N.F. Dummer et al. / Journal of Catalysis 243 (2006) 165–170
167
fied with an appropriate alkaloid. Commercial cinchona alka-
loids used included cinchonine (Cn; Fluka, 98%), hydroquini-
dine (HQd; Aldrich, 95%), hydroquinidine 4-chlorobenzoate
7.35 (1 H dd), 6.81 (1 H d), 3.95 (3 H s), 3.36 (1 H aq), range
2.95–2.65 (4 H m), 1.96 (1 H tt), 1.76 (1 H s), range 1.59–1.47
(6 H m), 0.88 (3 H t). m/z 431 (M + H ).
+
(
(
(
HQdClB; Aldrich, 98%), hydroquinine 4-chlorobenzoate
HQnClB; Aldrich, 98%), hydroquinidine 9-phenanthryl ether
HQdPAE; Aldrich, 96%), and hydroquinidine 4-methyl-2-
2.3.2. Cinchonine acetate (CnA)
Cinchonine (100 mg, 0.34 mmol) and triethylamine (51.4
mg, 0.51 mmol) were dissolved in dichloromethane (20 ml).
quinolyl ether (HQdMQE; Aldrich, 97%). Premodification of
catalyst samples was conducted on a small batch scale, giv-
ing sufficient catalyst for several experiments. Catalyst samples
◦
The solution was cooled to 0–5 C, and acetyl chloride (26.6
mg, 0.34 mmol) was added dropwise. The product was recov-
ered and recrystallised as described previously.
(
200 mg) were stirred in a slurry of alkaloid and dissolved in
dichloromethane (25 ml) in air for 10 min. Control experiments
using nitrogen in place of air showed no differences in catalytic
efficacy. The catalyst was filtered and dried under vacuum be-
δH (500 MHz, CDCl3) = 8.80 (1 H d), 8.13 (1 H d), 8.05
(1 H d), 7.64 (1 H dd), 7.52 (1 H dd), 7.31 (1 H d), 6.50 (1 H d),
5.95 (1 H m), range 5.20–5.00 (2 H m), 3.22 (1 H qa), range
2.85–2.59 (4 H m), 2.19 (1 H qa), 2.05 (3 H s), range 1.80–1.73
fore being transferred to the glass reactor tube for reaction at
◦
+
1
0 C. Catalyst samples (25 mg) were preconditioned in a He
(2 H m), range 1.48–1.41 (3 H m). m/z 337 (M + H ).
−1
flow (80 ml min ) for 10 min, followed by a 3:1 He:H2 flow
80 ml min ) for a further 10 min. This mixture was then di-
−1
(
2.3.3. Cinchonine phenyl acetate (CnPA)
◦
verted through a saturator (3 C) containing the reactant. The
effluent reaction gas was analysed by on-line gas chromatogra-
phy at regular intervals during the reaction period. (See [20] for
full details and demonstration that the conditions used do not
involve gas/liquid/solid systems due to capillary condensation.)
Cinchonine (100 mg, 0.34 mmol) and triethylamine (51.4
mg, 0.51 mmol) were dissolved in dichloromethane (20 ml).
◦
The solution was cooled to 0–5 C, and benzoyl chloride
(47.6 mg, 0.34 mmol) was added dropwise. The product was
recovered and recrystallised as described previously.
δH (500 MHz, CDCl3) = 8.76 (1 H d), 8.28 (1 H d), range
.2. Addition of Bi3 to the platinum surface
+
8.07–8.00 (4 H m), range 7.64–7.35 (5 H m), 6.81 (1 H d), 5.93
2
(
1 H ddd), 5.04 (1 H d), 5.00 (1 H d), 3.37 (1 H aq), range 2.99–
A 1-g sample of 2.5% Pt/SiO2 was placed in a round-
bottomed flask to which aqueous Bi(NO3)3 (5 ml, 0.15–
.75 mM; Aldrich, 99.999%) dissolved in ultra-pure water was
2.63 (4 H m), 2.23 (1 H aq), 1.94 (1 H t), 1.79 (1 H s), range
1.57–1.47 (3 H m). m/z 399 (M + H ).
+
0
added. The slurry was stirred for 3 h and filtered under vacuum
and washed with ultra-pure water (500 ml) to remove any re-
maining nitrate species. The catalyst was dried under vacuum
and used immediately in the alkaloid premodification proce-
dure.
2.4. Transmission electron microscopy
Bright-field (BF) and high-resolution transmission electron
microscopy (HRTEM) was performed at 200 kV using a JEOL
2200FS transmission electron microscope with a point-to-point
resolution of 0.19 nm. Samples were prepared for TEM analysis
by dispersing the catalyst powder in high-purity ethanol, then
allowing a drop of the suspension to evaporate on a holey car-
bon film supported by a 300-mesh copper TEM grid.
2
.3. Synthesis of cinchonine and hydroquinidine derivatives
All reactions were carried out under a dried N2 atmosphere.
All starting materials but cinchonine (98%, Fluka) were pur-
chased from Aldrich and used as received. APCI mass spectra
were recorded on a Fisons Platform II spectrometer at a cone
3. Results and discussion
1
voltage of 20 V. H NMR spectra were recorded in CDCl3 us-
Recently, we showed that the enantioselective hydrogenation
of pyruvate esters can be carried out in the absence of solvent
at the gas–solid interface [20]. The significance of this obser-
vation is that enantioselective experiments can be conducted in
the absence of solvent effects, which are known to be a po-
tentially dominant feature [21]. In view of this, we carried out
an extensive set of experiments using gas-phase reactants and
investigated the use of various cinchonine and cinchonidine-
derived modifiers for the hydrogenation of ethyl pyruvate at the
gas–solid interface over a 2.5% Pt/SiO2 catalyst using two dif-
ferent concentrations of the modifiers. The results are given in
Table 1. Under these conditions, 100% conversion was achieved
irrespective of the choice of modifier or the presence of alka-
loid; this was stable throughout the reaction. Time-on-line data
for the two concentrations are given in Fig. 1 for the enan-
tioselectivity of Cn (squares), HQnClB (diamonds), HQdClB
(circles), HQdPAE (triangles), and no modifier (crosses). It is
ing a 500-MHz Bruker Avance spectrometer.
2
.3.1. Hydroquinidine phenyl acetate (HQdPA)
Hydroquinidine (200 mg, 0.61 mmol) and triethylamine
(
(
92.8 mg, 91.7 mmol) were dissolved in dichloromethane
20 ml). The solution was cooled to 0–5 C, and benzoyl chlo-
◦
ride (85.6 mg, 0.61 mmol) was added dropwise. The reac-
tion mixture was allowed to warm to room temperature and
stirred at this temperature overnight. The reaction mixture was
poured to ice water (50 ml) and extracted with dichloromethane
(
3×30 ml), the combined organic layers were washed with sat-
urated aqueous NaHCO3, dried over Na2SO4, and evaporated
to dryness. The product was recrystallised from a solution of
1
0% diethyl ether/light petroleum 40–60.
δH (500 MHz, CDCl3) = 8.70 (1 H d), 8.08 (2 H dd), 7.99
(
1 H d), 7.58 (1 H t), 7.51 (1 H d), 7.45 (2 H dd), 7.40 (1 H d),