Description of the prior art A multitude of valuable products are produced by the chemicals industry today including fine chemicals, pharmaceuticals, agrochemicals, foodstuffs and bulk chemicals, generating billions of dollars per annum.

The vast majority (about 90%) of these chemicals come into contact with a catalyst at some stage of their production. The field of catalysis is thus vital for the economic competitiveness and prosperity of industrialised nations.

A catalyst, which is usually an enzyme or a transition metal complex, is added to reactions in order to increase the reaction rate. In some cases, the length of time that a given chemical reaction would be expected to take can be changed from hundreds of years

to a few minutes, or even seconds, by the addition of a catalyst. Catalysts are typically classified as either homogeneous or heterogeneous. Homogeneous catalysts are those in which the catalyst, reactants and products are all present in the same phase, for example where all species are in solution and there is no particulate solid. With heterogeneous catalysts, the catalyst is in a different phase to the reactants and/or products.

Typically, the catalyst is a solid and the reactants and products are liquids or gases or solids dissolved in solution.

A major objective in the design of new catalysts is to provide for efficient catalyst recovery thus minimising both catalyst cost and reaction product contamination (Applied Homogeneous Catalysis with Organometallic Compound, B. Cornils and W. A. Herrmann (Eds. ), VCH, Weinheim, 1996). In general, for heterogeneous catalysts it is easy to separate the catalyst from the products, for example, by filtration, because the catalyst is usually a solid and the product is usually in a liquid phase. Typically, therefore, reaction product contamination with the catalyst arises where it is difficult or very expensive to remove the catalyst from the product because they are in the same phase.

The problem occurs commonly in homogeneous catalysis and is the most serious disadvantage of the technique.

For transition metal catalysts based on expensive metals such as rhodium, palladium or platinum, the issue of cost is a very important factor as even losses in the parts per million range can render the entire process economically unfeasible. For catalytic

processes in the food and pharmaceutical industries, product contamination must be completely eliminated.

One particularly attractive approach to addressing the two related issues of catalyst recovery and product contamination is to immobilise catalytically active transition metal complexes onto suitably stable solid supports. This approach potentially combines the advantages of both homogeneous catalysis (e. g. selectivity, and/or specificity for one particular reaction product over any others) and heterogeneous catalysis (i. e. facile catalyst recovery post- reaction).

In the past some progress in this area has been made by employing either porous solids (Catalytic Chemistry by B. C. Gates, Wiley-Interscience, New York, 1992), including zeolites (J. M. Andersen and A. W. S. Currie, J. Chem.Soc., Chem. Commun., 1996, 1543) and aluminium phosphonates (S. P. Elangovan and V. Murugesan, J.

Molec. Catal. A - Chemical, 1997,118, 301), inorganic oxides such as silica (S. Shinoda, T. Kojima and Y.

Saito, Stud. Surf. Sci. Catal., 1981, 7,1504) or alumina (P. M. Lausarot, G. A. Vaglioand M. Valle, J.

Organomet. Chem., 1981, 215,111), or organic polymer supports (T. Uematsu, T. Kawakami, F. Saitho, M. Miura and H. Hashimoto, J. Molec. Catal., 1981, 12, 11).

However, there are often serious drawbacks associated with the use of such systems. One is the dissociation (leaching away) of the catalytically active transition metal from the solid support during catalysis and a second is the partial or total degradation of the

support material under the reaction conditions where, typically, high pressures and temperatures are employed.

The use of organic polymers as solid supports offers considerable potential in catalyst design because there is the possibility of structurally modifying the support material itself. Such modifications could be introduced to deliberately alter the behaviour of the catalyst in such a way as to be beneficial to the reaction in question. Structural modification is much more difficult, or even impossible, to achieve with supports based on inorganic materials such as oxides or zeolites.

Previously, the use of organic polymer supports in the design of immobilised catalysts has been largely restricted to immobilising phosphine ligands. Here the structural modification has been most commonly introduced at the level of the phosphine moiety rather than to the actual support material. While the immobilised polydentate phosphine systems (which co- ordinate through more than one position) show less susceptibility to metal dissociation than the monodentate versions (which co-ordinate through only one position), some dissociation is always observed (W.

H. Lang, A. T. Jurewicz, W. 0. Haag, D. D. Whitehunt and L. D. Rollmann, J. Organomet. Chem., 1977, 134,85).

Furthermore, side reactions which prevent the phosphine P-atom from serving as a donor of electron density to the catalytically active metal ion, for example, oxidation in the presence of atmospheric oxygen and/or

water, are particularly vexatious, compared to the situation in solution where excess phosphine is used.

Essentially the problem is acute because the stoechiometric ratio of the metal to polymer support is directly dependent upon the state of the resin-bound phosphine ligands.

Dinuclear rhodium (II) complexes are particularly interesting as they are able to catalyse a rich diversity of hydrogenation, carbonylation and cycloalkylation reactions (Homogeneous Catalysis with Complexes of Rhodium and Iridium, R. S. Dickson, Kluwer Academic Publishers, New York, 1983). While phosphines are often added to generate a catalytic rhodium (I) species, there is much evidence to suggest that rhodium (II) can also catalyse such chemistry (C. Claver, N.

Ruiz, P. Lahuertaand E. Peris, Inorg. Chim. Acta, 1995, 233,161). Dirhodium (II) tetraacetate, which is obtained as the crystalline solvate 1 (see scheme I below) from methanol, is a particularly convenient source of Rh (II) in that it is easy prepare and air- stable. Its structure is interesting in that the four carboxylate groups form -bridges such that each donates one oxygen atom to each of the two metal ions.

In order to overcome some of these problems it is now proposed to use either functionalised ligands or functionalised polymer supports in which the donor atoms, derived from either the ligand or the polymer support, are both redox stable and are positioned carefully to allow very tight binding to the complex.

