Diphosphine ligands for metal complexes The invention relates to ditertiary diphosphine ligands for metal complexes as hydrogenation catalysts which are soluble in a polar reaction medium, for example water and organic solvents such as methanol, due to solubilizing groups and which can easily be separated from the reaction mixture after the hydrogenation, possibly by changing the pH.

Hydrogenation processes, which may be enantioselective or diastereoselective, using metal complexes of, for example, rhodium, iridium or ruthenium and achiral or chiral ditertiary diphosphine ligands have become important processes for the synthesis of intermediates or active substances such as pesticides and pharmaceuticals. In addition to reactions in organic solvents, aqueous or polar reaction media are often the medium of choice because of the solubility of starting compounds, but this then also requires water-soluble hydrogenation catalysts which can easily be separated from the reaction product by extraction.

Some hydrogenation catalysts soluble in aqueous media are already known. Thus, K. Wan et al. in J. Chem.

Soc., Chem. Com., pages 1262-1264 (1993) and I. Toth et al.; Tetrahydron: Asymmetry, Volume 1, pages 913-930 (1990) describe ditertiary diphosphine ligands for such catalysts. In these ligands, the phenyl radicals in the phosphine groups are substituted by either amino or -SO3H. However, these substituents change the catalytic properties so that the desired result can often no longer be achieved.

EP-A-0 329 043 describes pyrrolidine diphosphines bound to polyethylene glycol via a diisocyanate bridge; the rhodium complexes of these are used for the hydrogenation of 2,3-dehydrophosphinothricin

derivatives in aqueous solution. The stoichiometric composition of the complexes is difficult to control, so that reproducibility is not always ensured. Pyrrolidine diphosphines bound to a polyacrylic acid suffer from the same problem: although their complexes are water-soluble, the loading with the diphosphine is not sufficiently controllable [T. Malstrom et al., J. of Mol. Cat. A: Chemical 139, pages 259-270 (1999)]. In addition, the enantioselectivity depends on the loading with the diphosphine and the catalyst activity is relatively low.

The use of rhodium complexes of trimellitic acid diphosphinopyrrolidinamide as hydrogenation catalysts in aqueous medium (0.1 mol of Na2HP04) has also already been proposed. However, the catalyst activity and the optical yields achieved in the hydrogenation of acetamidocinnamic acid are too low.

There is a need for ligands for hydrogenation catalysts which are soluble in a polar reaction medium, have a good catalyst activity, enable high optical yields to be achieved when the hydrogenation catalysts are chiral, can be used in the acidic to basic range of an aqueous reaction medium to set optimized reaction conditions, can be used in polar solvents such as alcohols, have a high catalyst activity even at high ratios of substrate to catalyst, for example above 100, lead to high optical yields and can easily be separated from the reaction mixture.

It has now surprisingly been found that these objects are achieved when a group which has a solubilizing effect in polar reaction media and may be hydrolyzable is bound to the framework of a ditertiary diphosphine via a urethane or carbamate group.

The invention provides, firstly, compounds of the formulae I and Ia,

A- [ [Si- (C1-C4alkyl) )2]x-CmH2m-X1-C(O)-X1'- CnH2n-Y1-(R-Y)y]r(I), A1-C (0)-Xi'-CnH2n-Yi- (R-Y) y (Ia), in which A is a monovalent to tetravalent framework of a ditertiary diphosphine bound via a C atom, A1 is a monovalent framework of a 3-8-membered N- heterocyclic ditertiary diphosphine bound via the N atom, X1-NH-or-N-(C1-C4alkyl)-,-O-, -NH-or-N-(C1-C4alkyl)-;orX1'is-O-whenX1isX1'is -N-(C1-C4alkyl)-,-NH-or Y1 is a direct bond,-O-,-N=,-CH=, (y+1)-valent benzene radical, or a (y+1)-valent radical of a 3-8-membered cycloalkane, where n is from 1 to 4 when Y1 is-O-or-N=, R is a linear or branched, (y+1)-valent radical of a Cl-Cl2alkane which may be uninterrupted or interrupted -N=or-N(C1-C4alkyl)-,orby-O-, R is a direct bond and Y1 is not-O-or-N=, Y -N(C1-C4alkyl)2-,-N(C1-C4alkyl)2H+X2-or-CO2M, -N(C1-C4alkyl)3+X2-, x is 1 and m is from 2 to 4, or x is 0, m is 0 or an integer from 1 to 4, n is 0 or an integer from 1 to 4, r is an integer from 1 to 4, y is an integer from 1 to 5, M is H, an alkali metal or a monovalent hydrocarbon radical having from 1 to 18 C atoms, and X2 is the anion of an inorganic or organic acid.

A large number of achiral and chiral ditertiary diphosphines for hydrogenation catalysts are described in the literature, cf., for example, H. Brunner and

W. Zettlmeier, Handbook of Enantioselective Catalysis, Vol. II: Ligand References, VCH Verlagsgesellschaft mbH Weinheim (1993). Ditertiary diphosphines functionalized with hydroxyl or amine groups are likewise known or can be prepared by known or analogous methods. If the hydrogenation catalysts contain chiral ditertiary diphosphines as ligands, asymmetric inductions for enantioselective or diastereoselective hydrogenations can frequently be achieved. According to the invention, the radicals A and A1 can be either chiral or achiral ditertiary diphosphines.

The ditertiary diphosphines can be ones in which the phosphine groups are bound (a) to various C atoms of a chain having from 2 to 6 C atoms, preferably a carbon chain having from 2 to 4 C atoms, or (b) directly or via a bridging group of the formula-CRaRb-either in the ortho positions of a cyclopentadienyl ring or each to a cyclopentadienyl ring of a ferrocenyl group, where Ra and Rb are identical or different and are H, Ci- C8alkyl, Cl-C4fluoroalkyl, C5-C6cycloalkyl, phenyl, benzyl, or phenyl or benzyl substituted by from 1 to 3 Cl-C4alkyl or Cl-C4alkoxy groups. Rb is preferably hydrogen. Ra is preferably hydrogen or C,-C4alkyl.

The individual phosphine groups may contain identical or different hydrocarbon radicals. The individual phosphine groups preferably contain two identical hydrocarbon radicals, with various phosphine groups of this type being able to be bound to the framework of the diphosphine. The hydrocarbon radicals can be unsubstituted or substituted and contain from 1 to 20, preferably from 1 to 12, C atoms. Among the ditertiary diphosphines, particular preference is given to ones in which the individual phosphine groups contain two identical or different radicals selected from the group consisting of linear or branched Ci-Cl2alkyl; unsubstituted or Cl-C6alkyl-or Cl-C6alkoxy- substituted C5-Cl2cycloalkyl or C5-Cl2cycloalkyl-CH2-; phenyl and benzyl; and phenyl or benzyl substituted by

halogen (for example F, Cl and Br), Cl-C6alkyl, Cl-C6haloalkyl (for example trifluoromethyl), C1- C6alkoxy, Cl-C6haloalkoxy (for example trifluoro- methoxy), (C6H5) 3Si, (Cl-Cl2alkyl) 3Si or-C02-Cl-C6alkyl (for example-CO2CH3).

