Asymmetric hydrogenation of methyl pyruvate in the continuous gas phase using Supported Ionic Liquid Phase (SILP) catalysis

Article abstract of DOI:10.1016/j.molcata.2013.04.022

Based on the recently reported asymmetric hydrogenation of methyl acetoacetate in a continuous-flow, gas phase reaction using Supported Ionic Liquid Phase (SILP) catalysis, this study deals with the even more challenging asymmetric reduction of the α-keto ester methyl pyruvate. Different support materials, ligands and ionic liquids were screened to identify an optimized Ru-based SILP catalyst. With Ru-BINAP dissolved in the ionic liquids 3-hydroxypropylpyridinium bis(trifluoromethylsulfonyl)imide on silica 100, stable catalyst performance could be obtained for more than 50 h time-on-stream. At 95 C and with a residence time in the reactor of less than 3 s a stable product yield of 80-84% was obtained with moderate but stable enantiomeric excess of 26-30%.

Full text of DOI:10.1016/j.molcata.2013.04.022

Accepted Manuscript
Title: Asymmetric hydrogenation of methyl pyruvate in the
continuous gas phase using Supported Ionic Liquid Phase
(SILP) catalysis
Author: Martin Johannes Schneider Marco Haumann Peter
Wasserscheid
PII:
S1381-1169(13)00170-2
DOI:
Reference:
http://dx.doi.org/doi:10.1016/j.molcata.2013.04.022
MOLCAA 8741
To appear in:
Journal of Molecular Catalysis A: Chemical
Received date:
Revised date:
Accepted date:
12-9-2012
3-4-2013
18-4-2013
Please cite this article as: M.J. Schneider, M. Haumann, P. Wasserscheid, Asymmetric
hydrogenation of methyl pyruvate in the continuous gas phase using Supported Ionic
Liquid Phase (SILP) catalysis, Journal of Molecular Catalysis A: Chemical (2013),
http://dx.doi.org/10.1016/j.molcata.2013.04.022
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

*Graphical Abstract (for review)
Suggested graphical abstract:
Page 1 of 19

*Research Highlights
Research Highlights:



The asymmetric reduction of the -keto ester methyl pyruvate in continuous-flow, gas
phase reaction using Supported Ionic Liquid Phase (SILP) catalysis is demonstrated
Different support materials, ligands and ionic liquids were screened to identify an
optimized Ru-based SILP catalyst.
With Ru-BINAP dissolved in the ionic liquids 3-hydroxypropylpyridinium
bis(trifluoromethylsulfonyl)imide on silica 100 a stable catalyst performance could be
obtained for more than 50 hours time-on-stream.

At 95°C and with a residence time in the reactor of less than 3 seconds a stable
product yield of 80-84% was obtained with moderate but stable enantiomeric excess
of 26-30%.
Page 2 of 19

*Manuscript
Asymmetric hydrogenation of methyl pyruvate in the continuous
gas phase using Supported Ionic Liquid Phase (SILP) catalysis
Martin Johannes Schneider, Marco Haumann, Peter Wasserscheid*
Lehrstuhl für Chemische Reaktionstechnik, Friedrich-Alexander-Universität Erlangen-
Nürnberg, Egerlandstrasse 3, D-91058 Erlangen, Germany
Email: wasserscheid@crt.cbi.uni-erlangen.de
* Corresponding author. Tel.: +499131 8527420, fax: +499131 8527421, E-mail address:
wasserscheid@crt.cbi.uni-erlangen.de
Keywords: Asymmetric hydrogenation; gas phase; methyl pyruvate; Supported catalyst; SILP
catalyst; ionic liquid
Abstract
Based on the recently reported asymmetric hydrogenation of methyl acetoacetate in a
continuous-flow, gas phase reaction using Supported Ionic Liquid Phase (SILP) catalysis,
this study deals with the even more challenging asymmetric reduction of the -keto ester
methyl pyruvate. Different support materials, ligands and ionic liquids were screened to
identify an optimized Ru-based SILP catalyst. With Ru-BINAP dissolved in the ionic liquids 3-
hydroxypropylpyridinium bis(trifluoromethylsulfonyl)imide on silica 100, stable catalyst
performance could be obtained for more than 50 hours time-on-stream. At 95°C and with a
residence time in the reactor of less than 3 seconds a stable product yield of 80-84% was
obtained with moderate but stable enantiomeric excess of 26-30%.
1. Introduction
Chiral molecules, pharmaceuticals and agrochemicals produced via multistep synthesis are
often relatively small scale but high value products. In order to shorten “time to market” the
level of development in the applied production process is often rather limited and typical
standard procedures using multipurpose batch equipment is frequently applied in these
productions. The associated massive use of solvents is considered the biggest contributor to
waste in pharmaceutical production. Associated to the excessive solvent use, energy
consuming work-up procedures and downstream processing have to be applied. [1] [2]
The use of continuous flow techniques offers huge potential to intensify these production
processes, especially in the field of asymmetric catalysis, with improved heat and mass
1
Page 3 of 19