Objects of the invention A first object of the invention is a functionalised support for use in the preparation of a supported metallic complex which comprises a polymer backbone bearing at least one functionalised site able to react with and bind to at least one metallic species. By mmetallic species" we mean one or more metal atoms or one or more metal ions (having any possible oxidation state) and also includes such metal atoms and ions in the form of a complex.

It is particularly preferred that the functionalised site comprises at least two branched groups X and Y, each bearing a functionalised group able to react with and bind to at least one metallic species, and preferably to a bimetallic complex.

It is further preferred that each of the functionalised groups (attached to said X or Y groups) are each able to chelate at least two metallic species and each functionalised group comprises two donor atoms. The term"donor"'is intended to describe atoms having a good ability to transfer electron (s) to other atoms in order to form chemical bonds therewith.

It is particularly preferred to use as functionalised groups carboxylate or acetate moieties for binding to dinuclear complexes.

It is further preferred that the X and Y groups are separated by a spacer which is a non-reactive group.

Advantageously, the spacer is an aromatic ring. While the spacer is preferably a trifunctionalised derivatised benzene ring other trifunctionalised molecules could be used as a substitute, for example, glycerol derivatives.

Advantageously a functionalised group Z is attached to said spacer and is the only connection of the spacer and the X and Y groups to the polymer backbone. It is also preferred that the group Z is an heteroatom chosen from the group consisting of oxygen, nitrogen and sulphur to allow easy attachment of the resin material to the ligand-linker. Z could also contain other atoms that connect the spacer of the ligand-linker to the resin material. For example, Z could also be an aromatic hydroxymethyl or aminomethyl group where, preferably, the heteroatom would be linked to the polymer. However, other types of connection are not excluded.

It is preferred that the functionalised groups attached to said X and Y groups are carboxylate groups. Also, X and Y can be disposed in a meta- or ortho- position relative to each other. The meta- position is preferred. X is advantageously chosen from the group consisting of -CH2O-, -CHR-O-, -CH2CH2- and -CHRCH2-.

Y is advantageously chosen from the group consisting of -CH2O-, -CHR-O-, -CH2CH2- and -CHRCH2-.

An especially preferred functionalised site comprises a bis-carboxylalkylated meta-substituted dihydroxybenzene

group. Systems based upon bis-carboxyalkylated meta- substituted dihydroxybenzenes, e. g. where X and Y have the structure (-CH2O- or -CHR-O-) are particularly preferred because the introduction of the carboxylate moieties into the ligand-linker is easier preparatively.

Also, it is preferred that the metallic species be one of a transition metal like rhodium or ruthenium.

It is further preferred that the polymer backbone is a Merrifield resin. Advantageously the polymer support may be a Merrifield resin, a modified polytetrafluoroethylene or any other organic polymeric material that would survive the reaction conditions needed to support a given catalytic reaction.

Another object of the invention is a metallic species attached to a support as described above, and preferably having the structure A shown below.

A further object of the invention is a method of synthesis of a supported metallic complex, the method comprising the steps of : - reacting under suitable conditions a metallic species with a functionalised support of the invention described above.

A further object of the invention is a metallic complex comprising at least one metallic atom and a ligand capable of being attached to a polymer support. More

particularly the ligand may comprise at least two branched groups X and Y, each bearing a functionalised group which is bound to said metallic atom.

It is preferred that the functionalised groups attached to both X and Y groups each comprise two donor atoms which chelate at least two metallic atoms. Thus, it is preferred that said ligand is able to react with and bind to a bimetallic complex. The metal preferred is rhodium or ruthenium and it is especially preferred that the functionalised groups attached to said X and Y groups are carboxylate groups. Such groups are particularly for binding to dinuclear complexes derived from metal ions and/or metallic complexes like, especially, dirhodium tetraacetate.

It is further preferred that the X and Y groups are disposed in a meta- or ortho- position relative to each other. It is also preferred that X is chosen from the group consisting of -CH2O-, -CHR-O-, -CH2CH2- and - CHRCH2- and that Y is chosen from the group consisting of -CH2O- -CHR-O-, -CH2CH2- and -CHRCH2-.

In a specially preferred aspect X and Y groups are separated by a spacer which is a non-reactive group.

The spacer is preferably an aromatic ring like benzene, and the ligand thus comprises a trifunctionalised derivatised benzene ring. However other trifunctionalised molecules could be used as substitutes, for example glycerol derivatives.

Advantageously the spacer comprises a functionalised group Z'which can easily react with and bind to said polymer support. More particularly the spacer can be aromatic ring and the group Z'may a hydroxyl moiety.

Thus when linked to the polymer backbone the connecting group Z (derived from Z') is an oxygen atom. Also the group Z may be another heteroatom like nitrogen or sulphur to allow easy attachment of the resin material to the ligand-linker. Z could also contain other atoms that connect the third functionalised site of the ligand-linker to the resin material. For example, Z could also be an aromatic hydroxymethyl or aminomethyl group where, preferably, the heteroatom would be linked to the polymer and other types of connection are not excluded.

A particularly preferred ligand comprises a bis- carboxylalkylated meta-substituted dihydroxybenzene group. Systems based upon bis-carboxyalkylated meta- substituted dihydroxybenzenes, e. g. where X and Y have the structure (-CH2O- or -CHR-0-) are particularly preferred because the introduction of the carboxylate moieties into the ligand-linker is easier preparatively.

Another object of the invention is a metallic complex attached to a polymer support as described above, and preferably having the structure A shown below.

Advantageously the polymer support may be a Merrifield resin, a modified polytetrafluoroethylene or any other

organic polymeric material that would survive the reaction conditions needed to support a given catalytic reaction.

A further object of the invention is a method of synthesis of a supported metallic complex of the invention, said method comprising the step of : - reacting under suitable conditions a metallic complex described above with a polymer support.

