Elimination-addition mechanism for nucleophilic substitution reaction of cyclohexenyl iodonium salts and regioselectivity of nucleophilic addition to the cyclohexyne intermediate

Article abstract of DOI:10.1021/ja0496672

The reaction of 4-substituted cyclohex-1-enyl(phenyl)iodonium tetrafluoroborate with tetrabutyl-ammonium acetate gives both the ipso and cine acetate-substitution products in aprotic solvents. The isomeric 5-substituted iodonium salt also gives the same mixture of the isomeric acetate products. The reaction is best explained by an elimination-addition mechanism with 4-substituted cyclohexyne as a common intermediate. The cyclohexyne formation was confirmed by deuterium labeling and trapping to lead to [4 + 2] cycloadducts and a platinum-cyclohexyne complex. Cyclohexyne can also be generated in the presence of some other mild bases such as fluoride ion, alkoxides, and amines, though amines are less effective bases for the elimination. Kinetic deuterium isotope effects show that the anionic bases induce the E2 elimination (k _H/k_D > 2), while the amines allow formation of a cyclohexenyl cation in chloroform to lead to E1 as well as SN1 reactions (k _H/k_D ≈ 1). Bases are much less effective in methanol, and methoxide was the only base to efficiently afford the cyclohexyne intermediate. Nucleophiles react with the cyclohexyne to give regioisomeric products in the ratio dependent on the ring substituent. The observed regioselectivity of nucleophilic addition to substituted cyclohexynes is rationalized from calculated LUMO populations, which are governed by the bond angles at the acetylenic carbons: The less deformed carbon has a higher LUMO population and is preferentially attacked by the nucleophile.

Full text of DOI:10.1021/ja0496672

Published on Web 05/26/2004
Elimination-Addition Mechanism for Nucleophilic
Substitution Reaction of Cyclohexenyl Iodonium Salts and
Regioselectivity of Nucleophilic Addition to the Cyclohexyne
Intermediate
Morifumi Fujita,*,† Wan Hyeok Kim,† Yuichi Sakanishi,† Koji Fujiwara,†
Sayaka Hirayama,† Tadashi Okuyama,*,† Yasuhiro Ohki,‡ Kazuyuki Tatsumi,‡ and
Yasunori Yoshioka§
Contribution from Graduate School of Material Science, Himeji Institute of Technology,
UniVersity of Hyogo, Kohto, Kamigori, Hyogo 678-1297, Japan,
Research Center for Materials Science, Nagoya UniVersity, Furo-cho,
Chikusa-ku, Nagoya 464-8602, Japan, and Chemistry Department for Materials,
Faculty of Engineering, Mie UniVersity, Kamihama, Tsu, Mie 514-8507, Japan
Received January 19, 2004; E-mail: fuji@sci.u-hyogo.ac.jp; okuyama@sci.u-hyogo.ac.jp
Abstract: The reaction of 4-substituted cyclohex-1-enyl(phenyl)iodonium tetrafluoroborate with tetrabutyl-
ammonium acetate gives both the ipso and cine acetate-substitution products in aprotic solvents. The
isomeric 5-substituted iodonium salt also gives the same mixture of the isomeric acetate products. The
reaction is best explained by an elimination-addition mechanism with 4-substituted cyclohexyne as a
common intermediate. The cyclohexyne formation was confirmed by deuterium labeling and trapping to
lead to [4 + 2] cycloadducts and a platinum-cyclohexyne complex. Cyclohexyne can also be generated
in the presence of some other mild bases such as fluoride ion, alkoxides, and amines, though amines are
less effective bases for the elimination. Kinetic deuterium isotope effects show that the anionic bases induce
the E2 elimination (kH/kD > 2), while the amines allow formation of a cyclohexenyl cation in chloroform to
lead to E1 as well as SN1 reactions (kH/kD ≈ 1). Bases are much less effective in methanol, and methoxide
was the only base to efficiently afford the cyclohexyne intermediate. Nucleophiles react with the cyclohexyne
to give regioisomeric products in the ratio dependent on the ring substituent. The observed regioselectivity
of nucleophilic addition to substituted cyclohexynes is rationalized from calculated LUMO populations, which
are governed by the bond angles at the acetylenic carbons: The less deformed carbon has a higher LUMO
population and is preferentially attacked by the nucleophile.
are R-elimination (eq 1) and SN2-type (SNVσ)3 substitution (eq
2) at the R position.4-7 Some nucleophiles form a hypervalent
Introduction
Vinyl iodonium salts are a class of very reactive compounds,
which undergo nucleophilic substitution and base-induced
elimination.1,2 Main reactions of simple alk-1-enyliodonium salts
† University of Hyogo.
‡ Nagoya University.
§ Mie University.
(1) For reviews of reactions of iodonium salts, see: (a) Stang, P. J. Angew.
