Stereoselective catalytic hydrogenation of sorbic acid and sorbic alcohol with new Cp*Ru complexes

Article abstract of DOI:10.1039/a909355i

The new Cp*Ru complexes [Cp*Ru(η⁴-MeCH=CHCH=CHCO₂H)]⁺X- (X- = CF₃SO₃- or [B{C₆H₃(CF₃)₂-3,5}₄]^-) are very effective catalysts for the hydrogenation of sorbic acid to cis-hex-3-enoic acid and of sorbic alcohol to cis-hex-3-en-1-ol (leaf alcohol) under mild conditions in liquid two-phase systems.

Full text of DOI:10.1039/a909355i

Stereoselective catalytic hydrogenation of sorbic acid and sorbic alcohol with
new Cp*Ru complexes†
a
b
a
Stephan Steines, Ulli Englert and Birgit Drießen-Hölscher*
a
Institut f. Technische Chemie und Petrolchemie, RWTH Aachen, Templergraben 55, D-52056 Aachen, Germany.
E-mail: drihoel@itc.rwth-aachen.de
Institut f. Anorganische Chemie, RWTH Aachen, Templergraben 55, D-52056 Aachen, Germany
b
Received (in Cambridge, UK) 26th November 1999, Accepted 6th January 2000
4
The
new
(CF
Cp*Ru
complexes
[Cp*Ru(h -
The complexes have been obtained as orange powders or
+
2
2
2
MeCHNCHCHNCHCO
2
H)] X
(X
=
CF
3
SO
3
or
crystals in 72% (5a) and 41% (5b) yield and have been
2
1
13
[
B{C
6
H
3
3
)
2
-3,5}
4
] ) are very effective catalysts for the
characterized by H NMR, C NMR, secondary ion mass
4
hydrogenation of sorbic acid to cis-hex-3-enoic acid and of
sorbic alcohol to cis-hex-3-en-1-ol (leaf alcohol) under mild
conditions in liquid two-phase systems.
spectrometry and a crystal structure analysis.ठThe h -bonding
mode of the sorbic acid in the complexes has been determined
1
by a strong up-field shift of the olefinic proton signals in the H
NMR spectrum in comparison to free sorbic acid and by the
crystal structure. In solution, the complexes may exist as
monomers whereas Fig. 1 shows that complex 5b is a dimer in
the crystalline state which is bridged by the carbonyl oxygens of
the acid. The hydroxy groups form hydrogen bridges to a THF
molecule, the BARF anions do not interact with the cationic
ruthenium complex. 5a,b are the first sorbic acid–ruthenium
complexes.
When sorbic acid 1, a widely used preservative, or sorbic
alcohol are hydrogenated, different products might occur. The
mono-unsaturated products are of technical interest for the
production of fragrances and vitamins. This work focuses on
the stereoselective preparation of cis-hex-3-enoic acid 2 [eqn.
1
(
1)] and of cis-hex-3-en-1-ol 3, a fragrance commercialized as
‘leaf alcohol’.
In early studies Frankel and coworkers showed that me-
thylsorbate and sorbic alcohol can be catalytically hydrogenated
by chromium(0) carbonyl compounds with high selectivities to
methyl-cis-hex-3-enoate and to cis-hex-3-en-1-ol, respective-
2a–d
ly.
Unfortunately the reaction only works with sorbic acid
esters and sorbic alcohol but not with the cheaper sorbic acid 1,
and one must use toxic chromium catalysts and high catalyst/
substrate ratios. Since the start of our studies on stereoselective
3
,4
hydrogenations of
MeCN) ]Tf 4, we came to the conclusion that the [Cp*Ru]
fragment is essential for the stereoselectivity of the reaction.
Naked’ Cp*Ru complexes without inhibiting ligands should
thus be more active and very selective.
We thus synthesized the model complexes [Cp*Ru(h -
1
with complexes like [Cp*Ru-
+
(
3
‘
4
Fig. 1 Crystal structure of the cationic part of 5b.
+
2
MeCHNCHCHNCHCO
2
H)] X (X = Tf 5a, X = tetrakis[3,5-
bis(trifluoromethyl)phenyl]borate (BARF) 5b), in which the
Complex 5a is soluble in polar organic solvents such as
alcohols, nitromethane or sulfolane, but it is insoluble in water
and nonpolar organic solvents like ethers or alkanes. 5a can thus
be used as catalyst in liquid two-phase systems such as
nitromethane–dibutyl ether, ethylene glycol–MTBE (methyl
tertiary butyl ether) or sulfolane–MTBE, in which the complex
remains in the polar phase. After the reaction the complex can
be separated easily by decantation. Table 1 presents the results
of the hydrogenation of sorbic acid in different solvents at
60 °C. At 60 °C the solvents in the systems nitromethane–
dibutyl ether and sulfolane–MTBE become miscible, and at
room temperature the two solvents form two phases.
+
4
[
Cp*Ru] fragment is stabilized by a h -bonded sorbic acid
molecule. We assumed, that these complexes should have a
4
5
similar structure as [Cp*Ru (h -H
2
2
CNCHCHNCH )I], but
should bear weakly coordinating ligands instead of an iodide
ligand. The new complexes 5a,b have been synthesized from
6
[
(Cp*Ru(m-OMe))
2
]
which was cleaved by 2 equivalents of
trifluoromethanesulfonic acid or tetrakis[3,5-bis(trifluoro-
methyl)phenyl]boric acid (HBARF) in the presence of a slight
excess of sorbic acid [eqn. (2)].
The results obtained with the nitromethane systems (Table 1,
entries 1 and 2) show that 5a is about 30 times more active than
[
Cp*Ru(MeCN)
3
]Tf 4, which shows, that a naked [Cp*Ru]+
fragment is more active. 5a,b are active at room temperature
whereas 4 is active only above 60 °C. Nitromethane is normally
used as a weakly coordinating, aprotic solvent, therefore we
expected good catalytic results. However, in this case the
hydrogenation activities and the selectivities to cis-hex-3-enoic
acid are lower in comparison to the stronger coordinating
sulfolane and in comparison to the protic ethylene glycol.
†
Dedicated to Professor Wilhelm Keim on the occasion of his 65th
birthday.
Chem. Commun., 2000, 217–218
This journal is © The Royal Society of Chemistry 2000
217