A further object of the invention is the use of the metallic complexes described above as catalysts. Such catalysed reactions are, for example, hydroformylation, isomerisation and reduction reactions. z\eya ""-y spa= 'o M=NUal Ion Structure A Preferred embodiments of the invention will now be described below in more detail.

Example A: Synthesis of a resin-bounded dirhodium catalyst using dirhodium tetraacetate as starting material and use of the supported catalyst obtained in hydroformylation and isomerisation reactions.

In this example two of the carboxylate ligands of the dirhodium tetraacetate of structure 1 are replaced by a templated dicarboxylate which positions each carboxyl group in a position similar to adjacent carboxyl groups in the tetraacetate complex of structure 1 (i. e. at approximately 90° one from each other as shown in structure 2). This results in very tight binding of the complex. Additionally, appropriate functionalisation of the dicarboxylate moieties through linker Z allows to anchor the dinuclear metal ion assembly to the polymer via the dicarboxylates, in a form close to the natural structure of corresponding dirhodium tetraacetate complex of structure 1. The optimum metallocycle sizes suitable for a series of dicarboxylate chelates can be assessed using computer assisted molecular modelling. Meta-disubstituted benzenes used in structure 3 are capable of facilitating the formation of two -coordinated 11- or

12-membered rings and cause little perturbation to the positions of the other carboxylate groups in dirhodium (II) tetraacetate.

A. 1. Synthesis of a functionalised Z=O-Polymer support As shown in Scheme 1 below, the readily available 3, 4- dihydroxy- or 3, 5-dihydroxy- phenylacetic acids 4 are a suitable starting materials for the preparation of supported catalysts of type 3 where Z = 0-polymer. C02H C02Me C02Me + Regioisomer for the material C02R derived from the OH' OH' OJ 3, 4-dihydroxyphecrylacetic ester. 4 5 6, R = Me or t-Bu COabAe COaH C02R iv. C02H 3 / J/pJ Z = O-Polymer Polyme 7 Polyme 8 7 8 Scheme 1 Reagel1ts i.MeOH,H2SO4,reflwc3h; ii.NaHBrCH2CO2RTHE,25°C, 16h,thenSiO2chromatog.; iii. choromethylated polystyrene, NaR DMF, 60°C, 3h; iv. a) 0.1 mol dm~3 NaOH in 9:1 THF/H 2° 25°C, 72h, or, step iv. a) followed by 10% TFA in DCM, 25°C, 2h ; v. Rh2(OAc)4(MeOH)2

In summary, the methyl esters 5 can be mono O-alkylated with either methyl- or t-butyl- bromoacetic ester and the isolated phenolic diesters 6 are reacted with chloromethylated polystyrene (Merrifield resin) in the presence of sodium hydride (D. Stones, D. J. Miller, M.

W. Beaton, T. J. Rutherford and D. Gani, Tetrahedron Lett. , 1998, 39, 4875) to give the resin immobilised

bis-carboxylic esters 7. These esters 7 are obtained in quantitative yield from the diesters 6 as judged using a fluorophenol-based gel-phase NMR assay (D. Stones, D.

J. Miller, M. W. Beaton, T. J. Rutherford and D. Gani, Tetrahedron. Lett., 1998, 39,4875), as well as by the mass increase of the resin material, and displayed the expected strong IR stretching absorbances at about 1750 cm~1. Hydrolysis of the two ester functionalities gives the resin-bound diacids 8 which upon treatment at 25 °C with rhodium acetate in THF, a green solution, slowly darkened in colour over many days. Upon refluxing, the uptake of Rh (II) is completed within 48 hours and control experiments performed using underivatised resin showed no Rh (II) uptake whatsoever.

Detailed synthesis Preparation of compound 5 (Methyl 3, 4- dihydroxyphenylacetate). To a stirred solution of 3, 4-dihydroxyphenylacetic acid (2. 00 g, 11. 9 mmol) in methanol (50 cm3) was added a drop of concentrated sulphuric acid and the solution was heated under reflux for 3 h. After cooling, the methanol was removed under reduced pressure and the residue was redissolved in ethyl acetate (50 cm3).

This solution was washed with 5% sodium hydrogencarbonate solution (50 cm3), water (50 cm³) and brine (50cm³). The separated organic layer was dried(MgSO4), filtered and concentrated to give a dark- brown oil that was purified by flash chromatography on silica gel (eluting with 2 : 1 petroleum ether : ethylacetate) to give the required compound as a colourless oil that crystallised upon standing (1. 86 g, 86%).

Spectroscopic data : AH (300 MHz ; CDC13) 3. 51 (2H, s, CH2), 3. 71 (3H, s, C02CH3), 6. 02 and 6.22 (2 x 1H, 2 x br s, 2x OH), 6. 62-6. 65 and 6. 72-6. 75 (3H, m, Ar-H);8 (75. 4 MHz ; CDC13) 40. 35 (CH2), 52. 34 (C02CH3), 115. 52, 116. 36 and 121. 78 (Ar-CH), 126. 09 (quaternary ArC), 143. 21 and 143. 90 (quaternary Ar C-OH) and 173. 93 (C02CH3) ; vmax (thin film)/ cm 1 3384s (OH stretch), 1714s (C=O stretch), 1610m, 1519s, 1442s, 1347s, 1284vs, 1198s, 1148s, 1112s, 1012w, 962w, 872w and 795w.

Preparation of compound 6: To a stirred solution of methyl 3, 4-dihydroxyphenylacetate (203 mm³, 1.38 mmol) and tert-butylbromoacetate (500 mg, 2. 75 mmol) in dry

tetrahydrofuran (20 cm³) at 5 °C under a dry nitrogen atmosphere was added sodium hydride (60% dispersion in mineral oil, 220 mg, 5. 50 mmol) and the mixture was stirred and allowed to warm to room temperature over 16 hours. Water (5cm3) was added and the solvent was removed under reduced pressure. The residue was partitioned between ethyl acetate (30cm³) and water (30 cm3) and the separated aqueous layer was further extracted with ethyl acetate (30 cm³). The pooled organic extracts were washed with brine (60 cm³), dried (MgSO4), filtered and concentrated to give a brown oil that was purified by flash chromatography on silica gel (eluting with 2 : 1 petroleum ether : ethyl acetate) to give the required compounds as a colourless oil (1 : 1 mixture of regio isomers, 115 mg, 28%).