The two radicals in the phosphine groups can also together be unsubstituted or halogen-, Cl-C6alkyl-or Cl-C6alkoxy-substituted dimethylene, trimethylene, tetramethylene, pentamethylene or 3-oxapentane-1,5- diyl. The substituents are preferably bound in the two ortho positions relative to the P atom.

The phosphine groups can also be groups of the formulae in which o and p are, independently of one another, an integer from 2 to 10, and the sum of o+p is from 4 to 12 and preferably from 5 to 8. Examples are [3.3.1]- and [4.2.1]-phobyl of the formulae Examples of alkyl substituents, preferably containing from 1 to 6 C atoms, on P are methyl, ethyl, n-propyl, i-propyl, n-butyl, i-butyl, t-butyl and the isomers of pentyl and hexyl. Examples of unsubstituted or alkyl- substituted cycloalkyl substituents on P are cyclopentyl, cyclohexyl, methylcyclohexyl and ethylcyclohexyl and dimethylcyclohexyl. Examples of alkyl-, alkoxy-, haloalkyl-and haloalkoxy-substituted phenyl and benzyl substituents on P are methylphenyl, dimethylphenyl, trimethylphenyl, ethylphenyl, methylbenzyl, methoxyphenyl, dimethoxyphenyl,

trifluoromethylphenyl, bistrifluoromethylphenyl, tristrifluoromethylphenyl, trifluoromethoxyphenyl and bistrifluoromethoxyphenyl.

Preferred phosphine groups are those containing identical or different, preferably identical, radicals selected from the group consisting of Cl-C6alkyl, unsubstituted or Cl-C4alkyl-or Cl-C4alkoxy- monosubstituted to-trisubstituted cyclobutyl, cyclopentyl or cyclohexyl, benzyl and particularly phenyl which is unsubstituted or substituted by from 1 to 3 Cl-C4alkyl, Cl-C4alkoxy, F, Cl, Cl-C4fluoroalkyl or C1-C4fluoroalkoxy substituents.

The diphosphine radicals A and A1 in the formulae I and Ia preferably correspond to the formulae II, IIa and IIb, in which Ri R2, R4 and R5 are each, independently of one another, a hydrocarbon radical having from 1 to 20 C atoms which is unsubstituted or substituted by halogen, Cl-C6alkyl, Cl-C6haloalkyl, Cl-C6alkoxy, Cl-C6haloalkoxy, (C6H5) 3Si, (C1-C12alkyl) 3Si or -CO2-C1-C6alkyl ; or Ri and R2 or R4 and R5 are in each case together unsubstituted or halogen-, Cl-C6alkyl-or Cl-C6alkoxy-substituted tetramethylene, pentamethylene or 3-oxapentane-1,5-diyl; and R3 is unsubstituted or Cl-C6alkyl-, C1-C6alkoxy-, C5Cycloalkyl-or C6cycloalkyl-, phenyl-, naphthyl-or

benzyl-substituted C2-C4alkylene; unsubstituted or Cl-C6alkyl-, phenyl-or benzyl-substituted 1,2- or 1,3- cycloalkylene, 1,2- or 1,3-cycloalkenylene, 1,2- or 1,3-bicycloalkylene or 1,2- or 1,3-bicycloalkenylene having from 4 to 10 C atoms; unsubstituted or Cl-Csalkyl-, phenyl-or benzyl-substituted 1,2- or 1,3- cycloalkylene, 1,2- or 1,3-cycloalkenylene, 1,2- or 1,3-bicycloalkylene or 1,2- or 1,3-bicycloalkenylene having from 4 to 10 C atoms to which methylene or C2-C4alkylidene is bound in the 1 and/or 2 positions or in the 3 position; 1,4-butylene which is substituted by R6R, C (O-) 2 in the 2,3 positions and is unsubstituted or substituted by Cl-C6alkyl, phenyl or benzyl in the 1 and/or 4 positions, where R6 and R7 are, independently of one another, hydrogen, Cl-C6alkyl, phenyl or benzyl; 3,4- or 2,4-pyrrolidinylene or methylene-4-pyrrolidin- 4-yl whose N atom is substituted by hydrogen, Cl-Cl2alkyl, phenyl, benzyl, Cl-Cl2alkoxycarbonyl, Cl-C8acyl, Cl-Cl2alkylaminocarbonyl; or unsubstituted or halogen-,-OH-, Cl-C6alkyl-, Cl-C6alkoxy-, phenyl-, benzyl-, phenyloxy-or benzyloxy-substituted 1,2- phenylene, 2-benzylene, 1,2-xylylene, 1,8-naphthylene, 2,2'-binaphthylene or or R3 is a radical of the formulae in which R8 is hydrogen, Cl-Csalkyl, Cl-C4fluoroalkyl, unsubstituted phenyl or phenyl substituted by from 1 to 3 F, Cl, Br, Cl-C4alkyl, Cl-C4alkoxy or fluoromethyl substituents; q in formula IIa is 0 or an integer from 1 to 6, where the radical of the formula IIa is unfused or fused with

cycloaliphatic, heterocycloaliphatic, aromatic or heteroaromatic radicals, and the radical of the formula IIa is unsubstituted or substituted by halogen, <BR> <BR> <BR> Ci-C6alkyl, Cl-C6haloalkyl, Cl-C6alkoxy, Cl-C6haloalkoxy, (C6H5) 3Si, (C1-C12alkyl) 3Si or-CO2-Cl-C6alkyl; and Rg and Rlo are, independently of one another, C1-C4alkylene.

Rg and Rlo are preferably methylene, ethylene or linear or branched propylene or butylene. The chain formed by R9, Rlo and N preferably contains from 3 to 5 atoms.

Rl, R2, R4 and R5 are preferably identical or different, in particular identical, radicals selected from the group consisting of branched C3-C6alkyl, unsubstituted or Cl-C4alkyl-or Cl-C4alkoxy-monosubstituted to -trisubstituted cyclopentyl or cyclohexyl, unsubstituted or Cl-C4alkyl-or Cl-C4alkoxy- monosubstituted to-trisubstituted benzyl and in particular unsubstituted or Cl-C4alkyl-, Cl-C4alkoxy-, NH2-, OH-, F-, C1-, Cl-C4fluoroalkyl-or Cl-C4fluoro- alkoxy-monosubstituted to-trisubstituted phenyl.

In the formulae II and IIa, Y1 is preferably-N=,-CH= or a quaternary C atom.

In the formulae II and IIa, R is preferably Cl-C8alkyl and particularly preferably Cl-C6alkyl which may be interrupted by from one to three -O- or -N (Cl-C4alkyl)- groups.

The alkyl in the group N (Cl-C4alkyl) 2 is preferably linear and can be methyl, ethyl, n-propyl or n-butyl.

