Catalytic generation of arynes and trapping by nucleophilic addition and iodination

Article abstract of DOI:10.1002/anie.201108415

A fair exchange: In the title reaction, alkynyllithium serves as an initiator for benzyne generation through an iodine-lithium exchange (see scheme; Tf=trifluoromethanesulfonyl). When performed in the presence of stoichiometric amounts of a nucleophile, the generated benzyne undergoes attack by lithio nucleophiles to generate aryllithium, which is then iodinated by iodoalkyne to give the iodoarenes 1. Copyright

Full text of DOI:10.1002/anie.201108415

.
Angewandte
Communications
DOI: 10.1002/anie.201108415
Arynes
Catalytic Generation of Arynes and Trapping by Nucleophilic Addition
and Iodination
Toshiyuki Hamura,* Yu Chuda, Yuya Nakatsuji, and Keisuke Suzuki*
Benzyne is a highly reactive species with broad synthetic
potential.[1] Among various methods for generating this
species through elimination of 1,2-difunctionalized benzene
precursors,[2] we have exploited 2-iodoaryl triflate (1) as an
excellent benzyne precursor, which is immediately converted
into benzyne even at ꢀ958C upon reaction with nBuLi
(Scheme 1).[3] Two rapid consecutive processes are relevant:
Scheme 2. Slow benzyne generation.
complex formation, and 2) excellent leaving ability of triflate.
Even though the aryllithium A may be generated in a small
amount from an unfavorable equilibrium, it would be
converted into B at a certain rate. If B was suitably linked
to a product-forming processes, we could exploit the catalyti-
cally generated benzyne species for organic reactions.
Scheme 3 shows the catalytic cycle we envisioned. The
alkynyllithium D serves as an initiator for the benzyne
Scheme 1. Rapid benzyne generation. Tf=trifluoromethanesulfonyl.
generation by the iodine–lithium exchange (1!A!B). When
1) The rapid and virtually irreversible halogen–lithium
exchange by the formal anion transfer from nBuLi to the
aryllithium (sp3 carbanion!sp2 carbanion),[4] and 2) the elim-
ination of lithium triflate, which is also rapid because of the
excellent leaving ability of triflate. An essential element in
step 1 is the strong electron-withdrawing effect of the triflato
ꢀ
(CF3SO3 ) group that reinforces the electrophilicity of the
iodine atom in 1, thus facilitating formation of the ate
complex C and leading to the iodine–lithium exchange.
Recently, we became interested in the slower generation
of benzyne, hoping to form the basis of developing catalytic
benzyne chemistry. Along these lines, we came up with an
Scheme 3. Catalytic generation of benzyne and nucleo-iodinaton.
idea to use alkynyllithium D, instead of nBuLi, with the
expected outcome depicted in Scheme 2. In spite of the poor
ability of an alkynyl anion (sp carbanion) for the halogen–
lithium exchange,[4] we expected that the slow release of
benzyne (B) would be possible because of two prominent
features of the iodo triflate 1: 1) high susceptibility to the ate
lithio nucleophile F (Li-Nu) is present in a stoichiometric
amount, B would be intercepted by F to generate aryllithium
G, which may be “counter iodinated” by iodoalkyne E (that is,
a favorable process by sp2 carbanion!sp carbanion), thus
giving the nucleo-iodination product H,[5] and regeneration of
the alkynyllithium D completes a catalytic cycle. The
alkynyllithium D, if employed in a stoichiometric amount,
acts also as a nucleophile which attacks B (i.e., D = F).