transfer, easier scale-up, on-demand reaction analysis and even continuous automated on-
stream optimization of reaction parameters. [3] Therefore, a lot of recent publications have
focused on the field of asymmetric reactions in continuous flow, some of which have been
reviewed by Mak et al. [4] Besides the intensification of enantioselective transformations by
the use of continuous flow processes, the immobilization and recyclability of homogeneous
and heterogeneous asymmetric catalysts has been a major interest of the scientific
community in the recent years. [5], [6], [7]
Especially covalent tethering of active homogeneous catalysts to various support materials
and immobilization by electrostatic interactions have resulted in highly stable catalyst
systems capable of reuse. While immobilized catalysts often show lower catalyst leaching,
their activity and stereoselectivity is generally lower than in the respective homogeneous
systems, with a few exceptions. [6] Furthermore, the majority of immobilization techniques
require multistep synthesis to prepare the immobilized catalyst system. [8] Easy and cost
effective immobilization of highly elaborated transition metal complexes remains a critical
issue for potential industrial applications of asymmetric catalysis. [9], [4], [10]
A particularly easy technique of catalyst immobilization that has shown limited applicability in
liquid systems due to leaching issues is the adsorption of homogeneous catalysts and liquid
catalyst solutions on solid support material. This technique, however, appears very promising
when combined with alternative media – more precisely ionic liquids or supercritical fluids. In
supported ionic liquid phase (SILP) materials a homogeneous catalyst is dissolved in a thin
film of ionic liquid immersed on a highly porous support material. [11], [12] The resulting,
macroscopically solid SILP compound is easy to handle and combines the advantages of
homogeneous and heterogeneous catalysis. [13] SILP catalysts have shown excellent
stability and recyclability in various types of reactions, especially when applied in gas phase
reaction set-ups. The scope of successful SILP catalyzed gas phase reactions includes
hydroformylation [14], methanol carbonylation [15], hydroamination [16], hydrogenation [17]
and ultralow temperature water–gas shift reaction [18]. Just recently, we demonstrated that
even reactions with demanding selectivity problems like asymmetric hydrogenation can be
realized in gas phase using SILP catalysts. [19] In this earlier publication by Öchsner et al.
we showed the successful enantioselective hydrogenation of the -ketoester methyl
acetoacetate. The proposed mechanism of BINAP-Ru catalyzed hydrogenation of -keto
esters includes the coordination of the ester group towards the Ruthenium center. [20]
During this coordination step both carbonyl groups of the -keto esters experience the chiral
coordination sphere of the Ru catalyst resulting in a high enantiomeric excess. However, for
-keto esters the two carbonyl groups point towards opposite directions, thus leading to a
different coordination within the catalytic cycle compared to the -keto esters and therefore to
a much lower enantioselectivity of the hydrogenation reaction.
2
Page 4 of 19

In the present paper, we want to extend the scope of continuous-flow, gas phase asymmetric
hydrogenation to these more sophisticated -ketoester substrates. In detail, the asymmetric
hydrogenation of methyl pyruvate (MPY) to methyl lactate (ML) is investigated. Asymmetric
hydrogenation of MPY to ML has been identified in the past as a particularly demanding task
for batch-wise operated homogeneous transition metal catalysis. Typically
enantioselectivities between 80% and 88% could be realized at temperatures up to 50°C.
[21], [22], [23], [24], [25], [26], [27], [28], [29] Heterogeneous systems applying platinum
contacts with a chiral modification by cinchonidine or quinine (see Figure 1) showed
comparable [30], [31], [32] or even better [33], [34] activities and selectivities when used in a
slurry reaction system.
Figure 1: Chiral modifier cinchonidine
In 2003, Arx et al. [35] observed for the first time moderate enantioselectivities (25-51% ee)
in the continuous gas phase hydrogenation of -keto esters using heterogeneous chirally
modified supported platinum catalysts. However, a decrease of ee was observed during
operation and even the complete loss of enantioselectivity was encountered due to partial
hydrogenation of the cinchonidine ring system at elevated temperatures. At temperatures
above 50°C, the adsorption mode of the modifier changed and even desorption took place
limiting the use of such systems for hydrogenation reactions under harsher conditions. [32]
To the best of our knowledge, asymmetric hydrogenation of -keto esters has never been
realized in homogeneous reaction systems using a continuous-flow gas phase operation
mode so far. The present publication demonstrates such a process using Ru-based SILP
catalyst systems (see Figure 2). Obvious advantages of carrying out such reaction in
3
Page 5 of 19

