Full Paper
mediate en route to 14). For characterization purposes the
crude product was subjected to purification by flash chroma-
tography to give an analytically pure fraction (39% yield),
whose spectroscopic features were in accordance with the
structure of 14 (see the Supporting Information). Even though
the isolated yield was not high in this case, it should be point-
ed out that this was largely due to the need of discarding sev-
eral chromatographic fractions showing minor spots by TLC.
Given the nature of the contaminants and their low concentra-
tion in the crude product, it is not excluded, therefore, that
a more effective procedure could involve the direct “click” im-
mobilization of unpurified 14 (as for 12 and 13, see below).
Due to the large nitrogen content of the polymer support,
the alkaloid loading was not determined by elemental analysis.
Instead, it was estimated by the weight increase over the start-
ing azidomethyl resin Pb (Table 1).[23] This evidence confirmed
Table 1. Results in the “click” immobilization of 10–14.
Entry Soluble alkaloid deriva- Resin Organocatalyst loading
[c]
tive
[mmolgꢀ1
]
1
2
3
4
5
10
11
12
13
14
P10 0.77
P11 0.69
P12 0.25
P13 0.18
P14 0.20
“Click” immobilization of alkyne-Cinchona derivatives onto
[a] Alkaloid content in P10–P14, calculated from the weight increase.
azido-Merrifield resin and preparation of soluble models
Having established convenient routes to the alkaloid deriva-
tives provided with a terminal alkyne moiety, the next goal
was their immobilization onto an azido-functionalized insolu-
ble polymer. For this purpose, the gel-type azidomethylpolys-
tyrene material Pb was selected as the support. It was prepared
from the commercial Merrifield resin Pa (2.3 mmolCl gꢀ1) by
heating with an excess of sodium azide in dry DMSO
(Scheme 3).[19] The reaction was carried out under slow stirring
the successful immobilization of the chiral units onto Pb, al-
though to an extent that proved larger for the propargyl
mono-ethers 10 and 11 (Table 1, entries 1 and 2) than for the
bis-alkaloid derivatives 12–14 (Table 1, entries 3–5). Even
though no explanation can be provided at the moment for
this trend, according to our experience, this can be due in part
to the relatively small scale of the preparations carried out in
the present study.[24]
In order to carry out a proper comparison with the homoge-
neous phase, the soluble model compounds M10–M14
(Scheme 3) were synthesized by “click” addition of benzylazide
to 10–14. The fair to good yields obtained in these prepara-
tions (see the Supporting Information) provide validation of
the chemistry involved in the anchoring step.
Homogeneous and heterogeneous asymmetric dimerization
of ketenes
Scheme 3. (a) NaN3, DMSO, 608C, 3 d; (b) CuI (5 mol%), DIPEA (1 equiv),
CH2Cl2, rt, 2 d (for the structures of 10–14, see Figure 3).
To test the organocatalytic properties of the newly prepared
IPB Cinchona alkaloid derivatives in the asymmetric dimeriza-
tion of ketenes (Table 2), P10–P14 were employed under the
conditions described by Calter and co-workers for the in situ
generation of ketenes from acid chlorides.[7] As in the previous
studies,[5,25] the overall transformation which leads from the
starting material (23a–c) to the final Weinreb amide (25a–c)
was not carried out one-pot. Instead, after the time t1 for acid
chloride dehydrochlorination and ketene dimerization, the so-
lution containing the b-lactone intermediate (24a–c) was sepa-
rated from the IPB organocatalyst by filtration under an inert
atmosphere. The filtrate was then treated in a separate vessel
with N,O-dimethylhydroxylamine and a catalytic amount of 2-
pyridone (time t2) to achieve the ring opening of 24a–c to the
final product 25a–c. The latter was isolated by careful aqueous
work-up and, thanks to the lack of any organocatalyst contami-
nant, it generally displayed a high chemical purity after remov-
al of the volatiles. The results of these runs (Table 2) revealed
that all of the chiral IPB derivatives behaved as heterogeneous
enantioselective catalysts in the transformation under study.[26]
When the supported mono-quinidine ether P10 was em-
ployed in the reaction of propanoyl chloride (23a), the syn-
for three days in order to avoid mechanical damage of the
polymer beads. After filtration of the suspension, thorough
washing, and drying under reduced pressure, Pb was obtained
as an off-white solid in 95% recovery yield. The introduction of
N3 groups was confirmed by the observation of the strong
azide IR stretching (2095 cmꢀ1 [20]
and by the positive Kaiser
)
test with PPh3/ninhydrin.[21] According to nitrogen elemental
analysis (9.3 wt%, 2.2 mmolN3 gꢀ1) and negative 4-nitrobenzyl-
pyridine color test,[22] no significant amounts of residual chlor-
ine from the starting Merrifield resin Pa remained in the azido-
methyl material Pb.
The immobilization of 10–14 onto the polystyrene support
was achieved by the anticipated “click” reaction between the
alkyne-functionalized alkaloids (12 and 13 in the crude form)
and Pb in the presence of the CuI-DIPEA catalytic system
(Scheme 3). After two days of gentle shaking, the functional-
ized resins were filtered and washed exhaustively for removing
the copper salt and any free alkaloid species. Drying to con-
stant weight under vacuum afforded the insoluble materials
P10–P14, which were characterized by IR spectroscopy (see
the Supporting Information).
&
&
Chem. Asian J. 2014, 00, 0 – 0
4
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!