Table 1 Stereoselective hydrogenation of 1 with 4 and 5a,b as catalysts in various two-phase systemsa
Selectivity (S)
cis-hex-3-enoic trans-hex-3-enoic
TOF/h2
1
Entry
Solvent system
Catalyst
2
p(H )/bar Conv. 1 (%)
acid 2 (%)
acid (%)
n
1
2
MeNO
MeNO
2
–Bu
–Bu
2
O
O
4
60
60
60
10
60
16
45
95
94
78
68
85
93
66
86
96
71
96
6
34
7
1
29
3
3.1
92
300
97
580
1057
n
2
2
5a
5a
5a
5a
5b
b
3
4
Ethylene glycol–MTBE
Ethylene glycol–MTBE
Sulfolane–MTBE
MTBE
b
5
6
c
a
Reagents and conditions: 60 °C; 20 mmol sorbic acid; 0.06 mmol catalyst; catalyst phase: 30 ml solvent; nonpolar product phase: 44 ml solvent; Conv.
b
=
2
conversion; selectivity (S) = (n(product)/∑n (all products)) 3 100; TOF = turnover frequency = (∑n (all products))/(n(catalyst) h).; Some trans-hex-
-enoic acid is formed: (3) S = 7%, (4) 3%. c 40 mmol sorbic acid, 75 ml MTBE, 50 °C.
Table 2 Stereoselective hydrogenation of sorbic alcohol with 5a as catalyst in ethylene glycol–MTBEa
Conv. sorbic
alcohol (%)
Selectivity to leaf
alcohol 3 (%)
Entry
T/°C
2
p(H )/bar
n (5a)/mmol
t/h
TOF/h2
1
1
2
3
4
21
40
60
60
20
20
20
4
0.0424
0.0426
0.0219
0.0428
1.5
44
70
88
86
98.3
98.9
98.6
97.8
184
1055
2495
714
0.42
0.40
0.77
a
Reagents and conditions: 25–27 mmol sorbic alcohol; 25 ml ethylene glycol; 45 ml MTBE; in each experiment 1–2% trans-hex-3-en-1-ol is formed.
Conv. = conversion; selectivity = (n(product) / ∑n (all products)) 3 100; TOF = turnover frequency = (∑n (all products))/(n(catalyst) h).
The highest selectivities (S) (S(cis-hex-3-enoic acid) = 96%)
were obtained in ethylene glycol when the hydrogen pressure
Notes and references
‡
The complexes were synthesized in an argon atmosphere with dried
was reduced from 60 to 10 bar.
solvents. Before the hydrogenation experiments the solutions of the
complexes and the substrates were handled under argon.
The highest activities with 5a as catalyst (TOF = 580 h21)
were obtained in sulfolane, but the selectivity was not as high as
in ethylene glycol.
General procedure for the synthesis of 5a,b: 2.94 mmol [(Cp*Ru(m-
2 2 2
OMe)) ], which was prepared from 2.94 mmol [(Cp*RuCl ) ] according to
Since the BARF anion is much more lipophilic than the
triflate anion, complex 5b is more soluble in nonpolar solvents
than 5a. For this reason 5b was used as catalyst in an MTBE
solution instead of using it in a two-phase system. Experiments
the procedure described in the literature,6b were dissolved in 30 ml
dichloromethane and 10 ml diethyl ether. A solution of 3.44 mmol sorbic
acid and 3.15 mmol triflic acid (or HBARF) in 12 ml diethyl ether was
added to the stirred solution of the complex at 278 °C. The reaction mixture
changed immediately from deep red to brown. The reaction mixture was
stirred for 10 min at 278 °C and was slowly brought to room temperature.
After being stirred for a further 5 min at room temperature, the solvent was
evaporated under reduced pressure. The residue was treated with 10 ml ethyl
acetate to effect formation of an orange solid. For a better precipitation, 10
ml diethyl ether were added. The solid was filtered off washed twice with
10 ml diethyl ether and dried in high vacuum. 2.11 mmol 5a were obtained
5
and 6 in Table 1 can be compared because they both have been
carried out in one-phase systems at the same reaction tem-
perature. Evidently, 5b is a more active catalyst than 5a because
