Steric effects in the enantioselective transfer hydrogenation of 2- aroylacetates

Article abstract of DOI:10.1016/S0957-4166(99)00540-6

Benzoylacetate esters and aryl-substituted derivatives are efficiently reduced in 2-propanol using the readily available catalytic combination (1S,2R)-ephedrine/[RuCl₂(η⁶-p-cymene)]₂ to give the corresponding alcohols in high yields and enantioselectivities (up to 94% ee). (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0957-4166(99)00540-6

TETRAHEDRON:
ASYMMETRY
Pergamon
Tetrahedron: Asymmetry 10 (1999) 4663–4666
Steric effects in the enantioselective transfer hydrogenation of
2-aroylacetates
Kathelyne Everaere,a Jean-François Carpentier,a,∗ André Mortreux a and Michel Bulliard b
aLaboratoire de Catalyse de Lille associé au CNRS, ENSCL, B.P. 108, 59652 Villeneuve d’Ascq Cedex, France
bPPG-SIPSY, Z.I. La Croix Cadeau, B.P. 79, 49242 Avrillé Cedex, France
Received 12 November 1999; accepted 25 November 1999
Abstract
Benzoylacetate esters and aryl-substituted derivatives are efficiently reduced in 2-propanol using the readily
available catalytic combination (1S,2R)-ephedrine/[RuCl2(η6-p-cymene)]2 to give the corresponding alcohols in
high yields and enantioselectivities (up to 94% ee). © 2000 Elsevier Science Ltd. All rights reserved.
Despite significant advances in recent years, the catalytic enantioselective transfer hydrogenation of
ketones remains essentially limited to simple aryl alkyl ketones and α,β-acetylenic ketones.1 Studies by
us and others have shown that the reduction of functionalized ketones, such as the important β-ketoesters,
with Ru- and Rh-based catalysts leads in general to sluggish activity and/or poor enantioselectivities.2–4
The few notable exceptions to this trend are: (a) the reduction of PhCOCO2Me in iPrOH using Rh/C2-
diamine catalysts which affords methyl mandelate in up to 99% ee;5 (b) the reduction of PhCOCH2CO2Et
1a using the HCO2H/NEt3 system and the catalytic combination (S,S)-TsDPEN/[RuCl2(C6H6)]2 which
affords the corresponding alcohol 2a in 93% ee;6 (c) the recently reported transfer hydrogenation
of PhCO(CH2)nX derivatives (n=1–3, X=OR, OCOR, NRCO2R0) with a Ru–[cis-1-aminoindan-2-ol]
system in 38–97% yields and 79–88% ees.7 We here report that effective enantioselective transfer
hydrogenation of various 2-aroylacetates can be achieved using 2-propanol as the hydrogen source in
the presence of simple catalytic combinations of (1S,2R)-ephedrine with an adequate [RuCl2(η6-arene)]2
precursor (Scheme 1).
In order to assess the effect of phenyl substituents on catalytic performance, aroylacetate ethyl
esters 1c–f and 1h–i were prepared by condensation of the corresponding deprotonated acetophenone
derivative on diethylcarbonate.8 In addition, to probe the influence of the alkoxycarbonyl moiety, tert-
butyl esters 1b,g were synthesized by transesterification of the corresponding ethyl ester. The chiral
catalyst precursors were prepared ex situ by heating a mixture of [RuCl2(η6-arene)]2 and (1S,2R)-(+)-
ephedrine (2 equiv. vs. Ru) in 2-propanol, and transfer hydrogenation was carried out at 50°C under usual
reaction conditions.9 Representative results are reported in Table 1.
∗
Corresponding author. E-mail: carpentier@ensc-lille.fr
0957-4166/99/$ - see front matter © 2000 Elsevier Science Ltd. All rights reserved.
PII: S0957-4166(99)00540-6
tetasy 3175

4664
K. Everaere et al. / Tetrahedron: Asymmetry 10 (1999) 4663–4666
Scheme 1.
Table 1
Asymmetric transfer hydrogenation of 2-aroylacetates 1a–ia
In nearly all cases, virtually quantitative yields of 2 were obtained in reasonable times. A screening
of different [RuCl2(η6-arene)]2 precursors revealed that catalytic performances are strongly affected by
this parameter; i.e., the benzene complex induces significantly higher rates than its p-cymene equivalent
(as judged from the catalyst turnover frequency at half-reaction, TOF50) but the latter complex generates
much more enantioselective10 catalytic species (compare entries 1/2, 5/6 and 8/9). The present results
follow the same trend as that observed for the reduction of other types of functionalized ketones with
similar systems,2,4 and suggest that the bulkier the arene, the lower the catalytic activity, the higher

