Phosphoramidite accelerated copper(i)-catalyzed [3 + 2] cycloadditions of azides and alkynes

Article abstract of DOI:10.1039/b822994e

Monodentate phosphoramidite ligands are used to accelerate the copper(i)-catalyzed 1,3-dipolar cycloaddition of azides and alkynes (CuAAC) rapidly yielding a wide variety of functionalized 1,4-disubstituted-1,2,3- triazoles; Cu(i) and Cu(ii) salts both function as the copper source in aqueous solution to provide excellent yields.

Full text of DOI:10.1039/b822994e

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Phosphoramidite accelerated copper(I)-catalyzed [3 + 2] cycloadditions
of azides and alkyneswz
Lachlan S. Campbell-Verduyn,a Leila Mirfeizi,b Rudi A. Dierckx,b Philip H. Elsinga*b
and Ben L. Feringa*a
Received (in Cambridge, UK) 22nd December 2008, Accepted 5th March 2009
First published as an Advance Article on the web 19th March 2009
DOI: 10.1039/b822994e
Monodentate phosphoramidite ligands are used to accelerate the
copper(I)-catalyzed 1,3-dipolar cycloaddition of azides and alkynes
(CuAAC) rapidly yielding a wide variety of functionalized 1,4-
disubstituted-1,2,3-triazoles; Cu(I) and Cu(II) salts both function as
the copper source in aqueous solution to provide excellent yields.
have demonstrated strong ligand accelerating effects.13 They
are inexpensive, easily modified, remarkably stable and readily
accessible ligands.14 We report herein the first example of
dramatic rate acceleration of the 1,3-dipolar cycloaddition of
azides and alkynes using phosphoramidite ligands and the
application of the developed methodology to positron emis-
sion tomography imaging precursors. Phosphoramidites prove
to be excellent, high yielding, easily recovered ligands.
Following pioneering work by Huisgen, the discovery by
Sharpless et al. in 2002 that copper(I) catalyzes the 1,3-dipolar
cycloaddition of azides and alkynes to form 1,4-disubstituted
triazoles strongly contributed to the popularization of ‘click’
chemistry as a highly effective method for functionalization.1
Significant progress has since been made in the application of
this methodology to the areas of drug discovery, polymer chem-
istry, surface functionalization and to medicinal and biological
applications, amongst others.2 Our interest in applying this cyclo-
addition to time sensitive [18F]-radiolabelling methodology for
positron emission tomography (PET) led us to consider the recent
advances in ligand accelerated ligations of azides and alkynes. The
more general ligand-free ‘click’ reactions lack an appropriate rate
for [18F]-radiolabelling in the absence of high copper concentra-
tions or of numerous equivalents of nitrogen containing base.3
Several ligand systems have shown to promote rate
enhancement of the CuAAC. Most notably polytriazolylamine
(TBTA) has proven to sufficiently enhance the reaction rate to
allow for lower catalyst loadings.4 Although TBTA remains
the most widely used ligand,3,5 related examples have since
emerged including benzimidazoles,6 histidine derivatives,7
pybox ligands,8 phosphites,9 and NHC carbenes.10 Most of
these ligands show acceleration that, while significant, still falls
short of the required reaction times for labelling with short-
lived radioisotopes. Others require high catalyst loadings or
the assistance of higher temperatures11 and while NHC
carbenes have demonstrated very effective acceleration, their
use may be limited by their challenging synthetic route.12
Phosphoramidites are used as monodentate ligands for
copper in a number of stereoselective transformations and
A model reaction of benzyl azide with phenylacetylene was
used for preliminary screening of ligand effects. With 1 mol%
CuSO4ꢀ5H2O (reduced to Cu(I) with
5 mol% sodium
ascorbate), a vast array of phosphorus containing ligands
were tested (Table 1). All reactions were performed in a
DMSO–H2O (1 : 3) solution, combining the rate enhancing
power of water1 with the solvation effects of DMSO.
