Kinetic resolution by copper-catalyzed azide-alkyne cycloaddition

Article abstract of DOI:10.1016/j.tetlet.2005.05.019

The use of chiral pybox ligands imparts enantioselectivity to the Cu ^I-catalyzed azide-alkyne cycloaddition reaction, in the form of kinetic resolution of α-chiral azides and desymmetrization of gem-diazides. While levels of selectivity are modest, the results show unequivocally that the process benefits from ligand-accelerated catalysis.

Full text of DOI:10.1016/j.tetlet.2005.05.019

Tetrahedron
Letters
Tetrahedron Letters46 (2005) 4543–4546
Kinetic resolution by copper-catalyzed azide–alkyne cycloaddition
Jun-cai Meng, Valery V. Fokin* and M. G. Finn*
Department of Chemistry and The Skaggs Institute for Chemical Biology, The Scripps Research Institute,
10550 N.Torrey Pines Rd,. La Jolla, CA 92037, USA
Received 4 April 2005; revised 3 May 2005; accepted 6 May 2005
Available online 24 May 2005
Abstract—The use of chiral pybox ligands imparts enantioselectivity to the CuI-catalyzed azide–alkyne cycloaddition reaction, in the
form of kinetic resolution of a-chiral azidesand desymmetrization of gem-diazides. While levels of selectivity are modest, the results
show unequivocally that the process benefits from ligand-accelerated catalysis.
Ó 2005 Elsevier Ltd. All rights reserved.
The copper(I)-catalyzed 1,3-dipolar cycloaddition of
1
azidesand alkynes hasemerged asa useful procesto
substitution at or near the 5-position of each oxazole
ring appears to be required with substrate 1, since li-
gands lacking such a substituent were ineffective (4, 6,
and 10). The faster-reacting enantiomer of 1 wasfound
to be the same for all ligands except ent-7 and 8, suggest-
ing that the sense of asymmetric recognition is deter-
mined by the absolute configuration at the 5-position
of the oxazole.
connect diverse building blocks for chemical synthesis,
drug discovery, biological chemistry, and materials sci-
ence.2 In the course of our initial catalyst screening
and mechanistic efforts, we observed strong ligand-
accelerated catalysis of the process by a number of
different heterocyclic chelates.3 Such a situation is well-
suited for the imposition of enantioselectivity on a
metal-catalyzed process, and we report here the first
examples of asymmetric kinetic resolution for the
azide–alkyne cycloaddition reaction. Apart from an
expansion of the synthetic applications that this may
provide, the study is most useful for its relevance to
questions of reaction mechanism.
We surveyed the effect of various reaction conditions on
the kinetic resolution of phenethyl azide (13) using
ligand 12. The source of CuI wasfound to be important,
with cuprous iodide giving a much improved system
(s = 5.9). Increasing the concentration of the catalyst
provided more conversion but made relative little differ-
ence to the enantiomeric selectivity (Table 1). Catalytic
activity was found to be more sensitive to solvent than
is the case for standard Cu-catalyzed azide–alkyne
cycloaddition reactions, with dichloromethane provid-
ing the best rates and selectivities.6 Catalytic activity
was also substantial in acetonitrile and methanol solu-
tions, in which CuOTf was more effective than CuI.
Copper complexesof the b(oisxazolinyl)pyridine
(pybox) family of ligands are well established as catalysts
for a number of asymmetric transformations.4 While
most of these applications derive from the Lewis acidic
propertiesof Cu(II) complexes, two recent examplesin-
volve Cu(I)–acetylide species.5 A selection of pybox li-
gands, 4–12, was screened in the kinetic resolution of
racemic azide 1, ashsown in Figure 1. Each showed
an acceleration rate with respect to the reaction in the
absence of chelating ligand. A range of effectiveness in
kinetic resolution was observed up to maximum enan-
tiomer selectivity factors of approximately 3 for the 5-
substituted structures 5 and 12, showing that the ligand
isinvolved in the copper-catalyzed proce.s Aromatic
The existence of ligand-accelerated catalysis by 12 was
further revealed in an examination of kinetic resolution
with varying Cu:12 ratios( Table 2). The presence of ex-
cess copper eroded, but did not eliminate, kinetic resolu-
tion, demonstrating that Cu–12 complexesare more
active than the ligand-free catalyst. Reaction rates
peaked at Cu:12 ratiosof 1:2, and diminihsed in the
presence of greater amounts of ligand, more dramati-
cally with cuproustriflate than cuprousiodide. A 1:1
mixture of 12 and 4,40-dimethyl-1,10-bipyridine pro-
vided an s value of only 1.3. The Cu–bipy catalyst itself
*
Corresponding authors. Tel./fax: +1 8587848845; e-mail: mgfinn@
scripps.edu
0040-4039/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2005.05.019

4544
J.Meng et al./ Tetrahedron Letters 46 (2005) 4543–4546
0.4 equiv
CH3
N
CH3
N3
CuOTf
+
Ligand
CH2Cl2, iPr2NEt
N
N
1
2
3
RT
O
O
O
O
N
N
O
O
O
O
O
N
N
O
N
N
N
Ph
Ph Ph
N
N
Ph Ph
Ph
N
N
N
N
N
N
Ph
Ph
5
Ph
4
6
Ph
ent-7
19% conv.
s = –2.0
7
Ph
Ph
37% conv.
s = 1.0
39% conv.
s = 1.0
38% conv.
s = 3.2
31% conv.
s = 2.0
O
O
N
O
O
N
N
N
N
O
O
N
N
O
O
N
N
N
O
O
N
N
N
Ph
Ph
N
N
NH
HN
Me
Me
Ph
Ph
8
9
10
11
12
37% conv.
s = –1.8
34% conv.
s = 1.6
16% conv.
s = 1.0
38% conv.
s = 1.5
22% conv.
s = 2.8
Figure 1. Results for reactions of 1 and 2 using 10 mol% CuOTf and 20 mol% ligand with respect to 2, under the indicated conditions. For all data in
thisletter, the relative rate ratio for kinetic resolution isgiven by s = kfast/kslow with experimental error of 10% of the reported value; the maximum
percentage conversion is 40%, based on the limiting reagent 2.
Table 1. Dependence on CuI source and concentration for the reaction
of 13+2 in CH2Cl2 using 2 equiv of ligand 12 with respect to Cu
Table 2. Kinetic resolution for 13+2 in the presence of a constant
amount of CuI or CuOTf (0.1 equiv with respect to alkyne) and
varying amountsof 12 (CH2Cl2, room temperature)
N
N
H C
3
N
3
CuX : 12
1 : 2
CuI:12
% Conva
s
CuOTf:12
% Conva
s
0.4 equiv
H C
3
N
+
4:1
2:1
1:1
1:2
1:3
1:4
96
88
91
86
87
81
1.9
4.0
4.6
5.4
5.6
5.8
CH2Cl2, iPr2NEt
RT
13
2
14
1:2
1:4
1:6
96
30
1
3.4
3.1
Cu complexa
% Convb
s
mol%
CuI Æ 12c
% Convb
s
—
a % Conversion, maximum is 40% corresponding to limiting 2.
Cu(OTf)
Cu(MeCN)4PF6
CuBr
40
9
3.4
2.3
3
5
17
21
27
32
34
5.3
5.3
6.0
5.5
5.4
38
37
37
29
4.1 10
2.9 20
5.9 30
3.8
The maximum level of kinetic resolution thus far
achieved for the reaction of 13+2 (s = 8 0.4) was
observed for the CuIÆ(12)2 catalyst upon cooling to 0 °C,
near-complete reaction requiring approximately 24 h.8
The recovered starting material was compared with
enantiopure 13 prepared from commercially available
R-phenethyl alcohol9 to reveal R-13 to be the faster-
reacting enantiomer. The attempted extension of kinetic
resolution to substrates 15–17 using a selection of
CuCl
CuI
CuSO4
d
a [Cu] = 10 mol% with respect to azide and alkyne.
b % Conversion, maximum is 40% corresponding to the limiting
reagent 2.
c With respect to azide and alkyne; a 2:1 ratio of 12:CuI.
d With 2 equiv sodium ascorbate with respect to Cu; reaction in a 2:1
t-BuOH:H2O.
6
ligands was unsuccessful, with s valuesall <1.3. The
1-phenylpropargylic compounds 18–22 were even more
resistant to kinetic resolution, with no enantiomeric
discrimination observed whatsoever (s = 1.0).
showed approximately the same degree of reactivity as
Cu–12. These observations suggest that bipy is a
strongly competitive ligand for CuI and/or that the
species of the formulation (CuÆ12Æbipy) isnot enantio-
selective.7
The reactionsof geminal diazides 23 and 26 with 2, medi-
ated by the best catalyst in the above studies, have so far
Cpd
R
N
N
3
N
3
3
18
19
20
21
22
OH
OAc
R
CO Me
2
OEt
OCOPh
NHAc
15
16
17

J.Meng et al./ Tetrahedron Letters 46 (2005) 4543–4546
4545
Ph
Ph
2 (1.0 equiv)
CuI (0.1 equiv)
12 (0.2 equiv)
N
N
Me N3
N3
N
N
N
Me
+
Me
N
CH2Cl2
RT
Ph
N3
N
Ar
N
23
2 equiv
N
N
Me
H
24
25
60% yield
R
23% yield, 16% ee
N
O
N
O
N
N
N
Cu
Ph
N
Ph
N
R
2 (1.0 equiv)
CuI (0.1 equiv)
12 (0.2 equiv)
N3
N3
N
N
N
N
Ph
+
CH2Cl2
RT
Ph
N3
N
N
Cl
26
2 equiv
N
Cl
Cl
27
28
25% yield, 59% ee
63% yield
Figure 2. (Left) Desymmetrization of gem-diazides. (Right) Proposed interaction of ligand and azide stereogenic centers.
resulted in the formation of the corresponding bis(triaz-
oles) 25 and 28 as major products, even in the presence
of a deficiency of alkyne (Fig. 1). A similar phenomenon
hasbeen obesrved with 1,2- and certain 1,3-diazide,s
and isbelieved to be a function of the high reactivity
of a Cu–organometallic intermediate, as described else-
where.10 While the production of monotriazolesin this
type of reaction is not synthetically useful, a moderate
level of asymmetric induction was achieved in the
desymmetrization of 26. The absolute stereochemistry
of the chiral product hasnot yet been established.
Acknowledgements
We are grateful to The SkaggsIntsitute of Chemical
Biology at The Scripps Research Institute for support
of thiswork. J.M. isa SkaggsPostdoctoral Fellow.
Supplementary data
Synthetic details, a complete table of solvent effect stud-
ies, and a survey of substrate versus ligand are available
online with the paper in ScienceDirect. Supplementary
data associated with this article can be found, in the on-
line version at doi:10.1016/j.tetlet.2005.05.019.
The absolute sense of kinetic resolution of benzylic
azidesand the lack of resolution of enantiomeric alkynes
are consistent with an arrangement of reacting species
shown in Figure 2. We suggest that copper–pybox com-
plexesengage azide os asto place the azide
a-carbon
directly over the ligand plane and therefore in a position
to be influenced by itschiral environment. Such a Cu–
azide interaction hasbeen predicted on the basisof den-
sity functional theory calculations11 and implicated in
kinetic studies of the ligand-free reaction,10 although
the precise coordination environment of the metal center
and the role of counterion are not yet understood.12 Ori-
enting the aryl group away from the ligand plane and
the H-atom toward the more sterically demanding quad-
rant occupied by the ligand R-group definesa better
arrangement for the R-azide. The propargylic stereo-
genic center in such compounds as 18–22 will be farther
away from the chiral ligand if the reaction occurs
through a Cu–acetylide intermediate, providing poor
enantiomeric discrimination.
References and notes
1. (a) Tornøe, C. W.; Christensen, C.; Meldal, M. J.Org.
Chem. 2002, 67, 3057–3062; (b) Rostovtsev, V. V.; Green,
L. G.; Fokin, V. V.; Sharpless, K. B. Angew.Chem,. Int.
Ed. 2002, 41, 2596–2599.
2. (a) Wang, Q.; Chan, T. R.; Hilgraf, R.; Fokin, V. V.;
2003,
Sharpless, K. B.; Finn, M. G. J.Am.Chem.Soc.
125, 3192–3193; (b) Speers, A. E.; Adam, G. C.; Cravatt,
B. F. J.Am.Chem.Soc. 2003, 125, 4686–4687; (c) Link, A.
J.; Tirrell, D. A. J.Am.Chem.Soc.
2003, 125, 11164–
11165; (d) Deiters, A.; Cropp, T. A.; Mukherji, M.; Chin,
J. W.; Anderson, J. C.; Schultz, P. G. J.Am.Chem.Soc.
2003, 125, 11782–11783; (e) Fazio, F.; Bryan, M. C.; Blixt,
O.; Paulson, J. C.; Wong, C.-H. J.Am.Chem.Soc. 2002,
124, 14397–14402; (f) Breinbauer, R.; Ko¨hn, M. Chem-
BioChem. 2003, 4, 1147–1149; (g) Lummerstorfer, T.;
Hoffmann, H. J.Phys.Chem.B 2004, 108, 33966–39663;
(h) Wu, P.; Feldman, A. K.; Nugent, A. K.; Hawker, C. J.;
We have demonstrated here the first asymmetric induc-
tion in the Cu-catalyzed azide–alkyne cycloaddition pro-
cess, both in kinetic resolution and desymmetrization.
The achievement of more highly enantioselective sys-
tems, a target being currently pursued, will aid the syn-
thesis of useful biologically active compounds since the
triazole isemerging asa unit of isgnificant pharmaco-
phoric activity,13 and will assist in the development of
a better understanding of the mechanism of this highly
useful bond-forming process.
´
Scheel, A.; Voit, B.; Pyun, J.; Frechet, J. M. J.; Sharpless,
K. B.; Fokin, V. V. Angew.Chem., Int.Ed. 2004, 43, 3928–
3932.
3. Lewis, W. G.; Magallon, F. G.; Fokin, V. V.; Finn, M. G.
J.Am.Chem.Soc. 2004, 126, 9152–9153.
4. (a) Johnson, J. S.; Evans, D. A. Acc.Chem.Res. 2000, 33,
325–335; (b) Doyle, M. P.; Protopopova, M. N. Tetra-
hedron 1998, 54, 7919–7946; (c) Allais, F.; Angelaud, R.;
Camuzat-Dedenis, B.; Julienne, K.; Landais, Y. Eur.J.

4546
J.Meng et al./ Tetrahedron Letters 46 (2005) 4543–4546
Org.Chem. 2003, 1069–1073; (d) Yang, D.; Yang, M.;
Zhu, N.-Y. Org.Lett. 2003, 5, 3749–3752.
5. (a) Wei, C.; Li, C.-J. J.Am.Chem.Soc. 2002, 124, 5638–
5639; (b) Wei, C.; Mague, J. T.; Li, C.-J. Proc.Natl.Acad.
Sci. U.S.A 2004, 101, 5749–5754.
9. Thompson, A. S.; Humphrey, G. R.; DeMarco, A. M.;
Mathre, D. J.; Grabowski, E. J. J. J.Org.Chem. 1993, 58,
5886–5888.
10. Rodionov, V.; Fokin, V. V.; Finn, M. G. Angew.Chem,.
Int.Ed. 2005, 44, 2210–2215.
6. See Supporting Information for details.
7. Nonlinear effects in mixed-ligand systems are certainly
possible and are under investigation.
11. Himo, F.; Lovell, T.; Hilgraf, R.; Rostovtsev, V. V.;
Fokin, V. V.; Noodleman, L.; Sharpless, K. B. J.Am.
Chem.Soc. 2005, 127, 210–216.
8. Typical procedure (no precautionswere taken to exclude
oxygen): A mixture of ligand 12 (95 mg, 0.2 mmol) and
CuI (19 mg, 0.1 mmol) in 15 mL CH2Cl2 wassonicated at
room temperature for 10 min, and stirred until the
solution became clear. Alkyne 2 (102 mg, 1 mmol) in
10 mL CH2Cl2 was added and the solution was stirred at
room temperature for 30 min before cooling to 0 °C. A
cooled solution of azide 13 (368 mg, 2.5 mmol) in 10 mL
CH2Cl2 wasadded at 0 °C, followed by one drop of
iPr2NEt. The resulting solution was stored at 0–4 °C for
24 h. The solvent was removed and the mixture was
analyzed without purification to determine kinetic selec-
tivity. When desired, pure components were obtained by
silica gel column chromatography, eluting with hexanes to
recover the azide and with a 10:1 hexanes:EtOAc to elute
the product (14). Valuesof enantiomeric excesswere
determined by HPLC using a Chiralcel OD–H column.
12. Kinetic studies show the catalytic reaction to be second
order in copper, and azide to bind competitively to the
metal center.10 The Cu–pybox structure in Figure 2 is
meant only to illustrate the idea that the azide stereogenic
center is positioned close to a pybox ligand by virtue of Cu
interaction with itsproximal N atom, and not to represent
a proposed catalyst structure. Since CuI ismots often
found in tetrahedral coordination environments, bonds
from the pyridine and oxazoline nitrogen atomsare shown
asdotted linesto indicate that not all are expected to be
formed. A [Cu2(pybox)2]2+ system has been structurally
characterized, showing an intriguing helical structure in
the solid phase and in solution Gelalcha, F. G.; Schulz,
M.; Kluge, R.; Sieler, J. J.Chem.Soc.Dalton Trans. 2002,
2517–2521.
13. Kolb, H. C.; Sharpless, K. B. Drug Discovery Today 2003,
8, 1128–1137.