continuous gas-phase contact is the very simple and efficient immobilization of the catalyst
complex in a liquid matrix of negligible vapor pressure that avoids any additional synthetic
effort of chemical ligand modification or surface anchoring.
Figure 2: Supported Ionic Liquid Phase (SILP) catalyst
2. Experimental
2.1 Chemicals
The ruthenium precursor [bis(2-methylallyl)(1,5-cyclooctadiene)ruthenium(II)] as well as
methyl pyruvate (94%), methanol (99.9%, <50ppm H2O) and acetone (99.9%, <50ppm H2O)
were purchased from Acros Organics. The phosphine ligands (L) (R)-(+)-2,2'-
Bis(diphenylphosphino)-6,6'-dimethoxy-1,1'-biphenyl 97% (MeO-BIPHEP), (R)-(+)-2,2'-
Bis(diphenylphosphino)-1,1'-binaphthyl (BINAP) 98%, R-(-)-1,13-Bis(diphenylphosphino)-
7,8-dihydro-6H-dibenzo[f,h][1,5]dioxonin
(TUNEPHOS)
97%,
(R)-(+)-5,5'-
Bis(diphenylphosphino)-4,4'-bi-1,3-benzodioxole (SEGPHOS) 98%, and (+)-{4-[(1R,4S)-3-
(Diphenylphosphino)-1,7,7-trimethylbicyclo[2.2.1]hept-2-en-2-yl]-2,5-dimethyl-3-thien-3-
yl}bis(3,5-dimethylphenyl)phosphine (catASium T3) 95% were purchased from Strem
Chemicals. The monodentate ligand 4-phenyl-4,5-dihydro-3H-dinaphtho[2,1-c:1',2'-
e]phosphepine (phenylphosphepine) was provided by the group of M. Beller [36] (Leibniz-
Institut für Katalyse, University of Rostock) and 3-(2,5-(2R,5R)-dimethylphospholanyl-1)-4-di-
o-tolylphosphino-2,5-dimethylthiophene was provided by Evonik Degussa GmbH. [37] The
hydrobromic acid (48%) was purchased from Fluka. As support materials, calcinated silica
gel 100 (particle size 0.2–0.5mm, pore volume 0.98 cm3 g−1, BET surface area 359 m2 g−1,
Merck KgaA), aluminum oxide 90 (particle size 0.063–0.2mm, pore volume 0.28 cm3 g−1,
BET surface area 148 m2 g−1, Merck KgaA) and activated carbon (pore volume 0.997 cm3
g−1, BET surface area 1698 m2 g−1, Blücher). The ionic liquids 1-ethyl-3-methylimidazolium
4
Page 6 of 19

bis(trifluoromethylsulfonyl)imide [EMIM][NTf2] 99% (H2O< 500 ppm, Cl− < 100 ppm, Merck
KgaA), 1-ethyl-3-methylimidazolium trifluoromethanesulfonate [EMIM][OTf] 99% (H2O< 100
ppm, Cl− < 100 ppm, Merck KgaA) and 1-ethyl-3-methylimidazolium hexafluorophosphate
(H2O< 100 ppm, Cl− < 100 ppm, Merck KgaA) were used for the preparation of the SILP
materials as purchased. In addition, the ionic liquids tetraethylammonium
bis(trifluoromethylsulfonyl)imide [Et4N][NTf2], butylpyridinium bis(trifluoromethylsulfonyl)imide
[BPyr][NTf2], 3-hydroxypropyl pyridinium bis(trifluoromethylsulfonyl)imide [PrOHPyr][NTf2]
and 1-butyl-3-methylimidazolium hydrogensulfate [BMIM][HSO4] were synthesized in analogy
to literature procedures. [38] [39] [40] For the continuous gas phase experiments helium
(purity: 4.6, Linde AG) and hydrogen (purity: 5.0, Linde AG) were used.
2.2. Catalyst preparation
The applied general procedure to prepare the SILP materials has been published in the
literature [14],. The Ru-SILP catalysts were prepared in situ similar to literature procedure
[41] from one equivalent of [bis(2-methylallyl)(1,5-cyclooctadiene)ruthenium (II)] and two
equivalents of the bidentate phosphine ligands (four equivalents in case of the monodentate
phosphine ligand). The chemicals were placed in a Schlenk flask under argon atmosphere
and dissolved in 5 mL of dry acetone (5 mL). A methanolic solution (0.29M) of hydrobromic
acid (0.33 mL) was added and the solution was stirred 30 min at room temperature. The
solvent was removed under high vacuum over 1 h at 30°C. In this procedure the Ru-complex
RuLBr2 was obtained as a brown solid powder. For the preparation of the catalytic SILP
material, RuLBr2 was dissolved in 15 mL dry and degassed methanol and the porous support
material and the ionic liquid were added. The mixture was stirred under argon for 90 min and
subsequently the solvent was removed by means of a high vacuum treatment. The final SILP
catalyst was obtained as a yellow to brown powder.
2.3. Continuous gas phase experiments
Figure 3 shows the continuous gas phase rig applied in this work. The heart of the rig forms
the tubular reactor of 30 cm length (Vreactor= 35 mL). Helium and hydrogen were fed into the
rig via mass flow controllers (MFC (A) and (B), Bronkhorst F-201CV-200-RAD-11-V and F-
201CV-200-RAD-33-V). The carrier gas helium was used to evaporate the substrate MPY in
a CEM (controlled evaporation and mixing) unit (E) (Bronkhorst). The liquid substrates MPY
and MeOH were taken from reservoirs at room temperature and fed into the CEM-unit via
LIQUI-flows (C, D) (C: Bronkhorst L13-RAD-11-K-10S, D: Bronkhorst L13V02- RAD-11-K-
10S). The pressure was adapted by means of a back-pressure regulator valve (G). A
condenser unit and an autosampler ensured proper sampling at regular time intervals.
5
Page 7 of 19

