Copper(I)-Catalyzed Cycloaddition of Azides to Multiple Alkynes: A Selectivity Study Using a Calixarene Framework

Article abstract of DOI:10.1002/chem.201500946

Copper(I)-catalyzed addition of limited amounts of azides to multiple alkynes, which led to statistical mixtures of triazole/acetylene derivatives or, in other cases, resulted in preferred formation of multiple triazoles, was studied at pre-organizable calixarene platforms bearing up to four propargyl groups. Depending on calixarene structures and reaction conditions, the unprecedented specific or selective formation of exhaustively triazolated calixarenes or a complete loss of the selectivity were observed. Both autocatalytic copper activation and a local copper(I) concentration increase due to copper-triazole complexation were thoroughly studied as the most expected reasons for the selectivity and both were disproved. Mixed triazolated/propargylated calixarenes and their copper(I) complexes proved not to be involved in the cascade-like process that was modeled to be driven by an intramolecular transfer of two copper(I) ions from a just-formed binuclear copper intermediate to the adjacent acetylene unit. Get selective! Copper(I)-catalyzed addition of azides to multiple alkynes may process in a cascade-like manner through efficient intramolecular transfer of both copper atoms from a just-reacted dicopper triazolyl intermediate to an adjacent reacting triple bond, rather than through intermolecular exchange; this has been shown experimentally and theoretically by using a series of calixarenes with different arrangements of several propargyl groups.

Full text of DOI:10.1002/chem.201500946

DOI: 10.1002/chem.201500946
Full Paper
&
Click Chemistry
Copper(I)-Catalyzed Cycloaddition of Azides to Multiple Alkynes:
A Selectivity Study Using a Calixarene Framework
Alexander Gorbunov,[a] Dmitry Cheshkov,[b] Vladimir Kovalev,*[a] and Ivan Vatsouro*[a]
Abstract: Copper(I)-catalyzed addition of limited amounts of
azides to multiple alkynes, which led to statistical mixtures
of triazole/acetylene derivatives or, in other cases, resulted in
preferred formation of multiple triazoles, was studied at pre-
organizable calixarene platforms bearing up to four proparg-
yl groups. Depending on calixarene structures and reaction
conditions, the unprecedented specific or selective forma-
tion of exhaustively triazolated calixarenes or a complete
loss of the selectivity were observed. Both autocatalytic
copper activation and a local copper(I) concentration in-
crease due to copper–triazole complexation were thorough-
ly studied as the most expected reasons for the selectivity
and both were disproved. Mixed triazolated/propargylated
calixarenes and their copper(I) complexes proved not to be
involved in the cascade-like process that was modeled to be
driven by an intramolecular transfer of two copper(I) ions
from a just-formed binuclear copper intermediate to the ad-
jacent acetylene unit.
Introduction
has been converted into a polymer with a definite multitria-
zole–multipropragyl blocks sequence under CuAAC with a limit-
ed amount of benzyl azide.[3i]
Copper-catalyzed azide–alkyne cycloaddition (CuAAC) has rap-
idly become widely used in organic chemistry and related dis-
ciplines since the long-known Huisgen reaction has been
shown to be efficiently catalyzed by copper(I) salts and com-
plexes.[1] The CuI catalysis provided both rate enhancement
and regioselectivity of the cycloaddition under mild conditions
with wide range of copper sources and solvents, which has al-
lowed application of CuAAC for various molecular systems in-
cluding multifunctional or/and very sensitive ones.[2] First es-
tablished for conformationally constrained small bis(azides),[3a]
CuAAC between multiple azides and alkynes has been shown
to be clearly selective towards multiple triazoles in cases of oli-
goazide derivatives of cyclodextrins,[3b] calixarenes,[3c] silses-
quioxane,[3d] etc. In contrast, ‘inverted’ CuAAC reactions of mul-
tiple alkynes and azides have proceeded with a far less predict-
able outcome. Preferred formation of bis(triazoles)[3e–h] or com-
plete loss of the selectivity resulted in statistical mixtures of
mono- and bis(adducts),[3a] both have been published for bi-
s(alkynes), including structurally related ones. The only exam-
ple of clear selectivity towards adjacent multiple triazoles has
been recently published for poly(propargyl metacrylate) which
Though tentative rationalizations for the selectivity[3i] or for
its absence[3a] have been provided, they could not be cross-cor-
related even for the known cases and, thus, gave no general
understanding of the mechanism of CuI-catalyzed addition of
azides to multiple alkynes. Herein we studied equimolar reac-
tions between azides and two, three, and four propargyl
groups grafted at narrow rims of calixarenes known for their
power in the tunable pre-organization of multiple functional
units.[4] The reactions gave the full range of CuAAC outcomes
(from completely nonselective reactions to specific formation
of multitriazoles) that allowed the investigation of the selectivi-
ty in detail to provide a fully consistent explanation for it.[5]
Results and Discussion
Within the first set of reactions that initiated all further study,
equimolar mixtures of cone tetrakis(propargyloxy)calix[4]arene
1 and azides were subjected to CuAAC with several well-
known copper(I) catalysts (Scheme 1, Table 1). No reaction oc-
curred at room temperature, but, surprisingly, upon heating,
azides were spent nearly exclusively on exhaustive modifica-
tion of starting calixarene to form tetrakis(triazoles) 2–4,
~75 mol% of unreacted 1 returned, and no partially triazolated
calixarenes were detected in the samples after removal of Cu-
[a] A. Gorbunov, Prof. V. Kovalev, Dr. I. Vatsouro
Department of Chemistry
M. V. Lomonosov Moscow State University
Lenin’s Hills 1, 119991 Moscow (Russia)
E-mail: kovalev@petrol.chem.msu.ru
vatsouro@petrol.chem.msu.ru
1
salts (for representative H NMR spectra, see Figure 1, traces of
[b] D. Cheshkov
byproducts due to acetylene homocoupling or triazole iodina-
tion were detected in several cases). Kinetic measurements of
CuAAC between 1 and benzylazide showed no traces of par-
tially triazolated calixarenes even in early samples (see the Sup-
porting Information).
