature, and can be used to catalyse organic reactions.

Description

Supported BF₃ -complex solid acid catalyst, preparation and use

Background to the invention:

Tightening environmental legislation on the production of waste during homogeneously acid catalysed reactions has led to a demand for heterogenised systems that will aid recovery of the catalyst and minimise pollution. Conventional Lewis acids (e.g. AICI₃ and BF₃) catalysts are used extensively in a range of industrial processes, including aikylation, acylation, polymerisation, isomerisation and addition reactions. The major drawback of these catalysts stems from the need to purify the product to remove catalyst residues and the need to dispose of large volumes of contaminated waste produced during separation of the catalyst from the product/reactant mix. The activity of these catalysts is also difficult to control and in many systems they produce undesirable biproducts thus reducing reaction yield.

Solid acid catalysts suitable for the aikylation of isobutane with butenes have been proposed in Applied Catalysis A: General 107 (1994) 239 to 248. These catalysts consist of H₃PO₄-BF₃-H₂SO₄ supported on SiO₂ and ZrO₂. These catalysts however suffer from the disadvantage that they contain undesirable materials particularly phosphorous, they involve the use of mineral acids which are corrosive and they are proposed for use only is gas phase systems. Further the method of preparation uses the phosphoric acid to act as a bridge between the surface of the support and the BF₃ and the presence of water and sulphuric acid as well as the heating to dry the catalyst leaves doubts as to the nature of the catalytic species.

This invention specifically concerns the preparation of supported BF₃ catalysts for use in liquid organic reactions. The use of a heterogeneous catalyst in these reactions would offer ease of separation and catalyst recycling, and thus reduce the waste effluent currently produced which is essential if the chemical industry is to comply with ever tightening environmental legislation on the production of waste. However a suitable replacement supported system must also exhibit activities/selectivity's comparable to or superior to the existing homogeneous route.

A method for supporting AICI₃ on SiO₂ has been previously reported in US Patent 5 294 578 which involves the reaction of AI-CI bonds with hydroxyl groups on SiO to form Al-O-Si bonds liberating HCI in the process. Previous attempts to support BF₃ have been reported using Al₂0₃ (US Patent 4 407731 and WO 94/02243) or zeolites (US Patent 4 709110) as the support material but have required the use of gaseous BF₃. The BF₃ sites formed on Al₂0₃ are reported as being only partially supported when used in liquid reactions. Often solvents such as diethyl ether are used to make a slurry with the support material, gaseous BF₃ is then bubbled through the slurry. This can result in the formation of BF₃OEt₂ complexes which are known to be catalytically active in polymerisation reactions. There are however no reported patents or literature that involve the use of a liquid precursor to attach BF₃-cocatalyst complexes to the surface of inorganic oxide supports.

The present invention provides a new supported BF₃ catalyst system in which BF₃ is attached directly to the surface of the support and which incorporates an organic cocatalyst to control acidity and catalytic activity. The novel form of supported BFJcocatalyst complex exhibits Brønsted and Lewis acid properties that can be tuned by varying both the cocatalyst and the catalyst preparation techniques. Unlike the oxide supported solid acid catalysts described in earlier work which involve the use of gaseous BF₃ in their preparation, our route enables liquid BF₃ complexes to be supported directly on the support (preferably silica) surface. This simplifies the preparation, and also enables the loading of BF₃ to be tuned more precisely. BF₃ precursors are known to be unstable towards hydrolysis liberating HF. However, when attached to the support the hydrostability of the precursor is strongly dependent on the preparation route used.

The present invention therefore provides a supported BF₃ catalyst system comprising a BF₃/organic Brønsted base cocatalyst complex wherein BF₃ is attached directly to the surface of the support.

