One-pot and sequential asymmetric hydrogenation of β,δ-diketoesters into functionalized 1,3-diols: From anti- to syn-stereoselectivity

Article abstract of DOI:10.1002/(SICI)1099-0690(199912)1999:12<3421::AID-EJOC3421>3.0.CO;2-1

The asymmetric hydrogenation of methyl 3,5-dioxohexanoate (1) into mixtures of 3,5-dihydroxyesters (2) and 3-hydroxylactones (3) has been reinvestigated with a variety of ruthenium catalysts. Catalysts bearing diphosphanes which possess axial chirality such as (S)-MeO-Biphep give predominantly (3R,5S)-anti-2 in up to 78% de and 95% ee, affording an efficient synthesis of (S)-6-methyl-5,6-dihydro-2-pyrone [(S)-5] in up to 95% enantiomeric excess. On the contrary, some Ru-{amidophosphane-phosphinite} complexes catalyze sluggishly the formation of syn-2 in up to 92% de but with poor enantiomeric excesses. In all cases, methyl (R)-3-hydroxy-5-oxohexanoate [(R)-11] is the exclusive primary hydrogenation product, which can be isolated in high yields and enantiomeric excesses up to 78%. Further hydrogenation of enantiomerically enriched (R)-11 in a separate experiment affords (3R,5R)-syn-2 in high diastereomeric and enantiomeric excesses (up to 80% and 98%, respectively), provided a Ru-(R)-Binap-type catalyst is used.

Full text of DOI:10.1002/(SICI)1099-0690(199912)1999:12<3421::AID-EJOC3421>3.0.CO;2-1

FULL PAPER
One-Pot and Sequential Asymmetric Hydrogenation of β,δ-Diketoesters into
Functionalized 1,3-Diols: From anti- to syn-Stereoselectivity
[a]
[a]
[a]
´
´
Veronique Blandin, Jean-Francois Carpentier,* and Andre Mortreux
¸
Keywords: Asymmetric hydrogenation / 3,5-Dioxoester / Diols / Lactone / Ruthenium catalyst
The
asymmetric
hydrogenation
of
methyl
3,5-
syn-2 in up to 92% de but with poor enantiomeric excesses.
In all cases, methyl (R)-3-hydroxy-5-oxohexanoate [(R)-11] is
the exclusive primary hydrogenation product, which can be
isolated in high yields and enantiomeric excesses up to 78%.
dioxohexanoate (1) into mixtures of 3,5-dihydroxyesters (2)
and 3-hydroxylactones (3) has been reinvestigated with a
variety of ruthenium catalysts. Catalysts bearing
diphosphanes which possess axial chirality such as (S)-MeO– Further hydrogenation of enantiomerically enriched (R)-11 in
Biphep give predominantly (3R,5S)-anti-2 in up to 78% de
and 95% ee, affording an efficient synthesis of (S)-6-methyl-
5,6-dihydro-2-pyrone [(S)-5] in up to 95% enantiomeric
a separate experiment affords (3R,5R)-syn-2 in high
diastereomeric and enantiomeric excesses (up to 80% and
98%, respectively), provided a Ru-(R)-Binap-type catalyst is
excess. On the contrary, some Ru-{amidophosphane- used.
phosphinite} complexes catalyze sluggishly the formation of
outcome of the reaction was expected not to be simple. The
Introduction
conversion of a β,δ-diketoester such as 1 into the corre-
sponding 3,5-dihydroxyester 2 obviously proceeds in two
steps, and the stereochemistry of the hydrogenation of the
remaining carbonyl group should be affected, not only by
the chirality of the catalyst, but also by the chiral centre
formed in the first hydrogenation. In particular, the hydro-
genation of β-diketones is a typical example of such a
double asymmetric induction, which leads selectively to al-
most enantiopure anti-1,3-diols.[5,6b] Indeed, Saburi et al.
observed that the Љone-potЉ asymmetric hydrogenation of
β,δ-diketoesters [Scheme 1; R ϭ Me (1), Pr, BnOCH2] using
{RuCl2[(S)- or (R)-Binap]}2·NEt3 as the catalyst precursor
gave dominantly anti-3,5-dihydroxyesters. The monohyd-
rogenation intermediate was neither isolated nor identified
from the reaction mixture; however, the results of separate
asymmetric hydrogenations conducted on enantiomerically
pure tert-butyl 5-substituted 5-hydroxy-3-oxohexanoate
(8),[7] one of the two possible intermediates, suggest that
the hydrogenation of β,δ-diketoesters proceeds in fact via
first selective hydrogenation of the C-3 (β) carbonyl group
and subsequent hydrogenation of the C-5 (δ) carbonyl
group, under anti-selective stereocontrol as for the hydro-
genation of β-diketones.[5,6b]
Saburi et al. suggested some years ago that if the hydro-
genation of a β,δ-diketoester (such as 1) could be success-
fully performed to yield an enantiomerically pure syn-3,5-
dihydroxyester (2), it would provide a short and efficient
access to hydroxylactones (3),[1][2] which are the key build-
ing blocks for the synthesis of inhibitors of HMG-co-
enzyme A reductase, such as compactin (4a)[3] and mevino-
lin (4b)[4] (Scheme 1). The requirements for the success of
such a process are a high chemo-, enantio- and diastereosel-
ectivity for the formation of syn-3,5-dihydroxyester 2.
Scheme 1. Desired one-pot synthesis of chiral hydroxylactones and
some target molecules
We have reinvestigated this reaction with a variety of
well-established and new ruthenium-based catalyst precur-
sors using methyl 3,5-dioxohexanoate (1) as a model sub-
strate. It is shown that some catalytic systems allow the so
far unprecedented syn-diastereoselective formation of func-
tionalized diol 2, and that 3-hydroxy-5-oxohexanoate (11) is
indeed always the exclusive primary hydrogenation product
(Scheme 2). This intermediate can be isolated in high yields
and fair enantiomeric excesses under smooth reaction con-
ditions and when further hydrogenated in a separate experi-
ment in the presence of a chiral catalyst with adequate con-
figuration, syn-3,5-dihydroxyester 2 and anti-3-hydroxylac-
Chemoselectivity was not expected to be a crucial issue
since β,δ-diketoesters involve, in the same molecule, partial
structures of both β-diketone and β-ketoester, two kinds of
functionalized carbonyl compounds which are efficiently
hydrogenated by chiral rutheniumϪdiphosphane-type cata-
lysts.[5][6] However, for the same reasons, the stereochemical
[a]
Laboratoire de Catalyse, UPRESA CNRS 8010,
´
Ecole Nationale Superieure de Chimie de Lille,
B. P. 108-59652 Villeneuve dЈAscq, France
Fax: (internat.) ϩ 33-3/20436585
E-mail: carpentier@ensc-lille.fr
Eur. J. Org. Chem. 1999, 3421Ϫ3427
WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
1434Ϫ193X/99/1212Ϫ3421 $ 17.50ϩ.50/0
3421

V. Blandin, J.-F. Carpentier, A. Mortreux
FULL PAPER
tone 3 can be prepared in interesting diastereomeric ex- whole reaction mixtures containing 2 and 3 into lactone 5,
cesses and high enantiomeric excesses.