In the formulae II and IIa, Y is particularly -N(CH3)2-,preferably-CO2M, -N(CH3)3+X2-,or in which M is H, Li, Na, K or Cl-C4alkyl, and X2 is the anion of an inorganic or organic acid.

In the formulae II and IIa, m is preferably 0,1 or 2 and n is preferably 0,1 or 2.

In the formulae II and IIa, r is preferably from 1 to 3, particularly preferably 1 or 2 and very particularly preferably 1.

In the formulae II and IIa, y is preferably from 1 to 4, particularly preferably from 1 to 3 and very particularly preferably 1 or 2.

In the formulae II and IIa, a cycloalkane radical Y1 is preferably a 5-or 6-membered cycloalkane radical.

In the group Si (Cl-C4alkyl) 2, the alkyl is preferably methyl.

As alkali metals M, preference is given to Li, Na or K.

Anions X2 derived from an inorganic acid may be, for example, halides such as chloride, bromide and iodide, sulfite, sulfate, hydrogensulfate, carbonate, hydrogen- carbonate, nitrate, phosphite, hydrogenphosphite, phosphate, hydrogenphosphate, tetrafluoroborate or hexafluorophosphate.

Anions X2 derived from organic acids are the anions of carboxylic acids, sulfonic acids and phosphonic acids which may contain from 1 to 18, preferably from 1 to 12 and particularly preferably from 1 to 8, C atoms. Some preferred examples of anions of organic acids are acetate, propionate, butyrate, monochloroacetate, dichloroacetate and trichloroacetate, monofluoro- acetate, difluoroacetate and trifluoroacetate, hydroxyacetate, oxalate, malonate, and also the anions of cyclohexanemonocarboxylic and cyclohexane- dicarboxylic acids, benzoic acid, phthalic and terephthalic acids, trifluoromethylbenzoic acid, phenylacetic acid, phenylphosphonic acid, methyl-

sulfonic, ethylsulfonic, propylsulfonic, butylsulfonic, cyclohexylsulfonic, phenylsulfonic, methylphenyl- sulfonic, trifluoromethylphenylsulfonic, monochlorome- thylsulfonic, dichloromethylsulfonic and trichloro- methylsulfonic and monofluoromethylsulfonic, difluoro- methylsulfonic and trifluoromethylsulfonic acids. Particular preference is given to the anions of unsubstituted and substituted phenylsulfonic acids.

The group -[Si-(C1-C4alkyl)2]x-CmH2m-X1- in the radicals of the formula I is preferably-Si (CH3) 2-CH2CH2CH2-NH- or -Si (CH3) 2-CH2CH2CH2-0-.

Some specific embodiments of the group-CmH2m-are methylene, ethylene, 1,2- or 1,3-propylene, 1,2-, 1,3-, 1,4- or 2,3-butylene, ethylidene, propylidene and butylidene.

Some specific embodiments of the group-Xi-C (O)-X1'- CnH2n-Y1-(R-Y) y in the radicals of the formula I are, for example,-NH-C (O)-NH-C (CH2-CO2H) 3,-NH-C (O)-NH-CH (CH2- C02H) 2,-NH-C (O)-NH-CH2 (CH2-C02H),-NH-C (O)-NH-CH2-C (CH2- CO2H)3,-NH-C(O)-NH-CH2- CH2 (CH2-CO2H),-O-C (O)-NH-C (CH2-CO2H) 3,-O-C (O)-NH-CH (CH2- CO2H)2,-O-C(O)-NH-CH2-C(CH2- C02H) 3,-0-C (0)-NH-CH2-CH (CH2-C02H) 2, -O-C (O)-NH-CH2- CH2 (CH2-CO2H),-NH-C (O)-NH-C (CH2-CH2-C02H) 3,-NH-C (O)-NH- CH (CH2-CH2-C02H) 2,-NH-C (O)-NH-CH2 (CH2-CH2-CO2H),-NH- C(O)-NH-CH2-C(CH2-CH2-CO2H)3,-NH-C(O)-NH-CH2-CH(CH2-CH2- CO2H)2,-O-C(O)-NH- C(CH2-CH2-CO2H)3, -O-C(O)- NH-CH2(CH2-CH2-CO2H), -O- C(O)-NH-CH2-CH(CH2-CH2-CO2H)2,-O-C(O)-NH-CH2-CH2(CH2-CH2- -NH-C(O)-NH-CO2H),-NH-C(O)-NH-C-(CH2-O-CH2-CH2-CO2H)3, -O-CH(CH2-O-CH2CH2CO2H)2,-NH-C(O)-NH-CH2CH2-O-CH2CH2CO2H, C (O)-NH-C (CH2-O-CH2CH2CO2H) 3,-O-C (O)-NH-CH (CH2-O-CH2CH2- CO2H)2,-NH-C(O)-NH- CH2CH2-N (CH3) 2,-NH-C (O)-N [CH2CH2-N (CH3) 2] 2,-NH-C (O)-NH- CH2CH2-N (CH3)-CH2-CH2N (CH3) 2,-NH-C (O)-NH-CH2-CH2-N [CH2CH2-

N (CH3) 2]2, -O-C(O)-NH-CH2-CH2-N (CH3) 2,-0-C (0)-NH-CH2CH2- N(CH3)CH2CH2N(CH3)2, - NH-C (O)-NH-CH2CH [CH2-N (CH3) 2] 2,-0-C (0)-NH-CH2CH [CH2- N (CH3) 2]2, -NH-C(O)-NH-CH2C[CH2-N (CH3) 2] 3,-0-C (0)-NH- CH2C [CH2-N (CH3) 2] 3,-NH-C (0)-NH-CH2CH [CH2CH2N (CH3) 2] 2,-0- C (0)-NH-CH2CH [CH2CH2-N (CH3) 2]2, -NH-C(O)-NH-CH2C[CH2CH2- -NH-N(CH3)2]3,-O-C(O)-NH-CH2C[CH2CH2-N(CH3)2]3, C(O)-NHC6H3(CO2H)2, -O-C(O)-NHC6H3- -NH-C(O)-NHC6H9(CO2H)2,-(CO2H)2,-O-C(O)-NHC6H2(CO2H)3, -O-C(O)-NH-C(O)-NHC6H8(CO2H)3,-O-C(O)-NHC6H9(CO2H)2, -O-C(O)-NHC6H8-(CO2H)3,-O-C(O)-NHC6H3(N(CH3)2)2, NHC6H2(N(CH3)2)3, -NH-C(O)- NHC6Hg (N (CH3) 2) 3,-0-C (O)-NHC6H9(N(CH3)2)2, -O-C(O)-NHC6H8- (N (CH3) 2) 3. In the above groups, a (CH3) 2N group can be replaced by a (C2H5) 2N group.