The catalytic activity of homogeneous BF₃ complexes in many organic reactions is dependent on the ability of the complex [H⁺][X:BF₃ ^~] to act as a proton donor to olefins. The activity of the cocatalyst (HX) in homogeneous systems is observed to decrease in the order (see GA Olah "Friedel Crafts and Related Reactions", Volume 2, Interscience Publishers, 1940). HF> H 2 SO 4 >H₃°PO4 '>C6 H5JOH>H 2JO>RCOOH>ROH 1.1

By supporting different polarisable proton donating BF₃ complexes on supports such as Si0₂, the tuneable catalytic activity observed in the homogenous system should be obtained with the advantage of ease of catalyst recovery of a heterogeneous solid acid. Subsequent thermal treatment of the catalyst should also enable additional tuning of the relative amounts of Lewis and Brønsted acid sites present, as summarised below.

,._. . HF(g)

Brønsted acidity in solid acid catalysts normally arises from polarised "O-H⁺ sites. The observation of strong Brønsted acidity following attachment of a Lewis acid centre to an oxide support has been reported in other systems, and is attributed to polarisation of surface hydroxyl groups via an inductive effect of the electronegative halogen atoms on the acid site. In our model, additional Brønsted acidity is obtained from the Brønsted complex illustrated above. We have found that it is important that once formed the supported catalyst system should not be unduly heated since this can result in the loss of the cocatalyst. Low temperatures should therefore be used in catalyst preparation, care must be taken in any drying of the catalyst and calcination is preferably avoided.

By varying the complexing ligand HX, the Brønsted complex acidity can be varied. The organic Brønsted base is an organic molecule containing oxygen, nitrogen or sulphur basic centres capable of being protonated to produce the desired catalyst acidity typically a Pka value of +4 or lower, preferably -1 or lower. Preferred are those organic molecules containing an oxygen basic centre such as alcohols, ethers, ketones, aldehydes, ethoxylates, carboxylic acids or mixtures thereof. Particularly preferred are alcohols and carboxylic acids. Water may also be used in admixture with such organic molecules. The choice of solvent (which can also complex with the BF₃) used during the preparation will alter the acidic properties of the catalyst. The catalyst systems of this invention are usually produced by mixing the BF₃/Brønsted complex with the support in a solution. The solvent can therefore be chosen according to the acidic properties required in the catalyst system for example polar solvents (alcohols) will result in enhanced Brønsted acidity compared to nonpolar solvents (such as aromatic hydrocarbons). It is preferred that the solvent used in the preparation is predried to avoid hydrolysis of the BF3 complex.

Characterisation of these solid acids by MAS- N MR, DRIFTS and pyridine titration enable the Brønsted or Lewis acid characteristics of the different catalysts to be determined. We also expect that the thermal stability of these complexes will vary depending on the precursor used. Thus the relative amount of Brønsted:Lewis acid sites can be tuned by precalcining the samples under an inert atmosphere.

The catalyst when used in reactions can be recovered by filtration and reused in subsequent reactions.

The nature of the support used in the catalyst system of the present invention is also important. It must be able to react with the BF₃ and can be chosen according to the nature of the reaction to be catalysed and the cocatalyst. Examples of suitable supports are materials containing surface hydroxyl groups such as silica synthetic silicas (MCM) hexagonal mesoporous silica (HMS) as described in Nature 1992 359, page 710 and Science 267 page 865, and clay supports, including naturally occurring clay mineral such as kaolinite, bentonite, attapulgite, montmorillonite, clarit, Fuller's earth, hectorite, and beidellite; synthetic clay such as saponite and hydrotalcite; montmorillonite clay treated with sulphuric acid and/or hydrochloric acid; and modified clay such as Aluminium oxide pillared clay, cerium modified alumina pillared clay, and metal oxide pillared clay. The preferred supports have surface hydroxyl groups which can react with the Boron trifluroride, mesoporous silica and silica calcined at 500°C or above, preferably 600°C and above are particularly preferred support.

The support may also include at least one member selected from the group consisting of zeolite β, zeolite Y, seolite X, MFI, MEL, NaX, NaY, faujasite, mordenite, alumina, zirconia, titania and alumino sillicates.

