Dramatic Pressure Effects on the Selectivity of the Aqueous/Organic Biphasic Hydrogenation of trans-Cinnamaldehyde Catalyzed by Water-Soluble Ru(II)-Tertiary Phosphane Complexes

Article abstract of DOI:10.1002/adsc.200390006

In a water/chlorobenzene biphasic reaction, the hydrogenation of trans-cinnamaldehyde, catalyzed by water-soluble Ru(II)-phosphane complexes at pH 3.04 (phosphate buffer), produces a 61:39 mixture of cinnamyl alcohol and dihydrocinnamaldehyde at 1 bar H₂; however, the selectivity is increased to 93:7 by increasing the hydrogen pressure to 8 bar.

Full text of DOI:10.1002/adsc.200390006

COMMUNICATIONS
Dramatic Pressure Effects on the Selectivity of the Aqueous/
Organic Biphasic Hydrogenation of trans-Cinnamaldehyde
Catalyzed by Water-Soluble Ru(II)-Tertiary Phosphane
Complexes
a
a
b
c
a, b,
G a¬ bor Papp, J a¬ nos Elek, Levente N a¬ dasdi, G a¬ bor Laurenczy, Ferenc Jo o¬ *
a
Research Group of Homogeneous Catalysis, Hungarian Academy of Sciences, Debrecen 10, P.O.B. 7, 4010 Hungary
Fax: ( ‡ 36)-52-512915, e-mail: jooferenc@tigris.klte.hu
b
Institute of Physical Chemistry, University of Debrecen, Debrecen 10, P.O.B. 7, 4010 Hungary
c
Institut de Chimie Mol e¬ culaire et Biologique, Ecole Polytechnique F e¬ d e¬ rale de Lausanne, 1015 Lausanne, Switzerland
Received: August 2, 2002; Accepted: October 14, 2002
[
RuH (mTPPMS) ] proved to be a selective catalyst
2 4
Abstract: In a water/chlorobenzene biphasic reac-
tion, the hydrogenation of trans-cinnamaldehyde,
catalyzed by water-soluble Ru(II)-phosphane com-
plexes at pH3.04 (phosphate buffer), produces a
[7]
for the hydrogenation of the CˆO unit in aldehydes.
In addition to allowing the mild reduction of unsatu-
rated aldehydes with a pH-controlled selectivity, the
above findings also gave rationale to the selectivity
observed with aqueous sodium formate as reductant, in
which case the solutions typically had a pHaround 8.
Nevertheless, these observations were somewhat incon-
sistent with other literature reports. Namely, Grosselin
and Mercier, as well as Hernandez and Kalck have
found that crotonaldehyde and trans-cinnamaldehyde
could be hydrogenated to crotyl and cinnamyl alcohol,
respectively, with a selectivity up to 96% in unbuffered
aqueous-organic biphasic systems. The catalyst was
either prepared in situ from RuCl ¥ 3 H O and mTPPTS,
6
1:39 mixture of cinnamyl alcohol and dihydrocin-
namaldehyde at 1 bar H ; however, the selectivity is
2
increased to 93:7 by increasing the hydrogen pres-
sure to 8 bar.
[
8,9]
[10]
Keywords: alcohols; aldehydes; allylic compounds;
hydrogenation; regioselectivity; ruthenium com-
plexes
3
2
or added as [RuClH(mTPPTS)3] and cis-
[
RuH (mTPPTS) ], respectively [mTPPTS ˆ tris(m-
2
4
Aqueous/organic biphasic reactions are increasingly sulfonatophenyl)phosphane]. Since metal ion hydroly-
attractive in view of catalyst recycling and clean product sis and the reduction of Ru(III) to Ru(II) by mTPPTS
isolation.[
1,2]
Earlier, one of us has reported
[3,4]
the result in proton formation, unbuffered solutions of the
selective hydrogenation of unsaturated aldehydes to in situ prepared Ru- mTPPTS complexes are always
unsaturated alcohols with aqueous sodium formate as somewhat acidic, usually in the pHrange of 3 ± 5
hydrogen source and [{RuCl (mTPPMS) } ] ‡ excess depending on the metal concentration. According to
2
2 2
mTPPMS as catalyst (mTPPMS ˆ meta-monosulfonat- the results of our combined pH-potentiometric/NMR
ed triphenylphosphane). This aqueous/organic biphasic spectroscopic studies, in this pHrange the formation of
method allowed facile preparation of allylic alcohols in [RuClH(mTPPTS) ] and, consequently, that of the
3
close to quantitative yields. The outstanding selectivity saturated aldehyde products could be expected. (This
was rather unexpected, since one of the hydrogena- conclusion relies on the close similarity in the coordi-
tion products of [{RuCl (mTPPMS) } ], namely nation ability of mTPPMS and mTPPTS, which ± at least
2
2 2
[
RuClH(mTPPMS) ], had been previously found to in the case of hydridophosphaneruthenium(II) com-
3
[
11,12]
be an active catalyst for the hydrogenation of plexes ± seems well established.