Particularly preferred groups are -NH-C(O)-NH-C(CH2-O- CH2CH2C02H) 3,-0-C (0)-NH-C (CH2-0-CH2CH2CO2H) 3, -NH-C(O)- NH-CH2CH2-N [CH2CH2-N (CH3) 2]2, -NH-C(O)-NH-CH2-NH-CH2CH2- N [CH2CH2-N (CH3) 2] 2 and-NH-C (O)-N [CH2CH2-N (CH3) 2] 2, where the N (CH3) 2 group can be replaced by an N (C2Hs) 2 group.

Some specific embodiments of the group -C(O)-X1'-CnH2n- Y1- (R-Y) y are ones as listed above but in which-NH-or -O- is left off at the beginning. Particularly preferred groups are-C (O)-NH-C (CH2-0-CH2CH2C02H) 3, -C (0)-NH-CH2C (CH2-0-CH2CH2C02H) 3,-C (O)-NH-CH2CH2-N [CH2CH2- N (CH3) 2] 2 and-C (O)-NH-CH2CH2CH2-N [CH2CH2-N (CH3) 2] 2, where the N (CH3) 2 group can be replaced by an N (C2Hs) 2 group.

A preferred subgroup of tertiary diphosphines of the formulae I and Ia consists of those of the formulae

in which Rll and R12 are, independently of one another, linear or branched Cl-C4alkylene, and the chain-P-Rll-N-Rl2-P-has from 5 to 8 atoms, R13 and R14 are identical or different groups selected from among unsubstituted or Cl-C4alkyl-, Cl-C4alkoxy-or trifluoromethyl-substituted C5cycloalkyl or C6cycloalkyl and unsubstituted or Cl-C4alkyl-, Cl-C4alkoxy-, F-, C1- or trifluoromethyl-substituted phenyl, R15 and R16 are together C2-C6alkylene, Ri are, independently of one another, a bond or methylene or ethylidene, L is a radical of the formula-C (O)-X1'-CnH2n-Y1-(R-Y) y and Li is a radical of the formula -[[Si-(CH3)2]x-CmH2m-X1- C(O)-X1'-CnH2n-Y1-(R-Y)y]r, X1 -NH-or-NCH3-,-O-, X1' is -NH- or -N-(C1-C4alkyl)-; or X 'is-O-when Xl is -NH-or -N-(C1-C4alkyl)-, Y1 is a direct bond,-O-,-N=,-CH= or where

n is from 2 to 4 when Y1 is-O-or-N=, R is a linear or branched Ci-CBalkane radical which may be uninterrupted or interrupted by-O-,-N= or -N(CH3)-, Y -N(CH3)2,-N(CH3)2H+X2-or-N(CH3)3+X2-,-CO2M, x is 1 and m is 2 or 3, or x is 0, m is 0 or from 2 to 4, n is 0 or from 2 to 4, r is 1 or 2, y is an integer from 1 to 3, M is H, Li, Na, K or Cl-C4alkyl, and X2 is the anion of an inorganic or organic acid.

For X2, the above-described embodiments and preferences apply. X2 can be, for example, an anion selected from the group consisting of chloride, bromide, iodide, hydrogensulfate, dihydrogenphosphate, acetate, benzoate, tosylate and methylsulfonate.

Some specific examples of novel chiral ligands of the formulae I and Ia are

and also their Na and K salts and methyl or ethyl esters.

The compounds of the formulae I and Ia can be prepared in a manner known per se by reacting OH-or NH- functional ditertiary disphosphines with isocyanates or capped isocyanates of a solubilized compound. The invention further provides a process for preparing compounds of the formulae I and Ia, in which a) OH-or NH-functional ditertiary diphosphines of the formulae III and IIIa, A-[[Si-(C1-C4alkyl)2]x-CmH2m-X1H]r(III), Al-H (IIIa) are reacted with a compound of the formula IV, OCN-CnH2n-Yl- (R-Y) y (IV), or an activated urea derivative thereof, or b) OH-or NH-functional compounds of the formula IVb, H-Xi-CnH2n-Yi- (R-Y) y (IVb), are reacted with a compound of the formula IIIb

A- ( [si- (CI-C4alkyl) 2]'-CH2.-NCO], (IIIb), or an activated urea derivative thereof, or an activated urea derivative of the compound A1-H of the formula IIIa, in which A, Ai, Xi, Xl', Yl, Y, R, x, m, n, y and r are as defined above.

In the case of NH2-functional diphosphines of the formula A- [[Si-(Cl-C4alkyl) 2] x-CmH2m-NH2] r, it is also possible to prepare isocyanates of the formula A- [ [Si- (Cl-C4alkyl) 2] x-CmH2m-NCO] r or their capped isocyanates first and then to react these with a functional compound of the formula H-Xl-CnH2n-Yl- (R-Y) y. In the case of the NH-functional compounds A1-H of the formula IIIa, it is likewise possible to prepare capped isocyanates and to react these with a functional compound of the formula H-Xl-CnH2n-Yl- (R-Y) y.

Compounds of the formulae III, IIIa, IIIb, IVb and IV are known or can be prepared by known or analogous methods. In the preparation, it may be advantageous to use esters (for example methyl or ethyl esters) of carboxylic acids in place of the carboxylic acids (Y is -C02H) and subsequently to obtain the acids and salts by hydrolysis and salt formation.

Activated urea derivatives are also referred to as capped isocyanates. These can, for example, correspond to the formula IVa, Rlg- (0) C-HN-CnH2n-Yl- (R-Y) y] r (IVa), in which Y, Yi, R, n, y and r are as defined above and Pis is the N-bonded radical of an activating amine. Such amines are essentially cyclic secondary amines, preferably N-heteroaromatics such as pyrrole and in particular imidazole. The compounds of the formulae IV

and IVa are obtainable by reaction of appropriate ureas, for example carbonylbisimidazole, with compounds of the formula H2N-CnH2n-Y1- (R-Y) y] r. Here, mixtures of isocyanates and capped isocyanates in chemical equilibrium can be formed and can likewise be used. The reaction is generally carried out at room temperature or slightly elevated temperature in an inert solvent such as toluene. The reaction products can be used directly in further reactions without isolating the isocyanates/capped isocyanates. Syntheses using azolidene are described by H. A. Staab in Angew. Chem.

12,1962, pages 407 to 419.

The reaction of functional ditertiary diphosphine and isocyanate/capped isocyanate or vice versa is generally carried out in an inert solvent. Examples of suitable solvents are aliphatic, cycloaliphatic or aromatic hydrocarbons, halogenated hydrocarbons, ethers, carboxylic esters, lactones, N-substituted carboxamides, lactams and ketones. The reaction can be carried out at temperatures of from-20 °C to 100 °C.

The compounds of the formulae I and Ia are very useful ligands for metal complexes, in particular ones which can be used as hydrogenation catalysts and are soluble in polar reaction media such as water or aprotic or protic and polar organic solvents, e. g. alcohols, and can be separated off after the hydrogenation, for example by means of extraction, if appropriate after hydrolysis of ester groups and/or changing the pH.