Spectrometric data: AH(300 MHz; CDCl3) 1.48 (9H, s, C (CH3) 3), 3. 51 and 3. 52 (2H, 2 x s, CH2C02CH3), 3. 68 (3H, s, C02CH3), 4. 50 and 4. 53 (2H, 2 x s, OCH2CO2C (CH3) 3), 6. 71-6. 92 (3H, m, Ar-H), 7. 32 and7. 38 (2 x 1H, 2 x br s, Ar C-OH); #C (75.4 MHz; CDCl3) 27.87 (C(CH3)3), 40.29and 40.49 (CH2CO2CH3), 51.90 (CO2CH3), 69. 53 and 69.66 (OCH2CO2C(CH3)3), 83.31 (C(CH3)3), 115. 2, 116. 3, 117. 3, 118. 0, 120. 9 and 125. 01 (ArCH), 125. 7 and 130. 1 (quaternary Ar C), 145. 7, 146. 4, 147. 2 and 148. 0 (Ar C-OH and Ar COCH2), 170.0, 170.2, 172.1 and 172.3(CO2CH3and CO2C(CH3)3); vmax(thin film)/cm-1 3359br s (OH stretch), 2981s (CH stretch), 1736vs (C=O stretch).

Preparation of compound 7 : To a stirred mixture of Merrifield resin (1. 00 g, 1. 10 mmol substitution) and a 1 : 1 molar mixture of the preceding isomeric diesters 6 (4. 14 g, 5. 95 mmol) in dry 4-N,N-dimethylformamide (50 cm3) under a dry nitrogen atmosphere was added sodium hydride (60% dispersion in mineral oil, 286 mg, 7. 14 mmol) and the mixture was heated with stirring at 60°C for 3 hours and then allowed to stir at room temperature for 16 hours. There action was quenched by with water (1 cm3) and the resin was collected by filtration under. After exhaustive washing with water, followed by methanol and finally dichloromethane, the resin was dried under vacuum at 50 °C. The mass of resin isolated was 1. 28 g (expected mass 1. 28 g - 100% substitution by mass increase).

Spectrometric data : vmax (KBr disc)/cm-13057s, 2979s, 1944w, 1870w, 1752s (C=O stretch), 1599m, 1508s, 1491s, 1449s, 1375m, 1305m, 1247m, 1211m, 1115s, 1063s, 1000s, 963s, 940s, 903s, 840m, 817m and 741s.

Preparation of compound 8 : To a mixture of tetrahydrofuran (10 cm3) and 1 mol dm-3 sodium hydroxide solution (1 cm3) was added methanol, dropwise with shaking, until the solution became homogeneous. A 1 : 1 molar mixture of resin bound esters 7 (293 mg, 0. 226 mmol substitution) was then added and the mixture was shaken at room temperature for 3 days.

The resin was collected by filtration and then washed exhaustively with water, followed by methanol and finally dichloromethane. After drying at 50 °C under vacuum the resin was resuspended in a 10% solution of trifluoroacetic acid in dichloromethane (10 cm3) and this mixture was shaken at room temperature for 2 hours. Again the resin was collected by filtration and was thoroughly washed with water, followed by methanol and dichloromethane. After drying under vacuum at 50 °C, the resin was treated with rhodium (II) acetate as described below.

Spectrometric data : (vmax (KBr disc)/cm-1 3436s, 3021s, 2917s, 1943m, 1870m, 1779s, 1734s, 1600s, 1490s, 1450s, 1339s, 1265s, 1217s, 1132s, 1067s, 1026s, 904s, 818s and 745vs)

A. 2 Synthesis of a supported dirhodium complex catalyst (compound 3) Preparation of compound 3 (Z = 0-polymer) : ~-o solvent 0 t/ solvent/2 olystyrene To a solution of dirhodium (II) tetraacetate (20 mg, 40mmol) in dry tetrahydrofuran (25 cm3) was added a 1 : 1 molar mixture of resin bound diacids 8 (45 mmol substitution of required isomer). This mixture was heated under reflux for 4 days and upon cooling the resin was collected by filtration under reduced pressure. The resin was thoroughly washed with tetrahydrofuran and the pooled filtrates collected to recover any remaining rhodium (II) acetate dimer. The resin was then exhaustively washed with water then methanol and then finally dichloromethane, and then dried at 50 °C under vacuum. The resin recovered was a distinctive dark green colour and was isolated in a mass of 136 mg (maximum possible recovery was 151 mg).

Spectrometric data : vmax (KBr disc)/cm-1 3431m, 3019m, 2921s, 2852s, 1945m, 1872m, 1779s, 1739s, 1720m, 1636m, 1597s, 1489s, 1445s, 1342s, 1263s, 1214s, 1136s, 1067s, 1023s, 900m, 836m, 817m and748vs.

As a control experiment, to a solution of rhodium (II) acetate dimer (20 mg, 40 mmol) in dry tetrahydrofuran (25cm3) was added unsubstituted Merrifield resin (125mg, 102 mmol substitution). This mixture was also heated under reflux for 4 days but upon identical work up conditions to those described above, the resin showed no colour change or any mass increase.

