Notes on the asymmetric hydrogenation of methyl acetoacetate in neoteric solvents

Article abstract of DOI:10.1007/s10562-009-0233-3

Asymmetric hydrogenation of methyl acetoacetate to methyl (R)-3-hydroxybutyrate by [(R)- RuCl(binap)( p-cymen)]Cl has been studied in methanol-ionic liquid and methanol- dense CO₂ solvent systems. The ionic pairs triethylhexylammonium and 1-methylimidazolium with bis(trifluoromethane sulfonyl) imide and hexafluorophosphates were used. The role of ionic pairs on the kinetic parameters and (enantio)selectivity has been demonstrated. Although the CO₂ expanded methanol system suffered from a reduction in both reaction rate and product selectivity, this changed in the presence of water. The high selectivity of the optimized methanol-CO ₂-water-halide system was designed as a consequence of observed additive effects.

Full text of DOI:10.1007/s10562-009-0233-3

Catal Lett (2010) 134:279–287
DOI 10.1007/s10562-009-0233-3
Notes on the Asymmetric Hydrogenation of Methyl Acetoacetate
in Neoteric Solvents
•
•
•
Tomas Floris Petr Kluson Mark J. Muldoon
Helena Pelantova
Received: 15 July 2009 / Accepted: 21 November 2009 / Published online: 11 December 2009
Ó Springer Science+Business Media, LLC 2009
Abstract Asymmetric hydrogenation of methyl acetoac-
etate to methyl (R)-3-hydroxybutyrate by [(R)- RuCl(bin-
ap)(p-cymen)]Cl has been studied in methanol-ionic liquid
and methanol- dense CO2 solvent systems. The ionic pairs
triethylhexylammonium and 1-methylimidazolium with
bis(trifluoromethane sulfonyl) imide and hexafluorophos-
phates were used. The role of ionic pairs on the kinetic
parameters and (enantio)selectivity has been demonstrated.
Although the CO2 expanded methanol system suffered
from a reduction in both reaction rate and product selec-
tivity, this changed in the presence of water. The high
selectivity of the optimized methanol-CO2-water-halide
system was designed as a consequence of observed additive
effects.
Keywords Asymmetric hydrogenation ꢀ Ru-BINAP ꢀ
b-Ketoesters ꢀ Ionic liquid ꢀ Ionic pairs ꢀ
Dense gas expanded solutions
1 Introduction
It is well known that many organometallic complexes are
excellent homogeneous catalysts [1, 2] and those with
chiral centres may even provide multiple enantioselective
recognition. It qualifies them as attractive homogeneous
catalysts for various types of stereoselective reactions
[3–5]. Chiral complexes containing halogen with empirical
formula [RuX(aren)(BINAP)]Y (X = halogen, Y = halo-
gen or BF4) have dominated in asymmetric hydrogenations
yielding valuable optically pure intermediates for many
fine chemicals [6]. This type of the homogeneous hydro-
genation catalysts has revealed exquisite properties in
stereoselective transformation of molecules with keto
groups [7] located either in the b- or c- position to the
carboxyl function [5]. Of course the general problem with
homogeneous catalysts is the difficulty in efficiently sep-
arating and recycling. Recently many new alternatives
have appeared such as the three-dimensional entrapment of
the catalyst [8] (based on the electrostatic catalyst-poly-
electrolyte attraction), utilisation of magnetite nanoparti-
cles [9] or immobilization of the complex in thin films of
ionic liquids deposited on silica beads [10]. Also various
types of ionic liquids (IL) have been considered as reaction
co-solvents for selective accommodation of the catalytic
phase [11–15]. Due to their physical properties and vari-
able phase behaviour separation of the catalyst from the
reaction mixture is feasible. Currently the most popular
types of ILs used for such purposes are imidazolium
cations combined either with hexafluorophosphate or
Electronic supplementary material The online version of this
article (doi:10.1007/s10562-009-0233-3) contains supplementary
material, which is available to authorized users.
T. Floris
Faculty of Chemical Technology, Institute of Chemical
Technology – ICT Prague, Technicka 5, 166 28 Prague,
Czech Republic
P. Kluson (&)
Institute of Chemical Process Fundamentals, Academy of
Sciences of the Czech Republic, Rozvojova 2, 165 02 Prague,
Czech Republic
e-mail: kluson@icpf.cas.cz
URL: www.icpf.cas.cz/HANA
M. J. Muldoon
School of Chemistry and Chemical Engineering, David Keir
Building, Stranmillis Road, Belfast BT9 5AG, Northern Ireland,
UK
H. Pelantova
Institute of Microbiology, Academy of Sciences of the Czech
´ ˇ ´
Republic, Vıdenska 1083, 142 20 Prague 4, Czech Republic
123

280
T. Floris et al.
tetrafluoroborate, and quaternary ammonium cations (such
as n-alkyl-triethylammonium) combined with the bis(tri-
fluoromethane sulfonyl) imide anion [14–17].
(bmimPF6), triethylhexylammonium bis(trifloromethylsulfo-
nyl)imide [N6222][Tf2N]) or triethylhexylammonium hexa-
fluorophosphate ([N6222][PF6]), and finally with addition of
CO2. Despite quite a few scCO2 enhanced reaction systems
have been reported [26, 31–33] those including a chiral
Ru-BINAP complex in the asymmetric hydrogenation have
not been published yet.