Figure 3: Flow scheme of the applied continuous asymmetric hydrogenation reactor.
The obtained samples were diluted with diethyl ether and analyzed using gas
chromatography (Varian 3900, Lipodex E column, 25m inner diameter: 0.25mm, temperature
program 80°C for 20 min isothermal, 80°C  200°C, 10°C/min, 200°C for 10 min isothermal,
carrier gas: 0.4 mL/min helium, detector: FID, 250°C).
For the online experiments, 1.4 g of the SILP catalyst were filled into the tubular reactor (F)
under helium atmosphere, the complete rig was flushed and pressurized with 10 bar helium.
After an initial leak test, the rig was heated up to reaction temperature under continuous flow
of 20 NmL/min helium. When the reaction temperature was reached the respective flows of
helium, hydrogen, MPY and MeOH were adjusted to start the continuous experiment.
3. Results and Discussion
As a consequence of transferring the hydrogenation of MPY to the described continuous gas
phase reactor set-up, the reaction is carried out at elevated temperatures (up to 100°C) with
very short residence times (< 3 seconds). Compared to standard batch experiments these
reaction conditions are quite unusual for asymmetric hydrogenation. Therefore, our initial
experiments aimed at identifying a stable catalyst system under these demanding conditions.
For this purpose, a number of SILP materials have been prepared and screened using a
range of ligands (see Figure 4) in combination with the Ru-precursor and the ionic liquid
6
Page 8 of 19

[EMIM][NTf2]. We were especially interested in the performance of the ligand catASium-BoT
as the latter showed excellent stability in previous studies on gas phase asymmetric
hydrogenation of -keto ester methyl acetoacetate. [19] The phenylphosphepine ligand
represents the only monodentate ligand. This ligand is known from batch experiments to
form catalytic systems of enhanced activity and enantioselectivity at higher temperatures
(100-120°C). [36] The ligand structures Tunephos (dihedral angle 77°) [42] and Segphos
(dihedral angle 65°) [43] were chosen besides BINAP (73,5°) [43] to include a variation of
bite angle by means of different backbone structures.
Figure 4: Ligands screened for the asymmetric hydrogenation of MPY.
As can be seen in Figure 5, the various ligands applied in form of the SILP systems showed
very diverse activity and stability behavior. The yields of the desired enantiomer were rather
low and comparable to those obtained in gas phase reactions using chirally modified
heterogeneous systems. [35] While most of the systems experienced an initial activation
behavior, the SILP catalyst with catASium-T3 ligand showed very high activity right from the
start. However, similar to the other ligands, the catalyst significantly decreased in activity
after 20 to 30 hours time on stream. The generally low enantiomeric excess of all ligands and
the decrease over time on stream is attributed to the elevated temperatures and an
associated slow degradation of the chiral ligand. As already mentioned in the introduction,
the low enantiomeric excess is also an intrinsic problem of the asymmetric hydrogenation of
-keto esters. The ligand system with the best stability among the screened systems was
found to be the SILP system with the BINAP ligand. TOF values of this catalyst where in the
7
Page 9 of 19

-1
range of 15 to 22 molML molcat h-1 with a TON of 1037 after 50 hours time on stream.
Therefore this ligand was chosen for the following investigations.
Figure 5: Yield vs. time-on-stream profiles of the SILP catalyzed MPY hydrogenation in a continuous gas-
phase reaction - variation of chiral ligand. (Reaction conditions: L/Ru = 2:1, ionic liquid = [EMIM][NTf2],
 = 0.8, silica 100 (0.2–0.5 mm), 1 wt.% Ru, 95°C, 10 bar, 200 NmL/min He, 75 NmL/min H2, 0.25 g/h MPY,
0.475 g/h MeOH,  = 2.86 s)
One great advantage of continuous-flow, gas phase processes in catalysis is that detailed
kinetic investigations can be carried out very efficiently. Due to the short residence time and
absence of diluting solvents steady-state operation can be reached quickly allowing for very
precise measurement of kinetic data for various reaction conditions within one catalytic run.
In such a kinetic run (see Figure 6) the activation energy of the SILP catalyst with BINAP was
determined to be 82.6 kJ/mol. Additionally, partial reaction orders for hydrogen and MPY
were found to be 0.98 and 0.6, respectively. Slightly higher enantioselectivities were obtained
at lower temperatures and a visible drop in ee and a stronger deactivation was observed for
temperatures above 80°C. The activity dropped from 75% yield at 78% conversion after 40
hours time on stream to about 56% yield at 60% conversion after 200 hours time on stream.
In the same time span the enantiomeric excess decreased from 25% ee to about 10% ee.
Only minor amounts of sideproducts (≤ 3%), mainly aldol, hemi-ketal and ketal byproducts
were observed in all experiments. Moreover, the formation of high-boiling aldol or ketal
products could lead to a significant loss in catalytic activity and selectivity due to
accumulation in the SILP pore system.
8
Page 10 of 19