A. 3. Hydroformylation catalysis of hex-1-ene using the resin-bounded catalyst 3 Hydroformylation Catalysis Using compound 3 H O O 0 j o w + ms polyme-Rh 2(OAc) 2 (3) 1.33 1.00 PEt 3, tduerie) 1 (50. 70/.) (38. 1%) CtaH 2 40bar) B C 80'C, 10h + 0. 30 (11. 70/.) D To 70 mg of 3 (assuming maximum 20 % by weight of Rhodium) in a glass-lined autoclave was added 1 cm of substrate (hex-1-ene), 4cm3 of dry, degassed solvent (toluene) and 10 mm3 (about 3-fold molar excess) of phosphine (PEt3). The autoclave was sealed and pressurised to 40 bar with synthesis gas (1 : 1 CO : H2) and heated to 80 °C, with stirring. Stirring was continued at this temperature, under pressure, for 10 hours, after which time the autoclave was cooled to

room temperature, vented and the reaction products analysed by GC-MS and 1H and 13C NMR spectroscopy. The catalyst was recovered by filtration and dried under argon.

The catalytic hydroformylation of hex-1-ene was performed using synthesis gas (1 : 1 CO : H2), quantitative conversion to the corresponding heptanal isomers was observed after 10 hours, with a linear to branched chain ratio of about 1 : 1, as demonstrated in the above reaction scheme. Heptan-1-al and heptan-2-al derived from the hydroformylation of hex-1-ene, whereas heptan-3-al derived from the initial isomerisation of hex-1-ene to hex-2-ene, followed by the hydroformylation of hex-2-ene.

Table 1 : Hydroformylation of 1-hexene using polymer- Rh2(OAc) (compound 3) as catalyst Catalyst* T Rea- Pressure 2- Alde- Ratio of Ratio (°C) ction (bar) Rea- Hexene hydes Aldehydes N : iso Time ction % % B : C : D (h) 'Ya'80To"10100'01004.3:3.3:11:1 3a 120 4 40 50 50 50 5. 9 : 3. 7 : 1 1. 3 : 1 3b 80 10 40 100 0 100 5. 0 : 3. 3 : 1 1. 1 : 1 Recov'd 80 10 40 68 37 63 3. 9 : 1 : 0 3. 9 : 1 3a Recov'd 80 10 40 45 13 87 2. 7 : 1 : 0 2. 7 : 1 3b

*Typically, reactions contained 1-hexene (l. Og) and about 70mg of polymer supported catalyst, slurried in 4

cm3 of toluene with 10 mm3 of PEt3 (3a) or 20 mm3 of PEt3 (3b).

This result compares well with many catalyst systems based on Rh (I). Moreover, the new supported catalyst 3 (Z = 0-polymer) displayed several superior properties in selectivity compared to the control homogeneous catalyst derived from rhodium acetate 1. For example, the extent of substrate isomerisation by the supported catalyst 3 (Z = 0-polymer), to give 2-hexene, was significantly reduced, see Table 1. Also, the extent of product aldehyde hydrogenation to give the heptanols was completely suppressed. Interestingly, this latter observation accords with findings using catalysts supported within zeolites (J. M. Andersen, Platinum Metals Rev., 1997, 41,132). However, the rates of reaction for the polymer supported system 3 (Z = 0- polymer) were much higher than for the zeolite system.

Thus, it is evident from the results that the new system gives high rates of reactions, catalyses cleaner chemistry than the parent 1 and has the additional extremely desirable advantage over conventional homogeneous systems of being tightly bound to the polymeric support. Therefore, it is easy to separate at the end of there action by simple filtration. It should be noted that a resin bound mono carboxylic acid was able to form a complex with rhodium acetate but this was not stable under that hydroformylation reaction conditions. The tight binding properties displayed by the supported catalyst 3 (Z = 0-polymer) is expected to increase the conformational rigidity of catalytic intermediates, which could offer additional

advantage in asymmetric catalysis.

A. 4 Isomerisation Catalysis The polymer-supported catalyst 3 was also tested for catalytic activity in the isomerisation of hex-1-ene, and after 10 hours at 80 °C, 69% conversion to hex-2- ene was observed.

Isomerisation Catalysis Using Compound 3 polymer-Rh 2(0Ac) 2 (3) toluene 80 °C, 10 h To 70 mg of 3 in a Schlenk tube, under argon, was added 1 cm3 of substrate (hex-1-ene) and 4 cm3 of dry, degassed solvent (toluene). The resultant slurry was then heated (with stirring) to 80 °C and allowed to remain at this temperature for 10 hours. After this time, the reaction mixture was cooled to room temperature and the reaction products analysed by GC-MS and 1H and 13C NMR spectroscopy. The catalyst was recovered by filtration and dried under argon.

A. 5 Conclusion The dark green-coloured resin bound complexes 3 (Z = O- polymer) containing about 8-16 % Rh (II) was assessed as a catalyst for hydroformylation and isomerisation reactions. The results obtained from these experiments

demonstrate that the system serves as a catalyst which is not only extremely active, but is also very stable such that the rhodium complex remains tightly bound to the polymer and is thus easy to recover post-reaction.

Indeed, the recovered supernatant liquid was analysed for rhodium content using atomic absorption spectroscopy and the extent of leaching was negligible (<0. 01%). The catalyst system was re-used in several repeat reactions and gave essentially identical results.

Other resin-bound rhodium catalysts were prepared from other polystyryl methyloxyarenes possessing two appended carboxylic acids, see Example C.

Example B : Use of supported dirhodium (DIOP) catalyst in the reduction of acetophenone and study of the stereoisomerism of the reaction.

Asymmetric reduction of C=0 bonds forming chiral alcohols and amines is among the most fundamental molecular transformations (Asymmetric Synthesis, J. D.

Morrison, Academic Press, New York, 1983). In nature, oxidoreductases such as horse liver alcohol dehydrogenase (An Introduction to Enzyme and Coenzyme Chemistry, Tim Bugg, Blackwell Science, London, 1997) catalyse transfer hydrogenation of carbonyl compounds to alcohols using cofactors like NADH or NADPH. Such biochemical reactions are normally very stereoselective. However, organic synthesis needs economically and technically more beneficial methods that are more general. Hydride reduction provides a

useful complement to catalytic reduction using molecular hydrogen, particularly for small- to medium- scale reactions. Transfer hydrogenation is operationally simple, and the selectivities including functional group differentiation may be different from those of hydrogenation.