and direct measurement of its enantiomeric excess by chiral
GLC analysis, was used to confirm individual enantiomeric
excesses of each diastereomer as described above; as ex-
pected, it was found in each case that, within experimental
errors: ee(5) ϭ [0.01 ϫ%(syn-9) ϫ%ee(syn-9)] ϩ [0.01
ϫ%(anti-9) ϫ%ee(anti-9)]. Finally, the enantiomeric excess
of the intermediary hydrogenation product (11) was as-
sessed by chiral GLC analysis of the trifluoroacetate ester
obtained by treatment of samples of the reaction mixtures
with trifluoroacetic anhydride. With these analytical tools,
the hydrogenation of β,δ-diketoester 1 was investigated.
Scheme 2. Asymmetric hydrogenation of methyl 3,5-dioxohexa-
noate
Results and Discussion
Scheme 3. Conversion of primary hydrogenation products to lac-
tone and acetals
Analytical Procedures
In order to investigate the reaction paths and the chemo-
and stereoselectivity of the hydrogenation, different analyti-
cal procedures were used. As summarized in Scheme 2, the
asymmetric hydrogenation of 1 leads to a relatively complex
mixture of products. The analysis is hampered by the for-
Asymmetric Hydrogenation of Methyl 3,5-dioxohexanoate
(1) with Ruthenium Catalysts Bearing Chiral Atropisomeric
Diphosphanes
We first investigated the reaction in the presence of Ru-
mation of side products (in parentheses), which are some- Binap type catalysts by extending the study from SaburiЈs
times formed in significant amounts (vide infra). All of precursor, i.e. {RuCl2[(S)-Binap]}2·NEt3, to other diphos-
these products were unambiguously identified from their phanes which possess axial chirality as well as other com-
spectroscopic and analytical data after separation by col- plexes (Scheme 4). The most significant results are reported
umn chromatography and, in some cases, from comparison in Table 1. The hydrogenation with {RuCl2[(S)-
with authentic samples (see Experimental Section). Quanti- Binap]}2·NEt3 in methanol as the solvent was carried out
tative GLC analysis of the crude reaction mixtures allowed as a reference (entry 2); the results found are consistent with
the determination of the conversion of 1 and the respective those previously reported, although the enantiomeric excess
selectivity for the primary hydrogenation products (2, 3, 11) for lactone 5 is slightly higher (88 vs. 78%). The individual
and the side products (5, 12, 13).
enantiomeric excesses determined for each diastereomer,
For stereochemical features, SaburiЈs techniques were which turned out to be identical in this case, confirm the
used with important additions. The relative stereochemistry enantiomeric excess for 5. Attempts were conducted to im-
of the hydrogenation products of 1 was determined after prove the performances of this catalyst system by varying
conversion of 3,5-dihydroxyesters 2 and 3-hydroxylactones the reaction conditions and, in particular, the nature of the
3 into the corresponding acetonides anti-9 and syn-9, by solvent (entries 3, 4). The most interesting result in this way
treatment of small portions of the crude reaction mixtures has been obtained with dichloromethane, in which the reac-
with 2,2-dimethoxypropane (Scheme 3); the syn:anti dia- tion proceeds at a similar rate, but with a better diastereo-
stereomer ratio of 9, which corresponds to the diastereomer and enantioselectivity for anti-dihydroxyester 2. With the
ratio of the parent 1,3-diols 2, was determined by GLC aim of increasing further the selectivities for this com-
analysis on an achiral column. GLC analysis of the same pound, we examined other related catalyst precursors such
diastereomeric acetonides 9 on a chiral column allows the as trifluoroacetatoϪruthenium and dibromoϪruthenium
separation of the four enantiomers; this gives an additional complexes, readily prepared according to HeiserЈs[9] and
[10]
ˆ
confirmation of the syn:anti diastereomer ratio and, most GenetЈs
procedures, respectively (Scheme 4). It turned
importantly, the individual enantiomeric excesses for each out that none of these systems is really more efficient than
of the diastereomers.[8] In previous work, only the average {RuCl2[(S)-Binap]}2·NEt3; only RuBr2[(S)-MeO-Biphep]
enantiomeric excess at the C-5 (δ) carbonyl group was at- exhibits a higher diastereo- and enantioselectivity, but at
tainable after conversion of mixtures of 2 and 3 into 6- the expense of catalytic activity and chemoselectivity (entry
methyl-5,6-dihydro-2-pyrone 5 by treatment with para-tolu- 6). In fact, this catalyst precursor induces the formation of
enesulfonic acid (TsOH) (Scheme 3), and either measure- a noticeable amount of methyl 5-oxohexanoate (12). GLC
ment of the optical purity of purified 5 or HPLC analysis monitoring and sequential hydrogenation experiments (vide
after further derivatization. In this work, conversion of the infra) suggest that this side product arises from the hydro-
3422
Eur. J. Org. Chem. 1999, 3421Ϫ3427

One-Pot and Sequential Asymmetric Hydrogenation
FULL PAPER
genolysis of the primary hydrogenation product, i.e., methyl
5-oxo-3-hydroxyhexanoate (11).