The catalyst systems of the present invention may be used in the wide variety of reactions where acid catalysts have been traditionally used. The use of the catalysts brings the benefits of cleaner reaction (less product purification required, less waste disposal problems) and the ability to tune the catalyst to the reaction so increasing yields and specificity. Examples of reactions in which the catalysts may be used include aikylation, polymerisation, etherification, esterification and condensation reactions. We have found that the use of the catalyst systems of this invention in the aikylation of aromatics gives greater control over the position and number of alkylations on the aromatic nucleus.

The yield of the reactions are dependent on the accessibility of the reactants to the catalytic sites. Accordingly the reaction conditions should be adjusted according to the differing characteristics of the substrates. For example, we have found that in the aikylation of aromatics if the alkylating alkene is present in excess of the aromatic the yield is considerably reduced. We prefer therefore to operate under conditions in which the alkene is added gradually in a way which maintains a molar excess of the aromatic. Furthermore, in certain condensation reactions which can produce molecules capable of further condensation the conditions can be tailored to increase the yield of the first or second reaction products. An example is the condensation of 2-hydroxy acteophenone with benzaldehyde to 2¹ hydroxychalcone and the subsequent ring condensation of the 2- Hydroxychalcone to flavanone.

The catalysts of the present invention can therefore be used in the production of a wide range of useful chemicals and chemical intermediates.

The present illustration is illustrated by the following examples:

EXAMPLE 1

A series of catalysts were prepared using different cocatalysts and solvents.

CATALYST A: BF₃(H₂0)₂/Si0₂ prepared in ethanol

40 mmol BF₃(H₂O)₂ was added to a three necked flask purged with N₂, containing 100ml of absolute ethanol. 10 g of SiO₂ (K100 - predried at 300°C) was then added and the slurry stirred at 30°C for 2 hours. The slurry was then transferred to the rotary evaporator where the excess EtOH was removed under vacuum at 50°C.

CATALYST B: BF₃(H₂0)₂/Si0₂prepared in toluene 40 mmol BF₃(H₂O)₂ was added to a three necked flask purged with N₂, containing 100ml of anhydrous toluene. 10 g of Si0₂ (K100 - predried at 300^DC) was then added and the slurry stirred at 30°C for 2 hours. The slurry was then transferred to the rotary evaporator where the excess toluene was removed under vacuum at 50°C.

CATALYST C: BF₃.OEt₂ Si0₂ prepared in ethanol

40 mmol BF₃.OEt₂ was added to a three necked flask purged with N₂, containing 100ml of absolute ethanol. 10 g of Si0₂ (K100 - predried at 300°C) was then added and the slurry stirred at 30°C for 2 hours. The slurry was then transferred to the rotary evaporator where the excess EtOH was removed under vacuum at 50°C.

CATALYST D: BF^OEt SiO_j prepared in toluene

40 mmol BF₃.OEt₂ was added to a three necked flask containing purged with N₂, containing 100ml of anhydrous toluene. 10 g of Si0₂ (K100 - predried at 300°C) was then added and the slurry stirred under reflux for 2 hours. The slurry was then transferred to the rotary evaporator where the excess toluene was removed under vacuum at 50°C.

CATALYST E: BF₃(CH₃COOH)/Si0₂ prepared in ethanol

40 mmol BF₃(CH₃COOH) was added to a three necked flask purged with N₂, containing 100ml of absolute ethanol. 10 g of Si0₂ (K100 - predried at 300°C) was then added and the slurry stirred at 30°C for 2 hours. The slurry was then transferred to the rotary evaporator where the excess EtOH was removed under vacuum at 50°C.