Another possibility [18–20] is the use of carbon dioxide,
either dense phase (subcritical) or supercritical. If the
enantioselectivity parameter ee depends on the reaction gas
concentration then the CO2 phase may lead to significantly
increased optical yields [21, 22]. Carbon dioxide (even
when supercritical) is relatively non-polar and it is a poor
solvent for dissolving most transition metal complexes. Its
combination with other co-solvents, referred to as the CO2
expanded liquids (CXLs) [23], might be accepted as an
option. CXLs combine the compressed CO2 phase (often
subcritical) with another solvent, typically an organic sol-
vent or a suitable ionic liquid [24, 25]. Addition of such
solvents [26] allows the dissolution of homogeneous metal
catalysts, whilst CO2 brings an improved mass transfer and
increased solubility for other reactive gases. In the case of
organic solvents addition of pressurised CO2 can be added
at the end of the reaction and it acts as an ‘‘anti-solvent’’
for precipitation of the catalyst. Regarding involvement of
ILs, these are insoluble even in supercritical CO2, therefore
the reaction products can be simply extracted from the
ionic liquid (catalyst) phase [24, 27].
This paper is a continuation of our recent works focused
on the role of the alkyl chain length in the alkylammonium
salts ILs on the progress of MAA hydrogenation and on the
overall design of the asymmetric hydrogenation experi-
ment [15, 34, 35]. We believe that from the general point of
view the strategy described here might be useful for a
practical design of an effective asymmetric synthesis via
the hydrogenation reaction with a homogeneous chiral
Ru-BINAP complex.
2 Experimental
2.1 Chemicals
The organometallic catalytic complex Ru-BINAP, chemi-
cally ([2,20-bis (diphenylphosphino)-1,10-binaphthalene]
chloro(p-cymene)ruthenium chloride) was supplied as the
‘‘Noyori (R) type’’ (Aldrich). Other chemicals were
purchased in analytical p.a. grades––methyl acetoacetate,
1-bromohexane, 1-chlorobutane (Aldrich), KPF6, triethyl-
amine, and 1-methylimidazole and 1-butylimidazolium
hexafluorophosphate ([bmim][PF6]) (Fluka), Li bis
(trifluormethanesulfonyl)imide (LiTf2N) (SynQuest Labo-
ratories), methanol, dichloromethane, ethyl acetate, aceto-
nitrile, diethylether (Penta), CO2 (BOC, SCF grade).
In this paper we report on stereoselective hydrogenation
of achiral methyl acetoacetate (MAA) to optically pure
methyl-3-hydroxybutyrate (MHB) over the (R)-Ru-BINAP
catalyst ([(R)- RuCl(binap)(p-cymen)]Cl) in methanol,
MeOH-IL, MeOH-CO2, MeOH-CO2-H2O and MeOH-
CO2-H2O-QUATS (QUATS = quaternary ammonium
salts) reaction systems. Four different ionic liquids were
employed, each of them structurally resulting from com-
binations of the four appearing ionic pairs.
Quaternary ammonium and imidazolium based ILs
combining hexafluorophosphate and bistriflamide anions
were prepared. Depending on the choice of ions, the ionic
liquids reveal different physical properties such as melting
points, phase behaviour, solvation properties, etc. Upon a
contact with a particular IL the metallic centre of the cat-
alytic complex is re-coordinated with the Lewis anions of
the IL [28]. Alternatively metathesis of the complex
[29, 30] and the migration of the IL‘s anions might be
taken into account together with various steric effects. It
leads to conformation modifications of an active centre
and/or to blocking the substrate access. All these factors
could significantly contribute to the magnitude of the
achieved parameter of enantioselectivity. This kinetic
parameter is generally understood as the key indicator of
the stereoselective reaction progress.
2.2 Preparation of ILs
1-Butylimidazolium bis(trifloromethylsulfonyl)imide ([bmim]
[Tf2N]), triethylhexylammonium bis(trifloromethylsulfo-
nyl)imide ([N6222][Tf2N]) and triethylhexylammonium
hexafluorophosphate ([N6222][PF6]) were prepared. The
precursor salts, triethylhexylammonium bromide and 1-
methylimidazolium chloride were prepared as described
previously [36].
2.3 Preparation of [N6222][Tf2N] and [Bmim][Tf2N]
Synthesis of ILs was performed in 1-L Erlenmayer flask
equipped with a magnetic stirrer. The mixture of [N6222]Br
(0.51 mol) in water (50mL) was added drop wise to the
solution of LiTf2N in water (0.52 mol, 70mL). After
effective mixing (for 10 h), 150 mL of CH2Cl2 was added.
Next, the organic phase was twice washed with 70 mL of
ultrapure water to remove any water soluble impurities.
To summarise the performed reactions were carried out
separately in pure methanol, in methanol combined
with 1-butylimidazolium bis(trifloromethylsulfonyl)imide
(bmimTf2N), 1-butylimidazolium hexafluorophosphate
123

Notes on the Asymmetric Hydrogenation of Methyl Acetoacetate
281
Presence of halides was tested by AgNO3 (1M solution)
until no precipitation of AgBr was observed. Then the
organic phase was once washed with 30 mL of water.