Entry 1 shows that in the absence of ligand, the reaction
requires 30 h to reach completion. We were thus delighted to
see a dramatic accelerating effect upon addition of a phosphor-
amidite ligand. MonoPhos (entry 2) decreases the reaction time
to 2 h to yield the cycloadduct quantitatively. A diethyl
phosphoramidite (entry 3) gave similar acceleration but was
slightly lower yielding as was a sterically hindered ligand
(entry 4). Increasing the length of the amine alkyl chains
proved to decrease the rate of the reaction (entry 5) as did
the use of cyclic PipPhos (entry 6). Aromatic substituents at the
amine moiety also increased the reaction time. Curious as to
the effect of the diol backbone in the ligand upon the reaction
rate we investigated several variations and found that both a
simple catechol-based phosphoramidite and the less rigid
dibenzyl phosphoramidite gave less favourable results than
MonoPhos (entries 8 and 9). Oxidized MonoPhos (entry 10)
also gave comparatively long reaction times and lower yields.
Using triphenylphosphine and triphenyl phosphite as ligands
(entries 11 and 12) we tested the effect of phosphorus donor
properties and found modest acceleration. The use of more
rigid BINAP (entry 13) proved to have a stronger accelerating
effect, though still not as significant as that of the BINOL-
based phosphoramidites. We concluded that the rigid back-
bone is a key factor in the effectiveness of our ligand. In all
cases a single regioisomer of the product was obtained.
a Stratingh Institute for Chemistry, University of Groningen,
Groningen, The Netherlands. E-mail: b.l.feringa@rug.nl;
Fax: +31 50 363 4278; Tel: +31 50 363 4296
b Department of Nuclear Medicine and Molecular Imaging, University
Medical Center Groningen, University of Groningen, Groningen,
The Netherlands. E-mail: p.h.elsinga@ngmb.umcg.nl;
Fax: +31 50 3696750; Tel: +31 50 3614850
We proceeded with MonoPhos to investigate alternative
reaction parameters. Varying the concentration first of the
azide and then of the alkyne indicated that a fourfold excess of
either substrate initially improves rates; after 10 min conver-
sions were 47 and 80%, respectively. An excess of alkyne
proved to be a superior choice for overall rate enhancement,
reducing the reaction time from 2 h to 30 min. In view of the
w Dedicated to Prof. Roeland J. M. Nolte on the occasion of his 65th
birthday.
z Electronic supplementary information (ESI) available: General pro-
cedures for the synthesis of azides and triazoles, 1H and 13C NMR
data for all compounds. See DOI: 10.1039/b822994e
ꢁc
This journal is The Royal Society of Chemistry 2009
Chem. Commun., 2009, 2139–2141 | 2139

View Article Online
Table 1 Effect of ligand on the copper(I)-catalyzed [3 + 2] cyclo-
addition of phenylacetylene and benzyl azide
Table 2 Effect of catalyst loading upon reaction time
Cu (mol%) Liganda
Conv. at 10 min (%) t/bmin Yield (%)c
1
5
10
1
5
10
2 (1.1%) 21
2 (5.5%) 45
2 (11%) 60
3 (1.1%) 16
3 (5.5%) 38
60
40
30
100
80
40
91
93
92
90
95
89
3 (11%)
43
a
b
See entries 2 and 3 (Table 1). For completion. Isolated yield.
c
Table 3 Azide scope for the azide–alkyne cycloaddition
a
Isolated yield.
aIsolated yield after purification by column chromatography.
1-azido-4-fluorobutane also gave fast conversions and excellent
yields (entries 8 and 9). Unsaturated and ester functionalized
azides reacted very quickly with no detectable evidence of side
product formation (entries 7 and 10). Both resulting triazoles are
primed for further functionalization. Of particular interest to us
were the rapid conversions of the fluorine containing aryl and
alkyl azides to their corresponding triazoles (entries 2 and 9).
[18F] labelled analogues of these compounds have been prepared
to label target compounds using this rapid ‘click’ protocol.
A similar investigation of the alkyne scope revealed signifi-
cant differences in the rates of various substrates (Table 4)
likely due to electronic or steric effects (or both) upon the
initial formation of the copper acetylide species considered the
active species in the catalytic cycle.3
goal to achieve a fast method for radiolabelling we sub-
sequently work with 2.0 eq. of alkyne for this reason.
Increasing the catalyst loading had the expected effect of
enhancing the rate of the reaction. From 1 mol% to 10 mol%,
it was possible to significantly decrease the reaction time
to achieve full completion within 30 min with MonoPhos
(ligand 2) (Table 2). However, 1 mol% copper catalyst gives
sufficiently rapid reaction times, making it unnecessary to rely
on higher catalyst loadings to achieve the desired rates.