The invention further provides metal complexes of da metals with compounds of the formulae I and Ia, in particular metal complexes of the formulae V, Va and Vb A3MeXD (V), [A3MeX] +E- (Va), A3Ru (II) X3X4 (Vb), in which A3 is a compound of the formula I or Ia;

X is two monoolefin ligands or one diene ligand; Me is a de metal selected from the group consisting of Ir and Rh; D is-C1,-Br or-I; E is the anion of an oxo acid or complex acid; and X3 and X4 are identical or different and are as defined for D or E, or X3 and X4 are allyl or 2-methallyl, or X3 is as defined for D or E and X4 is hydride.

Preference is given to metal complexes in which X is 1,5-hexadiene, 1,5-cyclooctadiene or norbornadiene. In the metal complexes of the invention, D is-C1,-Br or -I. In the preferred metal complexes, E-is Cl04-, CF3SO3-, CH3SO3-, HS04-, BF4-, B (phenyl) 4-, PF6-, SbCl6-, AsF6~ or SbF6.

Further possible ruthenium complexes are known from the literature and described, for example, in US 4 691 037, US 4 739 085, US 4 739 084, EP 269395, EP 271310, EP 271311, EP 307168, EP 366390, EP 470756, JP 08081484, JP 08081485, JP 09294932, EP 831099, EP 826694, EP 841343, J. P. Genet, Arcos Organics Acta, 1 (1995) 4, N. C. Zanetti et al., Organometallics 15 (1996) 860.

The metal complexes of the formulae V, Va and Vb are prepared by methods known from the literature.

The metal complexes of the invention are very useful as catalysts for the hydrogenation of organic double and triple bonds. Examples are compounds containing the groups C=C, C=N, C=O, C=C-N or C=C-0 (cf., for example, K. E. König, The Applicability of Asymmetric Homogeneous Catalysis, in James D. Morrison (ed.), Asymmetric Synthesis, Vol. 5, Academic Press, 1985). In particular, the metal complexes of the invention are useful for the enantioselective hydrogenation of compounds containing prochiral carbon-carbon and carbon-heteroatom double bonds. Examples of such compounds are prochiral alkenes, imines and ketones.

A further aspect of the invention is therefore the use of the novel metal complexes of d8 metals as heterogeneous or homogeneous catalysts for hydrogenation, preferably for the asymmetric hydrogenation of prochiral compounds having carbon- carbon or carbon-heteroatom double bonds in a polar reaction medium. The metal complexes are preferably used for the asymmetric hydrogenation of prochiral compounds having carbon-carbon or carbon-heteroatom double bonds, in particular the Ir complexes for the hydrogenation of prochiral ketimines and the Rh complexes for the hydrogenation of prochiral ketimines and prochiral alkenes.

The invention further provides a process for the hydrogenation of compounds having carbon-carbon or carbon-heteroatom double bonds, which comprises reacting the compounds at a temperature of from-20 to 150 °C with hydrogen in the presence of catalytic amounts of one or more metal complexes of the invention in a polar reaction medium.

The metal complexes of the invention are hydrogenation catalysts or precursors of such catalysts which are activated under the hydrogenation conditions. The metal complexes can be prepared in situ and used as hydrogenation catalysts. The hydrogenations are advantageously carried out under protective gas such as noble gases (argon).

The process can be carried out under a hydrogen pressure of from 1 to 500 bar, preferably 1-150 bar, particularly preferably from 1 to 120 bar and in particular from 1 to 100 bar.

The reaction temperature can be, for example, from 0 to 150 °C, preferably from 10 to 120 °C and particularly preferably from 30 to 100 °C.

The amount of catalyst depends mainly on the desired reaction time and on economic considerations. Higher amounts of catalyst essentially promote shorter reaction times. Catalysts are preferably used in amounts of from 0.0001 to 10 mol%, particularly preferably from 0.001 to 10 mol% and in particular from 0.01 to 5 mol%, based on the compound to be hydrogenated. The molar ratio of substrate to catalyst can be, for example, from 30 to 100000, preferably from 50 to 50000, particularly preferably from 50 to 30000 and in particular from 100 to 2000.

A particularly advantageous aspect is that the hydrogenation can be carried out in polar reaction media, for example polar organic or aqueous media, and in the acid, neutral or basic range, so that reaction conditions optimized for a particular substrate can be set and the catalysts can easily be separated off, for example by means of extraction. If Y in the formulae I and Ia is an ester group, this can be hydrolyzed beforehand to form the acid group.

For the purposes of the invention, an aqueous reaction medium is one consisting of only water or water in admixture with a water-miscible or water-immiscible organic solvent. The proportion of water is preferably at least 30 per cent by volume, particularly preferably at least 50 per cent by volume and in particular at least 70 per cent by volume. The reaction medium very particularly preferably contains only water. Examples of suitable solvents are alcohols such as methanol, ethanol, propanol, butanol, ethylene glycol and ethylene glycol monomethyl ether; ethers such as diethyl ether, diisobutyl ether, tetrahydrofuran and dioxane; esters and lactones such as ethyl acetate, methyl acetate and valerolactone; sulfoxides and sulfones such as dimethyl sulfoxide, dimethyl sulfone, tetramethylene sulfone; ketones such as acetone or

methyl isobutyl ketone; halogenated hydrocarbons such as methylene chloride, chloroform, tetrachloroethane and chlorobenzene; and N-substituted carboxamides and lactams such as N-methylpyrrolidone and dimethylformamide. If the organic solvents are not miscible with water, a two-phase hydrogenation takes place.

Buffers, bases or acids can be added to the aqueous reaction medium. The reaction can, for example, be carried out at a pH of from 1 to 12, preferably from 2 to 10 and particularly preferably from 3 to 9. Suitable buffers are, in particular, phosphate buffers, but it is also possible to use carboxylic acids, carbonic acid, phosphoric acid and boric acid. Examples of suitable bases are alkali metal hydroxides and alkaline earth metal hydroxides, amines and alkali metal salts of carboxylic acids, carbonic acid, phosphoric acid and boric acid. Examples of suitable acids are HC1, HBr, HI, HBF4, HC104, carboxylic acids (unsubstituted, fluorinated or chlorinated acetic acid, benzoic acid, citric acid), boric acid, phosphoric acid, methanesulfonic acid, sulfuric acid and carbonic acid.

The bases and acids can also be soluble or insoluble polymers such as ion-exchange resins. The amount of bases, acids and/or buffers can be, for example, from 0 to 2, preferably from 0 to 1 and particularly preferably from 0 to 0.5, mol per litre of water.

Examples of suitable organic solvents are esters and lactones such as ethyl acetate and methyl acetate and valerolactone; ketones such as acetone or methyl isobutyl ketone; halogenated hydrocarbons such as methylene chloride, chloroform, tetrachloroethane and chlorobenzene; ethers such as diethyl ether, diisobutyl ether, tetrahydrofuran and dioxane; sulfoxides and sulfones such as dimethyl sulfoxide, dimethyl sulfone, tetramethylene sulfone; N-substituted carboxamides and lactams such as N-methylpyrrolidone and dimethyl-

formamide; and particularly alcohols.