Finally, the suspense was filtered off with help of a short
Brockmann alumina column, evaporated and dried under
vacuum at 333K (attained yield 98%). The BmimTf2N IL
was prepared analogously with an identical yield.
stirrer on (at 333 K temperature level). The constant
hydrogen pressure was maintained by a forward pressure
regulator connected to a hydrogen ballast vessel.
Turn-over frequencies (TOF), selectivities (S) to MHB
and enantioselectivities (ee) to the (R) isomer were eval-
uated as the principal kinetic parameters. The TOF, ee and
the selectivity were always determined at 90% conversion
(experiments with IL) or at the 75th min (experiments with
CO2) of the experimental run. In this case the continuous
sampling of reaction mixture was disabled to avoid chan-
ges of the liquid–gas reaction mixture composition. The
75th min was established due to the experimentally
observed higher level of MAA conversion. The ee
parameter is defined:
2.4 Preparation of [N6222][PF6]
The 250 mL flask was charged with KPF6 (60 mmol) and
water (30 mL), and CH2Cl2 (20mL). The [N6222]Br
(60 mmol) in water (20 mL) was added drop wise. The two
phase reaction mixture was stirred intensively for 8 h at
room temperature. The purification was performed as
described previously for [N6222][Tf2N] (attaining a yield of
98%).
ee ¼ ½Rꢁ ꢂ ½Sꢁ=ð½Rꢁ þ ½SꢁÞ ꢃ 100
ð1Þ
The general selectivity parameter S to MHB reflected
formation of acetals in the presence of methanol as
described previously [34, 37].
2.4.1 Hydrogenation Reactions
S ¼ ½Rꢁ þ ½Sꢁ=ð½Rꢁ þ ½Sꢁ þ ½ACꢁÞ ꢃ 100
ð2Þ
The catalytic reactions were carried out in two types of
reactors. The first reaction system enabled periodical
sampling. The second system was designed for experi-
ments in the dense CO2 phase. Hydrogenation experiments
with methanol and with methanol-IL phase were performed
in a stainless autoclave (25 mL, Parr; only ever used
for Ru-BINAP reaction tests). The typical amount of
Ru-BINAP introduced to the reaction was 10 mg in the
IL/methanol mixture (7 mL). It was loaded to the reactor
with help of Ar overpressure by means of a sampling tube.
The complex was activated for 1 h in hydrogen at 333 K
and 5 MPa. The same temperature and pressure were also
used in the reactions. MAA (2 g) was introduced to the
reactor in methanol (10 mL) similarly as the catalyst
through the sampling probe. Samples of the reaction mix-
ture were periodically withdrawn and analysed. Repeated
runs were performed analogously. MHB and other products
were washed out with water. The Ru complex remained
preserved in the IL phase and then introduced with the
appropriate amount of methanol and the substrate to the
reactor again.
2.4.2 Analytical
Samples of the reaction mixture were chromatographically
analysed with a chiral GC column MN Hydrodex-b-3P (50
m 9 0.25 mm 9 0.25 lm). NMR spectra were recorded on
a Bruker Avance III––400 MHz spectrometer (400.13 MHz
for 1H, 100.61 MHz for 13C) in CD3OD at 303 K. Residual
signal of the solvent was used as an internal standard (dH
3.330 ppm, dC 49.30 ppm). H, 13C, gCOSY and proton-
1
edited gHSQC spectra were recorded using the standard
manufacturer’s software. 1H NMR spectrum was zero filled
to fourfold data points and multiplied by a two-parameter
double-exponential Lorentz-Gauss function before Fourier
transformation to improve the resolution. The numbering
used is depicted below.
3
5
1
'
+
"
'
2
3
1
1
N
'
'
2
+
4
4
2
6
N
N
'
1
'
4
5
2
Experiments with CO2 were carried out in a 100 mL
Hastelloy C276 magnetically stirred pressure reactor
equipped with a large oblong quartz-window to enable
observation of the phase behaviour. The autoclave loaded
with Ru-BINAP (15 mg) had been flushed out with nitro-
gen for 15 min dried for 15 min before degassed MeOH
(15 mL) was added. The catalyst was pre-treated for 1 h in
the hydrogen pressure at 333 K followed by addition of the
degassed MAA (3 g) carried with liquid CO2 by high-
pressure piston pump. In the meantime the reactor was
pressurized. The reaction was started by switching the
[bmim][PF6]
1H NMR: 0.976 (3H, t, J = 7.4 Hz, H-40), 1.383 (2H, m,
H-30), 1.877 (2H, m, H-20), 3.914 (3H, s, H-100), 4.198 (2H,
t, J = 7.4 Hz, H-10), 7.486 (1H, dd, J = 1.8, 1.8 Hz, H-4),
7.546 (1H, dd, J = 1.8, 1.8 Hz, H-5), 8.697 (1H, m, H-2);
13C NMR: 13.87 (C-40), 20.52 (C-30), 33.12 (C-20), 36.62
(C-100), 50.77 (C-10), 123.70 (C-5), 125.00 (C-4), 137.84
(C-2)
123

282
T. Floris et al.
[bmim][Tf2N]
O
H
O
O
O
O
(R)-Ru-BINAP
1H NMR: 0.997 (3H, t, J = 7.4 Hz, H-40), 1.398 (2H, m,
H-30), 1.888 (2H, m, H-20), 3.933 (3H, s, H-100), 4.217 (2H,
t, J = 7.4 Hz, H-10), 7.532 (1H, dd, J = 1.9, 1.9 Hz, H-4),
7.599 (1H, dd, J = 1.9, 1.9 Hz, H-5), 8.831 (1H, m, H-2);
13C NMR: 13.89 (C-40), 20.62 (C-30), 33.24 (C-20), 36.70
(C-100), 50.86 (C-10), 121.45 (q, JC,F = 320.6 Hz, CF3)
123.86 (C-5), 125.15 (C-4), 137.9 (broad signal, C-2)
IL+meOH
or CO2+meOH
O
Scheme 1 Methyl acetoacetate hydrogenation
on the magnitude of the selectivity parameter S. Also the
mechanistic model of the Ru-BINAP catalytic cycle has
already been introduced [40], however, for the reaction
system without the ionic liquid phase.
