Tuning diastereoselectivity with the solvent: The asymmetric hydrogenation of simple and functionalized 1,3-diketones with ruthenium(amidophosphine-phosphinite) catalysts

Article abstract of DOI:10.1039/b000962h

Spectacular solvent effects in the asymmetric hydrogenation of methyl 3,5-dioxohexanoate (3) and 2,4-pentanedione (5) have been observed using Ru[(S)-Ph,Ph-oxoProNOP]X₂ complexes (X = η³-C₄H₇, IIa; CF₃CO₂, IIb; (R)-MTPA, IIc) as catalyst precursors. β-Diketones 3 and 5 are respectively reduced to the corresponding β-diols 4 and 6. In both cases, an almost complete reversal in the diastereoselectivity of the reaction is observed when changing the solvent from CH₂Cl₂ (syn-4 in up to 92% de; meso-6 in up to 84% de) to a 1:1 CH₂Cl₂-CH₃OH mixture (anti-4 and anti-6 in up to 84% de). The extent of this solvent effect is much less marked with Ru- atropisomeric diphosphine catalysts.

Full text of DOI:10.1039/b000962h

View Article Online / Journal Homepage / Table of Contents for this issue
Tuning diastereoselectivity with the solvent: the asymmetric
hydrogenation of simple and functionalized 1,3-diketones with
ruthenium(amidophosphine–phosphinite) catalysts
Veronique Blandin, Jean-FrancÓ ois Carpentier* and Andre Mortreux
 
