M. Marchetti et al. / Applied Catalysis A: General 373 (2010) 76–80
79
the fragrance Helional1 [19]. It is noteworthy to observe that, even
for a short reaction time and at low substrate conversions, Rh/HSA
maintains its activity practically unchanged in three recycle
experiments, while Rh/TPPTS begins to degrade after the first
recycle experiment.
Table 7
Aqueous biphasic hydrogenation of 3-phenyl-2-propenal (IX) catalyzed by Rh/HSA
and Rh/TPPTS for 1 h.
Run
Cat.
Conv.
(%)
X selectivity
XI selectivity
XII selectivity
(%)
(%)
(%)
1
2a
3a
4a
5
6a
7a
8a
Rh/HSA
12.5
13.3
11.6
11.3
76.9
54.0
53.3
48.0
100
100
100
100
87
–
–
–
–
–
–
–
–
–
–
Finally we tested the hydrogenation of two 3-aryl-2-methyl-2-
propenals, aryl being phenyl (compound XIII) and 1,3-benzo-
dioxol-5-yl (compound XVII) (Scheme 4), at 60 8C and 5 MPa of H2
for 22 h in the biphasic system H2O/2-MeTHF and H2O/toluene,
respectively. Hydrogenation products are intermediate and active
ingredients of drugs and fragrances [26–30]. In this case, Rh/HSA is
less active with respect to 3-phenyl-2-propenal (IX) hydrogena-
tion, either for steric hindrance of the methyl group on the carbon–
carbon double bond or perhaps for solubility reasons too. Despite a
decrease of activity, the selectivity towards the saturated
aldehydes XIV and XVIII was very high (86–92%) (Table 8).
The aldehydes XIV and XVIII present a stereogenic center,
therefore, after purification by flash-chromatography, they were
subjected to a polarimetric measurement but, disappointingly,
both compounds resulted to be nearly racemate.
Rh/HSA
Rh/HSA
–
Rh/HSA
–
Rh/TPPTS
Rh/TPPTS
Rh/TPPTS
Rh/TPPTS
13
12
10
7
88
90
93
Substrate = 5.2 mmol; Rh(CO)2acac = 0.0104 mmol; HSA = 12.2 mg; Rh(CO)2acac/
TPPTS (molar ratio) = 1/3; H2O = 5 mL; toluene = 2 mL; p(H2) = 5 MPa; T = 60 8C;
t = 1 h. Substrate/metal (molar ratio) = 500/1. TPPTS = (triphenylphosphine-3,30,300-
trisulfonic acid trisodium salt).
a
Reaction carried out by using the aqueous phase recovered from the previous
run.
This result was not quite surprising, as a similar outcome was
also found in styrene hydroformylation [18]. In order to explain
this phenomenon, we previously tried to get information on the
structure of the catalytic system Rh/HSA [18,31]. MALDI-TOFMS
measurements showed that the Rh/HSA complex is constituted by
4 molecules of HSA and 89 rhodium atoms; furthermore, the
presence of protein tetramers was revealed. We supposed that Rh
coordinates the sulfur atoms of the protein: this hypothesis was
supported by CD spectra and by SEM analysis that showed a
superficial distribution of rhodium very similar to that of sulfur.
Moreover, CD spectra suggest a different stereochemistry of the
HSA bound Rh(I) compound, most probably due to a different
conformation of the protein. Two limiting conformations are
actually reported, the N and the B, the N being the prevailing one
for pH values up to 7, and the B being more stable for pH ꢁ8
[32,33]. At physiological pH the HSA is expected to assume both
conformations [34]; therefore, at pH 7.4, the protein can assume
both the conformations, the prevailing one depending on the
structure of the ligand and the conformational equilibrium
between the N and B conformations must be taken into account.
Scheme 4. Aqueous biphasic hydrogenation of 3-aryl-2-methyl-2-propenals (XIII
and XVII).
Table 8
Aqueous biphasic hydrogenation of 3-aryl-2-methyl-2-propenals (XIII and XVII)
catalyzed by Rh/HSA.
Run
Substrate
Conv.
(%)
XIV (or XVIII)
XV (or XIX)
XVI (or XX)
selectivity (%)
selectivity (%)
selectivity (%)
1
2a
3a
4
5a
6a
7a
8a
XIIIb
XIIIb
XIIIb
XVIIc
XVIIc
XVIIc
XVIIc
XVIIc
47
42
45
24
31
32
48
36
92
94
91
86
90
89
88
89
3
2
5
4
6
–
–
–
–
–
3
14
10
11
12
11
4. Conclusive remarks
Substrate = 5.2 mmol; Rh(CO)2acac = 0.0104 mmol; HSA =12.2 mg; H2O = 5 mL;
The catalytic systems Rh/HSA and Ir/HSA showed to be active in
the hydrogenation of representatives substrates that present in
their molecules both C55C and C55O double bonds. In particular, the
rhodium based catalyst resulted to be interesting as far as
conversion and selectivity are concerned. Analogously to the
above mentioned rhodium complexes modified with water soluble
phosphanes [8–14], also our catalytic system Rh/HSA preferen-
p(H2) = 5 MPa; T = 60 8C; t = 22 h. Substrate/Rh (molar ratio) = 500/1.
a
Reaction carried out by using the aqueous phase recovered from the previous
run.
b
Solvent was 2-MeTHF (2 mL).
c
Solvent was toluene (2 mL).
Although data on the hydrogenation of substrate IX catalyzed
by some water soluble rhodium based catalysts were already
reported in the literature [8–12,14,25], we decided to perform
some experiments by using the catalytic system Rh(CO)2acac/
TPPTS in the same reaction conditions adopted for Rh/HSA to have
a better comparison: Rh/HSA resulted to be less active than Rh/
TPPTS but more selective towards the C55C hydrogenation
(Table 6).
In order to better compare the selectivity of these catalytic
precursors, we carried out some experiments under the same
reaction conditions but only for 1 h to obtain lower conversions:
Rh/HSA afforded exclusively the saturated aldehyde X while Rh/
TPPTS produced also alcohol XI, even if in small amounts (see
Table 7). This different chemo- and regioselectivity of Rh/HSA with
respect to Rh/TPPTS had been already noted also in the
hydroformylation of safrole and isosafrole, respectively, to produce
tially reduce the olefinic double bond of a,b-unsaturated carbonyl
compounds, even at high hydrogen pressure (5 MPa); in particular,
2-cyclohex-1-one, was exclusively hydrogenated to cyclohexa-
none, in all the experimental conditions employed. On the
contrary, but quite along with the data reported in the literature
for water soluble iridium complexes [8,12], the catalytic system Ir/
HSA is less selective towards the reduction of the olefinic double
bond and, under 5 MPa of H2, the hydrogenation of the carbonyl
moiety is the prevailing reaction. Very interestingly, by lowering
the hydrogen pressure, this catalytic system tends to promote the
reduction of the carbon–carbon double bond, as highlighted in the
hydrogenation of 2-cyclohex-1-one.
The catalytic system Rh/HSA showed to be very chemo- and
regioselective but, disappointingly, it was not able to induce any
enantioselection. At the moment we have not any experimental
evidence to explain this fact but we suppose that this lack of