Table 1. Solubility Tests of Copper(I) Halides in Organic
Solventsa
Table 2. Fixation Tests of Cuprous Iodide on Amberlyst A-21
in Acetonitrile at Room Temperaturea
salt EtOH MeOH DMSO DMF THF AcOEt CH3CN CH2Cl2
CuCl
CuBr
CuI
I
S
I
I
S
I
I
S
I
I
S
I
I
I
I
I
I
I
I
I
S
I
I
I
A-21b
entry mg (mmol) mg (mmol)
CuI
t
h
CuI fixedc
mmol (mg) mmol CuI/g
catalyst
a For 0.1 M solutions (0.5 mmol of CuX/5 mL of solvent); S ) soluble,
I ) insoluble or partially soluble.
1
2
3
4
5
6
250 (1.2)
500 (2.4)
500 (2.4)
500 (2.4)
1000 (4.8)
1000 (4.8)
95 (0.5)
95 (0.5)
95 (0.5)
190 (1.0)
95 (0.5)
190 (1.0)
24 0.221 (42)
0.195 (37)
24 0.368 (70)
24 0.805 (153)
24 0.274 (52)
24 0.726 (138)
0.76
0.36
0.65
1.23
0.26
0.64
1
We sought to develop a copper catalyst, as a heterogeneous
catalytic system for automation protocols, that can have
several advantages, such as fast and simple isolation of the
reaction products by filtration as well as recovery and
recycling of the catalyst.10 In this article, we report the
preparation of a polymer-supported catalyst and its prelimi-
nary evaluation in cycloaddition reactions between com-
mercially available terminal alkynes and synthetic azides.
To select our catalytic system, we took into account that
the copper salt has to be fixed onto a polymer, possibly by
chelation, and has to include a base to be complete. Because
of our previous experience with polymeric systems, we
selected to test the fixation and activity of copper(I) salts on
Amberlyst A-21. This well-known polymer is a dimethyl-
aminomethyl-grafted polystyrene and thus bears an amine
group that can act as both a chelatant and a base.
The solubility of simple copper(I) halides in various
organic solvents was first tested (Table 1). In usual solvents,
cuprous chloride is quite insoluble. Better results were
obtained with the corresponding bromide which was soluble
in four of them (EtOH, MeOH, DMSO, and DMF). Finally,
cuprous iodide was only soluble in acetonitrile. The solvent
used to fix the copper salt on the polymer has to be
compatible with the hydrophobic polystyrene network and
easily removed during the drying process following the
incubation and filtration.
a Performed in 5 mL of solvent at room temperaure for the indicated
time under agitation. b Dry Amberlyst A-21, 4.8 mmol N/g. c After filtration,
washings, and drying of the polymer in vacuo.
mmol of amine (entries 4 and 6), higher fixations were
observed at 1.0 mmol of CuI. Furthermore, the maximum
adsorption for all these experiments was reached for the 2.4
mmol of amine/1.0 mmol of CuI ratio (entry 4, 0.805 mmol
of CuI).
From all these experiments, the best fixation protocol was
the use of the ratio of 2.4 mmol of amine/1.0 mmol of CuI.
This gave a polymer with a final composition of 1.23 mmol
of CuI/g of resin. This catalyst was selected over the others
for its highest content in copper and a recalculated ratio of
3 amine groups for 1 copper atom, suggesting that at least
one amine function was free to act as the required base.
This polymer-supported catalyst was then tested in Huis-
gen’s cycloaddition between various terminal alkynes and
organic azides using a synthesis robot (Table 3).11 We first
conducted the experiments in the same solvent as that for
the fixation, i.e., acetonitrile (conditions A).
Even if all products were formed, LC-MS analyses showed
that they all needed to be purified afterward because of the
presence of colored impurities, some residual azide, and
traces of copper released from the polymeric catalyst. Quick
purifications on short pads of silica afforded the products
9-18 with purified yields ranging from low to good (27-
91%, 66% average). Under these conditions, the reaction of
benzyl azide (6) gave good yields from 74 to 91%, except
for the triazole 12 (50%). With propanol azide (7), the yields
obtained were modest (27-70%) with a good result for 18
(87%).
The reaction was then performed in dichloromethane
(conditions B). In this case, all products were isolated as
pure, except for a few, after a simple filtration to remove
the catalyst and evaporation of the solvent. No traces of
starting material or copper were detected on LC-MS analyses.
The use of a low polarity and nonchelating solvent such as
dichloromethane seems to prevent significant leaking of the
The use of CuBr in alcoholic or high boiling point solvents
was thus ruled out, and the CuI/CH3CN combination was
selected as the best alternative.
Fixation tests were conducted using various ratios of A-21
and CuI in the same volume of acetonitrile (Table 2).
Increasing the contact time between the CuI solution and
the resin (entries 2 and 3) from 1 to 24 h doubled the amount
of CuI fixed. Other tests were thus run overnight to ensure
a better fixation. Modifications in the amount of A-21 used
from 1.2 to 2.4 and 4.8 mmol of amine against 0.5 mmol of
CuI (entries 1, 3, and 5) showed that a better fixation was
reached for 2.4 mmol of amine/0.5 mmol of CuI. When
amounts of CuI were increased to 1.0 mmol for 2.4 and 4.8
(9) Feldman, A. K.; Colasson, B.; Fokin, V. V. Org. Lett. 2004, 6, 3897.
(10) For some examples of supported catalysts, see the following articles
and references therein: (a) Jansson, A. M.; Grøtli, M.; Halkes, K. M.;
Meldal, M. Org. Lett. 2002, 4, 27. (b) Gonthier, E.; Breinbauer, R. Synlett
2003, 1049. (c) Shibahara, F.; Nozaki, K.; Hiyama, T. J. Am. Chem. Soc.
2003, 125, 8555. (d) Chiang, G. C.; Olsson, T. Org. Lett. 2004, 6, 3079.
(e) Takeuchi, M.; Akiyama, R.; Kobayashi, S. J. Am. Chem. Soc. 2005,
127, 13096. (f) Kim, J.-H.; Kim, J.-W.; Shokouhimedr, M.; Lee, Y.-S. J.
Org. Chem. 2005, 70, 6714. (g) Liou, R.-M.; Chen, S.-H.; Hung, M.-Y.;
Hsu, C.-S.; Lai, J.-Y. Chemosphere 2005, 59, 117.
(11) Reactions performed in 13 mL glass reactors equipped with a
plunging filter on a Chemspeed ASW-2000. The reactions can also be done
in peptide synthesis glass reactors or in usual round-bottom flasks followed
by filtration on fritted glass funnels.
1690
Org. Lett., Vol. 8, No. 8, 2006