K. Everaere et al. / Tetrahedron: Asymmetry 10 (1999) 4663–4666
4665
the enantioselectivity. The reduction of tert-butyl esters 1b,g proceeds faster than that of their ethyl
equivalents 1a,f but is less enantioselective (compare entries 2/3 and 10/11). This increase in the apparent
rate going from ethyl to tert-butyl esters is most likely related to the bulkiness of the tert-butoxy group,
which possibly prevents the formation of less or non-reactive species; the decrease in enantioselectivity,
an opposite and more marked trend compared to the reduction of simple alkyl β-ketoesters,2,4 indicates
the relatively high sensitivity of this process to steric factors. In fact, when ortho substituents are
introduced on the phenyl ring, ees remain of synthetic interest but they are systematically lower than
for the reduction of 1a. It seems indeed that this drop in enantioselectivity reflects more steric than
electronic factors (and possible coordinating abilities) as the bulkier the substituent, the lower the ee
(Cl, 1f<Me, 1c<OMe, 1d<H, 1a) (compare entries 2/4/6/10 and 1/5). Despite there being no obvious
correlation of ees with Hammett parameters, electronic factors cannot be definitively ruled out in view of
the comparable decrease of ees upon the introduction of para substituents (entries 2/9/12/13 and 1/8).
In the course of this ongoing study, we also observed a relevant phenomenon. When transfer hydro-
genation of a 1:1 mixture of 1a and 1c was conducted under the usual reaction conditions (see Scheme
2), it turned out that the reduction of 1c barely proceeded before that of 1a was completed (Fig. 1).
So, in contrast with the nearly identical reactivities of these two aroylacetates in individual experiments
(see TOF50 values in Table 1), a catalytic kinetic resolution occurs on the mixture. The same parallel
experiment was repeated on a 1:1 mixture of 1a and 1d, knowing that 1d is reduced more than one order
of magnitude faster than 1a in separate experiments using the p-cymene–Ru precursor; in this case, the
transformation of both aroylacetates proceeded at a comparable apparent rate (yields of 2a and 2d=37
and 46% after 5 min, 72 and 90% after 10 min, under the conditions of Scheme 2), evidencing less
spectacularly the same phenomenon. In both parallel experiments, β-hydroxyesters 2 were formed with
the same ees as in individual experiments (Table 1). This kinetic resolution brings up some interesting
mechanistic features and reflects, in our opinion, the relative affinities of aroylacetates for the active
Ru species. Considering Noyori’s mechanism,6 we assume that this affinity corresponds to the easiness
of approach of the aroylacetate (one preferred enantioface) toward the 18 electron hydrido complex
[HRu(η1,η1-ONHMe)(η6-arene)] for hydrogen transfer. The fact that, in the mixtures of aroylacetates so
far investigated, the less hindered one, i.e. 1a, reacts faster is consistent with this hypothesis.
Scheme 2.
Efforts are currently under way to enlarge this study in order to confirm the latter hypothesis and to
rationalize steric effects on catalytic performance in transfer hydrogenation of β-ketoesters.
Acknowledgements
This work was supported by a grant (K.E.) from PPG-SIPSY SA and the CNRS.

4666
K. Everaere et al. / Tetrahedron: Asymmetry 10 (1999) 4663–4666
Fig. 1.
References
1. For a recent review, see: Palmer, M. J.; Wills, M. Tetrahedron: Asymmetry 1999, 10, 2045–2061.
2. Everaere, K.; Carpentier, J.-F.; Mortreux, A.; Bulliard, M. Tetrahedron: Asymmetry 1998, 9, 2971–2974.
3. Schwink, L.; Ireland, T.; Püntener, K.; Knochel, P. Tetrahedron: Asymmetry 1998, 9, 1143–1163.
4. We have found recently that significantly improved (with respect to well-established systems such as Ru–TsDPEN) catalytic
activities (up to 600 h−1 at 20°C) and enantioselectivities (56–89% ee) can be achieved for the reduction of simple alkyl
β-ketoesters, methoxyacetone and 2-acetylpyridine in 2-propanol by using catalytic combinations of an adequate [RuCl2(η6-
arene)]2 precursor and N-substituted derivatives of (1S,2R)-norephedrine such as (1S,2R)-N-benzyl-norephedrine. Everaere,
K.; Carpentier, J.-F.; Mortreux, A.; Bulliard, M. Tetrahedron: Asymmetry 1999, 10, 4083–4086.
5. Gamez, P.; Fache, F.; Mangeney, P.; Lemaire, M. Tetrahedron Lett. 1993, 34, 6897–6898.
6. (a) Noyori, R.; Hashiguchi, S. Acc. Chem. Res. 1997, 30, 97–102. (b) Fujii, A.; Hashiguchi, S.; Uematsu, N.; Ikariya, T.;
Noyori, R. J. Am. Chem. Soc. 1996, 118, 2521–2522.
7. Kenny, J. A.; Palmer, M. J.; Smith, A. R. C.; Walsgrove, T.; Wills, M. Synlett 1999, 1615–1617.
8. Comble, L.; Jones, R.; Palmer, C. J. Chem. Soc., Perkin Trans. 1987, 317–332.
9. A typical procedure is as follows (entry 2): In a 100 mL Schlenk tube placed under nitrogen, [RuCl2(η6-p-cymene)]2 (6.1
mg, 0.01 mmol) was added to a solution of (1S,2R)-ephedrine (6.6 mg, 0.04 mmol) in iPrOH (5 mL). The mixture was stirred
at 80°C for 20 min leading to an orange homogeneous solution. After cooling to room temperature, iPrOH (14 mL), ethyl
benzoylacetate (1a, previously distilled, 384 mg, 2.0 mmol) and iPrOK (0.12 m in iPrOH, 1.0 mL, 0.12 mmol) were added.
The reaction vessel was stirred at 50°C and the reaction was monitored by GLC analysis of aliquots (BPX5 and Chirasil CB
columns). After completion, the solution was concentrated under vacuum and the crude product was chromatographed on
1
silica to give 2a as a colourless oil (0.35 g, 92%). The identity of products 2a–i was established by MS, H and 13C{1H}
NMR, in agreement with literature data.
10. In every case, ee values were constant (±4%) throughout the whole reaction course, thus indicating that no significant
racemization occurred with the present catalytic systems.