L aboratoire de Catalyse de L ille, (CNRS UPRESA 8010), Ecole Nationale Supe

rieure de Chimie
de L ille, B.P. 108, 59652, V illeneuve dÏAscq cedex, France. Fax: ]33 320 436 585;
E-mail: carpentier=ensc-lille.fr
Received (in Strasbourg, France) 3rd February 2000, Accepted 10th March 2000
Published on the Web 13th April 2000
Spectacular solvent e†ects in the asymmetric hydrogenation of methyl 3,5-dioxohexanoate (3) and 2,4-pentanedione
(
5) have been observed using Ru[(S)-Ph,Ph-oxoProNOP]X complexes (X \ g3-C H , IIa; CF CO , IIb; (R)-
4 7
MTPA, IIc) as catalyst precursors. b-Diketones 3 and 5 are respectively reduced to the corresponding b-diols 4 and
2
3
2
6
. In both cases, an almost complete reversal in the diastereoselectivity of the reaction is observed when changing
the solvent from CH Cl (syn-4 in up to 92% de; meso-6 in up to 84% de) to a 1 : 1 CH Cl ÈCH OH mixture (anti-
2
2
2
2
3
4
and anti-6 in up to 84% de). The extent of this solvent e†ect is much less marked with Ru-atropisomeric
diphosphine catalysts.
enantiomeric excess could not be determined accurately. In a
recent thorough reinvestigation of this reaction in which we
set up efficient analytical tools,6 we have found that carrying
out the reaction with this catalyst precursor in CH Cl
1
. Introduction
Catalytic asymmetric hydrogenation of 1,3-dicarbonyl systems
mediated by Ru-diphosphine complexes is a highly efficient
tool in organic synthesis for the production of chiral function-
alized alcohols or diols, as useful synthons.1 Intensive and
successful developments have been made in this Ðeld in the
last decade, essentially by investigating the reactivity of new
functionalized ketones2 and by optimizing the catalyst per-
formance through tuning of the chiral diphosphine and the
ruthenium precursor.1c,d Another parameter that is known to
inÑuence the outcome of some of these processes is the nature
of the solvent. A classical example is the dynamic kinetic
resolution of b-ketoesters with epimerizable substituents at
the a position.1,3 In this process, diastereoselectivities are
always more efficient when the reductions are performed in
CH Cl compared to a protic solvent such as methanol. High
syn diastereomer formation is attributed to a transition state
that is stabilized by hydrogen bonding between the amide
hydrogen and ester moiety. Protic solvents would no longer
stabilize the hydrogen bonding in the transition state, which
results in dismal diastereoselectivities.1,3 Investigations on
solvent e†ects in the asymmetric hydrogenation of functional-
ized ketones are indeed very rare and have been almost exclu-
sively carried out with catalysts based on atropisomeric
bisphosphines. In this note, we wish to report on related, even
more marked solvent e†ects in the asymmetric hydrogenation
2
2
instead of CH OH allows a slight improvement of both dia-
3
stereoselectivity (from 52% to 72% de) and enantioselectivity
(
from 87 to 92% ee) for anti-(3R,5S)-4 (Table 1). Related
catalyst precursors such as RuBr [(S)-TolBINAP] and
2
RuBr [(S)-MeO-BIPHEP] a†ord slightly improved per-
2
formance (with respect to I) in terms of stereoselectivity [up to
7
8% de and 95% ee for anti-(3R,5S)-4 in CH Cl ], but exhibit
2
2
basically the same solvent dependence. In fact, the extent of
this solvent e†ect on the stereoselective outcome of the asym-
metric hydrogenation of 3 with Ru-atropisomeric diphosphine
systems remains rather limited (ca. [20 points in de from
CH Cl to CH OH), a trend that compares well with the
aforementioned example of dynamic kinetic resolution.
More interesting are the results obtained for this reaction
with some Ru(amidophosphine-phosphinite) complexes
2
2
3
2
2
(
(
hereafter denoted as Ru-AMPP) developed in our group
Scheme 2).7 Indeed, the catalytic species produced from these
precursors proved to have a much more versatile behavior
with respect to the nature of the solvent, allowing in certain
cases an almost complete reversal in the diastereoselectivity of
of
1,3-dicarbonyl
systems
in
the
presence
of
ruthenium(amidophosphine-phosphinite) catalysts.
2. Results and discussion
The Ðrst series of examples is concerned with the asymmetric
hydrogenation of methyl 3,5-dioxohexanoate (3). This model
reaction is aimed at evaluating a simple one-pot access to
syn-3,5-dihydroxyesters (i.e., syn-4; Scheme 1), which are key
building blocks for the synthesis of important inhibitors of
HMG-coenzyme A reductase.4 Saburi et al.5 have Ðrst investi-
gated this reaction with MRuCl [(S)-BINAP]N Æ NEt (I) as
2
2
3
catalyst precursor in methanol and shown that, despite excel-
lent chemoselectivity for the desired products,¤ it gives pre-
dominantly anti-diol 4 in ca. 60% de; the corresponding
Scheme 1 Asymmetric hydrogenation of methyl 3,5-dioxohexanoate
3 (for clarity only half of the stereoisomers are represented)
DOI: 10.1039/b000962h
New J. Chem., 2000, 24, 309È312
309
This journal is ( The Royal Society of Chemistry and the Centre National de la Recherche ScientiÐque 2000