Chem., Int. Ed. Engl. 1992, 31, 274-285. (b) Varvoglis, A. The Organic
Chemistry of Polycoordinated Iodine; VCH: New York, 1992. (c) Grushin,
V. V. Acc. Chem. Res. 1992, 25, 529-536. (d) Koser, G. F. In The
Chemistry of Functional Groups, Supplement D2; Patai, S., Rappoport,
Z., Eds.; John Wiley & Sons: Chichester, 1995; Chapter 21. (e) Stang, P.
J.; Zhdankin, V. V. Chem. ReV. 1996, 96, 1123-1178. (f) Varvoglis, A.
Tetrahedron 1997, 53, 1179-1255. (g) Zhdankin, V. V.; Stang, P. J. In
Chemistry of HyperValent Compounds; Akiba, K.-y., Ed.; Wiley-VCH:
New York, 1999; Chapter 11. (h) Ochiai, M. In Chemistry of HyperValent
Compounds; Akiba, K.-y., Ed.; Wiley-VCH: New York, 1999; Chapter
12. (i) Zhdankin, V. V.; Stang, P. J. Chem. ReV. 2002, 102, 2523-2584.
(j) Stang, P. J. J. Org. Chem. 2003, 68, 2997-3008. (k) Wirth, T., Ed.
Topics of Current Chemistry, 224, HyperValent Iodine Chemistry;
Springer: Berlin, 2003.
adduct, λ3-iodane, in a rapid equilibrium in solution, and the
adduct can lead to ligand coupling (SNVπ)3,8,9 or an intramo-
(3) For the designation of SNVσ and SNVπ of nucleophilic substitution with
inversion via σ* attack and that with retention via π* attack, see ref 2d.
(4) Ochiai, M.; Oshima, K.; Masaki, Y. J. Am. Chem. Soc. 1991, 113, 7059-
7061.
(5) Okuyama, T.; Takino, T.; Sato, K.; Oshima, K.; Imamura, S.; Yamataka,
H.; Asano, T.; Ochiai, M. Bull. Chem. Soc. Jpn. 1998, 71, 243-257.
(6) Okuyama, T.; Takino, T.; Sato, K.; Ochiai, M. J. Am. Chem. Soc. 1998,
120, 2275-2282.
(2) For reviews on vinyl iodonium salts, see: (a) Okuyama, T. ReV. Heteroatom
Chem. 1999, 21, 257-275. (b) Ochiai, M. J. Organomet. Chem. 2000, 611,
494-508. (c) Okuyama, T. Acc. Chem. Res., 2002, 35, 12-18. (d)
Okuyama, T.; Lodder, G. AdV. Phys. Org. Chem. 2002, 37, 1-56. (e)
Okuyama, T.; Fujita, M. Proc. Jpn. Acad. 2002, 78 (B), 167-172.
(7) Okuyama, T.; Imamura, S.; Ishida, Y. Bull. Chem. Soc. Jpn. 2001, 74,
543-548.
(8) Ochiai, M.; Oshima, K.; Masaki, Y. Chem. Lett. 1994, 871-874.
9
7548
J. AM. CHEM. SOC. 2004, 126, 7548-7558
10.1021/ja0496672 CCC: $27.50 © 2004 American Chemical Society

Cyclohexynes from Iodonium Salts
A R T I C L E S
Scheme 1
be generated in solution as transient species by elimination
reactions of appropriate precursors but readily undergo oligo-
merization and reactions with nucleophiles.13,20,23 Regioselec-
tivity of nucleophilic addition to cycloalkynes was previously
noted in some examples,13 and relatively careful examinations
were made for methyl-substituted cyclohexynes, generated from
the 1-halocyclohexenes with strong bases such as t-BuOK and
t-BuONa-NaNH2 (eq 5).25,26 Under these reaction conditions,
lecular â-elimination (Scheme 1).5-7 Cyclohex-1-enyliodonium
tetrafluoroborate 1 undergoes an SN1-type solvolysis via a
cyclohexenyl cation intermediate in protic solvents (eq 3).10
Reactions of this cyclic iodonium salt with nucleophiles/bases
are now further investigated, and a new type of substitution via
an elimination-addition (EA) mechanism with a cyclohexyne
intermediate to give ipso and cine products11 will be presented
in this paper (eq 4).12
cyclohexa-1,2-diene was also formed.30,31 Cycloalkynes are
known to isomerize to cycloalka-1,2-dienes under strongly basic
conditions. The ratio of ipso/cine substitution products was
found to change from 98/2 to 56/44 depending on the reaction
conditions25,26 probably due to the changing ratio of cyclohex-
yne/cyclohexa-1,2-diene intermediates. The intrinsic regiose-
lectivity of nucleophilic addition to cyclohexyne is unknown
for this reason. Thus, a mild and general method for selective
generation of cyclohexyne is needed to permit study of its
reactivity.32
Chemistry of small-ring cycloalkynes has been studied for
more than a century13,14 but still remains as a challenging
problem because of their constrained structure and high
reactivity.12-29 Among the angle-strained cycloalkynes, cyclo-
heptyne has some lifetime in solution at -25 °C,23a while
cyclohexyne is only isolable in matrixes at 77 K.24a They can
(19) For theoretical studies for cycloalkynes, see: (a) Bachrach, S. M.; Gilbert,
J. C.; Laird, D. W. J. Am. Chem. Soc. 2001, 123, 6706-6707. (b) Gilbert,
J. C.; Kirschner, S. Tetrahedron 1996, 52, 2279-2290. (c) Johnson, R. P.;
Daoust, K. J. J. Am. Chem. Soc. 1995, 117, 362-367. (d) Dewar, M. J.;
Gilbert, J. C.; Kirschner, S. J. Chem. Soc., Chem. Commun. 1994, 1105-
1106. (e) Gilbert, J. C.; Kirschner, S. Tetrahedron Lett. 1993, 34, 559-
602. (f) Gilbert, J. C.; Kirschner, S. Tetrahedron Lett. 1993, 34, 603-606.