[N6222][PF6]
1H NMR: 0.951 (3H, m, H-6), 1.304 (9H, tt, J = 7.3 Hz,
J(14N,1H) = 1.9 Hz, H-20), 1.406 (6H, m, H-3, H-4, H-5),
1.694 (2H, m, H-2), 3.184 (2H, m, H-1), 3.322 (6H, q,
3.1 The Role of Ion Pairs
13
J = 7.3 Hz, H-10); C NMR: 7.87 (C-20), 14.48 (C-6),
To elucidate the effect of combination of four ionic pairs,
[bmim][Tf2N], [N6222][Tf2N] and [N6222][PF6] were pre-
pared and carefully purified to avoid a role of the trace
contamination (mostly organic or inorganic halides,
metallic salts, triethylamine and 1-methylimidazole). The
kinetic data were compared with those obtained in pure
methanol (Table 1). It is seen the TOF at the 90% con-
version surpassed well the reaction systems involving ILs,
however, with distinctive differences among them. On the
other hand, as already introduced, due to high content of
alcohol, this reaction environment was very much affected
by the formation of acetals (reflected in the S90 parameter).
Evidently the bistriflamides based ILs revealed much less
negative effect on the reaction rate than the hexafluoro-
phosphates. This is an important finding due to the
broadband utilisation of [bmim][PF6] in reactions of this
type [11, 13]. The TOF was in average three times lower
with the [Tf2N] anion in comparison to the standard
experiment in methanol, and nearly six times lower in the
case of the [PF6] anion.
22.80 (C-2), 23.74 (C-5), 27.30 (t, J(14N,13C) = 1.9 Hz,
C-3), 32.56 (C-4), 54.11 (t, J(14N,13C) = 3.0 Hz, C-10),
58.39 (t, J(14N,13C) = 2.9 Hz, C-1)
[N6222][Tf2N]
1H NMR: 0.947 (3H, m, H-6), 1.302 (9H, tt, J = 7.3 Hz,
J(14N,1H) = 1.9 Hz, H-20), 1.401 (6H, m, H-3, H-4, H-5),
1.692 (2H, m, H-2), 3.183 (2H, m, H-1), 3.322 (6H, q,
13
J = 7.3 Hz, H-10); C NMR: 7.91 (C-20), 14.45 (C-6),
22.82 (C-2), 23.71 (C-5), 27.28 (t, J(14N,13C) = 1.3 Hz,
C-3), 32.53 (C-4), 54.16 (t, J(14N,13C) = 3.0 Hz, C-10),
58.44 (t, J(14N,13C) = 2.9 Hz, C-1), 121.48 (q, JC,F
320.6 Hz, CF3)
=
3 Results and Discussion
The primary intention of this work was to evaluate and
compare (positive and negative) aspects of various
arrangements of the (R)-methyl 3-hydroxybutyrate syn-
thesis from achiral methyl acetoacetate by means of the
stereoselective hydrogenation [11, 12, 14] with the chiral
Ru-BINAP complex. These arrangements comprised the
standard homogeneous experiment with methanol as a
solvent, the mixed solvent system of methanol-IL or
methanol-CO2 and finally the methanol-CO2 with additives
- water and organic chlorides. In parallel the role of the ion-
pairs combination was also elucidated.
Generally, there are two major reaction rate limiting
factors: hydrogen gas solubility and/or structural modifi-
cations of the active catalytic centre with the anionic pair
(Figs 1 and 2). Both factors are considered as similarly
important. Solubility of hydrogen is generally low in the
ionic liquids phase. It is believed that each Ru-BINAP
molecule is selectively entrapped by the IL molecules. The
effect of anion’s type of the IL predominated the effect of
the cation’s type as illustrated in the Fig. 3. The reaction
proceeded in generall faster with the quaternary amonium
based ILs, in comparison with the imidazolium based ILs.
The observed induction period might be attributed to the
expulsion of a ligand providing the catalytically active
centre [26].