For the purposes of the invention, an alcoholic reaction medium means the presence of an alcohol, either alone or in admixture with another organic solvent. Suitable alcohols are aliphatic, cyclo- aliphatic, cycloaliphatic-aliphatic and araliphatic alcohols. Some preferred examples are methanol, ethanol, n-or i-propanol, n-, i-or t-butanol, pentanol, hexanol, cyclohexanol, cyclohexanediol, hydroxymethylcyclohexane or dihydroxymethylcyclohexane, benzyl alcohol, ethylene glycol, diethylene glycol, propanediol, butanediol, ethylene glycol monomethyl ether or ethylene glycol monoethyl ether, and diethylene glycol monomethyl ether or diethylene glycol monoethyl ether. Preference is given to methanol and ethanol. The proportion of an alcohol is preferably at least 30 per cent by volume, particularly preferably at least 50 per cent by volume and in particular at least 70 per cent by volume. Very particular preference is given to using only an alcohol.

It is known that the catalytic properties of the diphosphine catalysts used can be influenced by addition of cocatalysts, modifiers, metal halides and ammonium halides. It can therefore be advantageous to add alkali metal or ammonium chlorides, bromides or iodides, for example LiCl, LiBr, LiI, NaI, NaBr or tetrabutylammonium iodide, to the reaction mixture. The amount can be, for example, from 0.001 to 5 mol per litre of solvent. Known modifiers are phthalimides and parabanic acid.

The hydrogenation can be carried out continuously or batchwise in various types of reactor. Preference is given to reactors which allow comparatively good mixing and good heat removal, e. g. loop reactors. This type of reactor has been found to be particularly useful when using small amounts of catalyst.

The hydrogenated organic compounds which can be prepared according to the invention are active substances or intermediates for preparing such substances, in particular in the production of pharmaceuticals and agrochemicals. Thus, for example, o, o-dialkylarylamine derivatives, in particular those having alkyl and/or alkoxyalkyl groups, act as fungicides, particularly as herbicides. The derivatives can be amine salts, acid amides, e. g. of chloroacetic acid, tertiary amines and ammonium salts (cf., for example, EP-A-0 077 755 and EP-A-0 115 470).

The following examples illustrate the invention.

A) Preparation of starting compounds Example Al: Preparation of (C2H502C-CH2CH2-0-CH2-) 3C-NCO (A1) A solution of 1 mmol of (C2H502C-CH2CH2-0-CH2-) 3C-NH2 (prepared as described by G. R. Newcome et al., Aldrichimica Acta, 25 (1992) 31-8)] in toluene (7 ml per mmol) is added to an equimolar amount of carbonyldiimidazole and the mixture is stirred overnight. A mixture of isocyanate and imidazolylurea is formed. The solution is used directly for the reactions below.

Example A2: Preparation of H2N-CH2CH2CH2-N [CH2CH2-N- (C2H5) 2 (A2) a) A mixture of 1 ml (3.5 mmol) of N, N, N', N'- tetraethyldiethylenetriamine in 5 ml of acetonitrile and 938 mg (3.5 mmol) of N- (3-bromopropyl) phthalimide and 977 mg (7 mmol) of potassium carbonate is stirred at room temperature over the weekend. The reaction mixture is extracted with methylene chloride/water, the organic phase is dried over sodium sulfate and evaporated on a rotary evaporator and the crude product is purified by chromatography (silica gel Merck 60, eluant: 50/1 mixture of ethanol/triethylamine). The

product is obtained as an oil in a yield of 80%.

1H-NMR (CDC13): 7.9-7.83 (m, 2H, phthalimide), 7.74- 7.68 (m, 2H, phthalimide), 3.74 (t, 2H, phthalimide- CH2), 2.53 (m, 18H, 9 x CH2-N), 1.85 (m, 2H, CH2-CH2- CH2), 1.02 (t, 12H, 4 x CH3). b) 1.52 g (3.79 mmol) of product as described in Example A2a in 10 ml of ethanol are treated with 1 ml of hydrazine hydrate solution and the mixture is stirred under reflux for 1 hour. After cooling to room temperature, 25 ml of methylene chloride are added and the mixture is filtered. The filtrate is evaporated to dryness on a rotary evaporator and the product A2 is dried over solid NaOH. (Yield 97%).

1H-NMR (CDC13): 2.76 (t, 2H, H2N-CH2-) 2.55 (m, 18H, 9 x CH2-N), 1.63 (m, 2H, CH2-CH2-CH2), 1.04 (t, 12H, 4 x CH3).

B) Preparation of diphosphine ligands Example B1: Preparation of ligand K1 a) Preparation of the triester A solution of 377 mg (0.83 mmol) of 2-diphenyl- phosphinomethyl-4-diphenylphosphinopyrrolidine (PPM) in 2.5 ml of toluene is added to a solution prepared as described in Example A1 (1.1 mmol of isocyanate triester) and the mixture is stirred overnight. After evaporation on a rotary evaporator and partial removal of the toluene under reduced pressure, the crude product is purified by chromatography (silica gel: Merck 60; eluant: ethyl acetate). 605 mg of product are obtained (yield: 81%). b) Preparation of the triacid 1 ml of water and 0.6 g of KOH are added to a solution of 590 mg of the triester prepared as described in Bla in 5 ml of ethanol and the mixture is stirred for 3 hours. The ethanol is then evaporated under reduced pressure and the mixture is dissolved in 25 ml of water. The solution is subsequently acidified using 2N

HC1 and extracted a number of times with ethyl acetate.

The organic phases are collected, washed with water, dried over sodium sulfate and finally evaporated to dryness under reduced pressure. The product is obtained as a white solid in a yield of 88%.

Example B2: Preparation of ligand K2 a) Preparation of the triester The procedure of Example Bla is repeated, except that 3,4-diphenylphosphinopyrrolidine (pyrphos) is used as starting compound. The reaction product is obtained in a yield of 63%. b) Preparation of the triacid The procedure of Example Blb is repeated. The product is obtained as a white solid in a yield of 95%.

Example B3: Preparation of ligand K3 a) Preparation of the triester The procedure of Example Bla is repeated, except that 2,2'-bis (diphenylphosphino)-5-methyl-5'- hydroxymethylbiphenyl (HO-biphemp) is used as starting compound. The reaction product is obtained in a yield of 82%. b) Preparation of the triacid The procedure of Example Blb is repeated. The product is obtained as a white solid in a yield of 92%.

Example B4: Preparation of ligand K4 a) Preparation of the triester (Amine ligand A, cf.