(g) Tseng, J.; McKee, M. L.; Shevlin, P. B. J. Am. Chem. Soc. 1987, 109,
5474-5477. (h) Olivella, S.; Perica`s, M. A.; Riera, A.; Sole´, A. J. Org.
Chem. 1987, 52, 4160-4163.
(9) Okuyama, T.; Takino, T.; Sato, K.; Ochiai, M. Chem. Lett. 1997, 955-
956.
(20) Gilbert, J. C.; Baze, M. E. J. Am. Chem. Soc. 1983, 105, 664-665. Gassman,
P. G.; Valcho, J. J. J. Am. Chem. Soc. 1975, 97, 4768-4770. Montgomery,
L. K.; Applegate, L. E. J. Am. Chem. Soc. 1967, 89, 5305-5307.
(21) Strozier, R. W.; Caramella, P.; Houk, K. N. J. Am. Chem. Soc. 1979, 101,
1340-13443. Rondan, N. G.; Domelsmith, L. N.; Houk, K. N.; Bowne,
A. T.; Levin, R. H. Tetrahedron Lett. 1979, 3237-3240. Hoffmann, R.;
Imamura, A.; Hehre, W. J. J. Am. Chem. Soc. 1968, 90, 1499-1509.
(22) Ng, L.; Jordan, K. D.; Krebs, A.; Ru¨ger, W. J. Am. Chem. Soc. 1982, 104,
7414-7416.
(10) Okuyama, T.; Takino, T.; Sueda, T.; Ochiai, M. J. Am. Chem. Soc. 1995,
117, 3360-3367.
(11) The ipso and cine substitutions are designated generally for aromatic
substitutions in which the entering group takes up the same position
occupied by the leaving group and a position adjacent to it, respectively.
Gold, V. Pure Appl. Chem. 1983, 55, 1281-1371.
(12) For a preliminary account of this paper, see: Fujita, M.; Sakanishi, Y.;
Kim, W. H.; Okuyama, T. Chem. Lett. 2002, 908-909.
(23) (a) Wittig, G.; Meske-Schu¨ller, J. Liebigs Ann. Chem. 1968, 711, 65-75.
(b) Krebs, A.; Cholcha, W.; Mu¨ller, M.; Eicher, T.; Pielartzik, H.;
Schno¨ckel, H. Tetrahedron Lett. 1984, 25, 5027-5030.
(13) (a) Hoffmann, R. W. Dehydrobenzene and Cycloalkynes; Academic
Press: New York, 1967. (b) Krebs, A.; Wilke, J. Top. Curr. Chem. 1983,
109, 189-233. (c) Gleiter, R.; Merger, R. In Modern Acetylene Chemistry;
Stang, P. J., Diederich, F., Eds.; VCH: Weinheim, 1995; pp 285-319. (d)
Hopf, H. Classic Hydrocarbon Chemistry; Wiley-VCH: Weinheim, 2000.
(e) Sander, W. Angew. Chem., Int. Ed. Engl. 1994, 33, 1455-1456.
(14) For very early attempts to generate cycloalkynes, see: Blumethal, M. Ber.
Dtsch. Chem. Ges. 1874, 7, 1092. Markownikoff, W. J. Prakt. Chem. 1894,
49, 409. Markownikoff, W. Liebigs Ann. Chem. 1903, 59, 327.
(15) Kitamura, T.; Kotani, M.; Yokoyama, T.; Fujiwara, Y.; Hori, K. J. Org.
Chem. 1999, 64, 680-681.
(24) (a) IR detection of cyclohexyne: Wentrup, C.; Blanch, R.; Briehl, H.;
Gross,G. J. Am. Chem. Soc. 1988, 110, 1874-1880. (b) Early study for
cyclohexyne formation: Favorsky, A.; Boshowsky, W. Liebigs Ann. Chem.
1912, 390, 122.
(25) Brunet, J. J. Tetrahedron 1971, 27, 3515-3526. Fixari, B.; Brunet, J. J.;
Caube´re, P. Tetrahedron 1976, 32, 927-934.