) The only major
olefins.[ Detailed NMR spectroscopic studies have difference in the reaction parameters was in the hydro-
revealed that in the presence of excess phosphane gen pressure: our studies were done at 1 bar, while those
the hydrogenation of [{RuCl (mTPPMS) } ] yields in the literature at 30 and 40 bar, respectively. We
5]
2
2 2
[
[
RuClH(mTPPMS) ] at low pH( <6) and cis- suspected that at sufficiently high pressures the in-
3
[
6]
RuH (mTPPMS) ] at high pH( >8). It was also creased hydrogen concentration in the solution can shift
2
4
established that at atmospheric H -pressure the equilibria of Equation (1) and Equation (2) in
2
[
RuClH(mTPPMS) ] catalyzed the selective hydroge- favour of the formation of cis-[RuH (mTPPMS) ]
3 2 4
nation of olefinic CˆC bonds, while cis- even in spite of the acidic conditions, and by this,
promote the formation of unsaturated alcohols.
172
¹ 2003 WILEY-VCHVerlag GmbH& Co. KGaA, Weinheim 1615-4150/03/3451+2-172 ± 174 $ 17.50-.50/0
Adv. Synth. Catal. 2003, 345, No. 1+2

Dramatic Pressure Effects on Selectivity of Biphasic Hydrogenation
COMMUNICATIONS
1/2
[{RuCl (mTPPMS) } ] + mTPPMS + H
2 2 2 2
+
–
[
RuClH(mTPPMS) ] + H + Cl
3
ꢀ
1†
[
RuClH(mTPPMS) ] + mTPPMS + H2
3
+
–
cis-[RuH (mTPPMS) ] + H + Cl
2
4
ꢀ
2†
Based on this assumption, the selectivity of the hydro-
genation of trans-cinnamaldehyde in water/chloroben-
zene biphasic reactions was studied asthe function of the
hydrogen pressure at constant pH(phosphate buffer) in
the acidic region. The pH-dependence of the reaction
was also reinvestigated, not only at 1 bar, but at 10 bar
Figure 1. The selectivity (S%) of the biphasic hydrogenation
of trans-cinnamaldehyde as a function of hydrogen pressure.
pHof the aqueous phase: 3.04, other parameters as in
Experimental Section. S% ˆ 100[(alc À ald)/(alc ‡ ald)],
alc ˆ cinnamyl alcohol, ald ˆ dihydrocinnamaldehyde.
H , as well. In all cases the reactions yielded only
2
cinnamyl alcohol and dihydrocinnamaldehyde, and no order dependence on P(H ) with related catalysts. It is
2
formation of the fully saturated product was observed more conceivable, therefore, that the increase in hydro-
(
Scheme 1).
gen pressure changes the molecular distribution of
ruthenium among its various hydride species, as shown,
for example, in reactions (1) and (2). At the reaction
H2
[
{RuCl2(TPPMS)2}2]
mTPPMS
temperature (80 8C), the solubility of H in water under
CH2OH
CHO
2
CHO
Ph
+
Ph
[16]
Ph
1 ± 10 bar pressure is approximately 0.7 ± 7 mM, which
water/chlorobenzene
0 °C
8
is in the range of the proton concentration in the solution
at pH3, so changes of the pressure can easily effect a
mobile equilibrium. Indeed, the cherry-red solution at
Scheme 1.
The pH-dependence measurements showed the ex- 1 bar H , containing [RuClH(mTPPMS) ], becomes
2
3
pected trend, i.e., an increasing yield of the unsaturated yellow at 10 bar, and cis-[RuH (mTPPMS) ] also has a
2
4
alcohol and a decreasing yield of the saturated aldehyde strong yellow colour. UV-visible spectrophotometric
with the increase in the pH. This trend was much more measurements in the range of 400 ± 600 nm under
expressed at 10 bar, where already at pH2.9 more than
increasing hydrogen pressure confirmed the reaction
9
0% of the products was represented by cinnamyl of [RuClH(mTPPMS) ] and H .
