Platinum Metals Rev., 1998, 42, (3), 108
Palladium-Mediated Heterogeneous Catalytic Hydrogenations
Selectivity of Liquid-Phase Reactions for the Fine Chemicals Industry
The fine chemicals industry, particularly the pharmaceutical industry, is the most developed of the chemical industries in Hungary, and for this reason our catalytic research work has been focused on liquid-phase hydrogenation and oxidation reactions. In a small country, such as Hungary, the chances of developing large catalytic technologies is minimal, so we have targeted our work on small scale processes, concentrating on investigating and developing hydrogenation processes and their catalysts for the synthesis of fine chemicals.
When developing catalysts, the most important properties are their activity, stability and selectivity. In liquid-phase batch processes, selectivity is of crucial importance, especially in the production of drugs, where the permissible amount of impurities allowed in the product has decreased dramatically in the last decades. This paper describes the various types of selectivity and looks at methods for their optimisation, using liquid-phase hydrogenation reactions as illustrations.
Types of Selectivity
The hydrogenation of 4-chloronitrobenzene in methanol using a catalyst of palladium supported on active carbon at room temperature and atmospheric pressure gives aniline as the product together with hydrochloric acid.
With a catalyst of skeletal nickel under the same conditions the process consumes one mole less of hydrogen and the product is 4-chloroaniline. This demonstrates that the Raney-nickel catalyst is chemoselective in the hydrogenation of halogen-containing aromatic nitro compounds, as it does not cleave the halogen from the aromatic ring.
Regioselectivity can be demonstrated by the palladium-catalysed hydrogenation of β-naphthol, which contains two similar aromatic rings to be saturated.
Under basic conditions the hydroxyl-substituted ring is hydrogenated, while in the presence of acids the unsubstituted ring is reduced producing aromatic tetralol in excess.
If the product molecule contains stereoisomers then the hydrogenation reaction can be characterised by its stereoselectivity. Palladium catalysts can be used to produce stereochemically-enriched products. The hydrogenation of thymol in water at 383 K and 6 bars of pressure results in an equilibrium mixture of menthone isomers (60 per cent menthone, 40 per cent isomenthone), which are themselves hydro-genated after the thymol is almost completely converted.
Under acidic conditions neomenthol is formed in excess, while under basic conditions menthol is formed with 75 per cent stereoselectivity as the main product (1, 2). In menthol the hydroxy group is in the equatorial position.
The most sophisticated type of selectivity is enantioselectivity, where a prochiral compound is converted into a chiral product, and the enantiomers are not produced in a 1:1 ratio:
Two well-known enantioselective heterogeneous hydrogenation reactions, which are carried out using Raney-nickel and platinum catalysts modified, respectively, with the chiral compounds tartaric acid and cinchonidine, are shown above. Under optimised conditions the enantioselectivity in both systems reaches 95 to 98 per cent (3).
In our laboratory, methods were developed for the asymmetric hydrogenation of isophorone using palladium-based catalysts. (S )-proline was used as the chiral auxiliary in one case and (−)-dihydro apovincaminic acid ethyl ester was used as the chiral modifier in the other:
In the presence of a stoichiometric amount of (S )-proline the highest enantiomeric excess was 80 per cent. However, when catalytic amounts of the vinca-alkaloid were used as the chiral modifier, the enantioselectivity reached 55 per cent (4, 5).
Factors Affecting Chemoselectivity
In developing industrial hydrogenation processes one important question is how the selectivity of a reaction can be increased. Ways in which this can be achieved are suggested in the next section.
Change of Solvent
When 4-chloronitrobenzene is hydrogenated in methanol at room temperature and atmospheric pressure using a palladium catalyst, the main products are aniline and hydrochloric acid. However, if the solvent is ethyl acetate, and an 0.2 per cent palladium/carbon catalyst is used with respect to the substrate, then chemoselectivity to 4-chloroaniline increases to over 90 per cent.
In the less polar, aprotic, ethyl acetate solvent the solvation changes, affecting the adsorption strength of the substrate and the products. Thus the hydrogenation of chloroaniline to aniline becomes a very slow process.
Another example of solvent effects is shown by the selective ring saturation of pyrrole derivatives which can be performed using a catalyst of palladium/carbon, but only in a solvent mixture; however, in methanol the reaction proceeds easily, if rhodium or ruthenium catalysts are used (6–9).