View Article Online
Table 1 Hydrogenation of methyl 3,5-dioxohexanoate (3)
Catalyst
Solvent
CH Cl
t/ha
Yield 4b
syn : anti 4b
ee syn-4b
ee anti-4b
I
91c
23
238
166
113
24
93
18
169
96
96
49d
74
14 : 86
24 : 76
61 : 39
42 : 58
10 : 90
68 : 32
14 : 86
96 : 4
22 (S, S)
87 (S, S)
14 (R, R)
33 (R, R)
nd
34 (R, R)
11 (R, R)
5 (R, R)
12 (R, R)
92 (R, S)
87 (R, S)
29 (S, R)
55 (S, R)
nd
60 (S, R)
12 (S, R)
nd
2
2
MeOH
IIa
CH Cl
2
2
CH Cl ÈMeOH (9 : 1)
2
2
CH Cl ÈMeOH (1 : 1)
78
2
CH Cl
2
2
IIb
IIc
[99
2
CH Cl ÈMeOH (1 : 1)
88
2
CH Cl
2
2
[98
72
2
CH Cl ÈMeOH (1 : 1)
8 : 92
5 (S, R)
2
2
a Non-optimized time for total conversion of 3. b Yield (mol %), syn : anti ratio and respective enantiomeric excesses of 4 as determined by GLC
analysis of crude reaction mixtures and corresponding acetonides. Letters in parentheses refer to the absolute conÐguration of the prevailing
enantiomer at C-3 and C-5, respectively. c S/C \ 500. d The monohydrogenation product at C-3 was formed in 51% yield and 8% ee (R).
the reaction. The readily attainable precursor MRu[(S)-Ph,Ph-
oxoProNOP](g3-C H ) N (IIa) and its dicarboxylato deriv-
(b) carbonyl group; this was clearly evidenced by GLC
analysis of some aliquot samples, which revealed that ees of
methyl 3-hydroxy-5-oxohexanoate (i.e., the sole intermediary
monohydrogenation product formed) typically range between
5 and 15%. The ine†ective enantioface di†erentiation of chiral
Ru-AMPP species is assumed to arise from a competitive liga-
tion of the functionalities of 3 onto the Ru center. The two
chelation modes, b-ketoester/b-diketone (also taking into
account the very likely involvement of b-ketoenol tautomers),
may have in fact opposite directions for the Ðnal stereoselecti-
vity at C-3.1c,5,6 In this regard, it is noteworthy that
Ru-AMPP systems such as IIa hydrogenate simple b-
ketoesters in 75È85% ee.9
4
7 2
atives MRu[(S)-Ph,Ph-oxoProNOP](O CCF ) N (IIb) and
2
3 2
MRu[(S)-Ph,Ph-oxoProNOP][(R)-MTPA] N (IIc) [(R)-MTPA)
2
\
(R)-a-methoxy-a(triÑuoromethyl)phenylacetate], both pre-
pared according to the procedure of Heiser et al.,8 are repre-
sentative examples of this unusual behavior. In pure CH Cl ,
2
2
all of these complexes induce the formation of syn-rich mix-
tures of diol 4 (Table 1); the reaction that proceeds sluggishly
with a poor de in the presence of IIa is signiÐcantly acceler-
ated with IIb and IIc with the latter complex leading to a
much better diastereoselectivity. In
a
1 : 1 (v/v)
CH Cl ÈCH OH mixture, the same catalyst precursors lead
2
2
3
this time to the diastereoselective formation of anti-4 in
2È84% de. Although it must be pointed out that the presence
Considering the above arguments, we investigated the
7
asymmetric hydrogenation of
catalyzed hydrogenations of symmetrical 1,3-diketones with
various ligands such as BINAP, MeO-BIPHEP,
a simple b-diketone. Ru-
of CH OH in the reaction medium also induces a decrease of
3
the reaction rate and some side reactions at the expense of
chemoselectivity (i.e., dehydration-hydrogenation, acetal for-
mation, etc., which hampers the study of these reactions in
pure methanol with Ru-AMPP systems, in contrast to Ru-
atropisomeric diphosphine systems),” the most striking feature
is undoubtedly the reversal in the diastereocontrol of the reac-
tion. The extent of this solvent e†ect is particularly important
in the case of complex IIc, for which the diastereomeric excess
switches from 92% for syn-4 in CH Cl to 84% for anti-4 in a
SKEWPHOS and Me-DuPHOS are well-known to a†ord the
anti diols in very high diastereo- and enantiomeric
excesses.1,10 Representative results obtained in the case of 2,4-
pentanedione (5) with Ru(Ph,Ph-oxoProNOP) systems IIaÈc
are summarized in Table 2. The latter reveal rather similar
trends as those found for the asymmetric hydrogenation of 3.
In fact, in pure CH Cl , the three Ru systems induce the for-
2
2
mation of meso-rich mixtures of diol 5; the bis(methylallyl)
and di(triÑuoroactetato) precursors IIa and IIb a†ord only
poor des while meso-5 is formed in up to 84% de in the pres-
ence of the MTPA complex IIc. Interestingly, ees as high as
93% can be reached in this solvent using catalyst precursor
IIb, which also proved the most efficient in terms of reaction
rate. In a 1 : 1 CH Cl ÈCH OH mixture, systems IIaÈc lead
2
2
1
: 1 CH Cl ÈCH OH mixture. As shown in the case of IIa
2
2
3
this e†ect can be modulated with the methanol content.
Despite high and versatile diastereoselectivities, Ru-AMPP
systems a†ord poor enantiomeric excesses. The highest ees so
far obtained for syn-4 and anti-4 reach only a modest 40%
and 60%, respectively, considering also that these values could
not be combined with high des. This general trend for Ru(Ph,
Ph-oxoProNOP) catalysts stems from their incapability to
enantioselectively promote the Ðrst hydrogenation at the C-3
2
2
3
this time to the diastereoselective formation of anti-5 in
74È84% de, but with poor ees. This low Ðnal enantio-
selectivity is ascribed again to the inefficiency of the Ðrst
hydrogenation step that proceeds in ca. 20% ee (see Table 2).
Thus, with respect to the solvent e†ect, this second series of
experiments reveal the same atypical behavior of Ru-AMPP
catalysts.
In conclusion, we have shown that some Ru(Ph,Ph-oxoP-
roNOP) complexes, such as the remarkable MRu[(S)-Ph,Ph-
oxoProNOP][(R)-MTPA] N (IIc), can induce the diastereo-
2
selective formation of either a syn-diol or an anti-diol just by
modifying the nature of the solvent. As aforementioned,
related solvent e†ects on the diastereoselectivity of the asym-
metric hydrogenation of epimerizable b-ketoesters with Ru-
Scheme 2 Chiral Ru(amidophosphine-phosphinite) catalyst precur-
sors used in this study
310
New J. Chem., 2000, 24, 309È312