Figure 6: Yield vs. time-on-stream profile of the SILP catalyzed MPY hydrogenation in a continuous gas
phase reaction – kinetic investigation. (Reaction conditions: L/Ru = 2:1, ligand: BINAP, ionic liquid =
[EMIM][NTf2],  = 0.8, silica 100 (0.2–0.5 mm), 1 wt.% Ru, 75-105°C, 10 bar, 200-250 NmL/min He, 75-100
NmL/min H2, 0.25-0.75 g/h MPY, 0.475 g/h MeOH,  = 2.84-2.86 s; partial pressures: A: pH2 = 3.78 bar, B: pH2
= 3.26 bar, C: pH2 = 4.92 bar, D: pH2 = 3.78 bar pMPY = 0.028 bar, E: pMPY = 0.055 bar, F: pMPY = 0.082 bar, G:
pMPY = 0.028 bar)
Further experiments aimed at the investigation of the influence of the support material, the
ionic liquid and the ionic liquid loading. Aluminum oxide and activated carbon were
investigated in addition to the silica 100 support material (see Figure 7). While ee values
were a bit higher on activated carbon, deactivation appeared after 30 to 40 hours time on
stream. Additionally leaching of the ionic liquid layer onto the glass wool below the catalyst
bed was observed for both carbon supports. Therefore the adsorption of the ionic liquid layer
seemed to be tighter on an inorganic support with surface oxygen and hydroxyl groups.
Additionally, the observed catalytic activities seem to correlate with the pore volume and
surface area of the support materials (see experimental section). Since the ionic liquid
loading  (VIL/Vpore) was kept constant, less IL is used for a support with a lower pore volume.
Aluminium oxide exhibits by far the lowest pore volume (0.28 cm3/g) and pore surface area
(148 cm3 g−1) of the support materials under investigation and therefore also experiences the
lowest catalytic activity. In previous studies surface silanol groups [19], protic solvents [44]
and also acidic additives [45] were found to increase the activity in the asymmetric
hydrogenation of methyl acetoacetate. These influences account for the high activity of the
SILP catalyst with silica support. The lower enantiomeric excess, however, could correlate to
9
Page 11 of 19

an undesired, competitive coordination of the phosphine ligand to the silica surface. This
effect was also observed for phosphine ligands in hydroformulation studies using silica
supported SILP catalysts. [14], [46]
Figure 7: Yield vs. time-on-stream profiles of the SILP catalyzed MPY hydrogenation in a continuous gas-
phase reaction - variation of support. (Reaction conditions: ligand = Xantphos, L/Ru = 2:1, ionic
liquid = [EMIM][NTf2],  = 0.8, 1 wt.% Ru, 95°C, 10 bar, 200 NmL/min He, 75 NmL/min H2, 0.25 g/h MPY,
0.475 g/h MeOH,  = 2.86 s)
Since the ionic liquid constitutes the solvent in the SILP catalysts it can have a tremendous
effect on activity and stability. [16], [17], [47] The viscosity, polaritiy or coordination strength
of the anion can be of particular importance for the respective reaction system. [13] Figure 8
shows activity and selectivity profiles of SILP catalysts with different ionic liquid anions. The
catalytic activity and selectivity is in line with the coordination strenght of the anion indicating
the importance of weakly coordinating anions. Stronger coordinating anions like [OTf]-
compete with the chiral ligand when coordinating to the metal and could lead to a less active
and selective system. Furthermore, a rather strong decrease in enantiomeric excess is
observed for the SILP system with [EMIM][PF6] after 30 hours time on stream. This decrease
could be associated to the hydrolysis of the anion and the production of HF which has been
observed previously. [13] Water can be formed as co-product from the side reactions aldol
condensation and ketal formation. Though being rather low in all ILs under investigation, the
hydrogen solubility is generally higher for [NTf2]- based ionic liquids compared to [OTf]- and
[PF6]- type ILs. This could additionally enhance the activity of the [EMIM][NTf2] SILP catalyst.
[48], [49]
10
Page 12 of 19