However, a stoichiometric amount of the nucleophile F could
be different from alkynyllithium D, given that suitable
reaction conditions are met.[6,7]
[*] Prof. Dr. T. Hamura, Y. Chuda, Y. Nakatsuji
Department of Chemistry, School of Science and Technology
Kwansei Gakuin University
and PRESTO, JST Agency
2-1 Gakuen, Sanda, Hyogo 669-1337 (Japan)
E-mail: thamura@kwansei.ac.jp
Prof. Dr. K. Suzuki
Department of Chemistry, Tokyo Institute of Technology
2-12-1 O-okayama, Meguro-ku, Tokyo 152-8551 (Japan)
E-mail: ksuzuki@chem.titech.ac.jp
Herein, a positive answer to this scenario is described
wherein the catalytic generation of benzyne species B from 2-
iodoaryl triflate (1) is enabled, followed by interception with
a stoichiometric amount of a nucleophile (halides or stabi-
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201108415.
3368
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 3368 –3372

Angewandte
Chemie
lized carbanions) and subsequent iodination. The overall
process corresponds to the nucleophilic addition and iodina-
tion (nucleo-iodination) of the catalytically generated ben-
zyne species, thus giving functionalized iodoarenes.
yield from the isomeric substrate 9, thus showing that the
common a-alkoxybenzyne I was involved. Note that, in the
latter case, the iodine atom formally migrated to the ortho
position of the benzyloxy group.[12,13]
Table 1 summarizes the initial feasibility study, which gave
The regioselectivity[11] is rationalized by the nucleophilic
attack at the more electrophilic C1 position of I by the
inductive effect of an alkoxy group (Scheme 4).[14] The
resulting aryl anion J[3c] is iodinated to give the product 8.
The iodoalkyne, generated by the initial iodine–lithium
exchange with iodo triflate 7, serves as an iodine source,
and is a favorable process (sp2 carbanion!sp carbanion).[15]
ꢁ
promising results. Upon treatment of 1 with LiC CSiMe3
(1.8 equiv) in Et2O (ꢀ78!258C), alkynyl iodobenzene 2 was
obtained, albeit in 10% yield (entry 1). The yield was
improved by using DME as a solvent (entry 2), and addition-
ally improved using THF at a lower temperature (ꢀ108C,
entry 3).[8–10] In contrast, toluene gave poorer results (entry 4).
Table 1: Feasibility study.
Entry[a]
Solvent
T [8C]
t [h]
Yield [%]
1
2
3
4
Et2O
DME
THF
25
0
12
3.5
2
10
67
82
11
ꢀ10
Scheme 4. Regioselectivity directed by an a-alkoxy group.
toluene
25
12
[a] 1.8 equiv of alkynyllithium was used. DME=dimethyl ether,
THF=tetrahydrofuran.
Such rigorous regioselectivity was seen also in the reaction
of a-fluoro-substituted benzyne K, which was generated from
10. Note that the iodine atom in the product 11 is formally
migrating to the position nearer to the fluorine atom
[Eq. (2)].
The reaction was applicable to various alkynyllithiums
(Table 2). Upon treatment of 1 with 1-hexynyllithium in DME
(ꢀ78!58C, 5 h), the iodoarene 3 was obtained in 81% yield.
The substituent (R) on the alkynyllithium could be either
cyclohexenyl or phenyl, thus giving good yields of the
respective products 4 and 5. The 2-bromophenyl group
stayed intact in the reaction to give product 6 in 65% yield.
Table 2: Iodoalkynylation with various substrate combinations.
Worth noting are the dual roles of the triflato group in the
substrate at the stage of benzyne generation. Before finally
serving as an excellent leaving group, it initially acts as an
activator of the neighboring iodine atom for the ate complex
formation en route to the iodine–lithium exchange. A telling
example is the reaction of bis(iodide) 12 [Eq. (3)]. Upon
R
Bu
product
3
4
5
6
yield [%][c]
81[a]
74[a]
76[b]
65[b]
[a] In DME. [b] In THF. [c] For details, see the Supporting Information.
Intermediacy of the benzyne species was firmly supported
by the following experiments [Eq. (1); TMS = trimethylsilyl].