CATALYST F: BF₃(MeOH)/Si0₂ prepared in methanol 40 mmol BF₃ MeOH was added to a three necked flask purged with N₂, containing 100ml of methanol. 10 g of Si0₂ (K100 - predried at 300°C) was then added and the slurry stirred at 30°C for 2 hours. The slurry was then transferred to the rotary evaporator where the excess methanol was removed under vacuum at 50°C.

CATALYST G: BF₃.OEt_ Si0₂ prepared in diethyl ether

40 mmol BF₃.OEt₂ was added to a three necked flask purged with N₂, containing 100ml of diethyl ether. 10 g of SiO₂ (K100 - predried at 300°C) was then added and the slurry stirred at 30°C for 2 hours. The slurry was then transferred to the rotary evaporator where the excess ether was removed under vacuum at 50°C.

CATALYST H: BF₃(H₂0)₂/Si0₂ prepared in ethanol (Si0₂precalcined 600° C)

40 mmol BF₃(H₂O)₂ was added to a three necked flask purged with N₂, containing 100ml of absolute ethanol. 10 g of Si0₂ (K100 - predried at 600°C) was then added and the slurry stirred at 30°C for 2 hours. The slurry was then transferred to the rotary evaporator where the excess EtOH was removed under vacuum at 50°C.

CATALYST I: BF₃(H₂0)_/HMS₂₄ prepared in ethanol

40 mmol BF₃(H₂O)₂ was added to a three necked flask purged with N₂, containing 100ml of absolute ethanol. 10 g of HMS₂₄ ^* (SA 1100m²g ¹) was then added and the slurry stirred at 30°C for 2 hours. The slurry was then transferred to the rotary evaporator where the excess EtOH was removed under vacuum at

50°C.

^*Preparation of HMS₂₄ HMS₂₄ was prepared according to the sol gel route of Pinnavaia et al by condensing 62.5g of tetraorthoethoxysilicate in a mixture of 123g ethanol and 160g H₂O with 15.3g n-dodecylamine as the surfactant template. The resulting gel was stirred for 24 hours, filtered, dried and then calcined at 600°C for 6 hours to remove the template.

CATALYST J: BF^OEt^S^ prepared in toluene at 30 ^<C

40 mmol BF₃.OEt₂ was added to a three necked flask purged with N₂, containing 100ml of anhydrous toluene. 10 g of Si0₂ (K100 - predried at 300°C) was then added and the slurry and was stirred at 30°C for 2 hours. The slurry was then transferred to the rotary evaporator where the excess toluene was removed under vacuum at 50°C.

EXAMPLE 2

Characterisation of Acid sites

Pyridine titration

- (a) Characterisation of the supported BF₃/Si0₂ catalysts was performed using DRIFTS in conjunction with pyridine titration which show that the catalysts A to C exhibit both Lewis and Brønsted acidity (Figure 1 ). This is determined by the absorption bands observed in DRIFT spectra at 1445 and

1461 cm- , (Lewis sites), 1638 and 1539 cm- (Brønsted sites) and 1611 and 1489 cm- (combined Lewis/Brønsted sites). There is a striking difference in the nature of the acid sites depending on catalyst preparation, with the catalysts prepared in ethanol exhibiting higher concentrations of Brønsted acid sites than those prepared in toluene. The lack of absorption at 1540 for catalyst D is believed to be due to either the use of reflux conditions in its preparation driving off the cocatalyst or that toluene in insufficiently polar to encourage bond polarisation.

(b) The corresponding DRIFTS/pyridine titrations of catalysts calcined to 200 and 400°C under Nitrogen, (shown in Figure 2) indicate that the Brønsted sites are gradually lost as the calcination temperature is increased. The most dramatic decrease in Brønsted acidity occurs once ethanol desorption is complete.