View Article Online
Table 2 Hydrogenation of 2,4-pentanedione (5)
Catalyst
IIa
Solvent
CH Cl
t/ha
Yield 6b
meso : anti 6b
ee anti-6b
172
148
24
24
49
62c
[99
[99
[99
49d
57 : 43
13 : 87
60 : 40
12 : 88
92 : 8
46 (R, R)
16 (S, S)
93 (R, R)
20 (R, R)
86 (R, R)
14 (R, R)
2
2
CH Cl ÈMeOH (1 : 1)
2
CH Cl
2
2
IIb
2
CH Cl ÈMeOH (1 : 1)
2
CH Cl
2
2
IIc
2
CH Cl ÈMeOH (1 : 1)
48
84e
8 : 92
2
2
a Non-optimized time for total conversion of 5, unless otherwise stated. b Yield (mol %), syn : anti ratio and enantiomeric excess of 6 as deter-
mined by GLC analysis. c 2-Hydroxy-4-pentanone was formed in 38% yield and 18% ee (R). d Conversion of 5 \ 66%; 2-hydroxy-4-pentanone
was formed in 17% yield and 23% ee (R). e 2-Hydroxy-4-pentanone was formed in 16% yield and 28% ee (R).
BINAP type catalysts have been discussed in terms of
stabilization of a transition state through hydrogen bonding
between the amide hydrogen and ester moiety.1,3 Such an
explanation could account also for our results in the asym-
metric hydrogenation of b,d-diketoester 3, as one can consider
a similar interaction between the b-OH group and the ester
moiety in the monohydrogenated intermediate. Nonetheless,
this is a very unlikely hypothesis considering that both asym-
metric hydrogenations of diketones 3 and 5 exhibit the same
solvent e†ects, that is a reversal in diastereoselectivity, and
that such an intramolecular interaction cannot exist in the
case of 5. An important, still open issue in our opinion is the
exact nature of catalytic species involved in ruthenium-
catalyzed hydrogenation of ketones and the inÑuence of the
solvent nature on it. We assume that the unusual features
brought by the NÈP and OÈP bonds confer to Ru-AMPP
species an even greater versatility compared to catalysts based
on traditional CÈP chiral diphosphines.
100 mL stainless steel autoclave equipped with a magnetic
stirrer bar. Hydrogen (99%, Air Liquide) was introduced (100
bar), the reactor was heated to 60 ¡C by circulating ther-
mostated 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 auto-
clave was cooled to room temperature, hydrogen was vented
and the solution was concentrated under vacuum to give an
oily residue.
Analytical procedures
Chemo-, diastereo- and enantioselectivities for the hydro-
genation products of methyl 3,5-dioxohexanoate (3) were
determined as described in ref. 6. In particular, des and ees
were determined by GLC analysis of the corresponding ace-
tonides of syn-4 and anti-4 (syn-9 and anti-9, respectively),
obtained by treatment of the Ðnal oily mixture with 2,2-dime-
thoxypropane in the presence of para-toluenesulfonic acid.
Yield, diastereo- and enantioselectivities for the hydro-
genation products of 2,4-pentanedione (5) were determined by
GLC analysis. The absolute conÐguration of the prevailing
enantiomer of anti-6 was established from the optical rotation.
3. Experimental
General techniques and materials
All the catalytic reactions were performed under anaerobic
conditions using standard Schlenk techniques. Hydrogenation
solvents were distilled from magnesium methoxide (methanol)
(CH Cl ), and degassed before use. GLC analyses
Notes and references
¤