ꢁ
Upon treatment of triflate 7 and LiC CSiMe3, a single
alkynylated product, 8, was obtained in 96% yield.[11]
Importantly, the same product 8 was also obtained in 92%
ꢁ
treatment with LiC CSiMe3, 12 underwent the reaction to
give product 13a in 74% yield. Notably, the isolated iodine
atom in 13a was fully retained without undergoing the lithio
exchange and protonation. Similarly, the reaction of 12 with
ꢁ
LiC CPh solely gave 13b in 82% yield.
The reactivity of the iodine atom in the substrate is
dependent on an even more subtle change of the neighboring
Angew. Chem. Int. Ed. 2012, 51, 3368 –3372
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3369

.
Angewandte
Communications
groups. We previously reported that the benzyne generation
of 14 by stoichiometric nBuLi occurs at low temperature
(ꢀ958C) at the triflato side, while that of the tosylato side
only proceeds at above ꢀ788C.[3c] Also under the present
reaction conditions, the behavior of bis(iodide) 14 proved
Nu2ꢀ attacks B to form aryl anion G. Note that the reactions
described so far included a single nucleophile, Nu1 = Nu2 =
ꢁ
LiC CR.
Could two nucleophilic species (Nu1ꢀ, Nu2ꢀ) be different?
We assumed that process 2 should be more facile considering
the extremely high reactivity of benzyne, and Nu2 could be
interesting [Eq. (4)].[16] Upon treatment of 14 with LiC
ꢁ
ꢁ
a stabilized anion. Knowing that LiC CR could serve as the
catalytic nucleophile for generating benzyne (Nu1ꢀ), we could
reasonably select stoichiometric nucleophile (Nu2ꢀ) that is
ꢁ
less reactive. A rough measure of the pKa is 25 (HC CH), and
we examined anionic nucleophiles derived from fairly acidic
conjugate acids with pKa values less than 25.[17]
As a proof-of-concept experiment, we employed halides
as weak nucleophiles (Nu2ꢀ). Indeed, upon treatment of iodo
triflate 7 with stoichiometric amounts of LiCl, LiBr, or LiI in
ꢁ
the presence of LiC CSiMe3 (10 mol%), smooth “haloiodi-
CSiMe3 (1.2 equiv, ꢀ78!ꢀ558C, Et2O, 2 h), the selective
reaction occurred to give product 15 as the sole product in
86% yield. No indication of the products that resulted from
the reaction at the iodo tosylate side of 14 was noted.
nation” of benzyne proceeded to give the corresponding
chloro iodide 18, bromo iodide 19, and iodo iodide 20 in
excellent yield (Table 3). The regiochemistries of 18 and 19
were assigned by NMR analysis (HMBC).[11] By contrast, in
ꢁ
ꢁ
However, by using LiC CSiMe3 in excess (2.5 equiv) at 08C
the absence of LiC CSiMe3 none of the products 18–20 were
obtained, even at higher reaction temperature (ꢀ78!258C);
two alkynyl units were smoothly introduced (Et2O, 0.1 h),
thus converting bis(iodide) 14 into bis(alkyne) 16 in 82%
yield.
instead, only the starting material 7 was recovered.
Furthermore, two different alkynyl units could be intro-
Table 3: Alkynyllithium-catalyzed haloiodination of benzyne.
ꢁ
duced in one pot. Upon treatment of bis(iodide) 14 with LiC
CSiMe3 (1.1 equiv) in Et2O (ꢀ78!ꢀ508C), the starting
material was gradually consumed, and formation of the
monoadduct 15 was observed by TLC. As the second
ꢁ
nucleophile, LiC CSi(iPr)3 (2.0 equiv) was added to the
Entry
X
T
t [h]
Product
Yield [%]
mixture (08C, 0.1 h), thereby giving the asymmetrically
bis(alkyn)ylated product 17 as a sole product in 63% yield
[Eq. (5); TIPS = triisopropylsilyl].