the BARF anion has much weaker coordinating properties than
7
the triflate anion. Thus, the BARF anion does not compete with
sorbic acid and hydrogen for free coordination sites at the
ruthenium center in the catalytic steps of the reaction.
(71.7% yield based on [(Cp*RuCl
§ Crystal data for C56
2.648(6), b = 14.323(3), c = 16.711(6) Å, a = 78.23(2), b = 80.09(3), g
2
)
2
]).
We also used hexa-2,4-diene-1-ol (sorbic alcohol) as sub-
strate, which can be directly hydrogenated to cis-hex-3-ene-1-ol
H51BF24
O
4
Ru 5b: M
r
1355.87, triclinic, a =
1
=
3
¯
84.53(3)°, V = 2914(2) Å , T = 203 K, Z = 2, space group P1 (no. 2),
3
(leaf alcohol). Preliminary results have shown that the
21
m(Mo-Ka) = 3.80 cm , 12258 independent reflections measured (Rint
0
=
catalytic hydrogenation is much faster with sorbic alcohol than
with sorbic acid. It was thus possible to reduce the hydrogen
pressure from 60 to 20 bar in the experiments shown in Table
2
.032). The final wR(F ) = 0.1497 (all data). The structure was solved
2
using direct methods and refined by full matrix least squares on F . Single
crystals of [Cp*Ru(m-O-(h:s-cis-2,3,4,5-Me(CH) CO(OH)))] [B(C
CF -3,5) ·4THF 5b were crystallized from THF–dibutyl ether and
4
2
6 3
H -
2.
(
)
3 2
4 2
]
The best results were obtained with 5a, which hydrogenates
washed with pentane.
CCDC 182/1513.
2
1
sorbic alcohol with a TOF of ca. 2500 h at 60 °C in the two-
phase system ethylene glycol–MTBE. Even at low hydrogen
pressures of 4 bar, which allows working in glassware reactors,
1
2
S. Arctander, Perfume and Flavour Chemicals II, Montclair, USA,
969.
(a) M. Cais, E. N. Frankel and A. Rejoan, Tetrahedron Lett., 1968, 1919;
b) E. N. Frankel, E. Selke and C. A. Glass, J. Am. Chem. Soc., 1968, 90,
1
2
1
the reaction rate stays fairly high (TOF = 714 h ). The very
high selectivity (98–99%) to leaf alcohol is virtually independ-
ent of the reaction temperature, while the hydrogenation activity
raises as expected with increasing temperature. Other than in
hydrogenations of sorbic acid the selecivity depends negligibly
on the conversion. While the selectivity to cis-hex-3-enoic acid
often decreases at conversion rates of > 90%, the selectivity to
cis-hex-3-enoic alcohol remains constant even at 100% con-
version.
(
2446; (c) A. Furuhata, K. Onishi, A. Fujita and K. Kogami, Agric. Biol.
Chem., 1982, 46, 1757; (d) A. A. Vasil’ev and E. P. Serebryakov,
Mendeleev Commun., 1994, 4.
3
J. Heinen and B. Drießen-Hölscher, J. Organomet. Chem., 1998, 570,
1
41.
4
5
J. Heinen, Ph.D. Thesis, 1997, RWTH-Aachen.
P. J. Fagan, W. S. Mahoney, J. C. Calabrese and I. D. Williams,
Organometallics, 1990, 9, 1843.
We have thus shown that the concept of using ‘naked’ Cp*Ru
complexes for stereoselective hydrogenations of functionalized
dienes to cis-olefins is successful. In further work we will try to
elucidate the mechanism and the kinetics of the reaction.
We thank the Bundesministerium für Bildung und Forschung
and the Ministerium für Wissenschaft und Forschung des
Landes NRW (Katalyseverbund NRW) for financial support.
6
(a) U. Kölle and J. Kossakowski, J. Chem. Soc., Chem. Commun., 1988,
5
49; (b) U. Kölle and J. Kossakowski, J. Organomet. Chem., 1989, 362,
383.
7 G. M. DiRenzo, P. S. White and M. Brookhart, J. Am. Chem. Soc., 1996,
118, 6225.
Communication a909355i
218
Chem. Commun., 2000, 217–218