State Research Institute for Chemistry and Technology
of Organoelement Compounds
Sh. Entuziastov 38, 105118 Moscow (Russia)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201500946.
Chem. Eur. J. 2015, 21, 9528 – 9534
9528
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 1. 1H NMR spectra of a) pure 1, b) the mixture obtained in equimolar
reaction of 1 with PhCH2N3 (Table 1, entry 4) after removal of Cu-salts,
c) pure 2; CDCl3, 600 MHz; *=residual solvent signals.
Table 2. Major products content in equimolar CuAAC reaction of partially
propargylated calixarenes with azides.
Reaction/conditions[a] Major product Major product content [mol.%]
Exp.[b]
Stat.[c]
Max.[d]
10+PhCH2N3/A
10+EtOC(O)CH2N3/A 16
11 + PhCH2N3/A 17
11 + EtOC(O)CH2N3/A 18
12+PhCH2N3/A 19
15
50
47
35
37
30
25
18
31.8
31.8
31.8
31.8
8.5
50
50
50
50
33
33
33
Scheme 1. Selective (top) and nonselective (bottom) routes of equimolar
CuAAC reactions between 1 and azides.
12+EtOC(O)CH2N3/A 20
21+EtOC(O)CH2N3/A 23
8.5
6.2
Table 1. Reaction conditions for selective conversion of 1 into 2–4 under
equimolar CuAAC.[a]
[e]
[e]
24+PhCH2N3/A
24+PhCH2N3/B
25+PhCH2N3/A
25+PhCH2N3/B
28+PhCH2N3/A
–
26
–
27
29
50
31.8
50
Entry Azide
Catalyst
Solvent
T [8C]
27
15
14
8.5
8.5
8.5
33
33
33
1
2
3
4
5
6
7
8
PhCH2N3
CuI·P(OEt)3
toluene
THF/H2O
toluene[c] 110
toluene[d] 110
toluene
toluene
110
65
PhCH2N3
PhCH2N3
PhCH2N3
CuSO4/sodium ascorbate[b]
[Cu(CH3CN)4]PF6
CuCl
28+EtOC(O)CH2N3/A 30
[a] Conditions A: CuI·P(OEt)3 (15%), toluene, reflux, 7 h. B: CuSO4·5H2O
(20%)/sodium ascorbate, THF/H2O, reflux, 7 h. c(calixarene)=c(azide)=
0.01m. [b] Measured by integration of calixarene aromatic signals in
1H NMR spectra of reaction mixtures after removal of Cu-salts. [c] Calculat-
ed from formal kinetic equations assuming equal rate constants for every
step. [d] Theoretical yield for the reaction furnishing the major product
specifically. [e] Complex mixture, no major product was detected.
EtOC(O)CH2N3 CuI·P(OEt)3
110
110
110
60[e]
EtOC(O)CH2N3 CuI[b]
EtOC(O)CH2N3 CuI/DIPEA (20 equiv per Cu+) toluene
PhN3
CuI·P(OEt)3
toluene
[a] 15 mol% of catalyst, reaction time 5–7 h, c(calixarene)=c(azide)=
0.01m. [b] Longer heating needed to complete the reaction. [c] No reac-
tion in boiling CH3CN even with excess of azide. [d] Wet toluene needed
to be used with this catalyst. [e] Lower temperature needed to prevent
side reactions of PhN3.
showed a high selectivity or even specificity towards bis-
(triazoles) 15 and 16 (Scheme 2). As a perfect example, equi-
molar CuAAC reaction of partial cone (paco) calixarene 21
gave predominantly tris(triazole) 23 with co-directed adjacent
propargyl groups involved in the reaction (but not the alter-
nating one as approved by ROESY), and just a limited amount
of tetrakis(triazole) 22 (Scheme 3).