Scheme 5. Possible competitive chelation modes of 1 onto a Ru-
(S)-Binap type catalyst
hydrogenation of simple β-ketoesters and β-diketones with
Ru-Binap-type catalysts yield the corresponding secondary
alcohols in high enantioselectivity (95Ϫ99% ee).[5][6] The
moderate enantiomeric excesses obtained for 11 show that,
for 1, like for other multifunctionalized ketones,[6b] there
is a competitive ligation of functionalities to the Ru atom
(Scheme 5). Because of the enolic structure of 1[11] and the
high final anti diastereoselectivity, it is reasonable to assume
that the hydrogenation of the C-3 (β) carbonyl group of 1
arises mainly from a β-diketone chelated intermediate,
which gives preferentially the (R)-enantiomer of 11 in the
case of a Ru-(S)-Binap type catalyst. On the contrary, the
hydrogenation of simple β-ketoesters with Ru-(S)-Binap
catalysts is known to lead to the corresponding (S)-β-
hydroxyesters.[6] These two chelation modes have, in fact,
opposite directions for the final stereoselectivity at C-3, and
the β-diketone/β-ketoester control ratio in the case of 1 can
be roughly estimated to 80/20. It is noteworthy that this
ratio does not necessarily account for the competitive chel-
ation modes of 1 onto the Ru species, as kinetic issues of
subsequent elementary steps of the catalytic cycle (e.g. H2
oxidative addition, H transfer) should also be taken into ac-
count.[12]
Scheme 4. Chiral atropisomeric diphosphanes and associated cata-
lyst precursors used in this study
Indeed, analysis of aliquot samples established that, for
all of the catalysts investigated in this study, the first step
of the process consists in the hydrogenation of the C-3 (β)
carbonyl group of 1 to yield 11; the other possible monohy-
drogenation regioisomer, i.e., methyl 3-oxo-5-hydroxyhexa-
noate, was never detected. Under the standard reaction con-
ditions, compound 11 enters in competition with 1 towards
the catalyst as soon as it is formed, and it is therefore not
possible to isolate it in combined good yield and selectivity.
This goal was achieved by carrying out the reaction at room
temperature in CH2Cl2 (which proved not to be possible in
CH3OH) and carefully monitoring the reaction by GLC.
Representative examples are given in Table 1 (entries 7Ϫ11).
The enantiomeric excesses of 11 obtained with (S)-catalysts
are moderate and range typically from 60 to 80% in the (R)-
enantiomer. These values highlight interesting issues about
Asymmetric Hydrogenation of Methyl 3,5-dioxohexanoate
(1) with Ru-AMPP Catalysts
We also investigated the catalytic precursor properties of
the mechanistic pathway. Actually, it is known that the some ruthenium complexes of an amidophosphane-phos-
Table 1. Asymmetric hydrogenation of methyl 3,5-dioxohexanoate (1) with Ru-Binap-type catalyst precursors[a]
Entry Catalyst Precursor
solvent temp. time[b] 11
2 ϩ 3
sel
5
Side Products
sel
(°C)
(h)
sel
ee
syn:anti[c]ee syn[c] ee anti[c] ee[d]
1
2
3
4
RuCl2(PPh3)3
CH2Cl2 60
88
10
4
0
87
96
96
21/79
24/76
14/86
20/80
0
0
0
5 (2), 12 (1)
[RuCl2((S)-Binap)]2.NEt3 MeOH 50
[RuCl2((S)-Binap)]2.NEt3 CH2Cl2 60
[RuCl2((S)-Binap)]2.NEt3 MeOH/ 60
iPrOH
23
nd
nd
87 (S,S) 87 (R,S) 88 (S)
-
91[e]
62
3
22 (S,S) 92 (R,S) 84 (S) 5 (1)
30 (S,S) 87 (R,S) 78 (S) 5 (2), 12 (1)
17
69 (R) 80
5
RuBr2((S)-Tol-Binap)
CH2Cl2 60
136
97
27
7
85
97
92
96
96
nd
73
86
14
3
13/87
11/89
nd
2 (S,S) 93 (R,S) 84 (S) -
6
RuBr2((S)-MeO-Biphep) CH2Cl2 60
nd
78 (S,S) 95 (R,S) 95 (S) 5 (1), 12 (6)
7
RuCl2(PPh3)3
CH2Cl2 50
2[f]
0
0
-
0
-
0
-
5 (1)
8
[RuCl2((S)-Binap)]2.NEt3 CH2Cl2 20
29[g]
89
78 (R)
69 (R)
78 (R)
74 (R)
nd
-
-
-
-
9
[RuCl2((S)-Binap)]2.NEt3 iPrOH 60
8
nd
-
-
-
10
11
RuBr2((S)-Tol-Binap)
CH2Cl2 40
40[f]
47
4
nd
-
-
-
RuBr2((S)-MeO-Biphep) CH2Cl2 40
4
nd
-
-
-
[a]
Reactions were conducted under 100 atm H2; [1]/[Ru] ϭ 200:1, [Ru] ϭ ca. 1 mmol·lϪ1. Conversion (mol-%) of 1 and selectivities (sel,
ee, de;%) for the products were determined by quantitative GLC analysis (see Experimental). Ϫ [b] Non optimized time for total conversion
[c]
of 1. Ϫ syn:anti ratio and enantiomeric excesses of 3,5-dihydroxyesters 2 as determined by GLC analysis of acetonides 9. The letters
[d]
into brackets refer to the absolute configuration of the prevailing enantiomer at C-3 and C-5, respectively. Ϫ Enantiomeric excess of
[e]
[f]
5 determined by GLC analysis after dehydration of crude reaction mixtures. Ϫ
[1]/[Ru] ϭ 500:1. Ϫ Conversion of 1 ϭ 97%. Ϫ
[g]
Conversion of 1 ϭ 94%.
Eur. J. Org. Chem. 1999, 3421Ϫ3427
3423

V. Blandin, J.-F. Carpentier, A. Mortreux
FULL PAPER
phinite (AMPP) ligand; i.e., (S)-Ph,Ph-oxoProNOP. Pre- turned out to be a general trend for all the Ru-AMPP cata-
vious studies in our group have shown that {Ru[(S)-Ph,Ph- lytic systems investigated, the highest enantiomeric excess
oxoProNOP](methylallyl)2} [(S)-A, Scheme 6] and its bi- for syn-3 reaching only a modest 40% (entry 4). This poor
s(trifluoroacetato)
analogue,
{Ru[(S)-Ph,Ph- final enantioselectivity for the 3,5-dihydroxyesters stems
oxoProNOP)(TFA)2}, catalyze (sluggishly) the hydrogen- from the incapability of the Ru-[(S)-Ph,Ph-oxoProNOP]
ation of β-ketoesters with moderate to good enantioselectiv- catalysts to promote enantioselectively the first hydrogen-
ities (75Ϫ85% ee).[13] In the present study, we sought to ation at the C-3 (β) carbonyl group; indeed, GLC analysis
observe peculiarities due to the unusual P(O)/P(N) feature of some aliquot samples revealed that the enantiomeric ex-
present in AMPP ligands and for that, we examined a vari- cesses of 11 typically range between 5 and 15% (R). This
ety of catalysts prepared ex situ by protonolysis of (S)-A significant decrease in the enantioface discrimination of the
with some carboxylic and sulfonic[14] acids (Scheme 6) (see chiral Ru-AMPP species going from simple β-ketoesters
Experimental Section).