1
2
3
Cl
Br
I
ꢀ78!ꢀ208C
ꢀ78!08C
ꢀ30!08C
3
3
1.5
18
19
20
90
97
91
Pleasingly, it was found that lithio carbanions, if more
ꢁ
stabilized than LiC CR, served as stoichiometric nucleo-
philes as well. A representative example is shown in
Equation (6), wherein the alkynyllithium 21 with a tert-butyl
ester moiety, thus more stabilized compared to the usual
alkynyllithium, took part in the reaction. Upon treatment of 7
Our goal in this project was to explore the possibility of
catalytic generation of benzyne, with an appropriate nucle-
ophile serving as a stoichiometric reaction partner. Let us
consider two roles played by the nucleophiles (Scheme 5; and
see Scheme 3): process 1: Nu1ꢀ attacks the iodine atom in 1 to
form the ate complex L, thus leading to B; and process 2:
ꢁ
with 21 (3 equiv) in the presence of 30 mol% of LiC CSiMe3
(ꢀ78!-608C), the carboiodination occurred smoothly to give
the product 22 in 82% yield [Eq. (6)]. No alkynylated product
8 was produced. It should be noted that the use of the
ꢁ
stoichiometric 21 in the absence of LiC CSiMe3 resulted in
no reaction, thus confirming the inability of 21 to effect the
benzyne generation.
Various other carbon nucleophiles could be employed
(Table 4). Indenyllithium and pentamethylcyclopentadienyl-
Scheme 5. Two roles played by nucleophilic species.
3370
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 3368 –3372

Angewandte
Chemie
b) I. Sapountzis, W. Lin, M. Fisher, P. Knochel, Angew. Chem.
Table 4: Alkynyllithium-catalyzed carboiodination of benzyne.
2004, 116, 4464 – 4466; Angew. Chem. Int. Ed. 2004, 43, 4364 –
4366; c) Z. Liu, R. C. Larock, Org. Lett. 2004, 6, 99 – 102; d) M.
Uchiyama, T. Miyoshi, Y. Kajihara, T. Sakamoto, Y. Otani, T.
Ohwada, Y. Kondo, J. Am. Chem. Soc. 2002, 124, 8514 – 8515;
e) M. Uchiyama, Y. Kobayashi, T. Furuyama, S. Nakamura, Y.
Kajihara, T. Miyoshi, T. Sakamoto, Y. Kondo, K. Morokuma, J.
Am. Chem. Soc. 2008, 130, 472 – 480.
Entry
1
NuLi
T [8C]
ꢀ60
t
[h]
Product
Yield
[%]
[3] a) T. Matsumoto, T. Hosoya, M. Katsuki, K. Suzuki, Tetrahedron
Lett. 1991, 32, 6735 – 6736; b) T. Hosoya, E. Takashiro, T.
Matsumoto, K. Suzuki, J. Am. Chem. Soc. 1994, 116, 1004 –
1015; c) T. Hamura, T. Arisawa, T. Matsumoto, K. Suzuki,
Angew. Chem. 2006, 118, 6996 – 6998; Angew. Chem. Int. Ed.
2006, 45, 6842 – 6844.
15
19
77
61
[4] For selected examples of halogen–lithium exchange of function-
alized arenes, see: a) J. Clayden, Organolithiums: Selectivity for
2
25
Synthesis, Vol. 23
, Pergamon, Oxford, 2002, pp. 111 – 147
(Tetrahedron Organic Chemistry Series); b) P. Beak, D. J.
Allen, J. Am. Chem. Soc. 1992, 114, 3420 – 3425; c) C. Nꢀjera,
J. M. Sansano, M. Yus, Tetrahedron 2003, 59, 9255 – 9303; d) M.
Da˛browski, J. Kubicka, S. Lulinski, J. Serwatowski, Tetrahedron
2005, 61, 6590 – 6595.
3
4
5
5
2
49
72
[5] a) D. Astruc, Modern Arene Chemistry, Wiley-VCH, Weinheim,
2002; b) C. B. de Koning, A. L. Rousseau, W. A. L. van Otterlo,
Tetrahedron 2003, 59, 7 – 36.