The reaction course expressed in the form of a formal
reaction scheme (Scheme 1) has been reported repeatedly
[34, 37, 38]. It comprises a series of steps and formation of
a series reaction intermediates. When the reaction is per-
formed in the alcohol environment appearance of acetals
cannot be omitted. It should be emphasized that especially
the acidic properties of Ru-BiNAP play a significant
important role toward the extent of their appearance [39]. It
is noteworthy there is a clear reflection of this phenomenon
Unlike the TOFs the ee values achieved with the
[TF2N]- based ILs were quite high, especially with
[bmim][TF2N] (Table 1). It should be emphasized that its
value was nearly comparable with the ee evaluated for the
123

Notes on the Asymmetric Hydrogenation of Methyl Acetoacetate
283
Table 1 The ionic pairs effect
IL
TOF90
(h-1
ee90 (%)
98
S90 (%)
79
eere-use (%)
)
MeOH
1100
F
F
F
S
F
O
N
O
N+
F
S
O
390
93
97
55
78
87
91
86
88
54
54
–
F
O
N6222Tf2N
F
F
F
N+
F
O
N
O
F
S
N
S
340
210
160
O
F
O
bmimTFN
F
P
F
F
F
F
N+
F
N6222PF6
F
P
F
N+
F
F
F
F
N
32
6
bmimPF6
Reaction conditions: 2 g MAA, 17 mL solvent IL-MeOH 1/1 wt., S/C = 1580, 333 K, 50 bar H2
(Ph)2
P
(Ph)2
Ru+
Ru+
P
Cl
Cl
P
(Ph)2
O
F
F
O
Cl
P
F
F
N
S
(Ph)2
F
O
F
F
S
S
O
N
O
N+
O
F
O
S
F
O
F
N+
F
F
Cl
Fig. 1 Proposed ‘‘weak’’ catalyst-IL ionic pairs interaction
Fig. 2 Proposed catalyst-IL ionic pairs re-coordination
standard methanol experiment. On the other hand these
parameters (for the PF6 anion based ILs) were rather low.
As introduced in the experimental section the structure
and the ‘‘origin’’ of ILs had been determined by NMR. It is
of note here that the J(14N,13C) and J(14N,1H) couplings
were observed due to the symmetry of the N6222 cation.
The NMR data displayed almost no change depending on
the anion’s type. It must be emphasized that no relation of
the ionic pairs to the catalytic course were observed from
the 1HNMR or 13CNMR point of view. Hence, it should be
noted that the extended positive charge delocalization of
the imidazolium’s cation may play a certain role in the
catalytic course when compared to the aromatic-less
ammonium IL, and considering the aromatic nature of the
BINAP and p-cymene ligands.
As might be seen in Fig. 4 the parameter of enantios-
lectivity typically revealed an increasing tendency with the
progress of the reaction and this phenomenon was even
123

284
T. Floris et al.
1
0.8
0.6
0.4
0.2
0
At higher conversions concentration of enols is lower and
thus ee increases. When the ionic liquid is present in the
methanol phase the reaction rate is much lower as already
described and also the enol and acetal intermediates are
formed at much lower extent (Fig. 5). This in turn may
cause indirect suppression of the magnitude of ee. In other
words the lower S90 at more rapid reactions may indicate
much higher ee. The maximum of acetals formation was in
all cases higher in the presence of imidazolium cation. It
might be attributed to the hydrogen acidic properties of
imidazolium [42, 43]. In spite of the lower acidity of IL in
comparison with methanol the imidazolium based IL are
more acidic than ammonium.
xMA
only Me OH
N6222Tf2N
bm im Tf2N
N6222PF6
bm im PF6
3.2 The Effect of CO2
To investigate the effect of CO2 the reactor with a large
oblong quartz window was used, enabling the in situ
observation of the phase behavior of the solvent system. As
described earlier the used metal complexes reveal a ten-
dency not to be well soluble in dense or scCO2. Therefore
the influence of added CO2 pressure to the methanol sys-
tem was examined. The addition of relatively small pres-
sure increment can lead to the improved solubility of other
gases such as H2. Furthermore, at the end of the reaction
further increase in the CO2 pressure can lead to the pre-
cipitation of the catalyst from the reaction mixture. The
expansion of the liquid (solvent methanol and MAA) by
CO2 resulted in a single liquid phase. The presence of
MeOH was essential due to the virtually zero catalyst’s
solubility in MAA and liquid CO2. It must be noted that
with this arrangement the mixture sampling was not con-
sidered to avoid the undesired composition changes. Hence
the reaction parameters were always elucidated at the 75th
min of the reaction run.
0
2
4
6
8
10
time,
h
Fig. 3 Effect of IL pairs IL on the conversion of MAA
1
0.8
0.6
e
0.4
only MeOH
N6222Tf2N
bm im Tf2N
N6222PF6
0.2
Under the studied conditions the incremental increase of
the solvent swelling was observed as a consequence of
increasing the partial pressure of CO2. However, it was also
expected that increasing the pressure could increase the
reaction rate due to improved mass transfer. However, it
was found that the reaction rate was actually suppressed in
the presence of CO2 (Table 2). (Note: The same trend was
indicated in [31] for a similar system but for synthesis of
naproxen via enantioselective hydrogenation). The expla-
nation might be potentially sought in terms of the poison-
ing of the catalyst with CO originated from CO2 reduction
observed [44] in the enantioselective hydrogenation of
ethyl pyruvate via heterogeneous Pt/Al2O3. However, such
deactivation has never been indicated for homogeneous
Ru-phosphines catalyzed reactions and thus we suppose it
only as a hypothetical explanation. The molecular com-
petitive interaction of the carbonyl groups of the substrate
and CO2 on the Ru-BINAP active centre is more plausible
bm im PF6
0
0
2
4
6
8
10
time,
h
Fig. 4 Ionic pairs effect on the enantioselectivity
more pronounced with reactions in the IL/methanol phase.