(26) Bottini, A. T.; Corson, F. P.; Fitzgerald, R.; Frost, II, K. A. Tetrahedron
1972, 28, 4883-4904.
(27) Atanes, N.; Escudero, S.; Pe´rez, D.; Guitia´n, E.; Castedo, L. Tetrahedron
Lett. 1998, 39, 3039-3040. Shakespeare, W. C.; Johnson R. P. J. Am. Chem.
Soc. 1990, 112, 8578-8579. Cunico, R. F.; Dexheimer, E. M. J. Organomet.
Chem. 1973, 59, 153-160. Wong, H. N. C.; Ye, X.-S. J. Org. Chem. 1997,
62, 1940-1954.
(16) Laird, D. W.; Gilbert, J. C. J. Am. Chem. Soc. 2001, 123, 6704-6705.
(17) (a) Tu¨mer, F.; Taskesenligil, Y.; Balci, M. J. Org. Chem. 2001, 66, 3806-
3810. (b) Fabian, J. J. Org. Chem. 2000, 65, 8940-8947. (c) Gilbert, J.
C.; Hou, D.-R.; Grimme, J. W. J. Org. Chem. 1999, 64, 1529-1534. (d)
Marchand, A. P.; Namboothiri, I. N. N.; Ganguly, B.; Bott. S. G. J. Am.
Chem. Soc. 1998, 120, 6871-6876. (e) Gilbert, J. C.; McKinley, E. G.;
Hou, D.-R. Tetrahedron 1997, 53, 9891-9902. (f) Marchand, A. P.;
Ramanaiah, K. C. V.; Bott, S. G.; Gilbert, J. C.; Kirschner, S. Tetrahedron
Lett. 1996, 37, 8101-8104. (g) Marchand, A. P.; Kumar, K. A.; Rajagopal,
D.; Eckrich, R.; Bott, S. G. Tetrahedron Lett. 1996, 37, 467-470. (h)
Taskesenligil, Y.; Kashyap, R. P.; Watson, W. H.; Balci, M. J. Org. Chem.
1993, 58, 3216-3218. (i) Gilbert, J. C.; Baze, M. E. J. Am. Chem. Soc.
1984, 106, 1885-1886. (j) Montgomery, L. K.; Clouse, A. O.; Crelier, A.
M.; Applegate, L. E. J. Am. Chem. Soc. 1967, 89, 3453-3457. (k)
Montgomery, L. K.; Applegate, L. E. J. Am. Chem. Soc. 1967, 89, 2952-
2960. (l) Erickson, K. L.; Wolinsky, J. J. Am. Chem. Soc. 1965, 87, 1142-
1143. (m) Montogomery, L. K.; Scardiglia, F.; Roberts, J. D. J. Am. Chem.
Soc. 1965, 87, 1917-1925.
(28) (a) Fujita, M.; Sakanishi, Y.; Okuyama, T. J. Am. Chem. Soc. 2001, 123,
9190-9191. (b) Fujita, M.; Sakanishi, Y.; Nishii, M.; Okuyama, T. J. Org.
Chem. 2002, 67, 8138-8146.
(29) Fujita, M.; Ihara, K.; Kim, W. H.; Okuyama, T. Bull. Chem. Soc. Jpn.
2003, 76, 1849-1855.
(30) For cycloalka-1,2-dienes, see: (a) Johnson, R. P. Chem. ReV. 1989, 89,
1111-1124. (b) Caube´re, P. Chem. ReV. 1993, 93, 2317-2334. (c) Balci,
M.; Jones, W. M. J. Am. Chem. Soc. 1980, 102, 7607-7608. (d) Lam, B.;
Johnson, R. P. J. Am. Chem. Soc. 1983, 105, 7479-7483. (e) Angus, R.
O., Jr.; Schmidt, M. W.; Johnson, R. P. J. Am. Chem. Soc. 1985, 107,
532-537. (f) Miller, B.; Shi, X. J. Am. Chem. Soc. 1987, 109, 578-579.
(g) Jamart-Gre´goire, B.; Brosse, N.; Inelli, S.; Nardelli, M.; Caube´re, P. J.
Org. Chem. 1993, 58, 4572-4578.
(31) Wittig, G.; Fritze, Angew. Chem., Int. Ed. Engl. 1966, 5, 846. Wittig, G.;
Fittze, P. Liebigs Ann. Chem. 1968, 711, 82-87.
(32) (a) Metal ion can also affect reactivity of cycloalkynes. Lithium complex
of cyclopentyne has been reported.17c,32b (b) Gilbert, J. C.; Hou, D.-R. J.
Org. Chem. 2003, 68, 10067-10072.
(18) For recent examples of synthetic application of cycloalkynes, see: (a) Adam,
W.; Bosio, S. G.; Fro¨hling, B.; Leusser, D.; Stalke, D. J. Am. Chem. Soc.
2002, 124, 8316-8320. (b) Iglesias, B.; Pena, D.; Pe´rez, D.; Guitia´n, E.;
Castedo, L. Synlett 2000, 486-488.