The Meerwein-Pondorf-Verley reduction (J. G. Aston, J.

D. Nekirk, J. Am. Chem. Soc., 1951, 73, 3900) is catalysed by aluminium alkoxydes and using iso-propanol as the hydride source, Scheme 3. This reaction takes place under mild conditions, room temperature, atmospheric pressures and uses iso-propanol, which is easy to handle and dissolves many organic compounds.

Also, since acetophenone 9 is a prochiral ketone, attempts have been made to develop asymmetric catalysts using chiral rhodium complexes (P. Gamez, F. Fache, M.

Lemaire, Tetrahedron : Asymmetry, 1995,6, 705 ; R.

Noyori and S. Hashiguchi, Acc. Chem. Res. , 1997,31, 97).

Use of this approach for the present invention is exemplified in Scheme 4.

Scheme3:MPVredllction of acetophenone

H CH Resin-DIOP*-Rh * + C CO 3 I CHg 2 9 10 * = chiral centre Scheme 4:Asymmetrlc reduction of acetop11enone using chiral rhodium complexes B. 1 Synthesis of Rhodium- (-) DIOP 2. 3-0-Isopropylidene- 2,3-dihydroxybis(diphenylphosphino) butane] and preparation of resin supported dirhodium- (-) -DIOP complex : Tests were carried out on the catalytic activity of the resin bound dinuclear rhodium chelated with (-) DIOP (2. 3-O-Isopropylidene-2, 3-dihydroxybis (diphenylphosphino) butane). For this ligand, proceed to attach the ligand (-) DIOP to the metal prior to attachment on the polymer support proved to be the more effective synthetic pathway. Such method could also be applied to the synthesis of compounds (3). For example compound 6 may bet treated with NaOH to cleave the ester protecting group. The free diacid thus obtained may be treated with the rhodium tetraacetate complex 1 to synthesis a bimetallic complex of the invention which can eventually be attached to a resin support.

One equivalent of (-) DIOP (2. 3-0-Isopropylidene- 2, 3-dihydroxybis (diphenylphosphino) ) butane were first coordinated to the rhodium tetraacetate by heating under reflux in tetrahydrofuran followed by the attachment onto the resin. Tests carried out using this chiral catalyst are summarised below in Table 2 below.

Rhodium tetraacetate (100 mg, 0. 226 mmol) were dissolved in dry tetrahydrofuran (100 cm3). The solution was heated under reflux overnight, and then concentrated to 20 cm3. Addition of petroluem ether (80 cm3) caused precipitation of the complex. The product was filtered, washed with petroleum ether and then dried under vacuum at 50 °C.

Activated resin (60 mg, 0. 066 mmol) and the above, dirhodium- (-) -DIOP complex (58. 1 mg, 0. 066 mmol) was suspended in the dry tetrahydrofuran (25 cm3) and the mixture heated under reflux for 4 days, with constant stirring. The resulting purple resin was filtered and then dried under vacuum at 50 °C (82 mg).

B. 2 Hydrogenation catalytic tests Rhodium- (-) -DIOP supported resin (19. 54 mg, 21.5 mol) and potassium hydroxide powder (7. 23 mg, 0. 129 mmol) were suspended in degassed iso-propanol (5 cm3) under an inert atmosphere and stirred for 15 min.

Acetophenone (50 mm3, 0. 43 mmol) was added to the solution and the resulting mixture was allowed to stir at 70 °C under inert atmosphere for 19 h.

B. 3. Measurement of enantiomeric excesses Reaction conversions and enantiomeric excesses were determined by gas chromatography using a chiral capillary column CPCHIRASIL-DEX CgTM.

Oven conditions 100 °C for 3 min, then 5 °C per min. up to 140 °C.

Injector : 240 °C.

Detector : FID, 250 °C.

Carrier gas : Helium.

Pressures : main valve : 18 psi ; column valve : 10 psi ; make up gas : 5 psi.

B. 4 Results Table 2 : Supported catalytic reduction of acetophenone Resin Solvent Temp. (°C) Yield max Enantiomeric (%) excess (%) Resin-Rh-DIOP(-)iPrOH Room 4 3 temperature Resin-Rh-DIOP(-)jPrOH/toluene Room 8 2 temperature Resin-Rh-DIOP(-)jPrOHTo964 Resin-Rh-DIOP(-) iPrOH/toluene 70 97

These supported catalyst shows a good activity in the hydride reduction of acetophenone using low metal/substrate ratios. It is easy to recover the catalyst post-reaction by simple filtration and it can be re-used successfully. The hydrogenation reaction described here takes place under mild conditions (low temperature and atmospheric pressure) and provides a useful complement to catalytic reduction using molecular hydrogen. Enantiomeric excesses, are comparable to those obtained in homogeneous experiments, indicating that the presence of the resin

does not improve the enantiotopic face differentiation of the acetophenone.

Example C : Use of a dirhodium based chiral-linker catalyst in the reduction of acetophenone and study of the stereoisomerism of the reaction.

New unsupported dimetallic chiral-linker compounds having a ligand suitable to be attached to a support and having a structure similar to structure A have been studied for their catalyst and their abilities for asymmetric reduction.