(75Ϫ85% ee) to β,δ-diketoester 1 is assumed to arise from
the aforementioned competitive ligation of functionalities
(Scheme 5). As described below, when an enantiomerically
enriched sample of 11 is hydrogenated with a Ru-AMPP
catalyst such as (S)-A, much better enantiomeric excesses
with comparable, good diastereoselectivity for syn-3 are ob-
tained.
Asymmetric Hydrogenation of Enantiomerically Enriched
Methyl (R)-3-hydroxy-5-oxohexanoate [(R)-11]
Scheme 6. Chiral Ru-{Amidophosphane-phosphinite} catalyst pre-
cursors used in this study
As described previously, we were able to isolate the pri-
The most striking feature among the results reported in mary intermediate of the hydrogenation process (11) in
Table 2 is the propensity of Ru-[(S)-Ph,Ph-oxoProNOP] good yields and fair enantiomeric excesses. Because the
catalysts to promote the formation of syn-rich mixtures of stereochemical outcome of the hydrogenation reaction de-
3,5-dihydroxyesters 2. This tendency is observed with the pends, not only, on the chiral catalyst but also on the chiral
bis(methylallyl) complex (S)-A and all of its dicarboxylato centre formed in the primary stage, it was worth evaluating
derivatives, provided the reaction is carried out in dichloro- some combinations which are not favoured in one-pot
methane or in 1,2-dichloroethane. In a 1:1 (v/v) methanol/ experiments. Thus, we next investigated the hydrogenation
dichloromethane mixture, the above catalyst precursors in- of enantiomerically enriched (R)-11 under the same con-
duce the selective formation of anti-2, accompanied by no- ditions as those employed for the hydrogenation of 1 with
ticeable amounts of side products (see for representative ex- various catalysts. In particular, for chiral catalysts, it is of
amples, entries 2 and 6).[15][16][17] The best results in terms obvious interest to focus on catalysts modified by ligands
of diastereoselectivity for the syn-3,5-dihydroxyesters are having an opposite configuration to the one which leads to
obtained with (S)-A itself (entry 1), and above all with the the prevailing enantiomer of 11 used in the experiment. The
ruthenium complexes prepared from (S)-A and (R)- or (S)- usual analytical procedures were used to determine the con-
α-methoxy-α-(trifluoromethyl)phenylacetic acid [(R)- and version and the different selectivities (Scheme 7). The most
(S)-MTPA; entries 8 and 9, respectively]. The combination significant results are summarized in Table 3.
of Ru(η3-C4H7)2[(S)-Ph,Ph-oxoProNOP] with (R)-MTPA
As expected, achiral catalysts do not direct the reaction
(1:4) proved to be remarkably active and diastereoselective towards the syn diastereomer (entries 1Ϫ3), but rather to
but the enantiomeric excess was extremely poor. This anti-2 or ca. 50/50 mixtures. It is noteworthy that the afore-
Table 2. Asymmetric hydrogenation of methyl 3,5-dioxohexanoate (1) with Ru-AMPP catalyst precursors[a]
Entry
Catalyst Precursor[e]
time[b]
(h)
11
sel
2 ϩ 3
sel
5
Side Products
sel
syn:anti[c] ee syn[c]
ee anti[c] ee[d]
1[f]
2[g]
3
(S)-A
235
113
23
<1
0
>99
78
87/13
10/90
65/35
72/28
64/36
14/86
68/32
96/4
14 (R,R) 29 (S,R) 22 (R)
nd nd 5 (S)
33 (R,R) 59 (S,R) 43 (R)
40 (R,R) 70 (S,R) 51 (R)
34 (R,R) 63 (S,R) 48 (R)
11 (R,R) 12 (S,R) 10 (R)
13 (R,R) 55 (S,R) 21 (R)
-
(S)-A
5 (15), 13 (7)
(S)-A ϩ CF3CO2H
(S)-A ϩ CF3CO2H
(S)-A ϩ CF3CO2H
(S)-A ϩ CF3CO2H
(S)-A ϩ PhOCH2CO2H
(S)-A ϩ (R)-MTPA
(S)-A ϩ (S)-MTPA
(S)-A ϩ CH3SO3H
(S)-A ϩ CF3SO3H
0
>99
97
-
4[h]
5[i]
6[g]
7
138
17
0
5 (3)
6
94
-
93
0
88
12 (12)
137
18
43
0
56
5 (1)
8
9
10
11
>98
92
< 5 (R,R) nd
<2 (R)
-
63
2
92/8
5 (R,R)
28 (S,S)
nd
42 (S,R) 7 (R)
10 (S,R) 8 (R)
5 (6)
17
0
72
9
23/77
nd
5 (9), 12 (10), 13 (9)
12 (86)
118[j]
5
nd
nd
[a] Unless otherwise stated, reactions were conducted in CH2Cl2 at 60°C under 100Ϫ130 atm H2; [1]/[Ru] ϭ 200:1, [Ru] ϭ ca. 1 mmol.lϪ1
.
[e]
Conversion and selectivities: see Table 1. Ϫ [b],[c],[d] See Table 1. Ϫ Catalyst pre-prepared by addition of 4 mol equiv. of acid to catalyst
precursor (S)-A. Ϫ [f] P ϭ 155 atm. Ϫ [g] Solvent: CH2Cl2/MeOH (1:1). Ϫ [h] T ϭ 20°C. [i] Solvent: (CH2Cl)2. Ϫ [j] Conversion of 1 ϭ 88%.
3424
Eur. J. Org. Chem. 1999, 3421Ϫ3427

One-Pot and Sequential Asymmetric Hydrogenation
FULL PAPER
experiments, the fair enantiomeric excesses obtained in the
hydrogenation of 11 with Ru(TFA)2L2 precursors and the
moderate diastereomeric excesses obtained with RuBr2L2
complexes are most likely due, at least in part, to the limited
enantiomeric purity of 11.
Scheme 7. Stereoselective hydrogenation of methyl (R)-3-hydroxy-
5-oxohexanoate
Conclusion
mentioned hydrogenolysis by-product 12 is formed in
higher amounts than during the corresponding one-pot
experiment (compare Table 1, entry 1 and Table 3, entry
1); this observation confirms that the formation of 12 is
subsequent to that of 11, which could be prevented by low-
ering the concentration of 11 in the reaction medium. The
hydrogenations of a 66% ee sample of (R)-11 using trifluo-
roacetato-Ru complexes bearing (R)-atropisomeric ligands
led to the formation of (3R,5R)-syn-2 in high dia-
stereomeric excesses (up to 80%) and fair enatiomeric ex-
cesses (entries 4Ϫ6). Similar stereoselectivities were ob-
tained with the trifluoroacetato derivative of the Ru-AMPP
complex (S)-A, although this proved to be significantly bet-
ter in terms of activity than the previous catalyst precursors
(entry 10). On the other side, dibromo-Ru complexes bear-
ing (R)-atropisomeric ligands, although somewhat less
chemoselective than their trifluoroacetato analogues, led to
(3R,5R)-syn-2 in lower diastereomeric excesses but very
high enantiomeric excesses (96Ϫ98%) (entries 7Ϫ9). As it
proved to be also for the one-pot experiments, RuBr2[(R)-
MeO-Biphep] is the most efficient catalyst precursor for this
sequential hydrogenation, leading to (3R,5R)-syn-2 in ca.