9
J. AM. CHEM. SOC. VOL. 126, NO. 24, 2004 7549

A R T I C L E S
Fujita et al.
Table 1. Reaction of 1 with Tetrabutylammonium Acetatea
Scheme 2
yield (%)
run
substrate
[acetate] (M)
3
4
PhI
5 (6)
3/4b
1
2
3
4c
5
6
7c
1a
1b
1b
1b
1c
0.01
0.10
0.01
0.01
0.10
0.01
0.01
0.10
0.01
50
82
2
15
6
3
7
7
6
11
9
24
20
37 44 69
38 44 70
25 27 66
21 33 74
20 32 81
20 32 70
20 52 89
15 44 73
46/54 (44/56)
46/54
48/52
39/61 (39/61)
39/61 (39/61)
38/62
28/72 (27/73)
25/75
Selective formation of ring-strained triple-bond compounds
has recently been achieved by using the 1,2-elimination of
â-silylvinyl iodonium salts15,16 or â-silylvinyl triflates.27 Kita-
mura15 and Gilbert16 independently showed that norbornyne
derivatives could be generated in a high yield from â-silyl
iodonium salts (Scheme 2). This method combines the advantage
of a facile fluoride-induced desilylation with that of the high
nucleofugality of the iodonio group.10 This strategy was also
applied successfully to the generation of benzynes.33,34 The
regioselectivity of nucleophilic addition to substituted benzynes
has been studied extensively to show a good correlation with
the inductive parameters of the substituents of benzyne.13a,35-38
This paper is also concerned with the regioselectivity of
nucleophilic addition to substituted cyclohexynes formed during
the EA reaction of cyclohex-1-enyliodonium salts. The regio-
selectivity is compatible with the calculated LUMO populations
at the acetylenic carbons of substituted cyclohexynes, which
are controlled by the bond angle.
1c
1c
8
9
1d
1d
1d
1d
2d
2d
1e
1e
2e
2e
1d′
1d′
10d
11e
12
13
14
15
16
17
18g
19e,g
0.03 (THF)d 10 23 76
30/70
25/75
0.1 (AN)e
0.10
0.01
0.10
0.01
0.10
0.01
0.01
0.01 (AN)e
13f 39f 65
15 55 78
(7) 21/79
(5) 20/80
28
29
(18) 51/49 (51/49)
(18) 48/52 (49/51)
7
9
37 76
36 27 67
33 27 64
31 30 60
20 22 55
11 45 76h
57/43 (57/43)
55/45 (57/43)
20/80
25/75
4
12 56h
29
a Reactions were carried out in chloroform at 60 °C for 2 h, unless
otherwise noted. b The ratio of 3/4 and the values in parentheses are those
of the cyclohexanones (11/12) after hydrolysis of the products. c In the
d
presence of methanol-O-d (1 vol %). Tetrahydrofuran is used as a solvent.
e Acetonitrile is used as a solvent. f Cyclohexanones were also obtained and
sums of the yields of 3d + 11d and 4d + 12d are given. g Mesitylene was
also detected in 6 and 29% yield for runs 18 and 19, respectively. h Yields
of iodomesitylene.
Results
Reaction with Acetate. The reactions of cyclohex-1-enyl-
(phenyl)iodonium tetrafluoroborate (1a) and 4- and 5-substituted
derivatives (1b-e and 2d,e) with tetrabutylammonium acetate
were carried out in chloroform at 60 °C. The products include
two regioisomers of 1-acetoxycyclohexene, 3 and 4, which
correspond to the ipso and cine substitution products of 1 or
the cine and ipso products of 2, respectively (eqs 6 and 7).
confirmed by the analysis of the cyclohexanones obtained upon
acid-catalyzed hydrolysis of the acetate products. The cine
substitution product was obtained in all the cases, and the
isomeric iodonium substrates, 1 and 2, resulted in similar product
ratios of 3/4 (runs 8-17). Formation of a mixture of the ipso
and cine products is in contrast to the selective formation of
the ipso substitution product observed in the acetolysis of 1 and
2 in unbuffered acetic acid (see Supporting Information for
experimental details).
In tetrahydrofuran or acetonitrile solution, the yields of the
acetate products were lower with increasing yields of iodocy-
clohexene 5 (runs 10, 11, and 19), while the regioisomeric ratio
of the acetate products 3/4 remained similar to that observed in
chloroform. The ratio 3/4 depends on the substituent but is not
much affected by the concentration of acetate ion, a small
amount of added methanol (1 vol %) (runs 4 and 7), or the
regiochemistry of the starting iodonium salt.