Some interpretation might be sought in terms of formation
of highly unstable enolic intermediate (Scheme 2) in the
presence of pure methanol. The enol form is much less
stereospecific than MAA and thus a lower ratio of the (R)
to (S) isomers could be expected [41]. By analogy to this
explanation the hydrogenation of acetals [36] might be also
considered. Obviously highly reversible enol and acetals
formation is significant especially at low MAA conversion.
123

Notes on the Asymmetric Hydrogenation of Methyl Acetoacetate
285
Scheme 2 Proposed reaction
intermediate formation
O
H
H
OH
CH2COOCH3
OH
H2
+
H3C
H3C
CH2COOCH3
MAA
H3C
CH2COOCH3
(S)-MHB
(R)-MHB
H3C C
CHCOOCH3
OH
+ 2 MeOH
- H2O
H2
Enol
OMe
MeO
H3C
CHCOOCH3
+ MeOH
- H2O
Acetal
- MeOH
H3C
OMe
C
CHCOOCH3
but no quantification of the relative binding strength of
CO2 vs. MAA (or ester) on ruthenium phosphines was
available. When considering the reduction of CO2 with H2
catalyzed by ruthenium phosphines complexes [50] formic
acid is generally formed. On one hand, in the presence of
acids [40], the TOF and the selectivity enhancements had
been achieved, on the other hand the TOF and ee decreased
dramatically in the presence of carboxylates N4444OAc or
0.25
0.20
0.15
cetals
0.10
0.05
0.00
N
4444OOPh (20mol. eq. to catalyst). In the presence of CO2
only Me OH
neither significant TOF increase nor drop of ee was indi-
cated. In principle, the esters (substrates) are not assumed
as deactivants (taking into account the possible formation
of methyl formate).
bm im Tf2N
bm im PF6
N6222Tf2N
N6222PF6
0
2
4
6
8
10
Another possible explanation is the dilution effect. It is
expected to be prevailing at higher pressures of CO2 [55,
56]. Experiments carried out at much lower pressures, it
means at the level when virtually no expansion took place,
did not point out on any contribution of diluting to the drop
of the reaction rate. We believe that at moderate pressures
used in our tests the contribution of dilution is plausible,
but likely accompanied with some other mechanism. The
effect of the CO2 presence on the selectivity was rather
negative, the ee value was influenced in minimal, rather
positively and specifically at higher carbon dioxide partial
pressures.
time,
h
Fig. 5 Ionic pairs effect on the formation of by-products
Table 2 Effect of the CO2 partial pressure
Entry
pH2
(bar)
pCO2
(bar)
x
TOF1.25
(h-1
ee1.25
(%)
S1.25
(%)
MAA
)
1
4.4
4.4
10
9.2
33
33
50
50
50
20
80
–
0.38
0.19
0.37
0.25
0.51
0.37
0.97
0.72
0.70
0.28
0.996
470
240
440
300
680
470
1300
930
830
350
1300
98.6
98.6
98.6
98.6
98.6
98.6
98.7
98.8
98.9
99.0
98.6
14
3
2
30
–
3
23
23
56
53
78
72
69
68
83
3.3 Optimization of the CO2 System with Water
4
25
–
5
A certain role of the pH in acceleration of the MAA
transformation over Ru-BINAP has been reported [40, 45].
In the presence of slightly acidic H2CO3––formed from
carbon dioxide pH decrease might be expected [46] as well
as the formation of methyl carbonic acid from CO2 and
methanol, reported by Eckert and Liotta [54]. The very
positive effect of water on the reaction rate and especially
selectivity had been described elsewhere [6, 34, 51–53].
The reversible formation of acetals, producing H2O, is in
6
30
–
7
8
30
51
80
20
9
10
11
Reaction conditions: 3 g MAA, 25 mL MeOH, S/C = 1580 (mol/
mol), 333 K, t = 1.25 h
123

286
T. Floris et al.
Table 3 Effect of the additives in H2–CO2 hydrogenation systems
Entry
pH2 (bar)
pCO2 (bar)
Additive/amount
x
TOF1.25 (h-1
)
ee1.25 (%)
S1.25 (%)
MAA
7
50
50
50
50
50
50
50
50
50
50
50
–
–
0.97
0.52
1
1300
660
98.7
98.3
99.2
98.9
98.6
98.6
98.1
97.4
99.1
98.9
98.8
78
55
90
70
94
99.6
96
98
90
77
98
12
13
9
–
H2O/0.4 mL
choline chloridea
–
51
50
50
50
50
50
50
50
–
0.70
0.79
1
830
960
14
15b
16
17
18
19
20
H2O/0.4 mL
H2O/0.4 mL
H2O/0.9 mL
0.90
0.76
1
1100
960
H2O/2.35 mL
Choline chloridea
Betaine hydrochloride
H2O/0.4 mL choline chloridea
0.86
1
1100
Reaction conditions: 3 g MAA, 25 mL MeOH, S/C = 1580 (mol/mol), 333 K, t = 1.25 h
a
Additive/catalyst = 20 (mol/mol)
b
3h
the presence of water suppressed, thus the ratio substrate/
acetals is higher such as the yield of the product.
increased by 0.5–99.2% when MeOH/choline chloride was
used. The selectivity increased to 12% in comparison with
the standard experiment. The CO2/MeOH system (entry
13) turned very positively when choline chloride was
added (entry 18). Comparing this system with MeOH/
choline chloride, the ee was slightly negatively influenced
by CO2 and the role of CO2 seemed to be negligible. We
attempted to design ‘‘a new reaction system’’ for the
studied hydrogenation: MeOH/water/CO2/choline chloride,
resulting from optimisation of all preceding systems.