3
2
alcohol. At the same pH, but at 1 bar H , the concen-
This deceptive coincidence of colours, UV-vis spectra,
2
tration of cinnamyl alcohol (alc) and dihydrocinnamal- equilibria and reactivity notwithstanding, the product of
1
dehyde (ald) were roughly the same (51% and 49%). the reaction is not cis-[RuH (mTPPMS) ]. HNMR
2
4
The effect of pressure on the selectivity, defined as spectra of [{RuCl (mTPPMS) } ] ‡ excess mTPPMS
2
2 2
oS %n Fˆig 1u 0r e0 [ 1( a. lc ± ald)/(alc ‡ ald)] at pH ˆ 3.04 is shown under > 5 bar hydrogen pressure did not show the
characteristic signal of cis-[RuH (mTPPMS) ] (d ˆ
2
4
It is seen that the selectivity readily increases with the À 10.3 ppm, pseudo-quartet), instead, a new signal
increasing pressure and already at 8 bar the product appeared, which could be best interpreted assuming
[
17]
1
1
mixture contains 93% cinnamyl alcohol (S ˆ 86%). This
a
trans-[RuH (mTPPMS) ] structure
( H d ˆ
2
4
31
shows that a relatively small increase in the pressure À 7.7 ppm, quintet, JP-H ˆ 8.6 Hz;
H{ P} d ˆ
3
1
1
changes the selectivity of this hydrogenation dramati- À 7.7 ppm, s; P{ H} d ˆ 58.5 ppm, br s). Despite many
cally and makes a highly selective reaction from a non- efforts, at present we do not have any definitive
selective one.
structural evidence for this compound, except that it
Marked pressure effects on selectivity are not unusu- does not seem to be a non-classical Ru-hydride (T1/
[
13,14]
al, especially in enantioselective hydrogenations.
400 MHz 430 ms at 40 8C). Further studies of this new
However, these effects are mainly observed in a wider compound are underway, and the results will be
pressure range. In our case, as well, one possible reported in due course. Nevertheless, based on the
explanation could be a different pressure dependence pressure effect, it cannot be excluded that minor
of the rate of CˆC hydrogenation, catalyzed by species containing molecular H , such as, e.g.,
2
‡
[
RuClH(mTPPMS) ] and that of CˆO hydrogenation, [RuH(H )(TPPMS) ] , play role either in the cis/trans
3
2
4
catalyzed by cis-[RuH (mTPPMS) ]. However, both isomerization of [RuH (mTPPMS) ] or in the catalysis
2
[15]
4
2
4
olefin[ and aldehyde hydrogenations showed first of trans-cinnamaldehyde hydrogenation itself. Non-
3]
Adv. Synth. Catal. 2003, 345, 172 ± 174
173

COMMUNICATIONS
G a¬ bor Papp et al.
classical phosphane-ruthenium(II) hydrides have been organic phase. Occasional use of an internal standard (naph-
thalene) showed that no loss of material occurred.
already suggested to be involved in the selectivity
changes with H pressure in the hydrogenation vs.
2
reductive imination of nitriles, albeit no such complexes
were characterized in that system.
[
18]
Acknowledgements
In conclusion, we have shown, that the selectivity of
trans-cinnamaldehyde hydrogenation in aqueous-or-
ganic biphasic systems with water-soluble Ru(II)-phos-
phane catalysts is strongly sensitive to the hydrogen
pressure and that this sensitivity depends on the pHof
the aqueous phase. Such effects are due to the low
Financial support by the National Science Fund of Hungary
(
OTKA T029934) and by the Swiss National Science Founda-
tion (grant 2100-061653) is gratefully acknowledged. We thank
Johnson Mathey Plc for a loan of hydrated RuCl . G.P. and J.E.
are grateful for travel grants within the COST D10/WG 0001
collaboration.
3
solubility of H in water and the acid-base equilibria of
2
transition metal hydrides in this solvent. The results call
attention to the special phenomena which may be
observed in aqueous organometallic catalysis in com-
parison with related reactions in purely organic solvents.
References and Notes
[
[
[
[
1] F. Jo o¬ , Aqueous Organometallic Catalysis, Kluwer,
Dordrecht, 2001.
2] B. Cornils, W. A. Herrmann, Eds., Aqueous-Phase Orga-
nometallic Catalysis, Wiley-VCH, Weinheim, 1998.
3] F. Jo o¬ , A. B e¬ nyei, J. Organomet. Chem. 1989, 363, C19 ±
C21.
Experimental Section
General Remarks
4] A. B e¬ nyei, F. Jo o¬ , J. Mol. Catal. 1990, 58, 151 ± 163.
mTPPMS and [{RuCl (mTPPMS) } ] were prepared according
¬
¬
2
2 2
[5] Z. Toth, F. Joo, M. T. Beck, Inorg. Chim. Acta 1980, 42,
53 ± 161.
[19]
to published procedures. trans-Cinnamaldehyde and chlor-
1
obenzene (both from Aldrich) were distilled prior to use.