Attempts to prepare the 5-methyl derivative of cyclohex-
enyliodonium salt (2b) from 3-methylcyclohexanone resulted
in a 1:1 mixture of 2b and the 3-methyl isomer 7b. The reaction
of this mixture with acetate (0.1 M) in chloroform gave four
methyl derivatives of 1-acetoxycyclohexene, 3b, 4b, 8b, and
Iodobenzene and small amounts of 1-iodocyclohexenes, 5 and
6, respectively from 1 and 2, were also obtained. Yields of the
products were determined by GC and are summarized in Table
1. The regioisomeric structures and the isomer ratios were also
(35) For recent reviews of arynes, see: (a) Wenk, H. H.; Winkler, M.; Sander,
W. Angew. Chem., Int. Ed. 2003, 42, 502-528. (b) Pellissier, H.; Santelli,
M. Tetrahedron 2003, 59, 701-730. (c) Hart, H. In The Chemistry of
Functional Groups, Supplement C2; Patai, P., Ed.; John Wiley & Sons:
Chichester, 1994; pp 1017-1134. (d) Kseear, S. V. In ComprehensiVe
Organic Synthesis; Trost, B. M., Fleming, I. Eds.; Pergamon Press: Oxford,
1991; Vol. 4, pp 483-515. (e) Gilchrist, T. L. In The Chemistry of
Functional Groups, Supplement C; Patai, P., Rappoport, Z. Eds.; John Wiley
& Sons: Chichester, 1983; pp 383-419.
(36) For spectroscopic detections of arynes, see: (a) Orendt, A. M.; Facelli, J.
C.; Radziszewski, J. G.; Horton, W. J.; Grant, D. M.; Michl, J. J. Am.
Chem. Soc. 1996, 118, 846-852. (b) Warmuth, R. Angew. Chem., Int. Ed.
Engl. 1997, 36, 1347-1350. (c) Moriyama, M.; Ohana, T.; Yabe, A. J.
Am. Chem. Soc. 1997, 119, 10229-10230.
(33) Kitamura, T.; Yamane, M. J. Chem. Soc., Chem. Commun. 1995, 983-
984. Kitamura, T.; Yamane, M.; Inoue, K.; Todaka, M.; Fukatsu, N.; Meng,
Z.; Fujiwara, Y. J. Am. Chem. Soc. 1999, 121, 11674-11679. Kitamura,
T.; Todaka, M.; Fujiwara, Y. In Organic Syntheses, Vol. 78; Roush, W.
R., Ed.; Wiley: 2002; pp 104-112. Kitamura, T.; Fukatsu, N.; Fujiwara,
Y. J. Org. Chem. 1998, 63, 8579-8581. Kitamura, T.; Meng, Z.; Fujiwara,
Y. Tetrahedron Lett. 2000, 41, 6611-6614.
(34) Akiyama, T.; Iwasaki, Y.; Kawanishi, M. Chem. Lett. 1974, 229-230.
Cadogan, J. I. G.; Rowley, A. G.; Sharp, J. T.; Sledzinski, B.; Wilson, N.
H. J. Chem. Soc., Perkin Trans. 1 1975, 1072-1074. Beringer, F. M.;
Huang, S. J. J. Org. Chem. 1964, 29, 1637-1638.
9
7550 J. AM. CHEM. SOC. VOL. 126, NO. 24, 2004

Cyclohexynes from Iodonium Salts
A R T I C L E S
Table 2. Product Yields in the Reaction of 1d with Piperidinea
9b in 13%, 20%, 14%, and 19% yield, respectively (eq 8). The
yield (%)
run
solvent
[base] (M) 11d 12d PhI 5d 13d 14d 15d 16d
1
2
3
4
5
6
7
CHCl3
CHCl3
CHCl3
CHCl3
CHCl3
THF
0.01
0.05
0.10
0.50
1.0
5
9
65
2
7
9
6
3
2
2
2
27 18
10 10
13
5
3
<1
0
20 36 67
25 45 69
20 37 57 29
19 35 55 31
7
14
7
<1
<1
10
5
4
0.10
0.10
14 71 28 17
57 30
CH3CN
4
6
a At 60 °C for 2 h at the concentration of 1d of 0.003 M.
to that of iodocyclohexene 5d, and acetoxybenzene was not
detected in the reaction of 1 (or 2) with acetate.
ratio of 3b/4b (39/61) found here is close to that obtained from
1b (46/54). The products 3b and 4b must be derived from 2b
of the substrate mixture, while 8b and 9b must be formed from
7b.
The acetate reaction of 6-methylcyclohex-1-enyliodonium salt
10b was also examined and found to give 8b and 9b in a ratio
of 37:63 (eq 9). This ratio agrees with that obtained from the
3-methyl substrate 7b of the above mixture (8b/9b )
42/58).
Deuterium-Labeling Experiments. The reaction of 1b-d
with acetate in chloroform was carried out in the presence of
CH3OD (1 vol %) at 60 °C. The product ratios of 3/4 are similar
to those obtained in the absence of CH3OD, but deuterium was
incorporated in the acetate products at the vinylic position (eq
11). The extents of deuterium incorporation of the products
from 1b, 1c, and 1d were determined as 76, 75, and 64 atom
%, respectively, by 1H NMR analyses of the product mixture 3
and 4. Although the separate NMR determinations of deuterium
contents of 3 and 4 were difficult due to poor separation of the
olefinic proton peaks, the GC-MS analysis of these product
mixtures indicated that the deuterium is distributed in both 3
and 4.