70% yield and 97% ee from a 59% ee sample of (R)-11.[18]
As previously mentioned, comparable asymmetric hydro-
genations have been carried out by Saburi et al. on the other
regioisomer; i.e., tert-butyl 5-hydroxy-3-oxohexanoate
(8).[1][7] The reaction of enantiopure (R)-8 with a Ru-(R)-
Binap catalyst gave dominantly the corresponding (3R,5R)-
syn-dihydroxyester, but in only 20% diastereomeric excess.
Noyori also reported that the hydrogenation of enantiopure
(R)-2-hydroxy-4-pentanone (the intermediate in the hydro-
genation of β-diketones) with a Ru-(S)-Binap catalyst led
The most striking result of this systematic study is that
some ruthenium[19] complexes associated with an amido-
phosphane-phosphinite ligand induce the highly dia-
stereoselective formation of syn-dihydroxyester 2. Although
the catalytic activity and above all the enantioselectivity are
poor, this is the first example of a one-pot catalytic syn-
thesis of a functionalized syn-1,3-diol; recent results in our
group show that this system is also effective for the dia-
stereoselective conversion of β-diketones into simple syn-
1,3-diols. Also, the asymmetric hydrogenation of an
enantiomerically enriched β-hydroxy-δ-ketoester offers an
attractive entry to the syn-dihydroxyesters; this alternative
procedure leads to either high enantiomeric excesses or dia-
stereomeric excesses, the combination of both being in part
hampered by the enantiomeric purity of the starting mate-
rial which is initially produced via selective hydrogenation
of the β,δ-diketoester. Further investigations in this direc-
tion are currently under way
Experimental Section
General: All the catalytic reactions were performed under anaerobic
conditions using standard Schlenk techniques. Hydrogenation sol-
vents were distilled from magnesium alkoxide (methanol, 2-propa-
nol), CaH2 (CH2Cl2, ClCH2CH2Cl) or sodium benzophenone ketyl
(toluene), and degassed before use. NMR spectra were recorded on
a AC-300 Bruker spectrometer at ambient temperature; chemical
shifts are reported in ppm downfield from TMS and coupling con-
stants are reported in Hz. Mass spectra were performed with a
Finnigan Mat at 70 eV. GLC analyses were performed on a Chrom-
pack apparatus equipped with a flame ionization detector and a
to isomeric anti and syn diols in 15:85 ratio.[6b] In our BPX5 (25 m ϫ 0.32 mm, SGE) or a chiral Cydex-B (25 m ϫ
Table 3. Asymmetric hydrogenation of optically active methyl 3-hydroxy-5-oxohexanoate (11) with Ru-Binap type catalyst precursors[a]
Entry Catalyst Precursor
11
time[b] 11
2 ϩ 3
sel
5
Side Products
sel
ee[e]
(h)
conv
syn:anti[c] ee syn[c]
ee anti[c] ee[d]
1
RuCl2(PPh3)3
Rh/C
66 (R)
66 (R)
66 (R)
138
40
>99
54
80
99
99
98
98
20/80
50/50
52/48
90/10
81/19
42 (R,R) 61 (R,S) 37 (S) 5 (2), 12 (15), 13 (3)
2[f]
3[f]
4
55 (R,R) 67 (R,S) <5
60 (R,R) 67 (R,S) <5
Ϫ
Ru/C
23
96
5 (1)
Ru(TFA)2((R)-Binap) 66 (R)
21
21
>99
100
75 (R,R) 70 (R,S) 70 (R) 12 (2)
67 (R,R) 63 (R,S) 45 (R) 12 (2)
5
Ru(TFA)2((R)-Tol-
Binap)
67 (R)
67 (R)
59 (R)
6
Ru(TFA)2((R)-MeO-
Biphep)
22
96
99
86/14
72 (R,R) 36 (R,S) 61 (R) 5 (1)
7
8
9
RuBr2((R)-Binap)
25
23
25
89
>99
100
90
93
95
55/45
58/42
72/28
96 (R,R) 19 (R,S) 45 (R) 5 (1), 12 (7), 13 (2)
98 (R,R) 23 (R,S) 53 (R) 5 (1), 12 (5), 13 (1)
97 (R,R) 32 (S,R) 79 (R) 5 (1), 12 (2), 13 (2)
RuBr2((R)-Tol-Binap) 67 (R)
RuBr2((R)-MeO-
Biphep)
59 (R)
59 (R)
10
(S)-A ϩ CF3CO2H
3
100
96
80/20
74 (R,R) 4 (S,R)
64 (R) 5 (4)
[a]
Unless otherwise stated, reactions were conducted in CH2Cl2 at 60°C under 100 atm H2; [11]/[Ru] ϭ 50:1, [Ru] ϭ ca. 0.8 mmol.lϪ1
.
[b]
[c],[d]
[e]
[f]
Ϫ
Non optimized time. Ϫ
See Table 1. Ϫ Enantiomeric purity of purified starting material. Ϫ Solvent: MeOH; T ϭ 80°C.
Eur. J. Org. Chem. 1999, 3421Ϫ3427
3425

V. Blandin, J.-F. Carpentier, A. Mortreux
FULL PAPER
0.32 mm, SGE) column. Optical rotations were measured on a Per-
Determination of Diastereo- and Enantioselectivities: For this pur-
kinϪElmer polarimeter in a 1-dm cell. IR spectra are expressed by pose, the final mixture containing dihydroxyesters 2 and lactones 3
wavenumber (cmϪ1).
was converted into the corresponding acetonides syn-9 and anti-9,
by refluxing a small portion of the oily mixture (ca. 20 mg) for 2 h
in 2,2-dimethoxypropane (5 mL) in the presence of para-tolu-
enesulfonic acid (2 mg). The authenticity of acetals 9 was con-
Materials: Methyl 3,5-dioxohexanoate (1) was prepared by the re-
ported method.[20] (R)- and (S)-Binap, (R)- and (S)-Tol-Binap, Ru-
(COD)(methylallyl)2, {RuCl2[(S)-Binap]}2·NEt3 and RuCl2(PPh3)3
were purchased from Acros and Strem. The catalyst precursors
Ru(TFA)2(L2)[9] and RuBr2(L2)[10] bearing chiral atropisomeric di-
phosphanes were prepared according to literature methods. AMPP
ligands (S)- and (R)-Ph,Ph-oxoProNOP[21] and the corresponding
1
firmed from their H- and 13C-NMR data.[1]
The resulting solution was first analysed by GLC on a BPX5 col-
umn (25 m ϫ 0.32 mm; 90°C iso, 0.2 bar N2); the retention times
for syn-9 and anti-9 are tR ϭ 11.0 min and tR ϭ 9.8 min, respec-
tively.