¬
[
6] F. Jo o¬ , J. Kov a¬ cs, A. Cs. B e¬ nyei, A. Kath o¬ , Angew. Chem.
Doubly distilled water was used throughout. H and Ar were
2
1998, 110, 1024 ± 1026; Angew. Chem. Int. Ed. 1998, 37,
1
31
acquired from Messer. Hand P NMR spectra were recorded
on Bruker AM360 and DRX400 instruments in D O or D O/
9
69 ± 970.
2
2
¬
[
[
[
7] F. Jo o¬ , J. Kov a¬ cs, A. Cs. B e¬ nyei, A. Kath o¬ , Catalysis
Today 1998, 42, 441 ± 448.
8] J. M. Grosselin, C. Mercier, J. Mol. Catal. 1990, 63, L25 ±
L27.
9] J. M. Grosselin, C. Mercier, G. Allmang, F. Grass,
Organometallics 1991, 10, 2126 ± 2133.
H O mixtures, using a thick-wall glass NMR tubes with a teflon
2
valve (Aldrich; up to 10 bar) or medium-pressure sapphire
NMR tubes (up to 100 bar). Chemical shifts are referenced to
3-(trimethylsilyl)-1-propanesulfonic acid Na salt (TSPSA,
Fluka) and 85% H PO , respectively. GC analyses of hydro-
3
4
genation product mixtures were carried out on a Chrom 5
chromatograph (2 m Carbowax 20M on 80/100 Chromosorb/
[
[
[
10] M. Hernandez, Ph. Kalck, J. Mol. Catal. A. 1997, 116,
31 ± 146.
11] E. Fache, C. Santini, F. Senocq, J. M. Basset, J. Mol.
Catal. 1992, 72, 331 ± 336.
1
3.5% KOHcolumn, FID).
12] M. Hernandez, Ph. Kalck, J. Mol. Catal. A. 1997, 116,
Hydrogenation Experiments
1
17 ± 130.
À5
1
1
9 mg (2 ¥ 10 mol) [{RuCl (mTPPMS) } ] and 22.8 mg (5.7 ¥
0
2
2 2
[13] R. Noyori, Asymmetric Catalysis in Organic Synthesis,
John Wiley & Sons, New York, 1994, pp. 32 ± 44.
À5
mol) mTPPMS were placed into a home-made heavy-
walled glass reactor equipped with a gas inlet, pressure gauge
and an inlet/sampling port. After closing the reactor it was
deoxygenated with several evacuation/refill (Ar) cycles. Then
[
[
[
14] Y. Sun, R. N. Landau, J. Wang, C. LeBlond, D. G.
Blackmond, J. Am. Chem. Soc. 1996, 118, 1348 ± 1353.
15] E. Fache, C. Santini, F. Senocq, J. M. Basset, J. Mol.
Catal. 1992, 72, 337 ± 350.
16] W. F. Linke, A. Seidell, Solubilities of Inorganic and
Metal-Organic Compounds, American Chemical Society,
Washington, D. C., 1958, Vol. 1, 1075.
3.0 mL of an aqueous phosphate buffer of the desired pHwere
injected through the inlet port, followed by 100 mL of trans-
cinnamaldehyde dissolved in 3.0 mL of chlorobenzene. The
reactor was evacuated and filled with H of the desired pressure
2
0
at room temperature, then placed into an oil bath of 80 C, the
[
[
17] For a related trans-[RuH P ] complex, see: M. T. Bautis-
2
4
temperature of which was controlled by a Lauda K4R
circulator. The reactions were started with starting the
magnetic stirring, and in typical experiments were run for
ta, E. P. Cappellani, S. D. Drouin, R. H. Morris, C. T.
Schweitzer, A. Sella, J. Zubkowski, J. Am. Chem. Soc.
1
991, 113, 4876 ± 4887.
18] R. P. Beatty, R. A. Paciello, (E. I. Du Pont de Nemours
Co.), US Patent 5,689,003, 1997; Chem. Abstr. 1996,
125, 222172w.
2
hours. Depending on the pHand pressure, conversions of
trans-cinnamaldehyde were obtained in the 5 ± 70% range,
however, complete conversions could be obtained in longer
reaction times. No 3-phenyl-1-propanol was detected. At the
end of the reaction the reactor was cooled in ice/water, and the
product mixture was analyzed by gas chromatography of the
&
¬
[19] F. Jo o¬ , J. Kov a¬ cs, A. Kath o¬ , A. C. B e¬ nyei, T. Decuir, D. J.
Darensbourg, Inorg Synth. 1998, 32, 1 ±8.
174
Adv. Synth. Catal. 2003, 345, 172 ± 174