In apolar toluene, when residual conformational motions in
bis- and tris(propargylated) calixarenes 24 and 25 were sup-
pressed by OH···OR bonding, extremely low yields of bis- and
tris(triazoles) 26 and 27 were observed in CuAAC, even with
an excess of benzylazide (Scheme 4).
The CuAAC reactions of 1 retained selectivity towards tetra-
kis(triazoles) in solvents of different polarity and in the pres-
ence of limited amounts of Cu-competitive ligand (Table 1,
entry 7), but turned to the nonselective route with a huge
amount of Et3N (240 equiv per Cu+) giving all possible mixed
propargylated/triazolated calixarenes 5–8.
Partially propargylated calixarenes were also studied in equi-
molar CuAACs (see Table 2 for numerical data). Propargylated/
propylated calixarenes 10–12 pre-organized (but not fixed) in
cone conformations converted into multiple triazoles with a se-
lectively governed by mutual arrangement of acetylene units
in the substrate: CuAAC reactions of proximal bis(alkyne) 11
were nearly nonselective, whereas those of its distal isomer 10
In contrast, in an H-bond competitive THF/H2O mixture, cal-
ixarenes 26 and 27 turned out to be major or even exclusive
products of equimolar CuAACs that resembled nicely the selec-
Chem. Eur. J. 2015, 21, 9528 – 9534
www.chemeurj.org
9529
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Scheme 4. CuAAC reactions of cone calixarenes 24, 25, and 28 with sup-
pressed residual conformational mobility.
The data presented cover all possible outcomes for CuI-cata-
lyzed addition of limited equivalents of azides to multiple al-
kynes. Selectivity of the reaction towards multiple triazoles did
appear or did not for structurally related, or even isomeric,
multiple alkynes (e.g. 10 and 11), and was not governed by
simply the number of acetylene units within a substrate (e.g.
tris(acetylene) 12 reacted less selectively than tetrakis(acety-
lene) 1 and bis(acetylene) 10), but also by their mutual ar-
rangement and flexibility.
Scheme 2. CuAAC reactions of cone calixarenes 9–12 with nonsuppressed
residual conformational mobility.
Looking for a general rationalization for our results, we first
explored theoretically an autocatalytic reason for the selectivity
of CuAAC. The idea of intramolecular triazole-promoted lower-
ing the activation barrier of CuAAC[7] (Figure 2) at additions of
second and further azide molecules to multiple alkynes
seemed attractive because of known rate-enhancing effects of
tris(benzyltriazolylmethyl)amine (TBTA) and related com-
pounds.[8]
Calculations applied for a TBTA-stabilized simple CuAAC re-
action showed the activation profit of ~6 kcalmolÀ1 provided
by the ligand. Nevertheless, analogous intramolecular stabiliza-
tion probed for diverse combinations of triazole and propargyl
Scheme 3. Equimolar CuAAC reaction between 21 (paco) and ethyl-2-azidoa-
cetate.
tive formation of related propylated/triazolated calixarenes 15
and 19 (Scheme 4) from 10 and 12, respectively. Tris(propargy-
lated) calix[6]arene 28 reacted well with an excess of azides
but showed limited selectivity towards tris(triazoles) in equimo-
lar CuAACs, which could arise from conformational restrictions
due to through-cavity CH3···p-interactions,[6] and could not be
avoided by changing the catalytic system.
Figure 2. Rate-determining step of CuAAC and formation of a six-membered
metallacycle.
Chem. Eur. J. 2015, 21, 9528 – 9534
www.chemeurj.org
9530
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
units at the calixarene backbone gave the maximum profit of
1.5 kcalmolÀ1 (see the Supporting Information for details). So
the activation of this type could not drive the CuAAC selectivi-
ty towards multiple triazoles (though might contribute to it),
and also gave no explanation for the differences in selectivity
of isomeric substrates.