[22]
complexes Ru[(S)- or (R)-Ph,Ph-oxoProNOP](methylallyl)2
[(S)-A and (R)-A] were synthesized according to reported
procedures.
The same mixture of syn-9 and anti-9 was then subjected to a GLC
analysis using a chiral column Cydex-B (25 m ϫ 0.32 mm; 90°C
iso, 0.75 bar H2); the retention times were as follows: (3R,5S), tR ϭ
15.2 min; (3S,5R), tR ϭ 16.5 min; (3R,5R), tR ϭ 17.0 min; (3S,5S),
tR ϭ 17.5 min.[8] TsOH was converted in 2,2-dimethoxypropane to
a product eluted at tR ϭ 16.1 min.
Catalyst precursors of the type Ru[(S)-Ph,Ph-oxoProNOP](X)2
(X ϭ OCOR or OSO2R) were generated ex situ using the same
procedure as described by Heiser.[9] A solution of (S)-A in the reac-
tion solvent (CH2Cl2 or CH2Cl2/CH3OH) was cooled to Ϫ78°C
and the appropriate acid (4.5 mol equiv.) was gently added via syr-
inge. The reaction mixture was stirred for 10 minutes at low tem-
perature, and then allowed to warm to room temperature. After
additional stirring for 0.5Ϫ1 h, the solution was directly used for a
catalytic experiment.
In parallel, the final mixture containing dihydroxyesters 2 and lac-
tone 3 was converted into lactone 5 by refluxing a small portion of
the oily mixture (ca. 20 mg) for 2 h in toluene (5 mL) in the pres-
ence of para-toluenesulfonic acid (2 mg). The resulting solution was
directly analyzed by GLC on a chiral column Cydex-B (25 m ϫ
0.32 mm; 90°C iso, 0.75 bar H2); the retention times for (S)-5 and
(R)-5 were tR ϭ 10.3 min and tR ϭ 10.8 min, respectively. The val-
idity of this analysis was assessed with racemic samples, and sca-
lemic samples of 5 prepared with catalysts of opposite configura-
tion. The enantiomeric excesses determined by polarimetry on
purified 5 are consistent with the chiral GLC analysis, but the latter
proved to be much easier to perform and more reliable.
Asymmetric Hydrogenations of Dioxoester 1 for Screening Purposes:
In a typical experiment (Table 1, entry 6), a solution of 1 (0.37 g,
2.4 mmol) in CH2Cl2 (10 mL) was degassed by two freezeϪthaw
cycles and then added under nitrogen to a solution of RuBr2[(S)-
MeO-Biphep] (10 mg, 0.012 mmol) in CH2Cl2 (10 mL). The re-
sulting solution was transferred to a 100 mL-stainless steel auto-
clave equipped with a magnetic stirrer bar. Hydrogen (99%, Air
Liquide) was introduced (100 bar), the reactor was heated to 60°C
by circulating thermostated water in the double wall, and stirring
was started. The reaction was monitored by quantitative GLC
analysis (BPX5 column) of some aliquots. After completion, the
autoclave was cooled to room temperature, hydrogen was vented
and the solution was concentrated under vacuum to give an oily
residue.
The enantiomeric excess of the primary hydrogenation product (11)
was determined on the trifluoroacetate derivative: trifluoroacetic
anhydride (3 drops) was added to an aliquot sample of the crude
reaction mixture, the solution was stirred at room temperature for
ca. 10 min, and then subjected to a GLC analysis using a chiral
column Cydex-B (25 m ϫ 0.32 mm; 95°C iso, 0.75 bar H2); the
retention times were as follows: (S)-TFA-11, tR ϭ 10.5 min; (R)-
TFA-11, tR ϭ 10.7 min.
Determination of Chemoselectivities: Chemoselectivities were deter-
mined by direct GLC analysis of the crude reaction mixture using
a BPX5 column (25 m ϫ 0.32 mm; 90°C iso, 0.2 bar N2). The
Synthesis of Methyl (S)-3-Hydroxy-5-oxohexanoate (11) via Selec-
tive Hydrogenation of 1: In a typical experiment, [Ru(COD)(methy-
lallyl)2] (41 mg, 0.13 mmol) and (R)-Tol-Binap (87 mg, 0.13 mmol)
were dissolved in acetone (15 mL). To this solution was added a
methanol solution of HBr (0.3 , 1.0 mL, 2.2 equiv). The resulting
solution was stirred at room temperature for 30 min and volatiles
were removed under vacuum to give a brown residue. A solution
of 1 (4.05 g, 25.6 mmol) in CH2Cl2 (50 mL) was degassed by two
freeze-thaw cycles and added under nitrogen to the brown residue
{RuBr2[(R)-Tol-Binap]}. The resulting solution was then trans-
ferred to a 100 mL-stainless steel autoclave equipped with a mag-
netic stirrer bar. Hydrogen was introduced (100 bar), the reactor
was heated to 40°C by circulating thermostated water in the double
wall, and stirring was started. The reaction was monitored by quan-
titative GLC analysis (BPX5 column) of some aliquots. After 5.5 h,
the autoclave was cooled to room temperature and hydrogen was
vented. GLC analysis of the crude solution showed that 96% of 1
was converted and that the selectivity of 11 was 84% [75% ee (S)].
The crude solution was concentrated under vacuum to give an oily
residue which was separated by chromatography on silica using
ethyl acetate/heptane (2:3) as eluent. The purest fractions were col-
lected and concentrated under vacuum to give analytically pure 11
retention times are as follows: 1 (diketoester and enol form), tR
ϭ
5.0 and 7.1 min; {syn-2 ϩ anti-2}, tR ϭ 11.1 min (immediate dehy-
dration occurred in the chromatograph injector to form
CH3CH(OH)CH2CHϭCHCO2CH3, as revealed by GC/MS); syn-
3, tR ϭ 19.2 min; anti-3, tR ϭ 20.5 min; 5, tR ϭ 4.1 min; 11, tR
8.1 min; 12, tR ϭ 4.4 min; 13, tR ϭ 4.7 min.
