Ruthenium-catalyzed azide-alkyne cycloaddition: Scope and mechanism

Article abstract of DOI:10.1021/ja0749993

The catalytic activity of a series of ruthenium(II) complexes in azide-alkyne cycloadditions has been evaluated. The [Cp*RuCl] complexes, such as Cp*RuCl(PPh₃)₂, Cp*RuCI(COD), and Cp*RuCl(NBD), were among the most effective catalysts. In the presence of catalytic Cp*RuCI(PPh₃)₂ or Cp*RuCl(COD), primary and secondary azides react with a broad range of terminal alkynes containing a range of functionalities selectively producing 1,5-disubstituted 1,2,3-triazoles; tertiary azides were significantly less reactive. Both complexes also promote the cycloaddition reactions of organic azides with internal alkynes, providing access to fully-substituted 1,2,3-triazoles. The ruthenium-catalyzed azide-alkyne cycloaddition (RuAAC) appears to proceed via oxidative coupling of the azide and alkyne reactants to give a six-membered ruthenacycle intermediate, in which the first new carbon-nitrogen bond is formed between the more electronegative carbon of the alkyne and the terminal, electrophilic nitrogen of the azide. This step is followed by reductive elimination, which forms the triazole product. DFT calculations support this mechanistic proposal and indicate that the reductive elimination step is rate-determining.

Full text of DOI:10.1021/ja0749993

Published on Web 06/21/2008
Ruthenium-Catalyzed Azide-Alkyne Cycloaddition: Scope
and Mechanism
Brant C. Boren,† Sridhar Narayan,† Lars K. Rasmussen,† Li Zhang,‡ Haitao Zhao,‡
Zhenyang Lin,*,‡ Guochen Jia,*,‡ and Valery V. Fokin*,†
Department of Chemistry, The Scripps Research Institute, 10550 North Torrey Pines Road, La
Jolla, California 92037 and Department of Chemistry, The Hong Kong UniVersity of Science
and Technology, Clear Water Bay, Kowloon, Hong Kong, China
Received July 6, 2007; E-mail: fokin@scripps.edu; chzlin@ust.hk; chjiag@ust.hk
Abstract: The catalytic activity of a series of ruthenium(II) complexes in azide-alkyne cycloadditions has
been evaluated. The [Cp*RuCl] complexes, such as Cp*RuCl(PPh3)2, Cp*RuCl(COD), and Cp*RuCl(NBD),
were among the most effective catalysts. In the presence of catalytic Cp*RuCl(PPh3)2 or Cp*RuCl(COD),
primary and secondary azides react with a broad range of terminal alkynes containing a range of
functionalities selectively producing 1,5-disubstituted 1,2,3-triazoles; tertiary azides were significantly less
reactive. Both complexes also promote the cycloaddition reactions of organic azides with internal alkynes,
providing access to fully-substituted 1,2,3-triazoles. The ruthenium-catalyzed azide-alkyne cycloaddition
(RuAAC) appears to proceed via oxidative coupling of the azide and alkyne reactants to give a six-membered
ruthenacycle intermediate, in which the first new carbon-nitrogen bond is formed between the more
electronegative carbon of the alkyne and the terminal, electrophilic nitrogen of the azide. This step is followed
by reductive elimination, which forms the triazole product. DFT calculations support this mechanistic proposal
and indicate that the reductive elimination step is rate-determining.
Introduction
The copper-catalyzed 1,3-dipolar azide-alkyne cycloaddition
(CuAAC) was an important advance in the chemistry of
1,2,3-triazoles.7,8 An enormous rate acceleration (107 to 108
comparing to the uncatalyzed process6), a remarkably broad
scope, a tolerance to aqueous and oxic conditions, and an
exclusive regioselectivity have enabled a number of applications
in the relatively short time since the reaction was discovered.9–14
The especially narrow reactivity profiles of azides and alkynes
and the efficiency of their copper-catalyzed union have been
recognized across disciplines, as evidenced by its the applica-
tions in bioorganic chemistry, organic synthesis, and materials
and polymer science.
While Cu(I) catalysis provides a reliable means for the
assembly of 1,4-disubstituted 1,2,3-triazoles, a general method
for the generation of 1,5-disubstitued regioisomers is lacking.
Among the available methods are the reactions of stabilized
phosphonium ylides15–20 or enamines2,21 with aryl azides and
the addition of magnesium acetylides to azides.22 However, these
methods have considerable limitations.
1,2,3-Triazoles have been known for over a century,1 yet they
have not been utilized as widely as other members of the azole
family. The conspicuous lack of 1,2,3-triazoles in the literature
is likely due to the limited repertoire of synthetic methods
leading to these heterocycles. Among those,2 the Huisgen 1,3-
dipolar cycloaddition3,4 of azides and alkynes prominently stands
out as the most direct way to assemble the 1,2,3-triazole
functionality.5 Although the reaction is highly exothermic (ca.