Table 3. Calixarene products of equimolar CuAAC reactions between
compounds 5–8 and benzylazide in toluene.[a]
Conditions
Substrate
5
6
7
8
[b]
[b]
[b]
[b]
Cu+ (15%), R-N3, 24 h, rt
–
–
–
–
An efficient entrapment of copper ion(s) by a substrate or/
and semiproducts might prevent the ions from leaving the re-
acting calixarene molecule unless all its triple bonds have been
converted to triazoles. In such a case, the local concentration
of Cu+ might be increased and this, in turn, may lead to a rate
enhancement for second and further CuAAC conversions of
multiple alkynes, which could thus explain the observed selec-
tivity. To check this proposal, equimolar reactions between
1 and azides were run at 1, 5, 10, 15, 20, 50, 100, 200 (two Cu+
per calixarene molecule), 400, 800 (two Cu+ per triple bond),
and 1000 mol% of loaded catalyst (CuI·P(OEt)3, toluene, 1008C,
identical calixarene and azide concentrations in all runs). In all
the cases, the highly selective formation of tetra(triazoles) 2
and 3 was observed, so the reaction outcome (but not rates
which were not monitored) was not dependent on Cu+ con-
centration, and a strong copper binding by reacting compo-
nents could not be responsible for the selective formation of
multitriazoles.
Cu+ (15%), R-N3, 7 h, 1008C 6 (10%) 8 (12%) 8 (26%) 2 (100%)
2 (6%)
2 (37%)
[b]
[b]
[b]
1) Cu+ (200%), 2 h, 1008C
or 72 h, RT
–
–
–
2 (100%)
2) R-N3, 24 h, RT
1) Cu+ (200%), 2 h, 1008C
2) R-N3, 8 h, 1008C
6 (20%) 8 (29%)
2 (25%) 2 (34%)
[a] CuI·P(OEt)3 was used as the Cu+ source, c(calixarene)=c(azide)=
0.01m. [b] No calixarene products were detected by 1H NMR spectrosco-
py.
(compare with 10) but complete conversion into calixarenes 8
and 2 under the CuAAC conditions. When copper complexes
of 5 and 6 were first prepared using 200 mol% of CuI·P(OEt)3,
the room-temperature equimolar reactions with benzylazide
failed, while heating resulted in complete conversion of the
substrates into mixtures of 2 and calixarenes with a single re-
acted triple bond adjacent to a triazole unit (5!6, 6!8). Simi-
lar behavior was observed for distally propargylated/triazolated
calixarene 7 already with 15% of Cu+. One could conclude
tentatively, that tris(triazolated) calixarene 8 often observed in
the reaction mixtures was less reactive in CuAAC than its
formal precursors 5–7, and even 1. But that was not the case,
as calixarene 8 reacted completely under CuAAC with one
equivalent of PhCH2N3 at 15% of Cu+ and was the only triazo-
lated calixarene reactive at room temperature after preliminary
copper complexation.
Still, a direct study of copper complexation and reactivity of
calixarenes 5–8, which were supposed to be semiproducts at
conversion of tetra(acetylene) 1 into tetra(triazole) 2, gave in-
teresting and unexpected results. While tetrakis(propargylated)
calixarene 1 formed no complexes when treated with CuI·
[9]
P(OEt)3 at room temperature or at heating (no complexation-
1
induced shifts were observed in H NMR spectra of [D8]toluene
solutions), all four mixed triazolated/propargylated calixarenes
5–8 formed internal copper complexes in which both the tria-
zole and acetylene unit took part in the ion stabilization (as fol-
1
lows from complexation-induced shift values in H NMR spec-
tra of 1:2 mixtures of the calixarenes and CuI·P(OEt)3 in
[D8]toluene, see the Supporting Information). Notably, mono-
and bis(triazolated) calixarenes 5–7 did form the complexes
rapidly at room temperature (5–10 min), whereas much more
time (up to 72 h) was needed to equilibrate the mixture of 8
and CuI·P(OEt)3. The equilibrium was reached much faster (<
2 h) at 1008C, but resulted in a different spectral pattern that
reflected a different structure of 8·(Cu+)n.[10]
Calixarenes 5–8 were studied in equimolar CuAACs with
benzylazide either with direct catalyst loading or with prelimi-
nary copper(I)-complex preparation (Table 3). The presence of
several triazole units in a substrate did not stimulate the
CuAAC reactivity of neighboring propargyl groups within the
same molecule at room temperature with 15% Cu+ loading.
Even more unexpectedly, at elevated temperature and 15%
Cu+, compounds 5 and 6 with one or two adjacent triazole
groups reacted significantly slower than respective propargy-
lated/propylated calixarenes 9 and 11, and showed no selectiv-
ity towards exhaustively triazolated adduct 2. Notably, calixar-
ene 5 was converted in low yield into proximal bis(triazole) 6,
but not to its distal isomer 7. The latter was not detected in
any reaction mixture that might result from its nonselective
Several competitive reactions were run to analyze directly
the difference in CuAAC reactivity of partially propargylated
calixarenes containing and not containing triazole groups
within the molecules. Surprisingly, from an equimolar mixture
of tripropargylated calixarenes 5 and 12 the only triazolated
one was involved in CuAAC reaction with one equivalent of
PhCH2N3 and converted in low yield into calixarene 6, whereas
12 returned unchanged, though was reactive under similar
conditions in the absence of 5 (Scheme 5). This showed that
copper complexation by 5 efficiently inactivated the metal ion
for catalysis of both intramolecular and intermolecular
CuAACs. When excess of the azide (weak copper-coordinating
ligand) was added, the complex lability was increased and
both 5 and 12 were completely converted into the corre-
sponding exhaustively triazolated calixarenes 2 and 19.