As shown in Scheme 1, the 3,5-dihydroxyesters syn-4 and anti-4 are
or CaH2
2 2
converted in situ to the corresponding anti- and syn-hydroxylactones,
respectively; the dihydroxyesters 4 and these lactones are equivalent
synthons.
were performed on a Chrompack apparatus equipped with a
Ñame ionization detector and a BPX5 (25 m ] 0.32 mm, SGE)
or a chiral Cydex-B (25 m ] 0.32 mm, SGE) column. Methyl
” Other solvents were also investigated using IIb as catalyst precursor
for the hydrogenation of 3. 1,2-Dichloroethane gave identical results,
in terms of reactivity and selectivity, as dichloromethane. In toluene
and methyl acetate, the reaction proved to be very sluggish and no
signiÐcant diastereo- and enantioselectivity were observed. When the
reaction was carried out in pure 2-propanol, unidentiÐed products
were also formed.
3
,5-dioxohexanoate (3) was prepared by the reported
method.11 2,4-Pentanedione was purchased from Aldrich and
distilled before use. MRuCl [(S)-BINAP]N Æ NEt (I) was pur-
2
2
3
chased from Strem. Ru[(S)-Ph,Ph-oxoProNOP](methylallyl)
2
(
IIa) was synthesized according to the reported procedure.7
Catalyst precursors of the type Ru[(S)-Ph,Ph-oxoProNOP]
(
OCOR) were generated ex situ using the same procedure as
2
1
For leading references, see: (a) R. Noyori, Asymmetric Catalysis
in Organic Chemistry, Wiley, New York, 1994, ch. 2; (b) H.
Takaya, T. Ohta and R. Noyori, in Catalytic Asymmetric Synthe-
sis, ed. I. Ojima, VCH, New York, 1993, ch. 1; (c) D. J. Ager and
S. A. Laneman, T etrahedron: Asymmetry, 1997, 8, 3327; (d) V.
described by Heiser et al.:8 a solution of IIa in the reaction
solvent (CH Cl or CH Cl ÈCH OH) was cooled to [78 ¡C
2
2
2
2
3
and the appropriate acid (4.5 mol equiv.) was gently added via
syringe. The reaction mixture was stirred for 10 min at low
temperature and then allowed to warm to room temperature.
After additional stirring for 0.5È1 h, the solution was directly
used for a catalytic experiment. Ex situ generated and pre-
formed catalyst precursors led to identical results in terms of
activity and stereoselectivity.
RatovelomananaÈVidal and J.-P. Gene
ü
t, J. Organomet. Chem.,
1
998, 567, 163.
2
3
4
See, for instance: (a) V. Blandin, J.-F. Carpentier and A. Mor-
treux, Eur. J. Org. Chem., 1999, 1787; (b) S. Schulz, Chem.
Commun., 1999, 1239.
R. Noyori, T. Ikeda, T. Okhuma, M. Widhalm, M. Kitamura, H.
Takaya, S. Akutagawa, N. Sayo, T. Saito, T. Taketomi and H.
Kumobayashi, J. Am. Chem. Soc., 1989, 111, 9134.
(a) A. Endo, M. Kuroda and Y. Tsujita, J. Antibiot., 1976, 29,
1346; (b) A. O. Alberts, J. Chen, G. Kuron, V. Hunt, C. Ho†man,
J. Rothrock, M. Lopez, H. Joshua, E. Harris, A. Patchett, R.
Monaghan, S. Currie, E. Stapley, G. AlbersÈSchonberg, O.
Hensens, J. HirshÐld, K. Hoogsteen, J. Liesch and J. Springer,
Proc. Natl. Acad. Sci. USA, 1980, 77, 3957.
Asymmetric hydrogenations
A
solution of diketone (2.4 mmol) in CH Cl or a
2
2
CH Cl ÈCH OH mixture (10 mL) was degassed by two
2
2
3
freeze-thaw cycles and then added under nitrogen to a solu-
tion of the catalyst precursor (0.012 mmol Ru) in the same
solvent (10 mL). The resulting solution was transferred to a
New J. Chem., 2000, 24, 309È312
311