A comparison of a range of catalysts prepared using different BF₃ complex precursors is shown in Figure 3. The Bronsted acidity of the catalysts can be determined by pyridine titration. The peak observed at 1540 cm^"1 is a unique vibrational mode of the pyridinium ion that forms via interaction of pyridine with Brønsted sites. Thus the intensity of this peak provides an indication of presence of Brønsted acid sites on the solid catalyst. Thus it can be seen that if BFa.OEySiO;, is prepared in toluene at either reflux or 30°C (Catalyst D and J) the Brønsted acidity is negligible. This is most likely due to a combination of decompostion of the etherate complex and low solvent polarity discouraging bond polarisation.

If however BF₃.OEt₂/SiO₂ is prepared in OEt₂ at 30°C, then Brønsted sites are detected due to the formation of [OBF₃]^~[Et-OH⁺-Et] complex. Likewise for the other BF₃ precursors tested a high degree of Brønsted acid character persists due to the presence of protic complexes. It should be noted however that these titrations simply indicate the presence of Brønsted sites, but are not related to the Brønsted acid strength.

TGIR

(b) The origin of the acid sites on the ethanol prepared catalysts A and C was investigated using thermogravimetric analysis coupled with evolved gas FTIR (TGIR), which allows molecules desorbing from the catalyst during thermal analysis to be identified by their vibrational spectrum. Heating both catalysts above 100°C results in significant weight loss and the observation of ethanol desorption in the IR. However the differential mass lost indicates that the ethanol desorption temperature from BF₃(H₂O)₂/SiO₂ is 10°C higher than from BF₃.OEt₂/Si0₂, and approximately twice the amount of ethanol is evolved. The uptake of short chain alcohols can be used as an indication of the strength and concentration of Brønsted acid sites on zeolites. These results therefore suggest that BF₃(H₂0)₂/Si0₂ possesses a higher coverage of stronger Brønsted acid sites compared to BF₃.OEt₂/Si0₂. Further heating beyond 400°C results in an additional weight loss which is accompanied by the evolution of HF from the catalyst.

MAS-NMR

The evolution of ethanol above 100°C coupled with the loss of Brønsted acidity indicates that Brønsted acid sites in the BF₃(H₂0)₂/Si0₂ catalyst may arise from the binding of ethanol to supported BF₃ centres resulting in the formation of a [SiOBF₃]^"[EtOH₂]⁺ complex.

Further evidence in support of this model comes from H MAS NMR of the as prepared catalyst which show resonances at 1.34, 4.01 and 8.16 ppm which are consistent with CH₃, CH₂ and OH₂ ⁺ of protonated ethanol respectively.

BET Surface Area:

Loss of surface area on calcining at 600°C is ascribed to formation of a borosilicate species. EXAMPLE 3

Phenol Aikylation by 1-octene -using solvent

(a) The catalytic activity of various supported BF₃ samples was tested using the reaction of 1-octene with phenol (performed at 85°C using 0.05 M of each reactant, in 100 ml of 1 ,2 dichlorethane with 1g of supported BF₃ cataiyst). Table 1 shows the phenol conversion and selectivities towards octyl-phenyl ether obtained after 23 hours reaction time. It is clear that the activity of the BF₃(H₂O)₂/Si0₂ catalyst prepared in ethanol is superior to the other samples. The activity can thus be correlated with the number and strength of Brønsted acid sites identified on these catalysts using TGIR.

(b) Following reuse of BF₃(H₂O)₂/Si0₂ samples, a decrease in conversion and selectivity towards ring aikylation products is observed relative to the fresh catalyst. The loss of activity on recycling the catalyst may result from organic residue deposited on the catalyst during reaction causing pore blocking and/or poisoning of active sites.

Table 1 EXAMPLE 4

Phenol Aikylation by cyclohexene - Optimisation of Reaction Conditions

A problem of using linear alkenes in aikylation reactions, is that they readily isomerise resulting in a wide range of aikylation products. To simplify the reaction conditions and assist with optimisation of performance of Catalyst A, cyclohexene was selected as the alkene as isomerisation will not occur. The range of products expected are shown in scheme 1.