The cyano-substituted cyclohexenyl iodonium salt was pre-
pared as a mixture of 4- and 5-substituted derivatives (1f and
2f) and the 61:39 mixture of 1f and 2f was employed for the
reaction with acetate ion (0.1 M) in chloroform to give 4- and
5-cyanocyclohex-1-enyl acetates (3f and 4f) in a ratio of 81:19
(eq 10).
2,6,6-Trideuterated iodonium salts, 1a-d3 and 1d-d3, were
prepared from the corresponding 2,2,6,6-tetradeuterated cyclo-
hexanone at the isotopic purity of about 90 atom % and were
employed for the reaction with acetate in chloroform without
additional methanol. The reaction of 1d-d3 with acetate ion (0.01
M) gave 3d and 4d in a ratio of 35:65, and the protium is
incorporated at the vinylic position of the acetate products (eq
1
12) to the extent of 84 atom % as deduced from the H NMR
spectra. The acetate reaction of 1a-d3 also gave the protium-
incorporated 3a, with the protium content at the vinylic position
being 84%.
The acetate reaction was also examined with a substrate of a
different leaving group, 4-tert-butylcyclohex-1-enyl(2,4,6-tri-
methylphenyl)iodonium tetrafluoroborate (1d′) (runs 18 and 19).
The results are similar to those obtained for 1d, but formation
of some reduction product, mesitylene, was confirmed. Forma-
tion of benzene was also observed during the reaction of 1d
with acetate in acetonitrile-d3 when the reaction was monitored
by 1H NMR. The amount of the reduction product corresponded
(37) For recent examples of the synthetic use of arynes, see: (a) Hamura, T.;
Hosoya, T.; Yamaguchi, H.; Kuriyama, Y.; Tanabe, M.; Miyamoto, M.;
Yasui, Y.; Matsumoto, T.; Suzuki, K. HelV. Chim. Acta 2002, 85, 3589-
3604. (b) Yoshida, H.; Sugiura, S.; Kunai, A. Org. Lett. 2002, 4, 2767-
2769. (c) Chatani, N.; Kamitani, A.; Oshita, M.; Fukumoto, Y.; Murai, S.
J. Am. Chem. Soc. 2001, 123, 12686-12687. (d) Rayabarapu, D. K.;
Majumdar, K. K.; Sambaiah, T.; Chemg, C.-H. J. Org. Chem. 2001, 66,
3646-3649. (e) Tripathy, S.; LeBlanc, R.; Durst, T. Org. Lett. 1999, 1,
1973-1975. (f) Pena, D.; Escudero, S.; Pe´rez, D,; Guitia´n, E.; Castedo, L.
Angew. Chem., Int. Ed. 1998, 37, 2659-2661.
Reaction with Piperidine and Methoxide. The reaction of
1d with a secondary amine, piperidine, was examined at varying
concentrations of the amine in chloroform (and THF and
acetonitrile) at 60 °C. A regioisomeric mixture of cyclohex-
anones 11d and 12d was obtained after aqueous workup (eq
13 and Table 2), along with various side products, including
iodide 5d, cyclohexene 13d, fluoride 14d, chloride 15d, and
(38) Regioselectivity of nucleophilic addition to benzynes has recently been
discussed using LUMO population: Hamura, T.; Ibusuki, Y.; Sato, K.;
Matsumoto, T.; Osamura, Y.; Suzuki, K. Org. Lett. 2003, 5, 3551-3554.
9
J. AM. CHEM. SOC. VOL. 126, NO. 24, 2004 7551

A R T I C L E S
Fujita et al.
Table 3. Reactions of 1d and 2d in Methanola
Table 4. Reactions of 1d and 2d with Bromide in the Presence of
Acetatea
yield (%)
PhI
concentration (M)
yield (%)
base (conc, M)
17d
18d
5d (6d)
run
Br-
AcO-
19d 20d 3d 4d
PhI
5d (6d)
(a) reaction of 1d
NaOAc (0.10)
(a) reaction of 1d
82
29b
35
17
0
12b
43
48
76
61
79
47
0
20
4
1
2
0.10
0
92
0
0
6
0
9
100
78
75
68
80
0
7
13
10
3
piperidine (0.10)
Bu4NCNc (0.03)
NaOMe (0.18)
0.094 0.007 23 27
0.070 0.029
0.070 0.030
3
5
7
6
9
15 34
10 26
7
4
4b
5c
(b) reaction of 2d
NaOAc (0.10)
Bu4NCNc (0.03)
NaOMe (0.18)
0.070 0.030 19 29
12
0
3
20
65
44
51
84
60
88
(0)
(0)
(3)
(b) reaction of 2d
6
7
0.10 78
0.094 0.005 12 48
0.096 0.005 13 54
0
0
0
1
1
4
1
0
4
2
16
6
100
98
100
61
(0)
(2)
(0)
(3)
(1)
8c
9
a At 60 °C for 1 h. b The yields of the hydrolysis products 11d (6%)
and 12d (9%) are included in those of 17d and 18d, respectively. c A
regioisomeric mixture of cyanide substituted cyclohexenes was also detected
but in <4% yield.