ϭ
The identity of all of the compounds issued from the hydrogenation
of 1 was formally established by isolation of the latter by column
1
chromatography and comparison of their H- and 13C-NMR and
GC/MS data with those reported in the literature.[1]
Methyl 5-Oxohexanoate (12):[23] 1H NMR (CDCl3): δ ϭ 3.66 (s, 3
3
3
H, CH3OCO), 2.50 (t, J ϭ 7.1, 2 H, CH2CO), 2.34 (t, J ϭ 7.1, 2
3
H, CH2COO), 2.13 (s, 3 H, CH3CO), 1.88 (tt, J ϭ 7.1 and 7.1, 2
H, CH2). Ϫ 13C{1H} NMR (CDCl3): δ ϭ 207.9 (CO), 173.2(COO),
51.1 (COOCH3), 42.0 (CH2CO), 32.6 (CH2COO), 29.5 (CH3CO),
18.4 (CH2). Ϫ MS (CI, NH3): 145 [MHϩ], 162 [M ϩ NH4ϩ]. Ϫ
MS (EI) (m/z,%): 144 [Mϩ] (2), 112 (20), 85 (22), 74 (34), 43
(CH3CO, 100).
6-Methyl-tetrahydropyran-2-one (13):[24] 13C{1H} NMR (CDCl3):
δ ϭ 172.0 (COO), 77.0 (CH(O)), 29.5 (CH2), 29.2 (CH2), 21.6 [1.8 g, 45% yield, 75% ee (S)]. Anal. C7H12O4: calcd. for C, 52.49;
1
(CH2), 18.5 (CH3). Ϫ MS (CI, NH3): 115 [MHϩ], 132 [M ϩ H, 7.55; found C, 52.9; H, 7.2. Ϫ H NMR (CDCl3): δ ϭ 4.43 (m,
NH4ϩ].
1 H, CHOH), 3.66 (s, 3 H, CH3OCO), 3.45 (d, 3J ϭ 3.8, 1 H, OH),
3426
Eur. J. Org. Chem. 1999, 3421Ϫ3427

One-Pot and Sequential Asymmetric Hydrogenation
FULL PAPER
[9]
2.64 (d, 3J ϭ 7.3, 1 H, CHHCO), 2.63 (d, 3J ϭ 5.1, 1 H, CHHCO),
2.48 (d, 3J ϭ 6.8, 1 H, CHHCOO), 2.47 (d, 3J ϭ 6.1, 1 H,
CHHCOO), 2.15 (s, 3 H, CH3CO). Ϫ 13C{1H} NMR (CDCl3):
δ ϭ 208.1 (CO), 172.0 (COO), 64.1 (CHOH), 51.6 (COOCH3), 49.0
(CH2CO), 40.4 (CH2COO), 30.5 (CH3CO). Ϫ MS (CI, NH3) m/z
(%): 160 [Mϩ] (100), 161 [MHϩ] (70), 178 [M ϩ NH4ϩ] (65). Ϫ
B. Heiser, E. A. Broger, Y. Crameri, Tetrahedron: Asymmetry
1991, 2, 51Ϫ62.
[10]
ˆ
J.-P. Genet, V. Ratovelomanana-Vidal, M. C. Cano De And-
rade, X. Pfister, P. Guerreiro, J. Y. Lenoir, Tetrahedron Lett.
1995, 36, 4801Ϫ4804.
According to 1H NMR data in CDCl3, compound 1 exists
mainly (85%) as its enol tautomer CH3C(O)CHϭ
C(OH)CH2CO2Me. β-Diketones show comparable enolization
equilibrium data, and it is therefore likely that 1 interacts with
the Ru species in a similar way as β-diketones do.[1]
[11]
20
[α]D (c 1.1, CHCl3) ϭ Ϫ10.0° [75% ee (S)].
[12] [12a]
Asymmetric Hydrogenation of β-Hydroxy-δ-oxoester 11: The hydro-
genation of 11 was carried out under similar conditions to those
used for the hydrogenation of 1, using samples prepared as de-
scribed above. The selectivities were determined by the procedures
described above.
A. S. C. Chan, J. J. Pluth, J. Halpern, J. Am. Chem. Soc.
[12b]
1980, 102, 5952Ϫ5953. Ϫ
C. R. Landis, J. Halpern, J. Am.
[12c]
Chem. Soc. 1987, 109, 1746Ϫ1754. Ϫ
J. M. Brown, P. A.
Chaloner, J. Chem. Soc., Chem. Commun. 1980, 344Ϫ346. Ϫ
[12d]
J. M. Brown, P. A. Chaloner, J. Am. Chem. Soc. 1980,
102, 3040Ϫ3048.
[13] [13a]
F. Hapiot, F. Agbossou, A. Mortreux, Tetrahedron: Asym-
metry 1997, 8, 2881Ϫ2884. Ϫ [13b] For a recent review on AMPP
coordination chemistry and their use in homogeneous catalysis
see: F. Agbossou, J.-F. Carpentier, F. Hapiot, I. Suisse, A. Mor-
treux, Coord. Chem. Rev. 1998, 180, 1615Ϫ1645.
Acknowledgments
[14]
[15]
Use of triflateϪruthenium complexes in asymmetric hydrogen-
We are grateful to Dr R. Schmid from Hoffman La Roche (Basel)
for valuable discussions and for the gift of samples of (R)- and (S)-
MeO-Biphep and (S)-Cy4-MeO-Biphep. We are indebted to Dr. F.
Hapiot for providing some samples of Ru[(S)-Ph,Ph-oxoProNOP]-
[14a]
ation:
N. W. Alcock, J. M. Brown, M. Rose, A. Wienand,
[14b]
Tetrahedron: Asymmetry 1991, 2, 47Ϫ50. Ϫ
J. M. Brown,
F. I. Knight, M. Rose, A. Wienand, Recl. Trav. Chim. Pays-Bas
1995, 114, 242Ϫ247.
A similar, although much less marked, solvent effect (CH2Cl2
vs CH3OH) on the diastereoselectivity has been reported for
the stereoselective hydrogenation via dynamic resolution of 2-
acetamido-3-oxoesters with RuϪBinap catalysts.[17] This effect
was ascribed to the existence in CH2Cl2 of a stabilizing intra-
molecular hydrogen bond between CONH an the ester OR,
which does not exist in CH3OH. In the present case, a compar-
able hydrogen bonding between the formed OH and the ester
OCH3 can be assumed.