Although the presence of the water-carbon dioxide com-
ponent depressed slightly the ee, this parameter was (98,
8%) was virtually on the same level as for the standard
experiment. The total conversion and very high selectivity
97.8% (as a consequence of CO2-water effect) might even
satisfy requirements of a practical process.
Thus to verify it the ‘‘catalytic amounts’’ of water (0.4,
0.9, 2.3 mL) were added to the reaction mixture containing
25 mL of methanol. The very impressive reaction rate
enhancement is showed in Table 3, revealing a clear
maximum for the addition of 0.9 mL of water, indicating an
optimum of pH in some cases. This ‘‘water effect’’ is
evidently very much limited to only low volumes of water.
The parameter of selectivity S increased significantly with
increasing the water content, up to the level of 98%. It
should be emphasized that such high values of the selec-
tivity are achieved in the standard (methanol) reaction
system at much longer reaction times (on the account of the
reversible formation of acetals). Nevertheless, the value of
ee decreased by 1.5% with increasing the concentration of
water as observed elsewhere [34] with CO2 free solvents.
The results obtained in the methanol/water system (without
CO2; catalyst’s precipitation signs) offer a clear evidence
of the positive effect of CO2––water.
4 Conclusion
Two homogeneous reaction systems were considered. The
first, methanol-ILs system is able to satisfy catalyst’s
recovery requirements, the latter, methanol-CO2 modified
system offers high reaction rate and (enantio) selectivity,
but it in parallel offers the possibility of the catalyst sep-
aration. Also the choice of an ionic pairs in the molecule of
IL plays an important role in the asymmetric hydrogenation
of MAA. Indeed, the reaction rate revealed to be strongly
influenced in the presence ionic liquids, but the ee decrease
as the most important parameter. The modified reaction
system was designed to explain ee increase accentuated in
the presence of ILs. In the second part the standard
(methanol) system was modified by scCO2. Even the sol-
vent swelling effect on the mass transport was diminished
3.4 Tuning the Reaction System
In transition metal catalysis the halide effect had been
described [47–49]. The preliminary experiments pointed
out the very essential role of organic halides positively
influencing [35] the TOF as well as the (enantio)selectivity.
The effect of cation’s structure of chlorides or bromides on
the reaction rate has been studied in every detail and this
subject will be covered in our next communication. Here
we wish briefly draw attention on the extraordinary role of
the choline chloride reaction additive. As shown in Table 3
in the presence of choline chloride all monitored kinetic
parameters were enhanced. It is seen that the value of ee
123

Notes on the Asymmetric Hydrogenation of Methyl Acetoacetate
287
22. Landis CR, Halpern J (1987) J Am Chem Soc 109:1746
23. Jessop PG, Subramaniam B (2007) Chem Rev 107:2666
24. Webb PB, Sellin MF, Kunene TE, Williamson S, Slawin AMZ,
Cole-Hamilton DJ (2003) J Am Chem Soc 125:15577
25. Kuhne E, Santarossa S, Perez E, Witkamp GJ, Peters CJ (2008)
J Supercrit Fluids 46:93
26. Dong X, Erkey C (2004) J Mol Catal A: Chem 211:73
27. Fredlake CP, Muldoon MJ, Aki SN (2004) Phys Chem Chem
Phys 6:3280
by catalyst’s inactivity in the presence of CO2, the reaction
was tuned by synergic water effect. Finally, the reaction
parameters were optimized with respect to the positive
effects of additives––CO2/water and halides. This high-
selectivity optimized system might be a challenge in
practical asymmetric b-ketoesters hydrogenations.
Acknowledgment Authors wish to acknowledge the Grant Agency
of the Academy of Sciences of the Czech Republic, project
KAN400720701, the Ministry of Education of the Czech Republic,
project NanoPin no. 1M4531433201 and TF specifically thanks the
Ministry of Education of the Czech Republic, project no. 203/08/
H032-1 for funding partly his PhD fellowship. We are also grateful to
the EU funded FP6 ‘Structuring the European Research Area’ pro-
gramme, which funded the access to CenTACat facilities at QUB.