Figure 8: Yield vs. time-on-stream profiles of the SILP catalyzed MPY hydrogenation in a continuous gas-
phase reaction - variation of ionic liquid anion. (Reaction conditions: ligand = Xantphos, L/Ru = 2:1,
 = 0.8, silica 100 (0.2–0.5 mm), 1 wt.% Ru, 95°C, 10 bar, 200 NmL/min He, 75 NmL/min H2, 0.25 g/h MPY,
0.475 g/h MeOH,  = 2.86 s)
The influence of different cations on the reactivity and selectivity was studied as well. Floris
et al. [50] utilized tetraalkylammonium ionic liquids in combination with methanol and found
good activities and ee values in the asymmetric hydrogenation of methyl acetoacetate.
However, when varying the amount of ionic liquid from 0.2 wt.% up to 100 wt.% a dramatic
decrease in activity and enantioselectivity was observed with increasing amount of ionic
liquid. This observation is in line with the work of Wolfson et al. on the importance of
alcoholic solvents [44] and acidic additives [45] in the asymmetric hydrogenation of -keto
esters. Alcoholic solvents and acidic additives were found to facilitate the protonation of the
product and hence increase the reaction rate. These findings were taken into account for the
ionic liquid cation screening. As shown in Figure 9 the tetraethylammonium ionic liquid
showed a significantly lower activity than the other ionic liquid cations under investigation.
Pyridinium based ionic liquids showed similar and even better activity and better stability in
terms of enantioselectivity. More protic solvents resulted in SILP catalysts with higher
activity. Particularly, this is observed for the protic ionic liquid [PrOHPyr][NTf2], which yielded
the highest activity and enantioselectivity. The protic character of the ionic liquid seems to be
of similar importance as observed in systems with molecular solvents [44].
To verify the effect of increasing ionic liquid loadings on catalyst activity compared to the
literature reports for liquid phase catalysis [50], the ionic liquid loading of the SILP catalysts
under investigation was varied.
11
Page 13 of 19

Figure 9: Yield vs. time-on-stream profiles of the SILP catalyzed MPY hydrogenation in a continuous gas-
phase reaction - variation of ionic liquid cation. (Reaction conditions: ligand = Xantphos, L/Ru = 2:1,
 = 0.8, silica 100 (0.2–0.5 mm), 1 wt.% Ru, 95°C, 10 bar, 200 NmL/min He, 75 NmL/min H2, 0.25 g/h MPY,
0.475 g/h MeOH,  = 2.86 s)
Figure 10 shows the variation of the ionic liquid loading . In contrast to the results in liquid
phase asymmetric hydrogenations [50], increased amounts of ionic liquid were found to
enhance catalytic activity, stability and selectivity. This stabilizing effect of higher ionic liquid
loadings was also observed in our previous studies on the asymmetric hydrogenation of the
-keto ester methyl acetoacetate. [19] Obviously, in the SILP system, the ionic liquid provides
a solvation layer for the asymmetric catalyst complex and a higher amount of IL stabilizes the
catalyst throughout the reaction in a better way. Notably, no mass transport limitation effects
were observed for high  values, which is due to the relatively slow reaction rate of the
asymmetric hydrogenation reaction. Another important result of the IL loading variation is the
fact that the relatively long catalyst activation time is independent on the amount of ionic
liquid loading. We assume that the extended time to full catalyst activation is linked to the low
equilibrium solubility of hydrogen in the applied ionic liquid leading to a slow formation of the
active species from the precursor complex.
12
Page 14 of 19

Figure 10: Yield vs. time-on-stream profiles of the SILP catalyzed MPY hydrogenation in a continuous
gas-phase reaction - variation of ionic liquid loading. (Reaction conditions: ligand = Xantphos, L/Ru = 2:1,
ionic liquid = [PrOHPyr][NTf2],  = 0 - 0.8, silica 100 (0.2–0.5 mm), 1 wt.% Ru, 95°C, 10 bar, 200 NmL/min
He, 75 NmL/min H2, 0.25 g/h MPY, 0.475 g/h MeOH,  = 2.86 s)
Finally, the influence of acidic additives was investigated since activity enhancing effects of
Brønsted acids had been reported for liquid phase systems. [45] For an addition of p-
toluenesulfonic acid as well as for the addition of the acidic ionic liquid [BMIM][HSO4] an
increase in catalytic activity was observed, while the enantiomeric excess slightly decreased
over time on stream (see Figure 11). Though the system with p-toluenesulfonic acid
achieved highest yields of about 95%, it deactivated more rapidly than the system with the
additive [BMIM][HSO4]. For the latter, yields of 88.7% and ee values of 25% were realized
over more than 50 hours time. Turn over frequencies were in the range of 25 molML molRu-1 h-
1
resulting in an overall TON of 1100. Hence, a moderate acidity of support material, ionic
liquid and acidic additives results in an active and quite stable SILP catalyst system.
Moderate but quite stable enantioselectivities were obtained in the continuous-flow, gas
phase asymmetric hydrogenation of the -keto ester methyl pyruvate.
13
Page 15 of 19