View Article Online
5
6
7
L. Shao, H. Kawano, M. Saburi and Y. Uchida, T etrahedron,
993, 49, 1997.
V. Blandin, J.-F. Carpentier and A. Mortreux, Eur. J. Org. Chem.,
999, 3421.
F. Hapiot, F. Agbossou, C. Me
9
F. Hapiot, F. Agbossou and A. Mortreux, T etrahedron: Asym-
metry, 1997, 8, 2881.
1
10 Leading references for b-diketone hydrogenation: (a) H. Kawano,
Y. Ishii, M. Saburi and Y. Uchida, J. Chem. Soc., Chem. Commun.,
1988, 87; (b) M. Kitamura, T. Okhuma, S. Inoue, N. Sayo, H.
Kumobayashi, S. Akutagawa, T. Ohta, H. Takaya and R.
Noyori, J. Am. Chem. Soc., 1988, 110, 629; (c) A. Mezetti, A.
Tschumper and G. Consiglio, J. Chem. Soc., Dalton T rans., 1995,
49; (d) D. Blanc, V. RatovelomananaÈVidal, A. Marinetti and
J.-P. Gene
11 (a) J. G. Batelaan, Synth. Commun., 1976, 6, 81; (b) G. Solladie
Colobert and D. Denni, T etrahedron: Asymmetry, 1998, 9, 3081.
1

liet, A. Mortreux, G. M. Rosair
and A. J. Welch, New J. Chem., 1997, 21, 1161; For a recent
review on AMPP coordination chemistry and their use in
homogeneous catalysis, see: F. Agbossou, J.-F. Carpentier,
F. Hapiot, I. Suisse and A. Mortreux, Coord. Chem. Rev., 1998,
1
80, 1615.
B. Heiser, E. A. Broger and Y. Crameri, T etrahedron: Asymmetry,
991, 2, 51.
ü
t, Synlett, 1999, 4, 480.
8

, F.
1
312
New J. Chem., 2000, 24, 309È312