Change of Catalyst
In the reduction of benzoic acid derivatives, containing similar functional groups, using a palladium catalyst, a simple change of solvent is not sufficient to preserve the chloro substituent.
To retain the chloro substituent on the benzoic acid, the palladium catalyst has to be substituted for platinum, iridium or nickel (10). The reactions were carried out at room temperature and atmospheric pressure.
Palladium is an especially selective metal for several hydrogenation reactions; for example, it can be used for the selective reduction of a C=C double bond next to an aliphatic carbonyl. A classic example where the selectivity of palladium has been increased is the Lindlar catalyst. This is a lead-poisoned palladium catalyst which can selectively saturate a triple bond to a double bond. Another example is the catalyst for the Rosenmund reaction, where after poisoning the activity of the palladium decreases, and only the selectivity is enhanced. Here palladium is applied to a low surface area BaSO4 support and poisoned with a sulfur-containing compound. This catalyst is effective for the selective hydrogenation of aromatic acid chlorides to the corresponding aldehydes.
Although this method of catalyst modification is applicable for many reactions, it is not sufficiently reproducible for industrial use. Therefore we have developed a preparation method for catalysts used in the Rosenmund reaction (11).
Alloying the Active Metal: According to our method the hydrogenation of acid chlorides is carried out at elevated temperature using a catalyst of palladium-copper alloy supported on active carbon, with aromatic hydrocarbons as solvents (11). The selectivities achieved for aldehyde production are between 70 and 90 per cent.
The alloying can be carried out during the preparation of the palladium catalyst or by metal adsorption at controlled potential onto the ready made palladium catalyst.
The following aromatic aldehydes were prepared by the hydrogenation of the corresponding acid chlorides using Pd-Cu/C catalyst:
Poisoning the Catalyst: In addition to their use in alloying, anions and cations can also increase catalyst selectivity by poisoning. Sometimes they work by causing a reduction in catalyst activity. However, when used in small controlled quantities they can improve the selectivity of palladium catalysts, such as in the hydrogenation of nitroso lupetidine:
Additional advantages of this process are that water can be used as the solvent and that the reaction takes place at low pressure and temperature.
In the next example, Cl- anions were used to increase the selectivity of the palladium catalyst. The reaction was first carried out in hydrochloric acid solution, but the reaction rates were slow. Then it was found that much higher reaction rates could be achieved in sulfuric acid solution containing small amounts of chloride ions, without any loss in the 95 per cent selectivity obtained with hydrochloric acid solution.
The reactions were carried out at 323 K and under 3 bars of pressure.
Optimised Reaction Conditions
Last, but not least, the optimisation of the reaction conditions is a good method to increase the selectivity:
In the palladium-mediated hydrogenation of dibenzo-18-crown-6 carried out at 423 K and under 10 bars of pressure, the proportion of products produced by hydrogenolysis was too high. By determining the optimal temperature, pressure, catalyst and substrate concentration, the selectivity was able to be increased from 40 to 65 per cent.
Factors Affecting Regieoselectivity
Optimised Catalyst Preparation
A new method of preparation was developed for palladium catalysts used in liquid-phase hydrogenations, in which a tetraalkyl ammonium palladate precursor is formed (12).
The resulting amine-containing catalyst has a dispersion > 50 per cent and gives higher selectivity (80 per cent) and activity for the hydrogenation of naphthol to aromatic tetralol:
The Role of pH in the Stereochemistry
It is known that the pH of a reaction can influence product distribution; for example, when cyclic ketones are reduced, the pH affects the ratio of the product stereoisomeric alcohols, demonstrated by the hydrogenation of thymol and menthone. This experience has been used to develop the stereoselective hydrogenation of 4-acetyl-amino phenol.
Using water as the solvent and a basic buffer, the stereoselectivity towards trans -4-acetyl-amino cyclohexanol was 80 per cent in the palladium-mediated hydrogenation of 4-acetyl-amino phenol. However, when the same reaction was performed with Adams’ platinum catalyst the stereoselectivity was only 40 per cent to the trans -4-acetyl-amino cyclohexanol. The trans product has two forms: the equatorial-equatorial and the axial-axial, with respect to the position of the two substituents. The form of the ketone containing the acetylamino substituent in the equatorial position can be adsorbed on the catalyst surface more easily. If the reduction of this form results in the equatorial alcohol, which is preferred in basic solution, the product will be an excess of the trans stereoisomer.
Factors Affecting Diastereoselectivity
The most important factors, which influence diastereoselectivity, are:
the properties of the molecule to be hydrogenated – the substituents of the unsaturated atoms
the catalytically active metal, the support that is used and its dispersion
the solvent (protic-aprotic, polar-apolar)
the preliminary reduction of the catalyst surface and its saturation with hydrogen
the pH of the reaction mixture.
The hydrogenation of the Schiff-base prepared by the condensation of acetophenone and optically active 1-phenyl-ethyl amine is a good example for demonstrating these factors (13):
Here, in an apolar solvent, such as benzene, the diastereoselectivity is over 90 per cent when a palladium catalyst of small dispersion is used. This is determined mainly by the substrate itself, where the bulky phenyl group substituents tend to take up positions as far from each other as possible and are attached to the catalyst surface by anchoring with their π electrons. Therefore the most favourable substrate conformation on the surface is fixed, and if the direction of attack of the entering hydrogen is determined, it results in high diastereoselectivity.
(S )-Proline as a Chiral Auxiliary
(S )-Proline has proved to be an excellent homogeneous chiral catalyst, for example, for the Robinson-type condensation of α, β-unsaturated ketones. We investigated whether this could also be applied to heterogeneous hydrogenations (14).
The hydrogenation of isophorone with a palladium catalyst, in the presence of stoichiometric amounts of (S )-proline and methanol as solvent at room temperature, resulted in the production of dihydroisophorone with enantiomeric excesses up to 80 per cent. However, after the uptake of more than one equivalent of hydrogen the chemical yield of dihydroisophorone decreased significantly, giving the alkylated proline as the major product (15).
In the hydrogenation of acetophenone, with palladium and stoichiometric amounts of (S )-proline and methanol as solvent, the enantiomeric excess was around 20 per cent (16).
Kinetic investigations of the reactions above, using circular dichroism spectroscopic studies of the methanol solution of isophorone and proline, and the characterisation of the side-product alkylated proline led us to conclude that these reactions, while appearing to be enantioselective, are in fact diastereoselective, see Scheme I (14-19).
The scope of enantioselective heterogeneous catalytic hydrogenations using chirally modified catalysts is rather limited; they can only be used for the reduction of α-keto esters and β-dike-tones or β-keto esters. The use of chiral auxiliaries for similar purposes is no more successful, as was demonstrated using (S )-proline, where the enantioselectivity was remarkable, but the chemical yield was low. In order to find new chiral modifiers and auxiliaries a screening was performed and α, α-dihydro apovincaminic acid ethyl ester, a vinca-type alkaloid, proved to be an effective chiral modifier (20-22):
This modifier is effective in the palladium-mediated hydrogenation of isophorone and in the platinum-mediated hydrogenation of ethyl pyruvate. According to circular dichroism spectroscopic investigations and studies of the enantioselective hydrogenations, the probable processes of enantiodifferentiation could be as shown in Scheme II.
Similar processes may take place in platinum-cinchona and nickel-tartaric acid mediated reactions. The enantiomeric excess of the product depends, in each case, on the equilibrium constants of adduct formation and adsorption reactions and the relative rates of chiral and racemic reactions, which are competing hydrogenations.
This new modifier is somewhat broader in scope than previously used systems, but the enantioselectivities are only moderate. However, its study has contributed to improving the understanding of the enantioselective reaction mechanisms.
Liquid-phase hydrogenation has been demonstrated to be a versatile method in organic syntheses. It can offer – even for asymmetric reactions – good catalytic systems which, in some cases, afford chiral products with high stereoselectivity. The scope of the reactions involving mostly palladium-mediated hydrogenations shows the usefulness of these type of reductions among organic synthetic methods. Studying asymmetric hydrogenations has a double impact: chiral compounds can be produced by such processes, in addition to giving detailed information about the elementary steps of surface reactions, for example about the processes involved in enantiodifferentiation.