28. Daguenet C, Dyson PJ (2004) Organometallics 23:6080
29. Daley CJA, Bergens SH (2002) J Am Chem Soc 124:3680
30. Abbel R, Rashid KA, Faatz M, Hadzovic A, Lough AJ, Morris
RH (2005) J Am Chem Soc 127:1870
31. Combes G, Coen E, Dehghani F, Foster N (2005) J Supercrit
Fluids 36:127
32. Berthod M, Mignani G, Lemaire M (2004) Tetrahedron Asym-
metr 15:1121
33. Hu Y, Birdsall DJ, Stuart AM, Hope EG, Xiao J (2004) J Mol
Catal A: Chem 219:57
34. Bartek L, Drobek M, Kuzma M, Kluson P, Cerveny L (2005)
Catal Commun 6:61
References
35. Floris T, Kluson P, Bartek L, Pelantova H (2009) Appl Catal A
366:160
36. Sun J, McFarlane DR, Forsyth M (1997) Ionics 3:356
37. Ochsner E, Etzold B, Junge K, Beller M, Wasserscheid P (2009)
Adv Synth Catal 351:235
38. Matteoli U, Beghetto V, Scrivanti A (1999) J Mol Catal A: Chem
140:131
39. Kou Y, Yamada N, Akutagawa S, Zhang G, Tanabe K (1991)
Bull Chem Soc Jpn 64:133
1. Akutagawa S (1995) Appl Catal A 128:171
2. Xia Y-Q, Tang Y-Y, Liang Z-M, Yu C-B, Zhou X-G, Li R-X, Li
X-J (2005) J Mol Catal A: Chem 240:132
3. Saito T, Yokozawa T, Ishizaki T, Moroi T, Sayo N, Muira T,
Kumobayashi H (2001) Adv Synth Catal 343:264
4. Huang HL, Liu LT, Chen SF, Ku H (1998) Tetrahedron Asym-
metr 9:1637
5. Ratovelomanana-Vidal V, Girard C, Touati R, Tranchier JP,
ˆ
40. Wolfson A, Vankelecom IFJ, Geresh S, Jacobs PA (2004) J Mol
Catal A: Chem 217:21
Hassine BB, Genet JP (2003) Adv Synth Catal 345:261
6. Wolfson A, Vankelecom IFJ, Geresh S, Jacobs PA (2003) J Mol
Catal A: Chem 198:39
7. Blaser HU, Spindler F, Studer M (2001) Appl Catal A 221:119
41. Sun Y, Wang J, LeBlond C, Landau RN, Laquidara J, Sowa JR,
Blackmond DG (1997) J Mol Catal A: Chem 115:495
42. Wang Y, Li H, Han S (2006) J Chem Phys 124:044504
43. Avent AG, Chaloner PA, Day MP, Seddon KR, Welton T (1994)
J Chem Soc Dalton Trans 3405
¨
8. Kockritz A, Bischoff S, Morawsky V, Pruße U, Vorlop K-D
¨
(2002) J Mol Catal A: Chem 180:231
9. Hu AG, Yee GT, Lin WB (2005) J Am Chem Soc 127:12486
10. Fow KL, Jaenicke S, Muller TE, Sievers C (2008) J Mol Catal A:
Chem 279:239
44. Minder B, Mallat T, Pickel KH, Steiner K, Baiker A (1995) Catal
Lett 34:1
45. King SA, Thompson AS, King AO, Verhoeven TR (1992) J Org
Chem 57:6689
11. Pugin B, Studer M, Kuesters E, Sedelmeier G, Feng X (2004)
Adv Synth Catal 346:1481
46. Toews KL, Shroll RM, Wai CM, Smart NG (1995) Anal Chem
67:4040
12. Guernik S, Wolfson A, Herskowitz M, Greenspoonc N, Geresh S
(2001) Chem Commun 2314
47. Fagnou K, Lautens M (2002) Angew Chem Int Ed 41:26
48. Jiang Q, Jiang Y, Xiao D, Cao P, Zhang X (1998) Angew Chem
Int Ed 37:1100
13. Wolfson A, Vankelecom IFJ, Jacobs PA (2005) J Organomet
Chem 690:3558
14. Starodubtseva EV, Vinogradov MG, Pavlov VA, Gorshkova LS,
Ferapontov VA (2004) Russ Chem Bull 53:2172
15. Cerna I, Kluson P, Drobek M, Cajthaml T, Bartek L (2007) Chem
Listy 101:994
¨
49. Vogl EM, Groger H, Shibasaki M (1999) Angew Chem Int Ed
38:1570
50. Jessop PG, Joo F, Tai C-C (2004) Coord Chem Rev 248:2425
´
51. Mashima K, Kusano KH, Sato N, Matsumura Y, Nozaki K,
Kumobayashi H, Sayo N, Hori Y, Ishizaki T, Akutagawa AS,
Takaya H (1994) J Org Chem 59:3064
52. Gelman F, Avnir D, Schumann H, Blum J (1999) J Mol Catal
146:123
53. Wan KT, Davis ME (1993) J Chem Soc Chem Commun 1262
54. West JN, Wheeler C, McCarney JP, Griffith JN, Bush D, Liotta
CL, Eckert CA (2001) J Phys Chem A 105:3947
55. Couchi D, Gourgouillon D, Courel M, Vital J, Ponte MN (2001)
Ind Eng Chem Res 40:2551
16. Brown RA, Pollet P, McKoon E, Eckert CA, Liotta CA, Jessop
PG (2001) J Am Chem Soc 123:1254
17. Malhotra SV, Kumara V, Parmarb VS (2007) Curr Org Synth
4:370
18. Olivier-Bourbigou H, Magna L (2002) J Mol Catal A: Chem
419:182–183
19. West KN, Hallett JP, Jones RS, Bush D, Liotta CL, Eckert CA
(2004) Ind Eng Chem Res 43:4827
20. Solinas M, Pfaltz A, Cozzi PG, Leitner W (2004) J Am Chem Soc
126:16142
56. Milewska A, Banet Osuna AM, Fonseca IM, Nunes da Ponte M
(2005) Green Chem 7:726
21. Pestre N, Meille V, Bellefon C (2006) J Mol Catal A: Chem
252:85
123