Figure 11: Yield vs. time-on-stream profiles of the SILP catalyzed MPY hydrogenation in a continuous
gas-phase reaction – addition of acidic additives. (Reaction conditions: ligand = Xantphos, L/Ru = 2:1,
ionic liquid = [PrOHPyr][NTf2],  = 0.8, acidic additive/Ru = 10/1 silica 100 (0.2–0.5 mm), 1 wt.% Ru, 95°C,
10 bar, 200 NmL/min He, 75 NmL/min H2, 0.25 g/h MPY, 0.475 g/h MeOH,  = 2.86 s)
4. Conclusion
In conclusion we have successfully demonstrated the application of SILP catalysts in the gas
phase asymmetric hydrogenation of -keto esters, a substrate class that is known to be
difficult to hydrogenate with high enantioselectivities. More precisely, the enantioselective
hydrogenation of methyl pyruvate was demonstrated utilizing Ru-based SILP catalysts.
Various ligands, support materials and ionic liquids were screened and evaluated with
respect to catalyst activity, selectivity and stability. With Ru-BINAP dissolved in the ionic
liquids 3-hydroxypropylpyridinium bis(trifluoromethylsulfonyl)imide on silica 100 an
enantiomeric excess of 26-30% could be obtained at 80-84% yield for more than 50 hours
time on stream. Moreover, high degrees of pore filling are necessary to achieve reasonable
enantioselectivity and good stability. Moderate acidity using silica support material and
adding p-toluenesulfonic acid or 1-butyl-3-methylimidazolium hydrogen sulfate increased the
product yield up to 95% yield. The reported SILP catalysts offer a simple yet very elegant
way to ensure proper catalyst/product separation and allow for catalyst recycling without any
need for ligand modifications.
Acknowledgements
This work was started in the context of the BMBF-sponsored network “Supported Ionic Liquid
Phase catalysis” and we thank our partners LIKAT (Prof. Beller) and Evonik (Dr. Kadyrov) for
ligand donations within this project. TWe gratefully thank the Deutsche
Forschungsgemeinschaft within the Erlangen Excellence Cluster “Engineering of Advanced
Materials” and the project was continued as part of the European Community's Seventh
Framework Programme [FP7/2007-2013] for SYNFLOW under grant agreement n° NMP2-
14
Page 16 of 19

LA-2010-246461 and this support is gratefully acknowledged. Infrastructural support was
granted by the Deutsche Forschungsgemeinschaft through the Erlangen Excellence Cluster
“Engineering of Advanced Materials”.
References
[1] A. D. Curzons, V. L. Cunningham D. J. C. Constable, Green. Chem., vol. 4, pp. 521-527,
2002.
[2] H. van Bekkum R. A. Sheldon, Fine Chemicals through Heterogeneous Catalysis.:
Wiley-VCH, Weinheim, 2001.
[3] R. A. Skilton, A. J. Parrott, D. J. Irvine, M. Poliakoff R. A. Bourne, Org. Process Res.
Dev., vol. 4, pp. 932-938, 2011.
[4] P. Laurino, P. H. Seeberger X. Y. Mak, Beilstein J. Org. Chem., vol. 5(19), 2009.
[5] S. Lee C. E. Song, Chem. Rev., vol. 102, pp. 3495-3524, 2002.
[6] G. J. Hutchings P. McMorn, Chemical Society Reviews, vol. 33, pp. 108-122, 2004.
[7] J. I. Garcia, J. A. Mayoral J. M. Fraile, Chem. Rev., vol. 109, pp. 360–417, 2009.
[8] M. A. Wegman, J. Rops, F. van Rantwijk, R. A. Sheldon M. A. O. J. Hacking, Journal of
Molecular Catalysis B: Enzymatic, vol. 5, pp. 155-157, 1998.
[9] D. P. Curran J. A. Glaadysz, , vol. 58, pp. 3823-3825, 2002.
[10] P. M. P. Gois, C. A. M. Afonso A. F. Trindade, Chem. Rev., vol. 109, pp. 418–514, 2009.
[11] R. A. Cook, N. C. Dispenziere, M. Afeworki C. P. Mehnert, J. Am. Chem. Soc., vol. 124,
pp. 12932-12933, 2002.
[12] P. Wasserscheid, R. van Hal, R. Fehrmann A. Riisager, J. Catal., vol. 219(2), pp. 452-
455, 2003.
[13] T. Welton P. Wasserscheid, „Ionic Liquids in Synthesis“, 2nd ed.: Wiley-VCH, Weinheim,
2007.
[14] R. Fehrmann, S. Flicker, R. van Hal, M. Haumann, P. Wasserscheid A. Riisager, Angew.
Chem. Int. Ed., vol. 44, pp. 815–819, 2005.
[15] B. Jørgensen, P. Wasserscheid, R. Fehrmann A. Riisager, Chem. Commun., pp. 994–
996, 2006.
[16] T.E. Müller, C. Sievers, A. Spirkl, J.A. Lercher O. Jimenez, Chem. Commun., pp. 2974–
2976, 2006.
[17] G. Laurenczy, P.J. Dyson, L. Kiwi-Minsker M. Ruta, J. Phys. Chem. C, vol. 112, pp.
15
Page 17 of 19