For the ‘opposed’ pair of compounds (8 and 9) each con-
taining a single acetylene unit, a quite different reactivity in
competitive CuAACs was observed (Scheme 6). With 15 or
200% of CuI·P(OEt)3 loaded (including precursive copper com-
plexation), limited PhCH2N3 did always react with triazolated/
propargylated calixarene 8 but not with propylated analogue
Chem. Eur. J. 2015, 21, 9528 – 9534
www.chemeurj.org
9531
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
to the reacting triple bond. As a copper(I) triazolide[13] could re-
lease just a single copper ion, another intramolecular source
for the second Cu+ was required to exclude undesired copper
recruitment from bulk solution. Within the catalytic cycle of
CuAAC, formation of a dicopper triazolyl intermediate might
be proposed at a six-membered metallacycle contraction step
(a cheap step with ~1.6 kcalmolÀ1 barrier as modeled for the
simplest CuAAC between propyne and methylazide) before its
fast conversion into copper triazolide at Cu+ release (Figure 3).
Scheme 5. Independent (top) and competitive (bottom) CuAAC reactions of
tripropargylated calixarenes 12 and 5.
Figure 3. Ring-contraction step of CuAAC. Formation of proposed transfer-
ready dicopper intermediate.
As no direct evidence for the formation and reactivity of dicop-
per intermediates of this type could be obtained, we applied
quantum-chemical calculations to model possible intramolecu-
lar copper-transfer complexes to get an explanation for all the
experimental data on the selectivity of CuAAC towards multi-
ple triazoles.
First, dicopper triazolyl intermediates were modeled for pre-
final CuAAC cycles at conversion of calixarene-based bis- and
tris(alkynes) 10–12 into multiple triazoles (CH3N3 was used as
the simplest azide, water molecules were used to complete
the environment of copper ions where appropriate). Next, the
yet nonreacted acetylene groups were forced to p-coordinate
intramolecularly to one of two copper atoms of the dicopper
triazolyl intermediates, and the resultant structures were sub-
jected to full geometry optimization with no constraints.[14]
Though the starting geometries were created manually, in
the energy-minimized complexes the triple-bond-to-copper p-
coordination retained and, less expected, the second copper
atoms were also ‘moved’ to a close proximity to alkyne units.
Thereby, the dinuclear complexes presented in Figure 4 might
be regarded as Cu-transfer ones. At the same time, the dis-
tance between the acetylene terminal carbon and the non-p-
bound copper atom in the complex of 10 (shown in dot-lines
in Figure 4a) was nearly 1.5 times shorter than that of 11 (Fig-
ure 4b). Reasonably, the distances reflected the different effi-
ciency of dicopper transfer in the two complexes as far as they
correlated well with the experimentally observed CuAAC be-
havior of 10 (showed perfect selectivity towards bis(triazoles))
and 11 (showed nearly no selectivity towards bis(triazoles)).
In general, the two structural features of starting oligoacety-
lene substrates seemed crucial for the formation of good
CuAAC-transfer complexes in which the transfer of both
copper atoms to a next reacting triple bond would not be in-
terfered by an intermolecular copper exchange. First, the
Scheme 6. Independent (top) and competitive (bottom) CuAAC reactions of
monopropargylated calixarenes 9 and 8.
9, and no traces of calixarene 13 were detected in reaction
mixtures.
Though the outstanding behavior of 8 was not yet rational-
ized, the data on reactivity of calixarenes 5–7 showed clearly
that (at least) one and two triazole units located near by a re-
acting triple bond within the same molecule did not assist the
selective conversion of multiple acetylenes into multiple tria-
zoles, and calixarenes 5–7 themselves and their copper com-
plexes did not intermediate the CuAAC-transformation of
1 into 2.
These observations proved that neither copper(I) activation
nor its strong binding by triazole units within reacting mole-
cules were responsible for the selectivity of CuAAC towards
multiple triazoles.[11] Thus, an efficient copper-ion(s) transfer
within the multifunctional substrate that moved the reacting
center to a neighboring triple bond intramolecularly rather
than intermolecularly, remained to be proposed as a main driv-
ing force for the observed CuAAC selectivity.[12]
To meet the experimental data on the invariance of multiple
CuAAC outcomes from outer copper concentration, both cop-
per(I) ions had to be transferred from an intramolecular source
Chem. Eur. J. 2015, 21, 9528 – 9534
www.chemeurj.org
9532
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
the selectivity provided by the addition of a huge amount of
a copper-targeted ligand (e.g., triethylamine); at lower concen-
tration such an additional ligand could not easily reach the
well-packed copper atom(s) within the transfer complexes or
their precursors to compete even with weak CuÀO bonding,
whereas at much higher concentrations (compared to that of
the solvent) it could solvate all the components and turn the
intermolecular copper exchange to prevail over the intramolec-
ular copper transfer.