Increasing the copper concentrations above 10 mol% gave
rise to copper clusters precipitating out of solution.
With optimized parameters in hand, we explored the
substrate scope.
Compared with the model substrate phenylacetylene, 1-ethynyl-
4-methoxybenzene and 1-ethynyl-4-fluoromethylbenzene had
slightly longer reaction times but still reacted quickly with good
yields (entries 2 and 7). An electron withdrawing group in the
ortho position of the ring seemed to have a negative effect on
the rate, however we noted in the course of this reaction that
the solubility of the substrate was poor. We were pleased to see
that an ester functionalized alkyne reacted smoothly (entry 4).
Reaction with propargyl amine (entry 6) showed very rapid
conversion to the corresponding triazole (70% in 1 h),
The reaction proceeded very smoothly with a wide variety of
azides (Table 3). We found that phenyl azide gave the slowest
reaction, taking 4 h to proceed to completion (entry 3) whereas
phenylazides with methoxy, bromo, and cyano substituents at
the para position reacted fully within 2 h (entries 4, 5 and 6).
This is consistent with the recently published discovery by Finn
et al. that electron withdrawing or donating substituents in place
of both hydrogen or methyl groups on the azide substrate
result in a rate increase.15 Alkyl substrates 1-azidooctane and
ꢁc
This journal is The Royal Society of Chemistry 2009
2140 | Chem. Commun., 2009, 2139–2141

View Article Online
Table 4 Alkyne scope for the [3 + 2] azide-alkyne cycloaddition
Table 5 [3 + 2] Cycloaddition with copper(I) salts
Copper salt
t/h
Yield (%)
CuCl
CuBr
CuI
CuBrꢀDMS
1
2
3.5
1
99
94
89
95
found that they enhance the rate of the reaction and stabilize
the catalytically active copper(I) oxidation state. The system is
versatile and functional group independent, a requirement in
the field of ‘click’ chemistry. The methodology has been
applied to the attachment of small [18F]-labelled prosthetic
groups to a model azide.
aIsolated yield after reaction completion (13 h). bReaction only reaches
70% conv.
however, no further conversion was detected. Closer consid-
eration of the resulting triazole led us to conclude that it is an
excellent ligand for Cu(I), and was used by Sharpless et al. to
catalyze the [3 + 2] cycloaddition of azides and alkynes.4 With
nearly one equivalent of the triazole product with respect to
catalytic copper, the metal center risks saturation and inhibi-
tion of any further catalysis. Propiolic acid (entry 5) showed
similar behaviour, reaching 55% conversion after 1 h and then
requiring a further 12 h to reach full conversion.
Financial support from the University of Groningen
(L. S. C.-V., B. L. F.) and University Medical Center Groningen
(L. M., R. A. D., P. H. E.) is gratefully acknowledged.
Notes and references
1 R. Huisgen, 1,3-Dipolar Cycloaddition Chemistry, ed. A. Padwa, Wiley,
New York, 1984, pp. 1–176; V. V. Rostovtsev, L. G. Green, V. V. Fokin
and K. B. Sharpless, Angew. Chem., Int. Ed., 2002, 41, 2596.
2 H. C. Kolb and K. B. Sharpless, Drug Discovery Today, 2003, 8,
1128; R. Breinbauer and M. Kohn, ChemBioChem, 2003, 4, 1147;
V. D. Bock, H. Hiemstra and J. H. van Maarseveen, Eur. J. Org.
Chem., 2006, 1, 51; K. B. Sharpless and R. Manetsch, Expert Opin.
Drug Discovery, 2006, 1, 525; J. E. Moses and A. D. Moorhouse,
Chem. Soc. Rev., 2007, 36, 1249; H. Struthers, B. Spingler,
T. L. Mindt and R. Schibli, Chem.–Eur. J., 2008, 14, 6173.
3 W. G. Lewis, F. G. Magallon, V. V. Fokin and M. G. Finn, J. Am.
Chem. Soc., 2004, 126, 9152; J. Marik and J. L. Sutcliffe, Tetrahedron
Lett., 2006, 47, 6681; M. Meldal and C. W. Tornøe, Chem. Rev., 2008,
108, 2952; M. Glaser and E. Arstad, Bioconjugate Chem., 2007, 18, 989.