17814–17819, 2008.
[18] N. Szesni, A. Bittermann, M.J. Schneider, P. Härter, M. Haumann, P. Wasserscheid S.
Werner, Appl. Catal. A, vol. 377, pp. 70–75, 2010.
[19] M. J. Schneider, C. Meyer, M. Haumann, P. Wasserscheid E. Öchsner, Applied
Catalysis A-general , vol. 399(1), pp. 35-41, 2011.
[20] R. Noyori, Angew. Chem. Int. Ed., vol. 41, pp. 2008-2022, 2002.
[21] T. Morimoto, K. Achiwa H. Takahashi, Chemistry Letters, pp. 855-858, 1987.
[22] K. Kusano, N. Sate, Y. Matsumura, K. Nozaki, H. Kumobayashi, N. Sayo, Y. Hori, T.
Ishizaki, S. Akutagawa, H. Takaya K. Mashima, J. Org. Chem., vol. 59, pp. 3064-3076,
1994.
[23] A. Pizzano, J. A. Martin M. J. Burk, Organometallics, vol. 19, pp. 250-260, 2000.
[24] I. Suisse, A. Mortreux, F. Agbossou, M. A. Ali S. Naili, Tetrahedron Letters, vol. 41, pp.
2867–2870, 2000.
[25] S. D. Debenham, E. B. Mackenzie, S. E. Large N. W. Boaz, Org. Lett., vol. 4(14), pp.
2421-2424, 2002.
[26] E. B. Mackenzie, S. D. Debenham, S. E. Large, J. A. Ponasik Jr N. W. Boaz, J. Org.
Chem., vol. 70, pp. 1872-1880, 2005.
[27] M. Alame, S. Pellet-Rostaing, M. Lemaire M. Jahjah, Tetrahedron: Asymmetry , vol. 18,
pp. 2305–2312, 2007.
[28] L. Xu, L. Feng, J. Ruan, Q. Fan, A. S.C. Chan K. H. Lam, Can. J. Chem. , vol. 83, pp.
903–908, 2005.
[29] W. Li, L. Zhou, X. Zhang X. Sun, Chem. Eur. J., vol. 15, pp. 7302 – 7305, 2009.
[30] A. G. Wilkinson P. B. Wells, Topics in Catalysis, vol. 5, pp. 39–50, 1998.
[31] H.-P. Jalett, M. Mtiller, M. Studer H.-U. Blaser, Catalysis Today , vol. 37, pp. 441--463,
1997.
[32]
P. Mäki‐Arvela, E. Toukoniitty, T. Salmi D. Yu. Murzin, Catalysis Reviews, vol. 47, pp.
175–256, 2005.
[33] H. Liu, M. Liu X. Zuo, Tetrahedron Letters , vol. 39, pp. 1941-1944, 1998.
[34] K. Balazsik, K. Felfödi, M. Bartok B. Török, Ultrasonics Sonochemistry, vol. 8, pp. 191-
200, 2001.
[35] N. Dummer, D. J. Willock, S. H. Taylor, R. P. K. Wells M. von Arx, Chem. Commun., pp.
16
Page 18 of 19

1926–1927, 2003.
[36] K. Junge, S. Enthaler, M. Michalik, T. Riermeier, A. Monsees, M. Beller B. Hagemann,
Adv. Synth. Catal. , vol. 347, pp. 1978 – 1986, 2005.
[37] R. Kadyrov, Chimica Oggi, no. 26(5), pp. 41-42, 2008.
[38] D.R. MacFarlane, M. Forsyth J. Sun, Ionics, no. 3, pp. 356-362, 1997.
[39] R. D. Singer, M. T. Garcia, P. J. Scammells J. R. Harjani, Green Chem., no. 11, pp. 83-
90, 2009.
[40] L. M. Ramenskaya, U. V. Chervonova, R. S. Kumeev M. S. Gruzdev, Russ. J. Gen.
Chem., no. 17(8), pp. 1720-1727, 2009.
[41] T. Ohkuma, M. Kitamura, H. Takaya, N. Sayo, H. Kumobayashi, S. Akutagawa R.
Noyori, J. Am. Chem. Soc. , vol. 109, pp. 5856–5858, 1987.
[42] H. Qian, J. Longmire, X. Zhang Z. Zhang, J. Org. Chem. , vol. 65, pp. 6223-6226, 2000.
[43] I. Nagasaki, K. Matsumura, N. Sayo, T. Saito H. Shimizu, Acc. Chem. Res. , vol. 40(12),
pp. 1385–1393..
[44] I F.J. Vankelecom, S. Geresh, P. A. Jacobs A. Wolfson, Journal of Molecular Catalysis
A: Chemical , vol. 198, pp. 39–45, 2003.
[45] I F.J. Vankelecom, S. Geresh, P. A. Jacobs A. Wolfson, Journal of Molecular Catalysis
A: Chemical, vol. 217, pp. 21–26, 2004.
[46] D. Hanna, S. Werner, A. T. Bell S. Shylesh, ACS Catal. , vol. 2, p. 487−493, 2012.
[47] M. Haumann, P. Wasserscheid S. Werner, Annu. Rev. Chem. Biomol. Eng. , vol. 1, pp.
203-230, 2010.
[48] G. Laurenczy, C. A. Ohlin, J. Vallance, T. Welton P. J. Dyson, Chem. Commun., pp.
2418–2419, 2003.
[49] J. L. Anderson, E. J. Maginn, J. F. Brennecke J. L. Anthony, J. Phys. Chem. B, vol. 109
(13), pp. 6366-6374, 2005.
[50] P. Kluson, L. Bartek, H. Pelantova T. Floris, Applied Catalysis A: General , vol. 366, pp.
160–165, 2009.
17
Page 19 of 19