Scheme 1

Five cases are compared

2.3g of phenol, 2.0 g of cyclohexene, 1.0g of dodecane (internal standard) and 25 ml 1 ,2 Dichloroethane was added to a three necked round bottomed flask and heated to 85°C. 0.6g of supported catalyst A was then added to the reaction and samples taken periodically for analysis by GC.

2.3g of phenol, 1.0g of dodecane (internal standard) and 25 ml of 1 ,2 Dichloroethane was added to a three necked round bottomed flask and heated to 85°C. 0.6g of supported catalyst was then added to the reaction then 2.0g of cyclohexene was added gradually using a peristaltic pump. Samples were taken periodically from the reaction for analysis by GC. III. 2.3g of phenol, 2.0 g of cyclohexene andl .Og of dodecane (internal standard) was added to a three necked round bottomed flask and heated to 85°C. 0.6g of supported catalyst A was then added to the reaction and samples taken periodically for analysis by GC.

IV. 2.3g of phenol, and 1.0g of dodecane (internal standard) was added to a three necked round bottomed flask and heated to 85°C. 0.6g of supported catalyst was then added to the reaction then 2.0g of cyclohexene was added gradually using a peristaltic pump. Samples were taken periodically from the reaction for analysis by GC.

V. 2.3 g of Phenol, 6.0g cyclohexene and 1.0g of dodecane (internal standard) was added to a three necked round bottomed flask and heated to 85°C. 0.6g of supported catalyst A was then added to the reaction and samples taken periodically for analysis by GC.

The results of these tests are shown in Figure 4, which illustrates that the conversion of phenol decreases from experiment

IV > III > II > I

Thus phenol conversion can be increased if the alkene is added gradually over the period of the reaction.

No Phenol conversion was observed with reaction V, illustrating that phenol aikylation is diffusion limited, and that excess alkene poisons the reaction, presumably due to the formation of oligomers at the active site.

¹⁹F NMR showed no detectable fluorine compounds present in the reaction mixture after filtration. Example 5

Comparison of Catalyst Performance

a. Effect of precursor/cocatalyst

2.3g of phenol, 2.0 g of cyclohexene and 1.0g of dodecane (internal standard) was added to a three necked round bottomed flask and heated to 85°C. 0.6g of supported catalyst was then added to the reaction and samples taken periodically for analysis by GC. The product distribution obtained after 8 hours reaction time using 3 catalysts A, E, and G are compared in Figure 5.

Ring aikylation requires stronger acid sites. From Figure 5 it can be seen that catalyst G results in higher conversions of Phenol, and higher selectivity towards ring alkylated products. These observations are consistent with the respective pKa's of the cocatalyst EtOHΕt and EtOH₂ ⁺ listed in table 2 . The exact nature of the cocatalyst formed with catalyst E is uncertain. Possible Bronsted acid species include CH₃COOHJ, CH₃C(OH⁺)OEt, EtOH₂ ⁺,and CH₃COOH. As the catalyst was prepared in EtOH, the EtOH₂ ⁺ complex may dominate .

Table 2

b. Effect of support material

The effect of precalcination temperature of K100 Si0₂, and the use of mesoporous silica HMS₂₄ is compared using the BF₃(H₂O)₂ precursor in EtOH. 2.3g of phenol, 2.0 g of cyclohexene and 1.0g of dodecane (internal standard) was added to a three necked round bottomed flask and heated to 85°C. 0.6g of supported catalyst was then added to the reaction and samples taken periodically for analysis by GC. The product distribution obtained after 8 hours reaction time using 3 catalysts A, H and I (all prepared using BF₃(H₂0)₂ precursor in EtOH) are compared in Figure 6.