ꢀ30
[6] For selected three-component coupling of arynes with anionic
nucleophiles and electrophiles, see: a) G. Wittig, R. W. Hoff-
mann, Chem. Ber. 1962, 95, 2729 – 2734; b) A. I. Meyers, P. D.
Pansegrau, Tetrahedron Lett. 1983, 24, 4935; c) A. I. Meyers,
P. D. Pansegrau, J. Chem. Soc. Chem. Commun. 1985, 690; d) C.-
J. F. Du, H. Hart, K.-K. D. Ng, J. Org. Chem. 1986, 51, 3162 –
3165; e) T. Ghosh, H. Hart, J. Org. Chem. 1988, 53, 3555 – 3558;
f) A. Saednya, H. Hart, Synthesis 1996, 1455; g) S. Tripathy, R.
LeBlanc, T. Durst, Org. Lett. 1999, 1, 1973 – 1975; h) S. Tripathy,
H. Hussain, T. Durst, Tetrahedron Lett. 2000, 41, 8401 – 8405;
i) H. Tomori, J. M. Fox, S. L. Buchwald, J. Org. Chem. 2000, 65,
5334 – 5341; j) F. Leroux, M. Schlosser, Angew. Chem. 2002, 114,
4447 – 4450; Angew. Chem. Int. Ed. 2002, 41, 4272 – 4274; k) W.
Lin, I. Sapountzis, P. Knochel, Angew. Chem. 2005, 117, 4330 –
4333; Angew. Chem. Int. Ed. 2005, 44, 4258 – 4261; l) I. Larrosa,
M. I. D. Silva, P. M. Gꢁmez, P. Hannen, E. Ko, S. R. Lenger, S. R.
Linke, A. J. P. White, D. Wilton, A. G. M. Barrett, J. Am. Chem.
Soc. 2006, 128, 14042 – 14043; m) D. Soorukram, T. Qu, A. G. M.
Barrett, Org. Lett. 2008, 10, 3833 – 3835.
[7] For selected three-component coupling of arynes with neutral
nucleophiles and electrophiles, see: a) H. Yoshida, E. Shirakawa,
Y. Honda, T. Hiyama, Angew. Chem. 2002, 114, 3381 – 3383;
Angew. Chem. Int. Ed. 2002, 41, 3247 – 3249; b) H. Yoshida, H.
Fukushima, J. Ohshita, A. Kunai, Angew. Chem. 2004, 116,
4025 – 4028; Angew. Chem. Int. Ed. 2004, 43, 3935 – 3938; c) Z.
Liu, R. C. Larock, J. Am. Chem. Soc. 2005, 127, 13112 – 13113;
d) C. D. Gilmore, K. M. Allan, B. M. Stoltz, J. Am. Chem. Soc.
2008, 130, 1558 – 1559.
lithium were found to serve as efficient carbon nucleophiles in
ꢁ
the presence of 10 mol% of LiC CSiMe3, selectively giving
iodoarenes 23 and 24 in 77 and 61% yields, respectively
(entries 1 and 2).[18] A precaution for the reaction of
indenyllithium was to maintain the temperature (< ꢀ608C),
otherwise a shift of the double bond occurred. The lithium
enolate 25[19] also worked nicely. Upon treatment of iodo
ꢁ
triflate 7 in the presence of LiC CSiMe3 (10 mol%, ꢀ78!
58C), the reaction occurred to give ketone 26 in 49% yield
(entry 3). Similarly, ester 28 was selectively obtained by using
this method (entry 4).[20]
In summary, the catalytic generation of a benzyne species
by alkynyllithium catalysis allowed efficient preparation of
various iodoarenes through nucleo-iodination. Further stud-
ies on expansion of the scope of the catalytic process are being
actively investigated.
Received: November 29, 2011
Revised: December 24, 2011
Published online: February 22, 2012
Keywords: arenes · arynes · iodine · lithiation · regioselectivity
.
[8] All new compounds were fully characterized by NMR analysis,
high-resolution MS, and/or combustion analysis. See the Sup-
porting Information.