WO 98/01457)

A solution of 1 g (1.5 mmol) of amine ligand A in 8 ml of methylene chloride is added at 0 °C to an equimolar amount of carbonyldiimidazole in 6 ml of methylene chloride and the reaction mixture is subsequently stirred at room temperature for 2 hours. 1.6 equivalents of H2N-C (CH2-O-CH2CH2C (O)-OCH2CH3 and 5 mg of dibutyltin dilaurate are then added and the mixture is stirred at 50°C for 48 hours. After purification by chromatography (silica gel: Merck 60; eluant: hexane/ethyl acetate, 1: 1), the product is obtained as a virtually solid, orange oil in a yield of 65%. b) Preparation of the triacid: 1 g of diphosphine triester is dissolved in 10 ml of ethanol and treated with 1 ml of 20% aqueous KOH solution. After stirring for 2 hours, the ethanol is evaporated under reduced pressure and the product is dissolved in 20 ml of water. The product is precipitated by addition of 2N HC1, filtered off, washed a number of times with water and finally dried at 50°C in a high vacuum. The product is obtained as an orange-yellow solid in a yield of 92%.

Example B5: Preparation of ligand K5 a) Preparation of the triester (Amine ligand B, cf.

WO 98/01457) A solution of 215 mg (0.34 mmol) of functionalized ligand in 2 ml of methylene chloride is added at 0°C to an equimolar amount of carbonyldiimidazole in 2 ml of methylene chloride and the reaction mixture is

subsequently stirred at room temperature for 2 hours. 2 equivalents of amine ligand B and 5 mg of dibutyltin dilaurate are then added and the mixture is stirred at 50°C for 48 hours. After purification by chromatography (silica gel: Merck 60; eluant: hexane/ethyl acetate, 1: 1), the product is obtained as a virtually solid, orange oil in a yield of 72%. b) Preparation of the triacid The procedure of Example B4b is repeated. The orange, solid product is obtained in a yield of 97%.

31P-NMR of the compounds K1 to K5 in CDC13 (Ph = phenyl, xyl = xylyl, cy = cyclohexyl): Ligand §Ligand 5 K1-7.6,-21.6 K1-7.96,-21.7 (ester) (acid) K2-11.4 K2 (ester) (acid) K3-13.6 (d, JPP (d, JPP 35.2 (ester Hz) (acid) Hz) -14.5 (d, JPP 35.7-14.5 (d, JPP 35.2 Hz) Hz) K4 17.0 (d, JPP 32 K4 16.4 (d, JPP 32 (ester) Hz) Pcy2 (acid) Hz) Pcy2 -25.0 (d JPP 32-25.5 (d JPP 32 Hz) PPh2 Hz) PPhz K5 7.3 (d, JPP 19 Hz) K5 6.6 (d, JPP 18 Hz) (ester) Pxyl2 (acid) Pxylz -25.0 (d JPP 19-24.5 (d JPP 18 Hz) PPh2 Hz) PPh2

1H-NMR: In all cases, the characteristic signals of the ethyl ester group (triplet at 8 = 1.2 and quartet at 8 = 4.1) have disappeared after saponification of the triester.

Example B6: Preparation of ligand K6 A solution of 83 mg (0.133 mmol) of the amine ligand B in 1.5 ml of methylene chloride is added at 0 °C to a solution of 22 mg (0.133 mmol) of 1, l'-carbodiimidazole in 2 ml of methylene chloride and the mixture is subsequently stirred at room temperature for 2 hours. 2 equivalents of N, N, N', N'-tetraethyldiethylenetriamine and 2 mg of dibutyltin dilaurate are then added and the mixture is stirred overnight. The methylene chloride is subsequently evaporated and the crude product is purified by chromatography (silica gel Merck 60, eluant: 100/1 mixture of ethanol/triethylamine).

The product is obtained as an orange, viscous oil in a yield of 87%.

1H-NNR (CDC13): Displays the typical signals for the

CH3-CH2-N groups at 1.02 and 2.6.

31P-NMR (CDC13): + 16.9 (d, Pcy2),-24.9 (d, PPh2).

Example B7: Preparation of ligand K7 97 mg (0.335 mmol) of tetraamine A2 in 1.5 ml of methylene chloride are added at 0 °C to a solution of 58 mg (0.35 mmol) of 1, l'-carbodiimidazole in 4 ml of methylene chloride and the reaction mixture is subsequently stirred at room temperature for 1 hour. A solution of 154 mg (0.34 mmol) of (2S, 4S)- (-)-4- diphenylphosphino-2- (diphenylphosphinomethyl)- pyrrolidine in 2 ml of methylene chloride is then added and the mixture is stirred under reflux for 6 hours.

Finally, the methylene chloride is evaporated and the crude product is purified by chromatography (silica gel Merck 60, eluant: 25/1 mixture of ethanol/ triethylamine). The product is obtained as a viscous colourless oil in a yield of 67%.

1H-NMR (CDC13): Displays the typical signals for the CH3-CH2-N groups at 1.02 and 2.55.

31P-NMR (CDC13):-7.6 (s),-21.5 (s) c) Use examples Test reaction: CL, OR H :--- \ HN'Chl lr l ( O O R= H or K, Na: acetamidocinnamic acid (AC) R= methyl: methyl acetamidocinnamate (MAC) All operations are carried out under argon as protective gas. The hydrogenations are carried out in a 25 ml glass flask (1 bar of hydrogen pressure) or a 50 ml steel autoclave (>1 bar of hydrogen pressure) provided with magnetic stirrer and a vacuum/gas connection. The hydrogenation solutions are introduced by means of a syringe and needle, in the case of the glass flask through a septum, in the case of the

autoclave under a countercurrent of argon. After addition of the hydrogenation solution, the inert gas is displaced by hydrogen (glass apparatus: 4 cycles of vacuum, hydrogen at atmospheric pressure; steel autoclave: 4 cycles of pressurization with 5-10 bar of hydrogen and depressurization). The hydrogenations are started by switching on the stirrer.

The following phosphate buffers are used: pH 7 = 0.041 mol of Na2HP04 and 0.028 mol of KH2PO4/litre; pH 8 = 0.063 mol of Na2HP04 and 0.004 mol of KH2P04/litre.

Conversion and optical yield (enantiomeric excess = ee) are determined on the reaction product using a GC (column: Chirasil-Val). In the case of hydrogenations of acetamidocinnamic acid, the reaction product is firstly esterified using trimethyloxonium tetrafluoroborate.

Examples C1-C15: Method a: Hydrogenation in water [Rh (norbornadiene) 2] BF4 and the ligand are dissolved in 2 ml of methanol (MeOH), the solution is stirred for 10 minutes and the MeOH is then evaporated under reduced pressure at 20-40°C. In a second vessel, acetamidocinnamic acid is slurried in 5 ml of water, and 2N NaOH is slowly added dropwise until a clear solution is obtained. The phosphate buffer (pH 7) is subsequently added. The pH is checked using pH paper or a pH electrode, and if necessary adjusted to pH 7 by means of O. 1N NaOH or methanesulfonic acid, and the solution is made up to a volume of 10 ml by addition of water. This solution is added to the previously prepared solid catalyst and the mixture is stirred until the catalyst has dissolved (dissolution can be accelerated by means of ultrasound). The solution obtained is transferred to a hydrogenation apparatus and hydrogenated.

Method b: Hydrogenation in water The ligand is placed in a vessel and dissolved while stirring with the aid of 0.1N NaOH (1 equivalent of NaOH per equivalent of carboxylic acid functions).

After addition of 1 ml of water, the ligand solution is added to [Rh (norbornadiene) 2] BF4 and the mixture is stirred until a clear solution is obtained. The substrate solution is prepared as described in Method a, but is made up to 10 ml with water only after addition of the catalyst solution. This solution is transferred to a hydrogenation apparatus and hydrogenated.

Method c: Hydrogenation in water The ligand is placed in a vessel, suspended in 1 ml of water and dissolved by addition of methanesulfonic acid (1% by volume in water) (1 equivalent of methane- sulfonic acid per equivalent of amine functions). The ligand solution is added to [Rh (norbornadiene) 2] BF4 and the mixture is stirred until a clear solution is obtained. The substrate solution is prepared as described in Method a), but is made up to 10 ml with water only after addition of the catalyst solution. This solution is transferred to a hydrogenation apparatus and hydrogenated.

Method d: Hydrogenation in methanol [Rh (norbornadiene) 2] BF4 and the ligand are dissolved in 2 ml of MeOH and the solution is stirred for 10 minutes. The substrate, dissolved in 8 ml of MeOH, is added to this catalyst solution. The resulting solution is transferred to a hydrogenation apparatus and hydrogenated.

Example of catalyst separation by extraction: The hydrogenation solution from Example C9 is, after the hydrogenation, evaporated to dryness under reduced pressure and the mixture is subsequently extracted with ethyl acetate and water (with phosphate buffer, pH 8).

The organic phases are collected, washed with water, dried over sodium sulfate and finally evaporated to dryness on a rotary evaporator. Analysis of the organic product indicates that more than 95% of the rhodium have been separated off.

The reaction conditions are indicated in Table 1. In Examples C5 and C9, MAC is used as substrate, while AC is used in all the others. The hydrogenation in Examples Cl and C10 is carried out by Method a), in Examples C2 to C7 by Method b), and in Examples C7, C8, Cll, C12 and C15 by Method c). In Examples C2 to C4 and C7, the hydrogen pressure is 15 bar, in all the other examples 1 bar. In Examples C1-C4, C6, C7 and C10, the solvent is water; in Examples C8, C9, Cll, C12 and C15 it is methanol; and in Example C5 it is a mixture of 10 ml of water (phosphate buffer pH 8) and 7 ml of ethyl acetate (two-phase hydrogenation). Reaction temperature: in Examples C1-C5 and C10-C11: 25 °C; in Examples C6 to C9 and C12-C15: 40 °C.

Table 1 Example mmol of pmol of Ligand llmol of Sub-Buffer no. sub-rhodium ligand strate: ml strate catalyst C1 2. 5 12. 5 K1 15. 6 200 1 C2 12. 5 12. 5 K1 15. 6 1000 2 1.3K11.6100002.5C312.5 C4 12. 5 0. 42 K1 0. 53 30000 2.5 C5 2. 5 12. 5 K1 15. 6 200 2.5 C6 2. 5 12. 5 K5 15. 6 200 1 C7 2. 5 2. 5 K5 3. 1 1000 2 C8 2. 5 12. 5 K5 15. 6 200- C92. 56. 3K57. 5400 12.5K412.52001C102.5 Cll 2. 5 12. 5 K4 12. 5 200 _ C12 2. 5 12. 5 K6 15 200 _ Example mmol of Rmol of Ligand Rmol of Sub-suffer no. sub-rhodium ligand strate: ml strate catalyst C13 K6 15 200 1 C14 K7 15 200 1 C15 K7 15 200 -

The results of the examples listed in Table 1 are shown in Table 2.

Table 2 Example Conversion Reaction time (ee) no. (h) (configuration) Cl 100 0. 4 95.5 (R) C2 100 0. 5 95.5 (R) C3 100 2 94 (R) C4 100 10 91.2 (R) C5 77 2 79 (R) C6 94 2. 8 61 (S) C7 100 0. 7 48 (S) C8 100 0. 16 78 (S) C9 100 0. 3 91 (S) C10 15 16 67 (R) Cil 100 0. 5 61 (R) C12 100 0. 5 84.5 (S) C13 60 5 55.7 (S) C14 100 2 91.5 (R) C15 100 0. 4 97.4

Examples C16-C23: Hydrogenation of folic acid 0.0025 mmol of ligand are dissolved in 5 ml of water and 0.5 ml of buffer pH 7 (0.041 mol of Na2HP04 and 0.028 mol of KH2PO4 in 1 1 of water). The carboxylic acid groups of the ligands are then reacted with 0.1N NaOH until a clear solution is formed. The resulting solution is added to 7.4 mg (0.02 mmol) of [Rh (NBD) 2] BF4 and stirred until a solution has formed (NBD is

norbornadiene). This solution is added to a solution of 2 mmol of folic acid disodium salt in 11 ml of water and 1.5 ml of buffer pH 7 and the mixture is transferred to a hydrogenation autoclave provided with a sparging stirrer by means of a syringe in a countercurrent of argon. The autoclave is closed, the argon is replaced by hydrogen and hydrogen is finally injected until the desired pressure has been reached.

The hydrogen pressure is maintained by further supplies from the reservoir via a reducing valve. The hydrogenation is started by switching on the stirrer.

In the following table, the hydrogenation time reported is the time until the reaction stops (no longer any uptake of hydrogen). Unless indicated otherwise, this corresponds to complete conversion of the folic acid.

The pressure is 80 bar and the reaction temperature is 70 °C (30 °C in Example C21). The molar ratio of substrate to catalyst (S/C) is 100 in Examples C16-C22 and is 1000 in Example C23. The optical yield DE is reported for the S, S configuration of folic acid. The results are summarized in Table 3.

Table 3: Example Ligand Time DE (%) Remarks No. (hours) C16 K5 17.5 36 25% folic acid C17 K1 4 47 C18 K2 2 18 C19 K3 3.2 46 C20 K4 0.5 31 C21 K4 12 49 C22 K4 0.6 36 Hydrogenation of folic acid suspension at pH 6.3 C23 K4 4 30 S/C 10002 Footnotes:

1) Buffer pH 6: 0.01 mol of Na2HP04 and 0.071 mol of KH2PO4 in 1 1 of water; at the end of the reaction, another 4 ml of 1N KH2PO4 are added.

2) 5 mmol of folic acid disodium salt, 0.005 mmol of [Rh (NBD) 2] BF4 and 0.00675 mmol of ligand are used in a total of 16 ml of water and 2 ml of buffer pH 7.