Catalysts H and I, prepared using 600°C calcined Si0₂ and HMS₂₄ materials respectively as the support are slightly more active for phenol aikylation than catalyst A. Catalysts H and I also exhibit higher selectivities towards ring alkylated products compared to the analogous catalyst prepared on K100 Si0₂ (catalyst A). These results may be ascribed to 600°C calcined K100 and HMS₂₄ having different surface polarities compared to conventional silica. This physical property will alter the diffusion rates of the reactants/products and in turn alter product selectivities. Another beneficial effect of high temperature calcination of silica, is that the surface becomes more hydrophobic, which will increase the stability of BF₃ sites towards hydrolysis.

c. Effect of Solvent used in catalyst preparation

Catalyst G (prepared from BF₃OEt₂ in OEt₂) exhibits the highest activity in phenol aikylation. In contrast Catalyst D (prepared from BF₃OEt₂ in refiuxing PhMe) exhibits very low activity. To understand whether this is due to toluene being a poor solvent for catalyst preparation, or if refiuxing is impairing catalyst acidity, catalyst J was prepared in toluene at 30°C as a control.

2.3g of phenol, 2.0 g of cyclohexene and 1.0g of dodecane (internal standard) was added to a three necked round bottomed flask and heated to 85°C. 0.6g of supported catalyst was then added to the reaction and samples taken periodically for analysis by GC. The product distribution obtained after 8 hours reaction time using 3 catalysts D, G and J are compared in Figure 7. Both catalysts D and J exhibit low phenol conversions, suggesting that reduced catalyst activity originates from using toluene as a solvent during catalyst preparation. Under these conditions, the most active catalyst results from using diethyl ether during catalyst preparation.

EXAMPLE 6

Effect of calcination on catalyst activity:

A sample of catalyst A was calcined at 150°C under Nitrogen for 2 hours.

2.3g of phenol, 2.0 g of cylcohexene and 1.0g of dodecane (internal standard) was added to a three necked round bottomed flask and heated to 85°C, then 0.6g of calcined catalyst was added to the reaction and samples taken periodically for analysis by GC. The results of this experiment after 24 hours reaction are compared with those for the uncalcined catalyst in Table 3.

Table 3

Following calcination the activity of the catalyst is considerably reduced, and the selectivity ring alkylated products decreases. This can be understood in terms of the loss of EtOH₂ ⁺ complex and reduced Brønsted acid strength of the calcined catalyst. EXAMPLE 7

Comparison with homogeneous BF₃(H₂0)₂:

2.3g of phenol, 2.0 g of cylcohexene and 1.0g of dodecane (internal standard) was added to a three necked round bottomed flask and heated to 85°C, then 0.3g of BF₃(H₂O)₂ was added to the reaction and samples taken periodically for analysis by GC. The results of this experiment are shown in Figure 8 and may be compared with those for the Catalyst A.

A comparison of phenol aikylation using the homogeneous and heterogenous versions of the BF₃(H₂O)₂ precursor is shown in Figures 8 and 9, from which it can be seen that the product distributions obtained are quite different. The heterogeneous catalyst forms ether products with a high selectivity that is maintained throughout the reaction. However for the homogeneously catalysed reaction the ether selectivity reaches a maximum, then decreases as the reaction proceeds with a concomitant increase in the yield of ring alkylated products.

These results can be interpreted in view of observations that unlike homogeneous BF₃(H₂O)₂ BF₃(H₂O EtOH/SiO₂ does not rearrange cyclohexyl- phenyl ether. However, ether rearrangement by the homogeneous system results in the formation of phenol and dialkylated products, suggesting that the rearrangement involves cracking of the C-0 bond yielding a cyclohexyl cation which can subsequently perform ring aikylation reactions.

EXAMPLE 8

Polymerisation of methyl styrene

20 ml of 4-methyl styrene was added to a flask containing 20 ml of anhydrous toluene. 0.2 g of Catalyst A was added and the reaction stirred for 1.5 hours at 20°C. The reaction was terminated by filtering off the catalyst. The light fractions were removed by stripping under vacuum, leaving a viscous white solid. The overall yield of polymer based on the original mass of monomer was 26%.