C. 1 Synthesis of the dirhodium based chiral-linker catalyst 11 The synthesis of the rhodium based chiral-linker solid support 11 is shown below in scheme 5. Using the method of Archer (C. H. Archer, N. R. Thomas and D. Gani, Tetrahedron : Asymmetry, 1993,4, 1141) (2S) -alanine 12 is diazotised in the presence of bromide ions to give (2S) -bromopropanoic acid. Treatment with diazomethane in diethyl ether then gives the bromoester 13 in 76% yield as a colourless oil, Scheme 5. The 3,5- dihydroxyphenylacetate is treated with sodium hydride in tetrahydrofuran and then reacted with the bromoester 13 to give the diester 14 as a pale yellow oil in 36% yield. The resulting diester is saponified with sodium hydroxide to give the free diacid 15, which is reacted, without further purification with dirhodium tetracetate in tetrahydrofuran to give the desired chiral support 11.

This complex 11 was then tested as chiral catalyst in the reduction of acetophenone, following the same reactional scheme than the one shown in scheme 4.

Although, 95% conversion was achieved after 6 h, the 3% enantiomeric excesses (ee) was small. The enantiomeric excesses were determined by gas chromatography using a chiral capillary column CP CHIRASIL-DEXTM as described above.

Scheme 5 Reagents & Conditionsi. NaNO2, HBr, KBr, 0°r. t., 12 h, 84%; ii-CH2 N2, Et O, 2 h, 90% iii. NaH, THF, 50 °C, 16 h, 36% ; iv. l M NaOH, MeOH, r. t., 12 h;v.Rh2(OAc)4, MeOH 65 °C, 12 h, 95% froml4

Synthesis of (25J-Bromopropanoic acid (2S) -Alanine 7 (4. 0 g, 45 mmol) was added to a saturated solution of potassium bromide (10 cm³), followed by the dropwise addition of hydrogen bromide (15 cm3 of a48% solution). The resulting mixture was then cooled to 0 °C and sodium nitrite (6. 21 g, 90 mmol) was added over 2 h. The reaction mixture was maintained below 5 °C for a further 1 hour, and then allowed to warm to room temperature overnight. The resulting solution was then extracted with diethyl ether (3x25 cm³). The combined ether extracts were then dried (MgSO4) and concentrated under reduced pressure to give a pale yellow oil. Distillation under reduced pressure gave pure (2S) -bromopropanoic acid as a colourless oil (5. 2 g, 84%) ; la D-17. 5 (c 2. 0 in CHC13) {lit., (C. Archer, N. Thomas and D, Gani, Tetrahedron : Asymmetry, 1993,4, 1141) -29. 1 (c 2. 0 in CHC13)}.

Spectroscoplc data : vmax(CHC13)/cm~l 3115(OH) and 1740 (CO); °H (300 MHz ; CDC13) 1. 82 (3H, d, J 7, CHCH3), 4. 38 (1H, q, J 7, CHCH3) and 11.93 (1H, br, COOH) ; 8 (50. 3 MHz ; CDC13) 23. 2 (CH3), 37. 9 (CHCH3) and 176. 5 (COOH) ; m/z (EI) 152 and 154 (22%, bromine isotopes, M+), 108 andllO (64%, bromlne isotopes, M - C02H +) and 45 (100%, C02H+).

Synthesis of the methyl (2S) -bromopropanoate 13 To a solution of (2S) -bromopropanoic acid (1 g, 6. 53mmol) in ether (10 cm3) was slowly added an ethereal solution of diazomethane until the yellow colour persisted. The reaction mixture was stirred for 10 min, and the excess diazomethane was removed in a stream of nitrogen. The solution was then concentrated under reduced pressure to give a pale yellow oil which was distilled from calcium hydride to give methyl (2S) - bromopropanoate as a colourless oil (981 mg, 90%).

Spectrometric data : vmax (CHC13)/cm-1 1755 (CO) 8g (300 MHz ; CDC13) 1. 75 (3H, d, J 7, CHCH3), 3. 72 (3H, s, OCH3) and 4. 33 (1H, q, J 7, CHCH3); #C (50.3 MHz; CDC13) 22. 5 (CH3), 40. 2 (CHCH3), 53.3 (OCH3) and 170. 5 (COOCH3).

Synthesis of the methyl 03-methyl (2R) -propanoate-3,5- dihydroxyphenylacetate 14 To a stirred solution of methyl3, 5- dihydroxyphenylacetate (1.92 g, 10. 55 mmol) and

methyl (2S) -bromopropanoate 13 (881 mg, 5. 27 mmol) in dry tetrahydrofuran (30 cm³) at room temperature under argon atmosphere was added sodium hydride (60% dispersion in mineral oil,210 mg, 5. 27 mmol) and the mixture was stirred and heated at 50 °C over 16 hours.

Water (5 cm³) was added and the solvent was removed under reduced pressure. The residue was partitioned between ethylacetate (50 cm³) and water (50 cm³) and the separated aqueous layer was further extracted with ethyl acetate (40 cm³). The pooled organic extracts were washed with brine (60 cm3), drled (MgSO4), filtered and concentrated to give a brown oil that was purified by flash chromatography on silica gel (eluting with toluene - ethyl acetate 5/1) to give the required compound 14 as a pale yellow oil (514 mg, 36%).

Spectrometric data : I (XID +15. 2 (c 2. 0 in CHCl3); AH (300 MHz, CDC13) 1. 54 (3H, d, J 7, CHCH3), 3. 51 (2H, s, CH2), 3.55 (3H, s, CH2CO2CH3), 3. 76 (3H, s, OCHCO2CH3), 4. 73 (1H, q, J 7, OCHCH3), 6. 2-6. 4 (3H, m, Ar-H) and 6. 67 (1H, s, Ar-OH); bC (50-3 MHz; CDCl3) 18.7(OCHCH3), 41-2 (CH2CO2CH3) 52-9 (CH2CO2CH3), 53.4 (OCHCO2CH3), 73. 3 (OCHC02CH3), 101. 3, 107. 1 and 110.0(Ar-CH), 135.8 (Ar-C quaternary), 157. 5 (Ar-C-OH), 158. 8 (Ar-COCH), 171. 5 and 173 (C02CH3) ; m/z (ES) 291. 1 (100%, [M + Na +).