`
(CF3CO2)2. This work was supported by the Ministere de lЈEnseig-
´
nement Superieur et de la Recherche and the CNRS.
[1]
L. Shao, H. Kawano, M. Saburi, Y. Uchida, Tetrahedron 1993,
49, 1997Ϫ2010.
[2]
For an easier reading of this manuscript, we used the same sys-
tem for numbering compounds as adopted by Saburi et al.[1]
[3]
[16]
A. Endo, M. Kuroda, Y. Tsujita, J. Antibiot. 1976, 29,
Experiments conducted with the (S)-A/TFA catalytic system in
other solvents showed that methyl acetate induces the forma-
tion of a ca. 50/50 mixture of syn-3 and anti-3, while the reac-
tion proceeds very slowly in toluene and chloroform and is not
chemoselective at all in pure methanol.
1346Ϫ1348.
[4]
A. O. Alberts, J. Chen, G. Kuron, V. Hunt, C. Hoffman, J.
Rothrock, M. Lopez, H. Joshua, E. Harris, A. Patchett, R.
Monaghan, S. Currie, E. Stapley, G. Albers-Schonberg, O. Hen-
sens, J. Hirshfild, K. Hoogsteen, J. Liesch, J. Springer, Proc.
Natl. Acad. Sci. USA 1980, 77, 3957Ϫ3961.
[17] [17a]
R. Noyori, T. Ikeda, T. Ohkuma, M. Widhalm, M. Kita-
mura, H. Takaya, S. Akutagawa, N. Sayo, T. Saito, T. Taketomi,
H. Kumobayashi, J. Am. Chem. Soc. 1989, 111, 9134Ϫ9135. Ϫ
[17b] K. Mashima, Y. Matsumura, K. Kusano, H. Kumobayashi,
N. Sayo, Y. Hori, T. Ishizaki, S. Akutagawa, H. Takaya, J.
Chem. Soc., Chem. Commun. 1991, 609Ϫ610.
[5]
[6]
[5a]
Leading references for β-diketone hydrogenation:
H. Ka-
wano, Y. Ishii, M. Saburi, Y. Uchida, J. Chem. Soc., Chem.
[5b]
Commun. 1988, 87Ϫ88. Ϫ
A. Mezetti, A. Tschumper, G.
Consiglio, J. Chem. Soc., Dalton Trans. 1995, 49Ϫ56.
Leading references for β-ketoester hydrogenation:
[6a]
[18]
[19]
R.
This reaction was also performed with starting materials of op-
posite absolute configuration; i.e., RuBr2[(S)-MeO-Biphep] and
(S)-11, and led to (S,S)-9 in almost identical diastereo- and en-
antioselectivities.
Noyori, T. Ohkuma, M. Kitamura, H. Takaya, N. Sayo, H. Ku-
mobayashi, S. Akutagawa, J. Am. Chem. Soc. 1987, 109,
5856Ϫ5858. Ϫ [6b] M. Kitamura, T. Ohkuma, S. Inoue, N. Sayo,
H. Kumobayashi, S. Akutagawa, T. Ohta, H. Takaya, R.
Attempts to perform the hydrogenation of 1 with rhodium-
based catalysts led to poor results. For instance, the in situ com-
bination [Rh(COD)Cl]2/(S)-Cy4-MeO-Biphep (1:2.2; [1]/Rh ϭ
200) in toluene at 60°C under 50 atm H2 gave after 47 h a 55/
45 mixture of syn-2/anti-2 (100% conversion, 99% combined sel-
ectivity into 2 and 3), with 4% ee (3R,5R) and 54% ee (3S,5R),
respectively; lactone 5 was then recovered in 25% ee (R).
Noyori, J. Am. Chem. Soc. 1988, 110, 629Ϫ631.
Ϫ
[6c]
R. Noyori, H. Takaya, Acc. Chem. Res. 1990, 23, 345Ϫ350.
[6d]
Ϫ
M. J. Burk, T. G. P. Harper, C. S. Kalberg, J. Am. Chem.
Soc. 1995, 117, 4423Ϫ4424. Ϫ [6e] J.-P. Genet, C. Pinel, V. Rato-
ˆ
velomanana-Vidal, S. Mallart, X. Pfister, L. Bischoff, M. C.
Cano De Andrade, S. Darsers, C. Galopin, J. A. Laffitte, Tetra-
[6f]
[20] [20a]
[20b]
hedron: Asymmetry 1994, 5, 675Ϫ690. Ϫ
K. Mashima, K.-
J. G. Batelaan, Synth. Commun. 1976, 6, 81Ϫ83. Ϫ
G.
´
H. Kusano, N. Sato, Y. Matsumura, K. Nozaki, H. Kumobaya-
shi, N. Sayo, Y. Hori, T. Ishizaki, S. Akutagawa, H. Takaya, J.
Org. Chem. 1994, 59, 3064Ϫ3076.
Solladie, F. Colobert, D. Denni, Tetrahedron: Asymmetry 1998,
9, 3081Ϫ3094.
[21]
A. Roucoux, L. Thieffry, J.-F. Carpentier, M. Devocelle, C. Mel-
iet, F. Agbossou, A. Mortreux, Organometallics 1996, 15,
2440Ϫ2449.
[7]
[8]
These compounds are readily prepared by reaction of an enanti-
opure β-hydroxyester with an alkyl acetate enolate; P. F. De-
schenaux, T. Kallimopoulos, H. Stoeckli-Evans, A. Jacot-Guil-
larmod, Helv. Chim. Acta 1989, 72, 731Ϫ737.
[22]
[23]
[24]
´
F. Hapiot, F. Agbossou, C. Meliet, A. Mortreux, G. M. Rosair,
A. J. Welch, New J. Chem. 1997, 21, 1161Ϫ1172.
B. W. Babcock, D. R. Dimmel, D. P. Graves, R. D. McKelvey,
J. Org. Chem. 1981, 46, 736Ϫ742.
The validity of this analysis was assessed using: (i) a racemic
sample of 2 (syn:anti ϭ 21/79) prepared by hydrogenation of 1
using RuCl2(PPh3)3 as catalyst precursor (Table 1, entry 1), and
(ii) scalemic samples of 2 prepared with catalysts bearing li-
D. B. Gerth, B. Giese, J. Org. Chem. 1986, 19, 3726Ϫ3729.
Received May 3, 1999
gands of opposite configuration; see note[18]
.
[O99252]
Eur. J. Org. Chem. 1999, 3421Ϫ3427
3427