4 T. R. Chan, R. Hilgraf, K. B. Sharpless and V. V. Fokin, Org.
Lett., 2004, 6, 2853; W. G. Lewis, F. G. Magallon, V. V. Fokin
and M. G. Finn, J. Am. Chem. Soc., 2004, 126, 9152.
5 P. S. Donnelloy, S. D. Zanatta, S. C. Zammit, J. M. White and
S. J. Williams, Chem. Commun., 2008, 2459.
As a copper source for the CuAAC, CuSO4ꢀ5H2O in
combination with the water-soluble reducing agent sodium
ascorbate is overwhelmingly favoured. Although Cu(I) salts
can also be used, they require an equivalent of nitrogen
containing base to promote the reaction.2 They cannot be
used in the presence of water due to the inherent thermo-
dynamic instability of copper(I), resulting in its easy oxidation
to copper(II).16 A greater likelihood of side-product formation
is observed in cases where copper(I) salts are used.2,17
We anticipated that the phosphoramidite might stabilize the
catalytically active Cu(I) oxidation state. We tested a range of
readily available copper(I) salts in aqueous solution (Table 5)
and indeed found that they give excellent reaction times
and yields, with no evidence of side product formation
(alkyne–alkyne homocoupling for instance) detected.2
6 V. O. Rodionov, S. I. Presolski, S. Gardinier, Y.-H. Lim
and M. G. Finn, J. Am. Chem. Soc., 2007, 129, 12696.
7 K. Tanaka, C. Kageyama and K. Fukase, Tetrahedron Lett., 2007,
48, 6475.
To test our methodology on the required time scale for
radiolabelling, we synthesized [18F]-fluorinated 1-ethynyl-4-
(fluoromethyl)benzene (Scheme 1). After fluorination, it was
ligated to benzyl azide in the presence of CuSO4ꢀ5H2O and
MonoPhos. Full conversion to the labelled triazole was de-
tected after 10 min (as determined by HPLC and radio-TLC).
Under identical conditions but in the absence of ligand, only
minor conversion to the triazole product was detected (o20%).
8 J.-C. Meng, V. V. Fokin and M. G. Finn, Tetrahedron Lett., 2005,
46, 4543.
9 F. Perez-Balderas, M. Ortega-Munoz, J. Morales-Sanfrutos,
F. Hernandez-Mateo, F. G. Calvo-Flores, J. A. Calvo-Asın,
J. Isac-Garcıa and F. Santoyo-Gonzalez, Org. Lett., 2003, 5, 1951.
10 S. Dıez-Gonzalez, A. Correa, L. Cavallo and S. P. Nolan,
Chem.–Eur. J., 2006, 12, 7559.
11 N. Candelon, D. Lastecoueres, A. K. Diallo, J. R. Aranzaes,
D. Astruc and J.-M. Vincent, Chem. Commun., 2008, 741.
12 A. J. Arduengo III, R. Krafczyk, R. Schmutzler, H. A. Craig,
J. R. Goerlich, W. J. Marshall and M. Unverzagt, Tetrahedron,
1999, 55, 14523.
13 D. J. Berrisford, C. Bolm and K. B. Sharpless, Angew. Chem., Int.
Ed. Engl., 1995, 34, 1059.
14 B. L. Feringa, Acc. Chem. Res., 2000, 33, 346.
15 V. O. Rodionov, S. I. Presolski, D. D. Dıaz, V. V. Fokin and
M. G. Finn, J. Am. Chem. Soc., 2007, 129, 12705.
Scheme 1 Synthesis of [18F]-labelled triazole (RCY = radiochemical
yield).
16 C. L. Merrill, L. J. Wilson, T. J. Thamann, T. M. Loehr, N. S. Ferris
and W. H. Woodruff, J. Chem. Soc., Dalton Trans., 1984, 2207.
17 C. W. Tornøe, C. Christensen and M. Meldal, J. Org. Chem., 2002,
67, 3057.
In conclusion, we have applied phosphoramidite copper
complexes to the azide–alkyne [3 + 2] cycloaddition and
ꢁc
This journal is The Royal Society of Chemistry 2009
Chem. Commun., 2009, 2139–2141 | 2141