Synthesis of the bis-rhodium-diacetate-03-methyl(2R)- propanoate-3,5-diSydroxyphenylacetate 11 To a stirred solution of 14 (157 mg, 0. 58 mmol) in methanol (25 cm3) was added sodium hydroxide (1 mol dm-3, 2. 3 cm3) and the mixture was stirred overnight at room temperature. The mixture was acidifie using 1 mol. dm-3 HC1, the solvent was removed under reduced pressure. The aqueous layer was extracted with ethyl acetate (60 cm3). The pooled organic extracts were washed wlth brlne (50 cm3), drled (MgSO4) and concentrated under reduced pressure to give a pale yellow oil (132mg, 95%). To this oil was added dirhodium tetraacetate (128 mg, 0. 29 mmol) and dry tetrahydrofuran (30 cm3). The mixture was heated under reflux overnight under argon atmosphere. The solvent was then removed under reduced pressure and the resulting brown solid was dried under vacuum at 50 °C.

C. 2 Catalytic reduction of acetophenone using catalyst 11 O H Rh-Ligand* I * 3 + (C42CO KOHPrOH The chiral complex (0. 0215 mmol) and potassium hydroxide powder (7. 23 mg, 0. 129 mmol) were dissolved in degassed iso-propanol (5 cm3) under an inert atmosphere and stirred for 15 minutes. Acetophenone (50 mm3, 0. 43 mmol) was then added to the solution, and the resulting mixture was allowed to stir at room temperature under inert atmosphere. After 6 h the conversion rate and the enantiomeric excess (ee) of the mixture is analysed by gas chromatography as described above. The analyse shows a good conversion rate of about 95%, but a poor enantiomeric excess (about 3%).

Poor enantiomeric excesses may be due to a lack of rigidity of the chiral catalytic intermediates.

However in this case the use of a chiral ligand which bind tightly to the rhodium and thus increase the enantiomeric excess was shown not to be very effective.

Compound 11 may easily be attached to the resin through the free OH group using the same chemistry as described for compound 3.

Example D : Use of a new diruthenium based chiral- linker catalyst both unsupported and supported in the reduction of acetophenone and study of the stereoisomerism of the reaction.

Ruthenium catalysts have been reported to give better enantiomeric excesses than rhodium catalysts (Applied Homogenous Catalysis with Organometallic Compounds, B.

Cornils and W. A. Hermann (Eds), VCH (Weinheim), 1996).

Therefore the chiral diruthenium tetraacetate complex 16 corresponding to the dirhodium complex 11 describes in example C was synthesised (see also structure A, where M = ruthenium) in order to obtain better enantiomeric excesses.

The chiral dirutheniumtetraacetate complex 16 was synthesised in an identical manner to that described for the chiral dirhodium tetraacetate complex 11, using the diacid 15 and dirutheniumtetraacetate chloride (R.

Mitchell, A. Spencer and G. Wilkinson, J. C. S. Dalton, 1973,846.

D. 1 : Synthesis of catalyst 16 Synthesis of the bis-ruthenium-diacetate-fO3- methyl(2R)-propanoate-3,5-dihydroxyphenylacetate 16 Note : the double bonds shown at the earboxylate moieties represent the delocalising of the negative charge.

To a stirred solution of 14 (157 mg, 0. 58 mmol) in tetrahydrofuran (25 cm3) was added sodiumhydroxide (1 mol dm-3, 2. 3 cm3) and the mixture was stirred overnight at room temperature. The mixture was acidifie using 1 mol dm-3 HCl, the solvent was removed under reduced pressure. The aqueous layer was extracted with ethyl acetate (60 cm³). The pooled organic extracts were washed with brine (50 cm³), dried (MgSO4) and concentrated under reduced pressure to give a pale yellow oil (132 mg, 95%). To this oil was added

dirutheniumtetraacetate chloride (138 mg, 0. 29 mmol) and dry methanol (30 cm3). The mixture was heated under reflux overnight under argon atmosphere. The solvent was then removed under reduced pressure and the resulting brown solid was dried under vacuum at 50 °C.

D. 2 Catalytic reduction of acetophenone using ruthenium-based catalyst 16 0 Ru-Ligand* n4 CH3 +(C)2CO U KOHPrOH 9 10 Scheme 6 The chiral complex (0. 0215 mmol) and potassium hydroxide powder (7. 23 mg, 0. 129 mmol) were dissolved in degassed iso-propanol (5 cm3) under an inert atmosphere and stirred for 15 min. Acetophenone (50 mm3, 0.43mmol) was then added to the solution, and the resulting mixture was allowed to stir at room temperature under inert atmosphere.

After 2 hours the mixture is analysed by chiral gas chromatography :). The conversion rate was low (about 4%) but the enantiomeric excess was unexpectedly excellent (about 96%). Therefore the catalyst compound is a particularly preferred embodiment of the invention together with its obvious structural equivalents. For example the linker Z', (here an hydroxyl moiety) might

be omitted and/or replace by another chemical moiety.

When this ruthenium complex 16 was attached to a resin and then tested as chiral catalyst in the reduction of acetophenone, the conversion rate reached 50% but the enantiomeric excess was surprisingly only 3%.

It should also be noted that these catalysts 11 and 16 attached or unattached should also be effective in the reduction of cinnamate derivatives (A. Corma, M.

Iglesias, C. del Pino and F. Sanchez, J. Chem. Soc., Chem. Commun., 1991, 1253) 17, to give the corresponding chiral a-amino acids 18 as described in Scheme 7. NEIQOCi3 H 40 bars COOME 3 MaOOC \==/ Resin-Ligand-Rh 17 18 Scheme7: Amino acid synthesis The scope of the present invention is not limited to the above examples but also covers all chemically equivalent substitutions which would be obvious to these skilled in the art.