To verify the proposed methodology at a noncalixarene mol-
ecule, the copper-transfer complex of 1,6-heptadiyne 31 was
modeled (Figure 5). This bis(alkyne) has been published to con-
Figure 5. Energy-minimized structure and key Cu–C interatomic distance of
dicopper transfer complexes at addition of second methylazide molecule to
1,6-heptadiyne 31.
vert into bis(triazoles) in equimolar CuAAC with benzylazide
predominantly (but not highly selectively) unless the reaction
medium has been changed from CH2Cl2 to ketones or, better,
to 2,5-hexandione.[3e] In the calculated structure, the key Cu–C
distance of 3.7 Š was not very short but still could allow a mod-
erately efficient intramolecular copper transfer in the complex
tightened by a solvent-breakable coordination bond between
the p-bound copper atom and amide oxygen atom.[16]
Though easily checkable (in the proposed simple version),
the dicopper transfer must not be regarded as the only key
process through all the cascade of CuAAC reactions, and other
processes might contribute significantly to rates of deeper
steps of multiple CuAACs. From them, the autocatalysis dis-
cussed above seemed the most plausible to rationalize the
outstandingly high CuAAC reactivity of calixarene 8 containing
a single triple bond surrounded by three adjacent triazole
units.
Figure 4. Energy-minimized structures and key Cu–C interatomic distances
of dicopper transfer complexes at final CuAAC additions of methylazide to
calixarenes a) 10, b) 11, c,d) 12 (at different order of triple-bonds reactions).
mutual arrangement and flexibility of the reacting acetylene
units must allow the proper orientation of the dicopper triazol-
yl unit and the next reacting acetylene unit (e.g., H-bonds in
24, 25, unless broken, and CH3···p-interactions in 28 restricted
the acetylene units’ motions and allowed no efficient Cu-trans-
fer to be developed). Second, additional donor atoms or/and
groups must be presented (or formed) within the substrate
molecule to assist the formation of tightened dicopper transfer
complexes through additional Cu stabilization. For instance, in
the complex of 11 (Figure 4b) the only ether oxygen atom of
the reacting propargyl group provided an additional stabiliza-
tion of the p-bound copper atom that gave no tightening of
the overall structure and left the second copper atom at 3.8 Š
distance from the targeted acetylene terminal carbon atom. In
contrast, in modeled complexes of 10 and 12 (Figure 4a,c,d),
the p-bound copper atom was additionally stabilized by sever-
al ether oxygen atoms from reacting and nonreacting calixar-
ene units as well as by triazole groups (for 12); such a multi-
dentate copper coordination resulted in tightened transfer
complexes with much shorter key Cu–acetylene distances of
2.8–3.2 Š, which allowed the intramolecular transfer of both
copper atoms rather than an intermolecular one.[15]
Conclusion
Being intrigued by the observed selective and condition-inde-
pendent (except for certain reasonable cases) copper(I)-cata-
lyzed conversion of cone tetrakis(propargyloxy)calix[4]arene
into tetrakis(triazoles) when reacted with one equivalent of an
azide, we performed a detailed study of equimolar CuAAC re-
actions between azides and a series of calixarenes with differ-
ent number, mutual arrangement, and flexibility of propargyl
groups attached to narrow rims. From the series of experi-
ments we got a full range of multiple CuAAC outcomes includ-
ing highly selective conversions of multiple acetylenes into
multiple triazoles as well as those resulting in nearly statistical
mixtures of triazolated/propargylated calixarenes. As reasons
for the drastically different reactivity of the quite similar oligoa-
cetylene molecules were not easily formulated, we studied ex-
The presented simple (and relatively fast) modeling and
analysis of dicopper transfer complexes gave the desired ex-
planation for all the experimental data on the highly selective
(for 1, 10, 12, 24, 25), and poorly or nonselective (for 11, 21,
28) conversions of multipropargylated calixarenes into the cor-
responding multiple triazoles in equimolar reaction with azides
under copper catalysis. It explained also the complete loss of
Chem. Eur. J. 2015, 21, 9528 – 9534
www.chemeurj.org
9533
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Ingale, J. Org. Chem. 2010, 75, 284–295; g) S. A. Ingale, S. S. Pujari, V. R.
Sirivolu, P. Ding, H. Xiong, H. Mei, F. Seela, J. Org. Chem. 2012, 77, 188–
199; h) H. Xiong, P. Leonard, F. Seela, Bioconjugate Chem. 2012, 23,
856–870; i) T. Kitaura, H. Tomioka, N. Fukatani, T. Kitayama, Polym.
