Iridium ferrocenyl diphosphine catalyzed enantioselective reductive alkylation of a hindered aniline

Article abstract of DOI:10.1055/s-1999-3106

The enantioselective reductive alkylation of 2-methyl-5-ethyl-aniline (MEA) with methoxyacetone - using a catalyst generated in situ from [Ir(cod)Cl]₂ and (R)-(S)-PPF-P(3,5-xyl)₂ - to give enriched (S)-N-(2- ethyl-6-methylphenyl)-N-(1'-methoxymethyl)-ethyl-amine is described. At 80 bar and 50 °C in the presence of iodide and methane sulfonic acid and with cyclohexane as solvent, complete conversion is reached within 14 h with a substrate to catalyst ratio of 10'000 and an ee of 76-78%. The effect of solvent and acid type were found to be important. To our knowledge, this is the first enantioselective reductive alkylation ever reported.

Full text of DOI:10.1055/s-1999-3106

LETTER
867
Iridium Ferrocenyl Diphosphine Catalyzed Enantioselective Reductive
Alkylation of a Hindered Aniline
a
b
a
a
a
Hans-Ulrich Blaser,* Hans-Peter Buser, Hans-Peter Jalett, Benoit Pugin , Felix Spindler
a
Novartis Services AG
b
Novartis Crop Protection; WRO 1055.627, CH-4002 Basel, Switzerland
E-mail: hans-ulrich.blaser@sn.novartis.com
Received 1 February 1999
MEA with methoxyacetone using the Ir-xyliphos catalyst
in presence of iodide ions and acid.
Abstract: The enantioselective reductive alkylation of 2-methyl-5-
ethyl-aniline (MEA) with methoxyacetone - using a catalyst gener-
ated in situ from [Ir(cod)Cl] and (R)-(S)-PPF-P(3,5-xyl) - to give
2
2
enriched (S)-N-(2-ethyl-6-methylphenyl)-N-(1’-methoxymethyl)-
ethyl-amine is described. At 80 bar and 50 °C in the presence of io-
dide and methane sulfonic acid and with cyclohexane as solvent,
complete conversion is reached within 14 h with a substrate to cat-
alyst ratio of 10'000 and an ee of 76-78 %. The effect of solvent and
acid type were found to be important. To our knowledge, this is the
first enantioselective reductive alkylation ever reported.
Key words: enantioselective reductive alkylation, hindered N-alkyl
aniline, Ir-ferrocenyl diphosphine, acid catalysis, (S)-metolachlor
®
Metolachlor is the active ingredient of Dual , one of the
most important grass herbicides for use in maize and a
number of other crops. It is an N-chloroacetylated, N-
alkoxyalkylated ortho disubstituted aniline. The commer-
cial product was introduced to the market in 1976 as a
mixture of four stereoisomers and is produced via a Pt cat-
alyzed reductive alkylation of 2-methyl-5-ethyl-aniline
Scheme
(
MEA) with aqueous methoxyacetone in the presence of
1
traces of sulfuric acid followed by chloroacetylation (see
upper part of the Scheme). Already in 1982 it was found
that about 95% of the herbicidal activity of metolachlor
Table 1 Reaction of MEA with methoxyacetone in presence of Ir-
xyliphos, acetic acid, and iodide ions: Effect of solvents on reaction
time, conversion and ee.
2
lies in the two (1’S)-diastereomers . In 1997, after years of
3
®
intensive research , Dual Magnum with a content of ap-
proximately 90% (1’S)-diastereomers and with the same
biological effect at about 65% of the use rate was intro-
duced in the USA. To make this ‘chiral switch’ possible,
a new technical process had to be found that allowed the
economical production of the enantiomerically enriched
precursor of metolachlor. Key step of this new synthesis
is the enantioselective hydrogenation of N-(2-ethyl-6-
methylphenyl)-N-(1'-methoxymethyl)-ethylidene-amine
(
see lower part of the Scheme). The optimized process op-
erates at 80 bar hydrogen and 50 °C with a catalyst gener-
ated in situ from [Ir(cod)Cl]2 and
the
ferrocenyldiphosphine ligand xyliphos at a substrate to
catalyst ratio (s/c) of >1'000'000. Complete conversion is
reached within 3-4 h, the initial tof exceeds 1'800'000 h
and enantioselectivity is approximately 80%.
Even though metal catalyzed reductive alkylation is a
method with a large preparative potential, no literature re-
ports on a successful enantioselective version could be
found. Therefore, in a first attempt, MEA and aqueous
-
1
It would be a further improvement to combine the advan- methoxyacetone were used as in the racemic process and
tages of both production processes and to use the chiral reaction conditions close to the optimized enantioselec-
homogeneous catalyst for carrying out a reductive alkyla- tive process with an s/c ratio of 10'000. In presence of ace-
tion without isolating and purifying the imine. This paper tic acid and TBAI (tetrabutyl ammonium iodide), the
describes the investigation of the reductive alkylation of desired reaction indeed took place with a remarkably high
Synlett 1999, S1, 867–868 ISSN 0936-5214 © Thieme Stuttgart · New York