[9] Bromoalkynylation of benzyne was reported by insertion of
benzyne into CuBr and subsequent alkynylation of the resulting
aryl copper species with bromoalkynes; see: T. Morishita, H.
Yoshida, J. Ohshita, Chem. Commun. 2010, 46, 640 – 642.
[10] Generation of benzyne by halogen – lithium exchange of lithi-
ated allyltrimethylsilane with 2-iodophenyl triflate has been
reported; see: A. Ganta, T. S. Snowden, Org. Lett. 2008, 10,
5103 – 5106.
[1] For reviews on arynes, see: a) R. W. Hoffmann, Dehydrobenzene
and Cycloalkynes, Academic, New York, 1967; b) S. V. Kessar in
Comprehensive Organic Synthesis, Vol. 4 (Ed.: B. M. Trost),
Pergamon, Oxford, UK, 1991, pp. 483 – 515; c) H. Pellissier, M.
Santelli, Tetrahedron 2003, 59, 701 – 730; d) H. H. Wenk, M.
Winkler, W. Sander, Angew. Chem. 2003, 115, 518 – 546; Angew.
Chem. Int. Ed. 2003, 42, 502 – 528.
[2] For selected methods for generating functionalized arynes, see:
a) T. Kitamura, M. Yamane, K. Inoue, M. Tokuda, N. Fukatsu, Z.
Meng, Y. Fujiwara, J. Am. Chem. Soc. 1999, 121, 11674 – 11679;
[11] The regiochemistry was assigned by the NMR analysis (HMBC).
For details, see the Supporting Information.
Angew. Chem. Int. Ed. 2012, 51, 3368 –3372
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3371

.
Angewandte
Communications
[12] For related “halogen migration” in benzyne chemistry, see
Refs. [6g] and [6j].
[13] Superior reactivity of iodo triflate 7 relative to its regioisomer 9
is ascribed to the ease of iodine–lithium exchange in the
presence of the neighboring benzyloxy group in 7.
[17] An additional hope was that most of the Nu1ꢀ used in a catalytic
amount would be converted into the corresponding iodide (Nu1-
I), so that the crossover attack on the benzyne species would be
minimized.
[18] The reaction of 7 with pentamethylcyclopentadienyllithium did
ꢁ
[14] For experimental and computational studies on the origin of the
regioselectivities in the nucleophilic addition to substituted
benzynes, see: a) P. H.-Y. Cheong, R. S. Paton, S. M. Bronner,
G.-Y. Im, N. K. Garg, K. N. Houk, J. Am. Chem. Soc. 2010, 132,
1267 – 1269; b) G.-Y. Im, S. M. Bronner, A. E. Goetz, R. S.
Paton, P. H.-Y. Cheong, K. N. Houk, N. K. Garg, J. Am. Chem.
Soc. 2010, 132, 17933 – 17944.
[15] As another possibility, triflate 7 may also work as an iodinating
agent. Further study is in progress.
[16] We recently developed a one-pot dual cycloaddition of a benz-
diyne equivalent; see: T. Arisawa, T. Hamura, H. Uekusa, T.
Matsumoto, K. Suzuki, Synlett 2008, 1179. See also Ref. [3c].
not occur at all in the absence of LiC CSiMe3. Details of the
catalytic reactions using other lithio-nucleophiles will be de-
scribed in a full paper.
[19] a) D. Seebach, M. Ertas, R. Locher, W. B. Schweizer, Helv.
Chim. Acta 1985, 68, 264 – 281; b) M. Ertas¸, D. Seebach, Helv.
Chim. Acta 1985, 68, 961 – 969.
[20] Lithium enolate 27 was generated by reaction of the ketene silyl
acetal, prepared from tert-butyl acetate, and MeLi. For related
reactions, see: G. Stork, P. F. Hudrlik, J. Am. Chem. Soc. 1968,
90, 4464 – 4465.
3372
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 3368 –3372