Chem. 2013, 4, 887–890.
perimentally and theoretically several most attractive rationali-
zations for the selectivity and its absence.
Neither autocatalytic rate enhancement of CuAAC nor one
provided by Cu+-entrapment were proved to control the se-
lectivity of the multiple reactions. Still, the consistent rationali-
zation for all the experimental data came from the modeling
of an intramolecular transfer that moved not a single but both
copper atoms from a just-formed triazole unit to an adjacent
triple bond within the same oligoacetylene molecule. The
model assumed the dicopper triazolyl intermediate as the
copper source, and Cu–C-distance and the metal atoms hin-
drance as indicators for the efficiency of the dicopper intramo-
lecular transfer process. Though the simplified modeling gave
no exact parameters for a ‘perfect’ dicopper-transfer coordina-
tion environment, it worked fine at the comparative level and
allowed us to arrange the oligoalkynes according to their abili-
ty to form the corresponding multiple triazoles in a cascade of
CuAAC reactions.
[4] a) Calixarenes 2001 (Eds.: Z. Asfari, V. Bçhmer, J. Harrowfield, J. Vicens),
Kluwer Academic, Dordrecht, 2001; b) V. Bçhmer in The Chemistry of
Phenols (Ed.: Z. Rappoport), John Wiley & Sons, Ltd, Chichester, 2003,
pp. 1369–1454; c) Calixarenes in the Nanoworld (Eds.: J. Vicens, J. Har-
rowfield, L. Baklouti), Springer, Dordrecht, 2007; d) C. D. Gutsche, Calix-
arenes, An Introduction, Royal Society of Chemistry, Cambridge, 2008.
[5] No clear data on the selectivity has been published for CuAAC-modifi-
cations of propargylated calixarenes though the reaction has been
widely applied in calixarene chemistry. For a recent partial review, see:
M. Song, Z. Sun, C. Han, D. Tian, H. Li, J. S. Kim, Chem. Asian J. 2014, 9,
2344–2357.
[6] a) V. V. S. Mummidivarapu, A. Nehra, V. K. Hinge, C. P. Rao, Org. Lett.
2012, 14, 2968–2971; b) J. P. M. van Duynhoven, R. G. Janssen, W. Ver-
boom, S. M. Franken, A. Casnati, A. Pochini, R. Ungaro, J. de Mendoza,
P. M. Nieto, P. Prados, D. N. Reinhoudt, J. Am. Chem. Soc. 1994, 116,
5814–5822.
[7] a) B. T. Worrell, J. A. Malik, V. V. Fokin, Science 2013, 340, 457–460; b) D.
Cantillo, M. valos, R. Babiano, P. Cintas, J. L. JimØnez, J. C. Palacios, Org.
Biomol. Chem. 2011, 9, 2952–2958.
[8] T. R. Chan, R. Hilgraf, K. B. Sharpless, V. V. Fokin, Org. Lett. 2004, 6, 2853–
2855.
[9] CuI·P(OEt)3 has been selected for all further experiments because of its
solubility in toluene that allowed taking of exact equivalents of the cat-
alyst by using aliquots of stock solutions.
The results presented are easily applicable for the analysis of
the CuAAC reactivity of nearly all known types of multipropar-
gylated calixarenes and related macrocycles widely used nowa-
days for constructing various molecular receptors/sensors. It is
expected also that the experimental treatment of CuAAC acti-
vation through copper-to-multitriazole complexations along
with the theoretical modeling of the intramolecular dicopper
transfer made in this work will allow the analysis and, more im-
portantly, the prediction of the selectivity of multiple CuAAC
reactions of diverse oligoalkynes.
[10] The two copper complexes of 8 were separated from toluene solutions
1
by addition of hexane. The samples were analyzed by H NMR spectros-
copy in [D6]acetone solutions in which they were stable against pro-
longed storage at room temperature. Though the general trends of
complexation-induces shift values were similar in both samples, the
spectra differed significantly, also by POEt-multiplet intensities (see the
Supporting Information), which might reflect their different nuclearities,
which needed to be investigated further.
Acknowledgements
[11] Still, these effects might assist deeper steps of multiple CuAAC (addi-
tion of the fourth azide in the case of 1), which followed both from ac-
tivation energy calculations and extended reactivity of 8.
[12] A possibility of copper transfer by a ‘copper-triazolyl’ intermediate to
an adjacent alkyne within multiacetylene molecules has been recently
mentioned, but not modeled and/or used to rationalize any feature of
CuAAC. S. Cecioni, S. E. Matthews, H. Blanchard, J.-P. Praly, A. Imberty, S.