Production of Enantiomerically Pure Products by Heterogeneous Catalytic Hydrogenations
- 1 A. Tungler,, T. Máthé,, Z. Bende and J. Petró, Appl. Catal ., 1985, 19, 365
- 2 A. Tungler,, T. Máthé,, J. Petró and T. Tarnai, Appl. Catal. A, 1991, 79, 161
- 3 Y. Izumi,, M. Imaida,, H. Fukawa, and S. Akabori,, Bull. Chem. Soc. Jpn ., 1963, 36, 21 ; Y. Orito,, S. Imai and S. Niwa, J. Chem. Soc. Jpn., 1979, 8, 1118
- 4 A. Tungler,, T. Máthé,, J. Petró and T. Tarnai, J. Mol. Catal, 1990, 61, 259
- 5 A. Tungler,, T. Tarnai,, T. Máthé,, G. Vidra and J. Petró, “ New Asymmetric Heterogeneous Catalytic Hydrogenation Reactions ”, 15th Conf. Organic Reactions Catalysis Society, 1994, Phoenix, Marcel Dekker, pp. 202 – 212
- 6 L. Hegedûs,, T. Máthé and A. Tungler, Appl. Catal. A, 1996, 143, 309
- 7 L. Hegedûs,, T. Máthé and A. Tungler, Appl. Catal. A, 1996, 147, 407
- 8 L. Hegedûs,, T. Máthé and A. Tungler, Appl. Catal. A, Gen ., 1997, 153, 133
- 9 L. Hegedûs,, T. Máthé and A. Tungler, Appl. Catal. A, Gen., 1997, 152, 143
- 10 T. Máthé and A. Tungler, Hungarian Patent 203,523; 1990
- 11 J. Petró, T. Máthé, A. Tungler and Z. Csûrõs,, Hungarian Patent 168, 073; 1973, U.S. Patent 4, 021, 376 ; 1974, British Patent 1, 510, 195 ; 1975 J. Petró, T. Máthé and A. Tungler, Hungarian Patent 169, 835; 1973
- 12 T. Máthé,, A. Tungler and J. Petró, Hungarian Patent 177, 860 ; 1979, U.S. Patent 4, 361, 500 ; 1980, British Patent 2, 052, 294B ; 1980, Russian Patent 1, 060, 096 ; 1982, Japanese Patent 21, 660 / 84 ; 1984, German Patent 3, 019, 582 ; 1986, Swiss Patent 644, 770 ; 1983
- 13 A. Tungler,, M. Àcs,, T. Máthé,, E. Fogassy,, Z. Bende and J. Petró, Appl. Catal, 1985, 17, 127
- 14 A. Tungler,, M. Kajtár,, T. Máthé,, G. Tóth,, E. Fogassy and J. Petró, Catal. Today, 1989, 5, 159
- 15 A. Tungler,, T. Máthé,, J. Petró, and T. Tarnai, J. Mol . Catal, 1990, 61, 259
- 16 A. Tungler,, T. Tarnai,, T. Máthé and J. Petró, J. Mol. Catal, 1991, 67, 277
- 17 A. Tungler,, T. Tarnai,, T. Máthé and J. Petró, J. Mol. Catal, 1991, 70, L5 – L8
- 18 G. Tóth,, A. Kovács,, T. Tarnai and A. Tungler, Tetrahedron: Asymmetry, 1993, 4, ( 3 ), 331
- 19 A. Tungler,, T. Tarnai,, A. Deák,, S. Kemény,, A. Gyôry,, T. Máthé and J. Petró, “Studies in Surface Science and Catalysis. Heterogeneous Catalysis and Fine Chemicals III, Poitiers”, Elsevier, Amsterdam, 1993, p. 99
- 20 T. Tarnai,, A. Tungler,, T. Máthé,, J. Petró,, R. A. Sheldon and G. Tóth, J. Mol. Catal. A: Chem ., 1995, 102, 41
- 21 A. Tungler,, T. Máthé,, T. Tarnai,, K. Fodor,, G. Tóth,, J. Kajtár,, I. Kolossváry,, B. Herényi and R. A. Sheldon, Tetrahedron: Asymmetry, 1995, 6, ( 9 ), 2395
- 22 A. Tungler,, T. Máthé,, K. Fodor,, R. A. Sheldon and P. Gallezot, J. Mol. Catal. A: Chem ., 1996, 108, 145