EXAMPLE 9

Etherification

7.0g of sec-phenylethyl alcohol was stirred at 120°C with 1.0g of catalyst A. After 2 hours the reaction was analysed by GC, which indicated that 72% of the alcohol was converted to diphenyl ether with a selectivity of 100%.

Phenylethyl-alcohol Di-phenlyethylether

EXAMPLE 10

Esterification Reactions

Phenylethyl-alcohol Acetic Acid

3.6 g of Phenethyl alcohol was added to 5g of acetic acid with 0.5g of dodecane as internal standard. The mixture was heated to 120°C, then 1 g of catalyst A added. Analysis of the reaction after 2 hours revealed 80% of the alcohol had reacted. The selectivities were as follows acetylated product = 86.5%, di-phenylthyl ether. = 12.5% other = 1 %

EXAMPLE 11

Claisen Schmidt Condensation reaction

40mmol of acetophenone, 40 mmol of benzaldehyde and 1g of dodecane (Internal standard) was added to a round bottom flask and heated to 70°C. 0.5g of catalyst A, D, E or F was added and samples taken periodically for analysis by GC.

Acetophenone Benzaldehyde

A comparison of the conversions and selectivity obtained after 8 hour reaction is shown in Table 4

Table 4 This activity of these catalysts increases with reaction temperature as illustrated in Table 5, after 24 hours using catalyst A

EXAMPLE 12

Condensation Reaction

acetophenone Benzaldehyde 2'-Hydroxychalcone

Further reaction to form the flavanone can also be observed.

40mmol of 2-hydroxyacetophenone, 40 mmol of benzaldehyde and 1g of dodecane (Internal standard) was added to a round bottom flask and heated to 150°C. 0.5g of catalyst A, was added and samples taken periodically for analysis by GC. Under these conditions a 66% conversion of 2-hydroxyacetophenone was obtained with the following product selectivity's hydroxychalcone = 65% flavanone = 22% other = 13%

Claims

CLAIMS:

1. A supported BF₃ catalyst system comprising a BF₃/organic Brønsted base cocatalyst complex wherein BF₃ is attached directly to the surface of the support.

2. A supported BF₃ catalyst system according to Claim 1 in which the organic Brønsted base is an organic molecule containing oxygen, nitrogen or sulphur containing basic centres capable of being protonated to produce a material having a Pka value of +4 or lower.

3. A supported BF₃ catalyst system according to Claim 1 or Claim 2 in which the Brønsted base contains an oxygen containing basic centre.

4. A supported BF₃ catalyst system according to Claim 3 in which the Brønsted base in an alcohol, an ether, a ketone, an aldehyde, an ethoxylate or a carboxylic acid.

5. A supported BF₃ catalyst system occurring to any of the preceding claims in which the support is selected from silica, alumina and zirconia.

6. A supported catalyst system according to any of the proceeding claims in which the support is mesoporus silica.

7. A supported catalyst system according to any of claims 1 to 5 in which the support is silica calcined at 500°C or above.

8. A process for the production of a catalyst system according to any of the preceding claims comprising mixing a solution of the BF₃/organic Brønsted base cocatalyst complex with the support and removing excess solvent under conditions which does not remove the organic Brønsted base cocatalyst.

9. A process according to Claim 8 in which the excess solvent is removed under vacuum at a temperature below the boiling point of the organic Brønsted base cocatalyst.

10. A process according to Claim 8 or Claim 9 in which the solvent is a polar organic solvent which is capable of complexing with BF₃.

11. A process according to Claim 10 in which the solvent is an oxy solvent.

12. A process according to Claim 11 in which the oxy solvent is an alcohol, an ester, a ketone, an aldehyde, an ethoxylate or a carboxylic acid.

13. The use of a catalyst system according to any of claims 1 to 7 in liquid phase reactions.

14. The use according to Claim 13 in which the reaction is selected from aikylation, polymerisation, etherification, esterification and condensation.