Vidal, Carbohydr. Res. 2012, 356, 132–141.
Financial support from the Russian Foundation for Basic Re-
search (projects: 11-03-92006, 12-03-31715) is gratefully ac-
knowledged. We also thank Dr. D. Sinitsyn and Dr. M. Kononets
for helpful discussions and assistance in data treatment.
[13] C. Nolte, P. Mayer, B. F. Straub, Angew. Chem. Int. Ed. 2007, 46, 2101–
2103; Angew. Chem. 2007, 119, 2147–2149.
[14] DFT B3LYP/def2-TZVP(-df)/LANL(10), COSMO (H2O); a) ORCA program
package, release 3.0.1; b) F. Weigend, R. Ahlrichs, Phys. Chem. Chem.
Phys. 2005, 7, 3297–3305 (def2-TZV basis sets); c) P. J. Hay, W. R. Wadt,
J. Chem. Phys. 1985, 82, 299–310 (LANL ECPs).
Keywords: calixarenes · click chemistry · copper complexes ·
multiple alkynes · reaction mechanism
[1] a) C. W. Tornøe, C. Christensen, M. Meldal, J. Org. Chem. 2002, 67, 3057–
3064; b) V. V. Rostovtsev, L. G. Green, V. V. Fokin, K. B. Sharpless, Angew.
Chem. Int. Ed. 2002, 41, 2596–2599; Angew. Chem. 2002, 114, 2708–
2711.
[15] The related copper(I)-templated grafting of acetylene and triazole units
(in the form of copper(I) triazolide) has been published to be responsi-
ble for exclusive formation of cone-shaped triazolated hexahomotrioxa-
calix[3]arenes from conformationally mobile propargylated precursor:
a) X.-L. Ni, S. Wang, X. Zeng, Z. Tao, T. Yamato, Org. Lett. 2011, 13, 552–
555; b) C.-C. Jin, T. Kinoshita, H. Cong, X.-L. Ni, X. Zeng, D. L. Hughes, C.
Redshaw, T. Yamato, New J. Chem. 2012, 36, 2580–2586. Such a templa-
tion during the copper-transfer process may also be applied for a more
precise rationalization of preferred formation of bridged compounds
and cyclodimers at CuAAC reactions of calix[4]arene 24 with diazides:
c) J. Morales-Sanfrutos, M. Ortega-MuÇoz, J. Lopez-Jaramillo, F. Hernan-
dez-Mateo, F. Santoyo-Gonzalez, J. Org. Chem. 2008, 73, 7768–7771.
[16] An additional sterical hindrance of the dicopper reacting centre by
chiral ligands used in the cited work might also decrease the availability
of the metal ions for intermolecular exchange.
[2] For recent reviews on CuAAC features and applications, see: a) M. Jurí-
ˇ
cek, P. H. J. Kouwera, A. E. Rowan, Chem. Commun. 2011, 47, 8740–
8749; b) L. Liang, D. Astruc, Coord. Chem. Rev. 2011, 255, 2933–2945;
c) B. Dervaux, F. E. Du Prez, Chem. Sci. 2012, 3, 959–966; d) J. M.
Palomo, Org. Biomol. Chem. 2012, 10, 9309–9318; e) D. Astruc, L. Liang,
A. Rapakousiou, J. Ruiz, Acc. Chem. Res. 2012, 45, 630–640; f) A. S. Gold-
mann, M. Glassner, A. J. Inglis, C. Barner-Kowollik, Macromol. Rapid
Commun. 2013, 34, 810–849; g) A. E. Fernandes, A. M. Jonas, O. Riant,
Tetrahedron 2014, 70, 1709–1731.
[3] a) V. O. Rodionov, V. V. Fokin, M. G. Finn, Angew. Chem. Int. Ed. 2005, 44,
2210–2215; Angew. Chem. 2005, 117, 2250–2255; b) G.-H. Park, J. M.
Baek, H. T. Le, C. W. Lim, T. W. Kim, Tetrahedron Lett. 2012, 53, 5911–
5915; c) A. Vecchi, B. Melai, A. Marra, C. Chiappe, A. Dondoni, J. Org.
Chem. 2008, 73, 6437–6440; d) M. E. PØrez-Ojeda, B. Trastoy, . Rol,
M. D. Chiara, I. García-Moreno, J. L. Chiara, Chem. Eur. J. 2013, 19, 6630–
6640; e) F. Zhou, C. Tan, J. Tang, Y.-Y. Zhang, W.-M. Gao, H.-H. Wu, Y.-H.
Yu, J. Zhou, J. Am. Chem. Soc. 2013, 135, 10994–10997; f) F. Seela, S. A.
Received: March 9, 2015
Published online on May 15, 2015
Chem. Eur. J. 2015, 21, 9528 – 9534
www.chemeurj.org
9534
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim