This is the peer reviewed version of the following article: Angew.Chem. Int. Ed. 2016, 55, 7152 –7156, which has
been published in final form at http://onlinelibrary.wiley.com/doi/10.1002/anie.201602569/full. This article may be
used for non-commercial purposes in accordance with Wiley Terms and Conditions for Self-Archiving."

NH-heterocyclic aryliodonium salts: Synthesis and Access to N1aryl-5-iodoimidazoles
Yichen Wu,[a] Susana Izquierdo,[a] Pietro Vidossich,[b] Agustí Lledós*[b] and Alexandr Shafir*[a]
Dedication ((optional))
Abstract: The synthesis of N-arylimidazoles substituted at the sterically
encumbered 5 position is a challenge for modern synthetic approaches. Here
we report a new family of imidazolyl aryliodonium salts that serve as stepping
stones on route to the selective formation of N1-aryl-5-iodoimidazoles; the
iodine can now act as a “universal” placeholder to be transformed into further
substituents. These new λ3-iodanes are produced by treating the NH-imidazole
with ArI(OAc)2, and are converted to N1-aryl-5-iodoimidazoles by a selective
Cu-catalyzed aryl migration. The method tolerates a variety of Ar fragments
and is also applicable to substituted imidazoles.

Imidazole is a ubiquitous heterocyclic core present in a wide
variety of biologically relevant molecules. [1] Although the
synthesis of imidazole derivatives is commonly accomplished
through a variety of cyclization routes, it is often desirable to
obtain a particular derivative starting from a preformed
heterocyclic ring. For this reason, imidazole derivatization has
been the focus of attention from a number of laboratories. A
particularly common challenge is the selective construction of
the 1,4- and 1,5-disubstituted imidazoles. Thus, the NH-arylation
of an imidazole substituted at the C4(5) position tends to
produce a mixture of isomers favoring the sterically less
encumbered NH position, i.e. that with a 1,4 substitution
pattern.[2,3] This bias was recently perfected by Buchwald et al.
through the use of highly bulky biaryl phosphine ligand in Pdcatalyzed imidazole N-arylation.[3b] A similar preference for the
less encumbered NH position can also be seen in the oxidative
Chan-Lam N-arylation of imidazole (Scheme 1A).[4]

A challenge, however, remains to access selectively the
corresponding 1,5-disubstituted imidazoles. Progress made in
recent years includes the usage of well-designed
protection/deprotection strategies,[5] and the C5-selective CHborylation[6] and CH-arylation[7] reactions.
Here we present a new route to a versatile class of
precursors for 1,5-disubstituted imidazoles. Specifically, the N1aryl-5-iodoimidazoles are produced via a relay in which a
hypervalent iodoarene fragment[8] serves as a trampoline for aryl
transfer to the proximal NH site (Scheme 1B). We reasoned that
if the iodane I could be generated, it can then undergo a phenyl
transfer to produce II, perhaps akin the intramolecular O- and Narylation observed in iodonium ylides.[9] Somewhat surprisingly,
the NH-heterocyclic λ3-iodanes have only received a limited
attention beyond the early work by Neiland et al in the
1970’s.[10,11] Recent reports, however, highlight the promise of
hypervalent iodine reactivity in azole functionalization, including
via heterocyclic λ3-iodanes.[12]
In particular, we found only a single precedent of an
imidazolyl-λ3-iodane; the species, however, was described as
containing the imidazole fragment bound to iodine through the N
atom.[13] A reaction between PhI(O2CCF3)2 and imidazole (2
equiv) in acetonitrile at room temp. produced a white precipitate
identified as [PhI(Imid)]TFA salt, 1a (58%). However, the
presence of just two imidazolic resonances in 1H NMR (1H each)
strongly suggested a CH rather than NH functionalization of the
imidazole; accordingly, X-Ray crystallography revealed a
classical T-shaped diaryliodonium environment, with the
imidazole bound to the iodine through the C4(5) carbon atom
(Scheme 2). An analogous acetate salt 2a was obtained by
employing PhI(OAc)2. A DFT analysis confirmed that both the
C2 and the N-bound isomer are higher in energy than the
observed C4(5) isomer. An N-bound species was found unlikely
even as an intermediate en route to 1a; rather, the reaction
appeared to proceed through a Wheland-type intermediate (see
Supporting Information).

Scheme 1. Examples of common imidazole N-arylation strategies (A) and the
relay arylation (B) proposed here.

[a]

[b]

Y. Wu, Dr. S. Izquierdo, Dr. A. Shafir
Institute of Chemical Research of Catalonia (ICIQ)
Barcelona Institute of Science and Technology
Av. Països Catalans 16, 43007, Tarragona, Spain
E-mail: ashafir@iciq.es
Dr. P. Vidossich, Prof. Dr. A. Lledós
Departament de Química
Universitat Autònoma de Barcelona
08193 Cerdanyola del Vallès, Spain
E-mail: agusti@klingon.uab.es
Supporting information for this article is given via a link at the end of
the document.

Scheme 2. Formation and structures of the imidazole-based λ3-iodanes and of
the neutral (betaine) 3. Gibbs Energies (kcal mol-1) in CH3CN.

While sparingly soluble in CDCl3, 1a and 2a dissolved well in
MeOH and water. They also underwent a facile deprotonation to
a zwitterionic 3, for which both the solid state and DFT
structures show an essentially “normal” single C imid-I bond (2.051
and 2.076 Å, respectively, vs 2.091 Å observed for in 1a). We
quickly discovered that the desired I-to-N phenyl transfer does
not take place upon heating 1a, 2a or 3 in CH2Cl2, with or
without Cs2CO3. Consistently, only a high energy transition state
(35.6 kcal mol-1) could be identified for the direct (non-catalyzed)
I-to-N 1,3 phenyl migration in 3 (Scheme 3).

observed for the 13C-I unit in 4, which is approx. 10 ppm lower
than in the corresponding 1,4 species 5 (82-85 ppm). Given the
synthetic potential of 4a, the method was extended to
structurally diverse aryl(imidazolyl)-λ3-iodanes (Table 2). The
most robust protocol involves the use of 20 mol% of N-Mebenzimidazole in combination with 5 mol% of Cu(OTf) 2.
Table 2. Scope of the relay synthesis of N1-aryl-5-iodoimidazoles 4.

Scheme 3. Reaction path modelled for uncatalyzed 1,3 phenyl migration.

Gratifyingly, the addition of 5 mol% of Cu(OTf)2 did allow for the
formation of two regioisomeric N-phenyl iodoimidazoles, with a
moderate selectivity towards the more hindered 4a achieved in
fluorinated alcohols (Table 1, runs 1-3, both isomers confirmed
by
X-Ray
diffraction).
The
use
of
Cs2CO3
in
hexafluoroisopropanol (HFIP) led to a combined yield of 86%
with a 4:1 ratio in favor of 4a (run 4). This ratio was further
improved by employing catalytic amounts of certain heterocyclic
additives (runs 5-7); e.g. the use of 20 mol% of N-Mebenzimidazole (run 6) led to an 8:1 selectivity and a 93% yield.
Table 1. Cu-catalyzed I-to-N phenyl transfer in 2a.[a]

Run

Base

Solvent

Additive

Yield(%)[b]

4a/5ab

1

---

CH2Cl2

--

39

0.1:1

2

---

CF3CH2OH

--

51

1.5:1

3

---

HFIP

--

53

4.2:1

4

Cs2CO3

HFIP

--

86

4.1:1

5

Cs2CO3

HFIP

4-methylimidazole

90

7.3:1

6

Cs2CO3

HFIP

benzimidazole

90

8.4:1

7

Cs2CO3

HFIP

N-Me-benzimidazole

93

8.0:1

[a] Using 0.5 mmol 2a, 5 mol% Cu(OTf)2 and 1.6 equiv of base (if any) in 2.6
mL of solvent. [b] Total yield (%4a+%5a) and the ratio as determined by GC.

It was subsequently found that the highest yields of 2 were
achieved in trifluoroethanol[14] and, notably, MeOH as solvents.
CH3CN, however, remained convenient for large scale
applications due to product precipitation, as seen in the
synthesis of a 23 g batch of 2a (Supp. Info). All the
aryl(imidazolyl)-λ3-iodanes, 2, exhibited the corresponding ArI(Imid)+ cation in the HR (ESI+) mass spectra. These species
were subsequently transformed into the N1-aryl-5-iodoimidazole,
4, with good selectivities. As previously observed for 4a, in all
cases a characteristic 13C resonance at 71-73 ppm was

The improved selectivity with these additives is likely due to the
formation of Cu-heterocycle complexes. Indeed, best results
were achieved by pre-mixing Cu(OTf)2 with the additive and
base for 20 min, presumably favoring complex formation. We

observed that while Cu(OTf)2 alone did not dissolve in HFIP, a
green solution formed upon addition of N-Me-benzimidazole.
Both electron-donating and mildly electron-withdrawing
substituents were well tolerated on the aryl fragment (4b-i, Table
2). In fact, even a di-ortho substitution was tolerated, as
illustrated in the successful synthesis of the highly hindered Nmesityl-5-iodoimidazole, 4j. We were particularly pleased with
the successful incorporation of a second heterocycle, as in the
2- and 3-thienyl derivatives 4k and 4l. The 4-iodobiphenyl and 2iodonaphthalene derivatives could also be obtained in 70% and
74% yield, respectively (4m and 4n). In the case of the 4-Meimidazolyl iodane 2o, a 13:1 4/5 selectivity was achieved,
affording the target 4o in an 87% yield, with the selectivity
benefiting from hindrance at the competing N site. The aryl
transfer in the 2-Me derivative 2p was less efficient, providing 4p
in 31% yield. The method was also applied to produce an 82%
of the 4,5-diiodo derivative 4q. In general, separation between 4
and 5 proved rather straightforward.
As mentioned earlier (see Scheme 1), the high selectivity
towards 4 would stem from an intramolecular aryl migration from
iodine to the proximal nitrogen.[16] Accordingly, a cross-over
experiment between 2a-d2 and 2c revealed a predominant
formation of 4a-d2 and 4c expected for an intramolecular
process (Scheme 4A).[15] Small amounts of the 1,4 isomers were
also produced, and for these, full aryl/imidazole scrambling was
observed, indicating their origin in a bimolecular process.
Indirect support for an intramolecular manifold was also obtained
from the poor performance of the pyrazole-derived iodane 6
(<15% yield, Scheme 4B) lacking a proximal NH site.

Scheme 4. Cross-over experiment (A), and the assay with pyrazol (B).

We envisaged that 3 (formed upon deprotonation of 2), binds
a Cu(I)-OTf fragment through N1 (Scheme 5).[17,18] Indeed,
despite employing a Cu(II) precatalyst, the true catalytic species
is likely a Cu(I) center.[18,19] The inclusion of MeOH in the
coordination sphere of Cu (as a stand-in for a solvent molecule)
was found to be beneficial to properly describe the Cu
intermediate, and, given that the process was already
moderately selective (up to 4:1) in the absence of an additive,
this initial DFT study was performed in the absence of an added
heterocycle. In the first step, the Ph group in A is transferred
from I to Cu, leading to a formal Cu(III)-phenyl intermediate
B.[19,20] This step features an activation barrier of 26.2 kcal mol -1
(ts-1). A Localized Orbital analysis supports the change in Cu
oxidation state and allows visualizing the flow of electrons (see
small green spheres of ts-1 in Scheme 5 and Supporting

Information). The final C-N bond is formed through an
essentially barrierless reductive elimination step (Scheme 5, ts2). Given the energetic proximity between B and ts-2, the
mechanism resembles a Cu-guided concerted I-to-N phenyl
migration. A preliminary investigation also revealed that the
coordination of N-Me-benzimidazole to the Cu(I) center may
disfavor the binding of two molecule of 3 to the same Cu center,
hence enforcing an intramolecular Ph transfer. [21]

Scheme 5. A DFT profile for the Cu(I)-catalyzed aryl migration. Relative Gibbs
energies in methanol (kcal mol-1).

In agreement with Scheme 5, the preformed zwitterionic 3 was
also an excellent substrate even in the absence of a base (Eq 1).

The reason for the poor performance of solvents such as CH2Cl2
is likely twofold. The deprotonation of 2 in CH2Cl2 appears
sluggish, which negatively affects the selectivity, giving rise to
by-molecular cross-over events (see Supporting Info). In
addition, while the use of 3 does render the reaction moderately
selective, the rate in CH2Cl2 remains low.
Iodine introduced at the C5 position ushers the synthesis of a
wide spectrum 1,5-imidazole derivatives (Scheme 6).

Scheme 6. Versatility of the 1-aryl-5-iodoimidazoles in the synthesis of 1,5substituted imidazoles.

Thus, the 5-alkynyl and 5-aryl derivatives 7 and 8 were prepared
via Pd-catalyzed C-C coupling reactions. In addition, a Cucatalyzed C-N bond formation was readily accomplished to give
9.[22] The 5-iodoimidazole 2a was also readily converted to an
organomagnesium species,[23] which served as precursor to the
5-formyl and the 5-borylderivatives 10 and 11.[23b,c]
In conclusion, we have shown that the new (NHimidazolyl)aryl iodonium cation, readily obtained from imidazole
and aryliodine diacetate, ArI(OAc)2, serves as an excellent
stepping stone for the formation of N-arylimidazoles bearing an
iodine substituent at the strategic C5 position. The method
complements common existing protocols known to produce the
sterically favored 1,4-derivatives. The method was tolerant of a
variety of aryl substitution patterns, including mono- or bis-ortho
substitution. Through subsequent transformation of the iodine
group, the newly formed N1-aryl-5-iodoimidazole constitutes a
valuable precursor to a wide range of products. Experimental
and DFT data suggest that the selectivity is likely the result from
an intramolecular copper-catalyzed iodine-to-nitrogen migration
of the aryl fragments.

[8]

[9]

[10]

[11]

[12]

Acknowledgements
This work was funded by Fundació ICIQ, MINECO (CTQ201346705-R, CTQ2014-54071-P and 2014-2018 Severo Ochoa
Excellence Accreditation SEV-2013-0319) and the Generalitat
de Catalunya (2014 SGR 1192). The CELLEX Foundation is
gratefully acknowledged for a post-doctoral contract to S. I. and
for support through the CELLEX-ICIQ HTE platform.

[13]
[14]

[15]

Keywords: imidazoles • hypervalent iodine • CH
functionalization • copper catalysis • C-N coupling • DFT

[16]

[1]

[17]

[2]
[3]

[4]

[5]

[6]
[7]

a) L. Zhang, X. M. Peng, G. L. V. Damu, R. X. Geng, X. H. Zhou, Med.
Research Rev., 2014, 34, 340–437; b) M. Gaba, C. Mohan, Med. Chem.
Res. 2016, 25, 173–210.
For a review on metal-catalyzed imidazole functionalization, see F.
Bellina, R. Rossi, Adv. Syn. Catal. 2010, 352, 1223-1276.
For examples, of catalytic imidazole N-arylation, see: a) R. A. Altman, E.
D. Koval, S. L. Buchwald, J. Org. Chem. 2007, 72, 6190–6199; b) S.
Ueda, M. Su, S. L. Buchwald, J. Am. Chem. Soc. 2012, 134, 700−706.
a) J. P. Collman, M. Zhong, Org. Lett. 2000, 2, 1233-1236; b) X. O. Yu,
Y. Yamamoto, N. Miyaura, Chem. Asian J. 2008, 3, 1517–1522; c) for a
review, K. Sanjeeva Rao, T.-S. Wu, Tetrahedron 2012, 68, 7735-7754.
a) B. Delest, P. Nshimyumukiza, O. Fasbender, B. Tinant, J. MarchandBrynaert, F. Darro, R. Robiette, J. Org. Chem. 2008, 73, 6816–6823; b)
E. Van Den Berge, R. Robiette, J. Org. Chem. 2013, 78, 12220–12223;
c) during the preparation of this manuscript, a method appeared for N1alkylation of unprotected 1,3-azoles (at Amgen): S. Chen, R. F. Grace,
A. A. Boezio, Org. Lett. DOI: 10.1021/acs.orglett.5b02994.
M. R. Smith, R. E. Maleczka, V. A. Kallepalli, E. Onyeozili, US
7,709,654 B2, May 4, 2010.
a) F. Bellina, S. Cauteruccio, L. Mannina, R. Rossi, S. Viel, J. Org.
Chem. 2005, 70, 3997-4005; b) F. Bellina, M. Lessi, C. Manzini, Eur. J.
Org. Chem. 2013, 5621–5630; c) F. Bellina, N. Guazzelli, M. Lessi, C.
Manzini, Tetrahedron 2015, 71, 2298-2305.

[18]
[19]

[20]
[21]
[22]

[23]

For reading on the chemistry of λ3-iodanes, see: a) Hypervalent Iodine
Chemistry. Modern Developments in Organic Synthesis, Editor: T.
Wirth, Springer 2003; b) A. Yoshimura, V. V. Zhdankin, Chem. Rev.
2016, 116, 3328−3435.
Examples include: a) I. Papoutsis, S. Spyroudis, A. Varvoglis,
Tetrahedron Lett. 1996, 37, 913-916; b) I. Papoutsis, S. Spyroudis, A.
Varvoglis, C. P. Raptopouloub, Tetrahedron 1997, 53, 6097-6112; for a
mechanistic study, see: c) R. M. Moriarty, S. Tyagi, D. Ivanov, M.
Constantinescu, J. Am. Chem. Soc. 2008, 130, 7564–7565.
a) B. Y. Karele, S. V. Kalnin', I. P. Grinberga, O. Ya. Neiland, Chem.
Heter. Comp. 1973, 9, 226-229; b) for a review, see: O. Neilands,
Chem. Heterocyclic Comp. 2003, 39, 1555-1569.
For the diaryliodonium renaissance: a) E. A. Merritt, B. Olofsson,
Angew. Chem. Int. Ed. 2009, 48, 9052–9070; for recent examples of
NH-pyrazole-based species, see b) M. Bielawski, J. Malmgren, L. M.
Pardo, Y. Wikmark, B. Olofsson, ChemistryOpen 2014, 3, 19–22.
a) K. Morimoto, Y. Ohnishi, A. Nakamura, K. Sakamoto, T. Dohi, Y. Kita,
Asian J. Org. Chem. 2014, 3, 382 – 386; also see: b) T. Dohi, K.
Morimoto, N. Takenaga, A. Goto, A. Maruyama, Y. Kiyono, H. Tohma,
Y. Kita, J. Org. Chem. 2007, 72, 109–116; c) D. Lubriks, I. Sokolovs, E.
Suna, J. Am. Chem. Soc. 2012, 134, 15436−15442; d) I. Sokolovs, D.
Lubriks, E. Suna, J. Am. Chem. Soc. 2014, 136, 6920−6928; e) G. L.
Tolnai, A. Székely, Z. Makó, T. Gáti, J. Daru, T. Bihari, A. Stirling, Z.
Novák, Chem. Commun. 2015, 51, 4488-4491; f) R. Samanta, R.
Narayan, J. O. Bauer, C. Strohmann, S. Sievers, A. P. Antonchick,
Chem.Commun. 2015, 51, 925; g) S. G. Modha, M. F. Greaney, J. Am.
Chem. Soc. 2015, 137, 1416−1419.
E. A. Veretennikov, A. E. Gavrilov, Chem. Heterocycl. Compd. 2007, 43,
1081-1082.
a) T. Dohi, M. Ito, K. Morimoto, Y. Minamitsuji, N. Takenaga, Y. Kita,
Chem. Commun. 2007, 4152–4154; b) T. Dohi, N. Yamaoka, I. Itani, Y.
Kita, Aust. J. Chem. 2011, 64, 529–535.
No scrambling between 2a-d2 and 2c was observed at 50 oC in the
absence of Cu catalyst; see: a) B. Wang, R. L. Cerny, S. Uppaluri, J J.
Kempinger, S. G. DiMagno, J. Fluorine Chem. 2010, 131, 1113, b) J.
Malmgren, S. Santoro, N. Jalalian, F. Himo, B. Olofsson, Chem. Eur. J.
2013, 19, 10334.
A 1,3 migration has also been proposed in pyrazole arylation by Ar2I+: Z.
Gonda, Z. Novák, Chem. Eur. J. 2015, 21, 16801–16806.
The DFT calculations show equi-energetic binding of Cu(I)OTf to either
of the two N site.
For Cu-catalyzed N-arylation of azoles using diaryliodonium salts: S. K.
Kang, S. H. Lee, D. Lee, Synlett 2000, 7, 1022-1024.
a) For an early mechanistic study, see: T. P. Lockhart, J. Am. Chem.
Soc. 1983, 105, 1940-1946; for a Cu(I)-Cu(III) cycle with hypervalent
iodonium, see b) R. J. Phipps, M. J. Gaunt, Science 2009, 323, 1593;
c) B. Chen, X. L. Hou, Y. X. Li, Y. D. Wu, J. Am. Chem. Soc. 2011, 133,
7668–7671; d) A. J. Hickman, M. S. Sanford, Nature, 484, 177–185; e)
N. Ichiishi, A. J. Canty, B. F. Yates, M. S. Sanford, Organometallics
2014, 33, 5525−5534; f) M. G. Suero, E. D. Bayle, B. S. L. Collins, M. J.
Gaunt, J. Am. Chem. Soc. 2013, 135, 5332–5335.
For Cu(III), see: A. Casitas, X. Ribas, Chem. Sci. 2013, 4, 2301-2318.
All solid state X-Ray structures have been deposited with Cambridge
Structural Database as CCDC entries 1465191 - 1465196
a) A. Klapars, J. C. Antilla, X. Huang, S. L. Buchwald, J. Am. Chem.
Soc. 2001, 123, 7727–7729; b) M.; Wang, Z.; Zhang, F.; Xie, W. Zhang,
Chem. Commun. 2014, 50, 3163-3165.
a) A. Krasovskiy, P. Knochel, Angew. Chem. Int. Ed. 2004, 43, 3333–
3336; for the use of iPrMgCl·LiCl in borylation, see: b) E. Demory, V.
Blandin, J. Einhorn, P. Y. Chavant, Org. Process Res. Dev. 2011, 15,
710-716; c) P. A. Bethel, A. D. Campbell, F. W. Goldberg, P. D.
Kemmitt, G. M. Lamont, A. Suleman, Tetrahedron 2012, 68, 5434-5444.

Entry for the Table of Contents (Please choose one layout)
Layout 1:

COMMUNICATION
A new family of imidazolyl
aryliodonium salts serves as
stepping stones on route to
the N1-aryl-5-iodoimidazoles;
the iodine substituent can now
act as a “universal”
placeholder to be transformed
into further substituents.
These new λ3-iodanes are
produced by treating the NHimidazole with ArI(OAc)2, and
are converted to N1-aryl-5iodoimidazoles by a selective
Cu-catalyzed aryl migration.

Yichen Wu, Susana Izquierdo, Pietro
Vidossich, Agustí Lledós* and Alexandr
Shafir*
Page No. – Page No.
NH-heterocyclic aryliodonium salts: a
stepping stone to N1-aryl-5iodoimidazoles