-50 to 65 kcal/mol), its high activation barrier (25-26 kcal/
mol for methyl azide and propyne6) results in exceedingly low
reaction rates for unactivated reactants even at elevated tem-
perature. Furthermore, since the difference in HOMO-LUMO
energy levels for both azides and alkynes are of similar
magnitude, both dipole-HOMO- and dipole-LUMO-controlled
pathways operate in these cycloadditions. As a result, a mixture
of regioisomeric 1,2,3-triazole products is usually formed when
an alkyne is unsymmetrically substituted.
We recently disclosed that pentamethylcyclopentadienyl
ruthenium chloride [Cp*RuCl] complexes catalyze the facile
† The Scripps Research Institute.
‡ The Hong Kong University of Science and Technology.
(1) Michael, A. J. Prakt. Chem. 1893, 48, 94.
(7) Tornoe, C. W.; Christensen, C.; Meldal, M. J. Org. Chem. 2002, 67,
3057.
(2) Tome, A. C. In Science of Synthesis; Storr, R. C., Gilchrist, T. L.,
Eds.; Thieme: Stuttgart, New York 2004; Vol. 13, pp 415.
(3) 1,3-Dipolar cycloaddition chemistry; Padwa, A., Ed.; Wiley: New
York, 1984.
(4) Synthetic applications of 1,3 dipolar cycloaddition chemistry toward
heterocycles and natural products; Padwa, A.; Pearson, W. H., Eds.;
Wiley: New York, 2002.
(5) Lwowski, W. In 1,3-Dipolar Cycloaddition Chemistry; Padwa, A.,
Ed.; Wiley: New York, 1984; Vol. 1, pp 559.
(6) Himo, F.; Lovell, T.; Hilgraf, R.; Rostovtsev, V. V.; Noodleman, L.;
Sharpless, K. B.; Fokin, V. V. J. Am. Chem. Soc. 2005, 127, 210.
(8) Rostovtsev, V. V.; Green, L. G.; Fokin, V. V.; Sharpless, K. B. Angew.
Chem., Int. Ed. 2002, 41, 2596.
(9) Bock, V. D.; Hiemstra, H.; van Maarseveen, J. H. Eur. J. Org. Chem.
2006, 51.
(10) Kolb, H. C.; Sharpless, K. B. Drug DiscoVery Today 2003, 8, 1128.
(11) Lutz, J.-F. Angew. Chem., Int. Ed. 2007, 46, 1018.
(12) Wu, P.; Fokin, V. V. Aldrichimica Acta 2007, 40, 7.
(13) Moses, J. E.; Moorhouse, A. D. Chem. Soc. ReV. 2007, 36, 1249.
(14) Nandivada, H.; Jiang, X.; Lahann, J. AdV. Mater. (Weinheim, Germany)
2007, 19, 2197.
9
10.1021/ja0749993 CCC: $40.75
2008 American Chemical Society
J. AM. CHEM. SOC. 2008, 130, 8923–8930 8923

A R T I C L E S
Boren et al.
cycloaddition of azides with terminal alkynes, regioselectively
leading to 1,5-disubstituted 1,2,3-triazoles.23 Furthermore, and
in stark contrast to the CuAAC, its sister ruthenium-catalyzed
process, RuAAC, readily engages internal alkynes in catalysis,
providing access to fully substituted 1,2,3-triazoles. In this
report, we summarize the results of our studies of the scope,
limitations, and mechanism of the RuAAC reaction.
sive number of ruthenium mediated reactions of alkynes have
been developed, especially those involving vinylidene, alle-
nylidene, metallacyclopentene, and metallacyclopentadiene in-
termediates. Direct addition processes include cyclotrimerization
to generate benzene rings,29–34 [2 + 2] cycloadditions of alkynes
with olefins,35–38 Alder ene reactions of alkynes and alkenes,39
and enyne metathesis.40 Formal cycloadditions in which the third
component is an oxygen nucleophile41 or carbon monoxide, as
in the Pauson-Khand reaction, are known as well.42,43 Ruthe-
nium-mediated propargylic substitution,44–46 anti-Markovnikov
addition,26 and related methods have also found broad utility.
However, catalytic dipolar cycloadditions are absent in this list.47
Keeping in mind the wide variety of transformations of alkynes
catalyzed by ruthenium species,48 we began our study with
ruthenium complexes known to engage alkynes in catalysis.
Catalysts, Ligands, and Experimental Conditions. The cata-
lytic activity of several ruthenium(II) complexes was evaluated
using the cycloaddition of benzyl azide and phenylacetylene.
Solvents such as dioxane, benzene, 1,2-dichloroethane, toluene,
or THF were employed. As previously reported, η5-pentameth-
ylcyclopentadienyl ruthenium [Cp*RuCl] complexes were par-
ticularly effective in promoting the reaction (Table 1). The
unique catalytic properties of these complexes can be ascribed
to the presence of the electron-rich Cp* ligand, which stabilizes
higher formal oxidation states of the metal center. Ruthenium(II)
complexes lacking this ligand, such as (COD)RuCl2, [(p-
cym)RuCl2]2, (p-cym)RuCl2(PPh3), RuCl2(PPh3)3, and RuH-
Cl(CO)(PPh3)3, showed no appreciable catalytic activity; CpRu-
Cl(PPh3)2, RuH2(CO)(PPh3)3, and Ru(OAc)2(PPh3)2 were
marginally effective in catalyzing the cycloaddition. It is
noteworthy that reactions catalyzed by RuCl2(PPh3)3, RuH-
Cl(CO)(PPh3)3, RuH2(CO)(PPh3)3, and Ru(OAc)2(PPh3)2 re-
sulted in the formation of the 1,4-substituted triazole regioiso-
mer, albeit in low yields. CpRuCl(PPh3)2 was a modestly active
Experimental Section
Unless stated otherwise, all reagents and solvents were purchased
from commercial suppliers and used without further purification.
Most manipulations were carried out under a nitrogen atmosphere
using standard Schlenck techniques unless otherwise stated. See
Supporting Information for complete details.
Procedure A. General procedure for Cp*RuCl(PPh3)2 catalyzed
reactions as exemplified for the synthesis of triazole 3: A solution
of 1-ethynylcyclohexanol (51 µL, 0.40 mmol) and ethyl 2-(2-
azidoacetamido)-3-hydroxypropanoate (86 mg, 0.40 mmol) in 0.5
mL of dioxane was added to Cp*RuCl(PPh3)2 (6.4 mg, 0.008 mmol)
dissolved in 2.5 mL of dioxane. The vial was purged with nitrogen,
sealed, and heated in an oil bath at 60 °C for 12 h, at which point
TLC and LC-MS analyses indicated complete consumption of the
alkyne and the azide starting materials. The mixture was adsorbed
onto silica and chromatographed with hexanes/ethyl acetate (1:1)
to remove nonpolar impurities, followed by ethyl acetate to elute
the product, which was isolated as a pale yellow oil. (Note: in all
reactions the alkyne was added first, followed by the addition of
the azide, or azide and alkyne were dissolved in the reaction solvent
and added to the solution of the catalyst.)
Procedure B. General procedure for Cp*RuCl(COD) catalyzed
reactionsasexemplifiedforthesynthesisoftriazole14:Cp*RuCl(COD)
(4.0 mg, 0.010 mmol) was added to a tube with a septa cap. The
tube was sealed, then evacuated, and filled with nitrogen three times.
Toluene (5 mL, degassed for 1 h with nitrogen purge) was added
followed by 2-methyl-4-phenylbut-3-yn-2-ol (80 mg, 0.50 mmol)
and 1-azido-4-methylbenzene (67 mg, 0.50 mmol). The reaction
was stirred at room temperature for 30 min, at which time TLC
analysis indicated complete consumption of the starting materials.
The mixture was adsorbed onto silica and chromatographed with
4:1 hexanes/ethyl acetate and then 2:1 hexanes/ethyl acetate to
afford the pure product as a white solid. (Note: in all reactions the
alkyne was added first, followed by the addition of the azide, or
azide and alkyne were dissolved in the reaction solvent and added
to the solution of the catalyst; azide should not be added first.)
Computational Details. See Supporting Information.
(29) Ura, Y.; Sato, Y.; Shiotsuki, M.; Kondo, T.; Mitsudo, T. J. Mol. Catal.
A: Chem. 2004, 209, 35.
(30) Yamamoto, Y.; Arakawa, T.; Ogawa, R.; Itoh, K. J. Am. Chem. Soc.
2003, 125, 12143.
(31) Yamamoto, Y.; Ishii, J.-i.; Nishiyama, H.; Itoh, K. J. Am. Chem. Soc.
2005, 127, 9625.
(32) Yamamoto, Y.; Kitahara, H.; Ogawa, R.; Kawaguchi, H.; Tatsumi,
K.; Itoh, K. J. Am. Chem. Soc. 2000, 122, 4310.
(33) Yamamoto, Y.; Takagishi, H.; Itoh, K. J. Am. Chem. Soc. 2002, 124,
28.
(34) Ruba, E.; Schmid, R.; Kirchner, K.; Calhorda, M. J. J. Organomet.
Chem. 2003, 682, 204.
Results and Discussion
In the past three decades, ruthenium-catalyzed transformations
of alkynes have received considerable attention.24–28 An impres-
(35) Mitsudo, T.; Naruse, H.; Kondo, T.; Ozaki, Y.; Watanabe, Y. Angew.
Chem., Int. Ed. 1994, 33, 580.
(36) Mitsudo, T.; Kokuryo, K.; Takegami, Y. J. Chem. Soc., Chem.
Commun. 1976, 722.
(15) Labbe, G.; Ykman, P.; Smets, G. Tetrahedron 1969, 25, 5421.
(16) Ykman, P.; Labbe, G.; Smets, G. Tetrahedron Lett. 1970, 5225.
(17) Ykman, P.; Labbe, G.; Smets, G. Tetrahedron 1971, 27, 5623.
(18) Ykman, P.; Labbe, G.; Smets, G. Tetrahedron 1971, 27, 845.
(19) Ykman, P.; Labbe, G.; Smets, G. Tetrahedron 1973, 29, 195.
(20) Ykman, P.; Smets, G.; Mathys, G.; Labbe, G. J. Org. Chem. 1972,
37, 3213.
(21) Fioravanti, S.; Pellacani, L.; Ricci, D.; Tardella, P. A. Tetrahedron-
Asymmetry 1997, 8, 2261.
(22) Krasinski, A.; Fokin, V. V.; Sharpless, K. B. Org. Lett. 2004, 6, 1237.
(23) (a) Zhang, L.; Chen, X.; Xue, P.; Sun, H. H. Y.; Williams, I. D.;
Sharpless, K. B.; Fokin, V. V.; Jia, G. J. Am. Chem. Soc. 2005, 127,
15998. (b) Rasmussen, L. K.; Boren, B. C.; Fokin, V. V. Org. Lett.
2007, 9, 5337.
(37) Jordan, R. W.; Khoury, P. R.; Goddard, J. D.; Tam, W. J. Org. Chem.
2004, 69, 8467.
(38) Jordan, R. W.; Tam, W. Org. Lett. 2000, 2, 3031.
(39) Trost, B. M.; Indolese, A. F.; Muller, T. J. J.; Treptow, B. J. Am.
Chem. Soc. 1995, 117, 615.
(40) Poulsen, C. S.; Madsen, R. Synthesis 2003, 1.
(41) Le Paih, J.; Derien, S.; Demerseman, B.; Bruneau, C.; Dixneuf, P. H.;
Toupet, L.; Dazinger, G.; Kirchner, K. Chem.sEur. J. 2005, 11, 1312.
(42) Chatani, N.; Morimoto, T.; Fukumoto, Y.; Murai, S. J. Am. Chem.
Soc. 1998, 120, 5335.
(43) Kondo, T.; Suzuki, N.; Okada, T.; Mitsudo, T. J. Am. Chem. Soc.
1997, 119, 6187.
(44) Ammal, S. C.; Yoshikai, N.; Inada, Y.; Nishibayashi, Y.; Nakamura,
E. J. Am. Chem. Soc. 2005, 127, 9428.
(24) Bruneau, C. Top. Organomet. Chem. 2004, 11, 125.
(25) Dixneuf, P. H. Pure Appl. Chem. 1989, 61, 1763.
(26) Dixneuf, P. H.; Bruneau, C.; Derien, S. Pure Appl. Chem. 1998, 70,
1065.
(45) Nishibayashi, Y.; Milton, M. D.; Inada, Y.; Yoshikawa, M.; Wakiji,
I.; Hidai, M.; Uemura, S. Chem.sEur. J. 2005, 11, 1433.
(46) Nishibayashi, Y.; Wakiji, I.; Hidai, M. J. Am. Chem. Soc. 2000, 122,
11019.
(27) Trost, B. M. Chem. Ber. 1996, 129, 1313.
(47) Schore, N. E. Chem. ReV. 1988, 88, 1081.
(48) Derien, S.; Dixneuf, P. H. J. Organomet. Chem. 2004, 689, 1382.
(28) Trost, B. M. Acc. Chem. Res. 2002, 35, 695.
9
8924 J. AM. CHEM. SOC. VOL. 130, NO. 28, 2008

Ruthenium-Catalyzed Azide-Alkyne Cycloaddition
A R T I C L E S
Table 1. Performance of Ruthenium(II) Catalysts in Cycloaddition
is evidenced by the activity of the former catalyst even at room
temperature. This feature of the Cp*RuCl(COD) catalyst is
particularly advantageous for the reactions involving internal
alkynes or aryl azides. The deactivation of the catalyst via the
formation of the tetraazadiene complex, discussed below, is
common for aryl azides, but it is minimized at room temperature,
and the triazoles are obtained in good to excellent yields and
with excellent regioselectivity.
of Benzyl Azide and Phenyl Acetylenea
Ru catalyst
1,5
1,4
[RuCl2(COD)]x
-
-
-
-
-
-
-
-
-
-
-
-
[(p-cym)RuCl2]2
[(p-cym)RuCl2(PPh3)
[Cp*Ru(H2O)(NBD)]BF4
RuCl2(PPh3)3
Ru(OAc)2(PPh3)2
RuHCl(CO)(PPh3)3
RuH2(CO)(PPh3)3
CpRuCl(PPh3)2
Cp*RuCl(PPh3)2
Cp*RuCl(COD)
Cp*RuCl(NBD)
[Cp*RuCl]4
<5%
46%
<5%
56%
1%
Scope of the [Cp*RuCl]-Catalyzed Cycloadditions. The
ruthenium-catalyzed reactions described herein exhibit excellent
scope with respect to both reactants, successfully engaging a
wide range of azides and alkynes in the catalysis. The steric
demands of the azide substituent significantly affect the outcome
of the cycloaddition. Thus, reactions involving primary azides
were most efficient and cycloadditions of secondary azides often
took longer and resulted in slightly lower yields, whereas tertiary
azides, with a few exceptions, reluctantly participated in
catalysis. On the other hand, the catalysis is not very sensitive
to the substituents on the alkyne. Worthy of note, certain alkyne
classes, such as 1,1-disubstituted propargyl alcohols, exhibited
especially high reactivity (Vide infra). Table 2 highlights
reactions of terminal alkynes with alkyl azides catalyzed by
Cp*RuCl(PPh3)2 and Cp*RuCl(COD) complexes (either catalyst
can be used in most cases). As seen from these examples,
terminal alkynes containing halide, alcohol, ether, acetal, nitrile,
ester, amine, sulfonamide, and pyridine functionalities readily
participated in the catalysis. In most cases 2 mol % of the
catalyst was used, and the reactions were carried out in THF or
dioxane at temperatures ranging from ambient to 80 °C. In all
cases involving terminal alkynes, only the 1,5-disubstituted
triazoles were formed. The only combination of the reactants
that gave a mixture of 1,4- and 1,5-regioisomers we have found
is benzyl azide and trimethylsilyl acetylene. Their reaction in
refluxing benzene in the presence of 2 mol % Cp*RuCl(PPh3)2
catalyst resulted in a mixture of 1,5- and 1,4-disubstituted
triazoles in 98:2 ratio.
13%
100%
100%
93%
100%
a Reaction conditions: refluxing THF (oil bath temp ) 75 °C), 2
mol% [Ru], azide/alkyne ) 1:1.2, 4 h. The yields are estimated based
on the integration of triazole and the unreacted azide. “-” means not
detected or trace.
catalyst as well, resulting in the formation of a 5.8:1 (in benzene)
or 10:1 (in THF) mixture of 1,5- and 1,4-regioisomers.
[Cp*RuCl] complexes, e.g., [Cp*RuCl]4, Cp*RuCl(PPh3)2,
Cp*RuCl(COD), and Cp*RuCl(NBD), were especially active
and selective catalysts, producing the 1,5-disubstituted triazole
product in excellent yields. Further studies described below
focused on the bis(triphenylphosphine) (1) and cyclooctadiene
(2) catalysts due to their ready synthetic availability and stability.
Both complexes can be conveniently obtained by treatment of
the [Cp*RuCl]4 precursor with an excess of PPh3 or cyclooc-
tadiene, respectively.26
As already mentioned, the [Cp*RuCl]-catalyzed reactions
appear to be more sensitive to the steric demands of the azide
than those of the alkyne. Thus, secondary azides reacted
significantly slower than primary ones, and tertiary azides were
even less reactive under the current conditions. Nevertheless,
as demonstrated by the examples in Table 2, cycloaddition
products of secondary azides were obtained in good yields.
Azides containing complex functional groups, such as 3′-azido-
2′-deoxythymidine (AZT), could also be readily coupled under
these conditions. As in the case of the Cu(I)-catalyzed cycload-
dition, the ability to carry out Ru-catalyzed triazole formation
from highly functionalized reactants is especially valuable, and
we anticipate that the present method may serve as a useful
platform for the generation of diversely substituted molecules
that are structurally distinct from the compounds available from
the CuAAC reaction.
A remarkable feature of the [Cp*RuCl] catalytic system is
its ability to engage internal alkynes in the cycloaddition with
organic azides resulting in the formation of 1,4,5-trisubstituted
1,2,3-triazoles. To date, these fully substituted triazoles were
only accessible from particularly activated, electron-deficient
alkynes (such as acetylene dicarboxylic acid derivatives) by
thermal cycloadditions, which are generally not regioselective
and often provide triazole products only in modest yields. In
contrast, in the presence of [Cp*RuCl], organic azides react with
a range of internal alkynes to provide 1,4,5-trisubstituted triazole
products (Table 3). Regioselectivity of the cycloaddition is
The [Cp*RuCl]-catalyzed cycloadditions proceed well in a
variety of aprotic organic solvents including tetrahydrofuran,
dioxane, toluene, benzene, dimethylformamide, and 1,2-dichlo-
roethane. Performing the reaction in protic solvents (MeOH,
i-PrOH) resulted in reduced yields and caused formation of
byproducts. Nevertheless, the presence of adventitious water
or protic functional groups in the reactants (Vide infra) usually
did not affect the performance of the catalysts; in other words,
the solvents employed need not be scrupulously dried. As
reported earlier, the cycloadditions with Cp*RuCl(PPh3)2 can
be carried out at temperatures ranging from ambient to 110 °C.
The reaction does not appear to be very sensitive to atmospheric
oxygen. Indeed, the cycloaddition of benzylazide and
Ph2C(OH)Ct CH in the presence of
1
mol
%
of
Cp*RuCl(PPh3)2 in benzene also proceeded smoothly even when
the reaction was performed without exclusion of oxygen (4 h
at reflux, >99% conversion). It should be noted that
Cp*RuCl(PPh3)2 is known to react with dioxygen to give
Cp*RuCl(O2)(PPh3). The dioxygen ligand in the latter complex
is not tightly bound and can be replaced with a phosphite ligand
L to give Cp*RuCl(L)(PPh3).28 Evidently, in our catalytic
reactions, atmospheric oxygen does not effectively compete with
the reactants.
The cyclooctadiene ligand in Cp*RuCl(COD) catalyst is more
labile than phosphines and is displaced more readily than
phosphine ligands in the bis(triphenylphosphine) complex. This
9
J. AM. CHEM. SOC. VOL. 130, NO. 28, 2008 8925

A R T I C L E S
Boren et al.
Table 2. [Cp*RuCl]-Catalyzed Reactions of Azides with Terminal Alkynes
influenced by several factors in these cases. Alkynes containing
a hydrogen bond donor group (e.g., propargylic alcohols and
amines) exhibit virtually exclusive regioselectivity: the new bond
is always formed between the ꢀ-carbon of the alkyne and the
terminal nitrogen of the azide. This directing effect can be
attributed to the formation of a strong H-bond between the
alcohol or amine and the chloride ligand on the ruthenium. In
the absence of such directing groups, regioselectivity seems to
be governed primarily by the electronic (and possibly by steric)
properties of the alkyne: the new bond in the metallacycle
intermediate 28 (Scheme 3) is formed between the more
nucleophilic carbon of the alkyne. In other words, the more
electronegative carbon of the alkyne becomes C-4 in the triazole.
Cp*RuCl(PPh3)2 readily catalyzes reactions of aliphatic azides
with internal alkynes at 60 °C, whereas the cyclooctadiene
analogue (2) exhibits excellent activity already at room tem-
perature. The advantages of the Cp*RuCl(COD) catalyst are
also evidenced by the reactions of aryl azides, which often suffer
from decomposition and catalyst deactivation at elevated
temperatures. In fact, we found that heating is often detrimental
in the reactions catalyzed by Cp*RuCl(COD), resulting in rapid
deactivation of the catalyst and low yields of the cycloaddition
product.
Possible pathways of deactivation of the ruthenium catalyst
were also examined. Although [Cp*RuCl] can catalyze cyclo-
trimerization of alkynes, we did not observe formation of the
benzenoid products. Likewise, ruthenacycles, whose interme-
diacy has been postulated in the catalytic alkyne trimerizations
and which are known to form quite rapidly even at 0 °C,49,50
were not detected among the species present in the reaction
mixture. Furthermore, ruthenacycle 24 prepared from the 1,6-
diyne shown in Scheme 1 using a slightly modified procedure
reported by Yamamoto et al.51 was not reactive with an organic
azide even after prolonged (several days) heating. Similarly,
other 1,6-diynes failed to produce triazole products when they
were added, regardless of the order of addition, to the solution
of the Cp*RuCl(COD) catalyst. It appears that the formation
of ruthenacyclopentatrienes similar to 24 still prevails when 1,6-
diynes are used as substrates, and the catalyst is effectively taken
out of the catalytic cycle. With other alkynes, the formation of
the ruthenacyclopentatriene complexes is suppressed and the
(49) Ernst, C.; Walter, O.; Dinjus, E. J. Organomet. Chem. 2001, 627, 249.
(50) Ernst, C.; Walter, O.; Dinjus, E.; Arzberger, S.; Gorls, H. J. Prakt.
Chem. 1999, 341, 801.
(51) Yamamoto, Y.; Kinpara, K.; Saigoku, T.; Takagishi, H.; Okuda, S.;
Nishiyama, H.; Itoh, K. J. Am. Chem. Soc. 2005, 127, 605.
9
8926 J. AM. CHEM. SOC. VOL. 130, NO. 28, 2008

Ruthenium-Catalyzed Azide-Alkyne Cycloaddition
A R T I C L E S
Table 3. [Cp*RuCl]-Catalyzed Cycloadditions of Azides and Internal Alkynes
Scheme 1. Syntheis of Ruthenacycle 24 and Its Reaction with
significant in itself (only a few similar ruthenium complexes
have been reported in the literature52,53), more important is the
realization that formation of 25 is a likely cause of diminished
catalytic activity of certain [Cp*RuCl] systems, especially in
cases involving organic azides which are prone to decomposition
and can readily engage in the formation of the imido complexes.
This finding leads to two practical conclusions: (a) reactions of
aryl azides, which are less stable than their aliphatic counterparts
and are prone to the formation of the ruthenium imido
complexes, are best performed at room temperature and (b) azide
should not be added to the catalyst before the alkyne. In our
experience, the best results are obtained when a mixture of the
azide and the alkyne is added to the solution of the catalyst.
(2-Azidoethyl)benzene
desired azide-alkyne cycloaddition takes place, likely as a
consequence of the strong σ-coordination of the azide to
ruthenium via the N-1 of the azide.
However, treatment of the Cp*RuCl(COD) catalyst at room
temperature with 2 equiv of (2-azidoethyl)benzene readily
produced ruthenium tetraazadiene complex 25 (or possibly
tetrazenide 26), shown in Scheme 2. This complex was isolated
by flash chromatography and appeared to be exceedingly stable.
Like ruthenacyclopentatriene 24, it failed to react further with
either alkynes or azides. While isolation of this compound is
(52) Danopoulos, A. A.; Hay-Motherwell, R. S.; Wilkinson, G.; Cafferkey,
S. M.; Sweet, T. K. N.; Hursthouse, M. B. J. Chem. Soc., Dalton
Trans. 1997, 3177.
(53) Danopoulos, A. A.; Wilkinson, G.; Sweet, T. K. N.; Hursthouse, M. B.
J. Chem. Soc., Dalton Trans. 1996, 3771.
9
J. AM. CHEM. SOC. VOL. 130, NO. 28, 2008 8927

A R T I C L E S
Boren et al.
Scheme 2. Formation of Catalytically Inactive [Cp*RuCl] Tetraazadiene Complexes
Scheme 3. Proposed Intermediates in the Catalytic Cycle of the
RuAAC Reaction
of the overall process. The new CsN bond is formed between
the more electronegative and less sterically-demanding carbon
of the alkyne and the terminal nitrogen of the azide. The
metallacycle intermediate then undergoes reductive elimination
(step C) releasing the aromatic triazole product and regenerating
the catalyst (step D) or the activated complex 27.
As noted above, the [CpRuCl] complexes, such as CpRu-
Cl(PPh3)2, also catalyze the reaction, albeit much less efficiently
and not regioselectively. The higher catalytic activity of
[Cp*RuCl] catalysts can be attributed to the lability of the
bystander ligands in such systems34 (thus facilitating the
formation of the activated complex 27), the more sterically
demanding nature of the Cp* ligand (which, in turn, facilitates
the reductive elimination, step C, in the catalytic cycle), and
prevents formation of the stabilized ruthenacycle IN2A (vide
infra).
The proposed reaction mechanism was examined computa-
tionally employing DFT calculations using the B3LYP hybrid
functional. In this computational study, methyl azide and
propyne were used as the model reactants with [CpRuCl] as a
catalyst. Methyl azide can, in principle, coordinate to the metal
center via the nitrogen proximal to carbon or via the distal
nitrogen atom. Both modes of coordination are known, although
the former is observed by far more commonly.55 The alkyne
can also coordinate to the metal center in a π-fashion in two
distinct orientations. Therefore, four activated azide/[Ru]/alkyne
complexes, PCA, PCB, PCC and PCD, are possible, as shown
in Figure 1. PCA and PCC complexes lead to the 1,5-
disubstituted triazole product, whereas PCB and PCD would
result in the formation of the 1,4-regioisomer. The computed
energies of the complexes are within 1 kcal/mol. The results
are consistent with those of a recent theoretical study.56
The energy profile for the reaction of PCA is illustrated in
Figure 2. The first step, oxidative coupling of methyl azide and
propyne, results in the formation of the six-membered ruthena-
cycle IN1A. This step is exothermic by 13.2 kcal/mol, and the
calculated barrier is only 4.3 kcal/mol. IN1A can undergo
reductive elimination via transition state TS3A (the rate-
determining step, 13 kcal/mol activation energy) forming the
metal-triazole complex PRA or can isomerize to a more stable
ruthenacycle IN2A with a very low barrier of 1.6 kcal/mol
(TS2A). In other words, the facility and the outcome of the
overall process are influenced by the relative energies and the
ease of interconversion of intermediates IN1A and IN2A. We
propose that the formation of the IN2A is disfavored in
[Cp*RuCl]-catalyzed reactions, and IN1A is directly converted
to PRA. The activation energy of the reaction is therefore
Mechanistic Considerations and DFT Studies. Participation
of both terminal and internal alkynes in catalysis suggests that
ruthenium acetylides are not involved in the catalytic cycle, and
our experiments with isolated ruthenium(II) acetylides, as well
as with vinylidene complexes, support this hypothesis. Since
different Ru(II) complexes, such as Cp*RuCl(PPh3)2, Cp*RuCl-
(COD), Cp*RuCl(NBD), and [Cp*RuCl]4 are competent cata-
lysts, we propose that the neutral [Cp*RuCl] is the catalytically
active species, especially in reactions mediated by Cp*RuCl-
(diolefin) and [Cp*RuCl]4. This hypothesis is supported by the
observations that (a) [Cp*RuBr] and [Cp*RuI] complexes are
significantly less active catalysts, (b) [Cp*Ru]+ cationic com-
plexes obtained by the removal of the chloride with Ag+ are
devoid of the catalytic activity altogether, and (c) chelating
diphosphines, such as bis(diphenylphosphino)ethane, completely
deactivate the catalyst.
Catalytic cyclotrimerization of alkynes catalyzed by [Cp*RuCl]
and similar complexes is, of course, well-known,29,30,32,33,51,54
and we believe that the RuAAC is related to it. Cyclotrimer-
ization catalyzed by the Cp*RuCl(COD) has been shown to
proceed via ruthenacyclopentadiene intermediates. Therefore,
we hypothesize that RuAAC represents a simple case that shunts
off the usual alkyne oligomerization sequence, as schematically
outlined in Scheme 3. The displacement of the spectator ligands
(step A) produces the activated complex 27, which is converted,
via the oxidative coupling of an alkyne and an azide (step B),
to the ruthenacycle 28. This step controls the regioselectivity
(54) Kirchner, K.; Calhorda, M. J.; Schmid, R.; Veiros, L. F. J. Am. Chem.
Soc. 2003, 125, 11721.
Figure 1. Structures and computed energies of the activated complexes.
9
8928 J. AM. CHEM. SOC. VOL. 130, NO. 28, 2008

Ruthenium-Catalyzed Azide-Alkyne Cycloaddition
A R T I C L E S
Figure 2. Schematic representation (energy vs reaction coordinate) of the reaction of organic azides and alkynes catalyzed by [CpRuCl].
Figure 3. Selected structural parameters (Å) calculated for species involved in the oxidative coupling steps of pathways A and B (from PCA and PCB,
respectively).
reduced from 24 kcal/mol for the uncatalyzed process6 to 13
kcal/mol, accounting for the observed rate acceleration. Pathway
B, leading to the 1,4-regioisomer (see Supporting Information
for complete details) is disfavored by ca. 3 kcal/mol in the
oxidative coupling step. Further examination of the structural
features of the species involved in the oxidative coupling steps
(Figure 3) reveals the steric repulsion between the methyl
substituent of the propyne and the hydrogens on the Cp ring in
the metallacycle intermediate IN1B, whereas it is avoided in
the intermediate IN1A. The Ru-Cꢀ bond distance in the
intermediate IN1B is 2.274 Å, longer than the Ru-Cꢀ distance
in IN1A, supporting the steric repulsion argument. Clearly, these
9
J. AM. CHEM. SOC. VOL. 130, NO. 28, 2008 8929

A R T I C L E S
Boren et al.
steric interactions are amplified in the [Cp*RuCl] complexes,
explaining the high regioselectivity of those catalytic systems.
Pathways beginning from the azide coordinated to ruthenium
via the terminal nitrogen (starting from PCC and PCD) were
also evaluated, revealing much higher activation barriers for
the formation of even the first intermediate, 20.9 and 17.8 kcal/
mol, respectively.
The computational results are in general agreement with the
observed performance of the CpRuCl(PPh3)2 catalyst: the
reactions are much slower than those with the Cp* analogue,
the regioselectivity does not exceed ca. 85:15 (in favor of the
1,5-regioisomer), and the yields are modest at best. Higher
sensitivity of the catalysis to the steric demands of the azide
substituent are also explained by this model: σ-coordination of
the azide to the ruthenium atom via its N-1 nitrogen is a required
event for the subsequent oxidative addition step, and the stability
of the complex PCA (and the strength of this interaction) is
directly affected by the size of the azide substituent.
Based on these results, we propose that the mechanism of
the Ru(II)-catalyzed cycloaddition reactions of azides with
alkynes involves an irreversible oxidative coupling, followed
by a rate-determining reductive elimination. The DFT calcula-
tions demonstrate that the oxidative coupling step is, in essence,
a nucleophilic attack of the coordinated alkyne on the terminal,
electrophilic nitrogen of the coordinated azide. Both steric and
electronic factors favor the pathway starting from the PCA
activated complex species which leads to the 1,5-regioselectivity
observed experimentally.
investigation of the catalytic activities of a series of ruthenium
complexes, [Cp*RuCl] compounds emerged as the most efficient
and regioselective catalysts. In the presence of the readily
synthesized and air stable Cp*RuCl(PPh3)2 catalyst, a range of
organic azides react with terminal alkynes containing a variety
of functionalities, usually at elevated temperature, to give
selectively 1,5-disubstituted triazole products. The Cp*RuCl(COD)
catalyst exhibits higher activity and is particularly suitable for
ambient temperature cycloadditions involving internal alkynes,
aryl azides, and generally thermally labile reactants. Neverthe-
less, both catalysts exhibit excellent activity as well as chemo-
and regioselectivity. Together with the CuAAC reaction, the
1
new RuAAC process provides ready access to all H-1,2,3-
triazole regioisomers.
Computational studies indicate that the [Cp*RuCl]-catalyzed
reactions of azides with alkynes involve an irreversible oxidative
coupling of azide and alkyne to give ruthenacycles, followed
by a rate-determining reductive elimination. The regioselectivity
is determined by the oxidative coupling step, which can also
be viewed as a nucleophilic attack of the activated alkyne at
the electrophilic terminal nitrogen of the azide.
Acknowledgment. This work was supported by the Skaggs
Institute for Chemical Biology, the Scripps Research Institute, and
Pfizer, Inc. (V.V.F.), and the Hong Kong Research Grant Council
(G.J.; Z.L). We are grateful to Dr. J. C. Loren and Prof. K. B.
Sharpless for stimulating discussions and advice.
Conclusion
Supporting Information Available: Experimental procedures,
compound characterization data, X-ray crystallographic data,
and details of computational studies. This material is available
free of charge via the Internet at http://pubs.acs.org.
A general catalytic system for the cycloaddition of azides
with both internal and terminal alkynes is now available. After
(55) Cenini, S.; Gallo, E.; Caselli, A.; Ragaini, F.; Fantauzzi, S.; Piangiolino,
C. Coord. Chem. ReV. 2006, 250, 1234.
(56) Zhou, Y. B.; Ke, Z. F.; Zhao, C. Y. Acta Chim. Sinica 2006, 64, 2071.
JA0749993
9
8930 J. AM. CHEM. SOC. VOL. 130, NO. 28, 2008