8
68
H.-U. Blaser et al.
LETTER
ee of 76% (Table 1, entry 2) but compared to the reference ence of cyclohexane there is still too much MEA and wa-
reaction with the isolated imine (Table 1, entry 1), the cat- ter present that are known to deactivate the Ir
alyst activity was quite low. When dry methoxyacetone ferrocenyldiphosphine catalyst. The reason for this behav-
was used, conversion increased significantly (Table 1, en- ior could either be a strong complexation, thereby block-
try 3). Since it was noticed that in both cases two liquid ing the access of the imine, or a slow dimerization of the
phases were formed, the effect of ethanol addition was in- Ir xyliphos complex to give an inactive species as ob-
5
vestigated. However, the resulting activity in a one-phase served for other Ir catalysts .
situation was even lower and the ee dropped to 15% (Ta-
ble 1, entry 4). As a consequence, cyclohexane was added
to the reaction mixture and the expected increase in con-
Acknowledgement
version and also a slightly better enantioselectivity was Robert Häusel, Beat Eng, Heidi Landert, Geneviève Thoma, Nadia
Vostenka and Andrea Holderer are thanked for their very careful
and skillful experimental work.
observed (Table 1, entry 5).
From these results it was concluded that best effects
would be obtained when as little water and MEA as pos-
References and Notes
sible were present in the organic phase. Since acetic acid
can also act as solvent mediator and is used in large
amounts, it was replaced by stronger acids that are effec-
tive at much lower concentrations. Thus, as shown in Ta-
ble 2, the catalyst activity was further improved. Best
results were achieved with small amounts of trifluoroace-
tic acid or of methane sulfonic acid while sulfuric acid
was not effective probably due to its low solubility .
(
1) Bader, R.R.; Blaser, H.U. Stud. Surf. Sci. Catal. 1997, 108, 17.
Bader R.R.; Flatt, P.; Radimerski, P., EP 605363-A1 (Ciba-
Geigy AG, 1992)
(2) Moser, H.; Ryhs, G.; Sauter, H. Z. Naturforsch. 1982, 37b,
451.
(3) For a case history of the discovery of the new catalyst system
see Spindler, F.; Pugin, B.; Jalett, H. P.; Buser, H. P.;
Pittelkow, U.; Blaser, H. U. Chem. Ind., 1996, 68, 153.
4) General hydrogenation procedure. The appropriate amounts
of 2-methyl-5-ethyl aniline and methoxyacetone were placed
(
Table 2 Reaction of MEA with methoxyacetone in presence of
cyclohexane, Ir-xyliphos, acetic acid, and iodide ions: Effect of acid
type on reaction time, conversion and ee.
in a 50 ml autoclave. Then, [Ir(cod)Cl] , xyliphos, tetrabutyl-
2
ammonium iodide, and (if stated) the solvent and/or acid were
added. The autoclave was sealed, first set under argon and
then under the required pressure of hydrogen. The reaction
mixture was intensively stirred with a magnetic stirring bar
(1200 rpm) for the time given at the specified temperature. Af-
ter removal of the solvent, the conversion was determined by
glc (column: DB 17/30W, 15 m (JCW Scientific Inc.), tempe-
rature program: 60 °C/1 min. to 220 °C, ∆T: 10°/min), and the
ee was measured by HPLC on Chiralcel-OD-H (hexane/2-
propanol: 99.7:0.3; flow rate 1ml/min; α=1.11).
Xyliphos. The ligand was prepared according to the procedure
6
described by Togni et al.
(
5) Blaser, H.U.; Pugin, B.; Spindler, F., Angew. Chem. 1990,
02, 561; Angew. Chem., Int. Ed. Engl. 1990, 29, 558.
Dorta, R.; Togni, A., Organometallics 1998, 17, 3423 and
441.
1
5
(
6) Togni, A.; Breutel, C.; Schnyder, A.; Spindler, F.; Landert, H.;
These results are quite remarkable because this is the first
enantioselective reductive alkylation of an aniline with a
ketone and the ee values and especially the s/c ratio of
Tijani, A., J. Am. Chem. Soc. 1994, 116, 4061.
1
0’000 are respectable. However, we clearly missed our
Article Identifier:
437-2096,E;1999,0,S1,0867,0868,ftx,en;W05499ST.pdf
objective to find a more economical process for the pro-
duction of the (1’S)-diastereomers of metolachlor because
the catalyst productivity is far too low compared to the
imine hydrogenation process. It seems that even in pres-
1
Synlett 1999, No. S1, 867–868 ISSN 0936-5214 © Thieme Stuttgart · New York