0.081 0.021
7
10
10c
0.079 0.018 18 40
99
a In chloroform at 60 °C. b In the presence of 0.1 vol % (0.025 M) of
methanol. c In the presence of 1 vol % (0.25 M) of methanol.
the Friedel-Crafts adduct 16d (mainly of the ortho isomer).
Table 5. Product Yields Determined by GC for the Reactions of
1b, 1d and 2d with Acetate in the Presence of R-Pyronea
substrate
21
3
4
5 (6)
PhI
1b
1d
2d
32
49
91
6
5
∼0
3
6
6
3
7
(3)
58
93
78
a Reactions were carried out at [substrate] ) ca. 0.004 M, [acetate] )
0.010 M, [pyrone] ) 0.10 M, and 60 °C.
The acetate reactions of 1 and 2 were carried out in the
presence of 2H-pyran-2-one (R-pyrone) as a trapping agent, and
the product mixtures were analyzed by GC (eq 16). Tetrahy-
dronaphthalene 21, the adduct with cyclohexyne, was obtained
together with some acetate products. The results are given in
Table 5.
Reaction of 1d with a stronger base was carried out in
methanol containing sodium methoxide (0.18 M) at 60 °C (eq
14). Both the ipso and cine methoxide substitution products 17d
and 18d were obtained in a ratio of 26/74 (Table 3). A similar
reaction of 2d also gave 17d and 18d in the ratio of 28/72.
Weaker bases were also employed for the reaction in methanol,
but more of the ipso product was formed (Table 3). The acetate
salt selectively gave the ipso methanolysis product, as was
observed during solvolysis of 1d in unbuffered methanol.10
Trapping experiments were also carried out on a preparative
scale using tetraphenylcyclopentadienone (TPC) under various
conditions with different bases and solvents. The adduct 22 (eq
17) was isolated, and the yields obtained are summarized in
Table 6. The yields are always good for acetate and fluoride in
aprotic solvents, but amine bases afforded lower yields (at most
52%) of 22 due to the extensive side reactions.
Trapping Experiments. The acetate reactions of 1d and 2d
in chloroform were carried out in the presence of bromide ion
as an additional nucleophile (Table 4). The reaction products
include two isomers of 1-bromocyclohexenes 19d and 20d, as
well as the acetate products 3d and 4d (eq 15). This is in contrast
A platinum complex, 23,39 of cyclohexyne was also obtained
(39) (a) Bennett, M. A.; Schwemlein, H. P. Angew. Chem., Int. Ed. Engl. 1989,
28, 1296-1320. (b) Bennett, M. A.; Robertson, G. B.; Whimp, P. O.;
Yoshida, T. J. Am. Chem. Soc. 1971, 93, 3797-3798. (c) Robertson, G.
B.; Whimp, P. O. J. Am. Chem. Soc. 1975, 97, 1051-1059. (d) Bennett,
M. A.; Yoshida, T. J. Am. Chem. Soc. 1978, 100, 1750-1759. (e) Bennett,
M. A.; Rokicki, A. Aust. J. Chem. 1985, 38, 1307-1318. (f) Lu, Z.; Abboud,
K. A.; Jones, W. M. Organometallics 1993, 12, 1471-1474.
to the exclusive formation of the ipso bromide in the absence
of acetate (runs 1 and 6), as summarized in Table 4. The addition
of methanol (1 vol %) as a proton source increased the fraction
of the bromide products over the acetate products (runs 3-5).
9
7552 J. AM. CHEM. SOC. VOL. 126, NO. 24, 2004

Cyclohexynes from Iodonium Salts
A R T I C L E S
Table 6. Isolated Yields of the Adduct 22 in the Reaction of 1a-e
and 2d in the Presence of Tetraphenylcyclopentadienone (TPC)a
Scheme 3
isolated yield of 22 (%)
base (conc, M)
solvent
1a
1b
1c
1e
1d
2d
AcONBu4 (0.010)
AcONBu4 (0.010)
AcONBu4 (0.010)
FNBu4(0.010)
FNBu4(0.010)
MeONa(0.18)
CHCl3
THF
MeCN
CHCl3
THF
MeOH
THF
CHCl3
CHCl3
THF
90
84
86
98
91
81
75
94
100
81
70
81
100
95
98
87
63
t-BuOK (0.020)b
piperidine (0.020)
piperidine (0.10)
piperidine (0.020)
Et3N (0.020)
89
36c
52c
20c
26c
29c
25c
CHCl3
CHCl3
CHCl3
Et3N (0.10)
Et3N (0.50)
a Reactions were carried out at [1 (or 2d)] ) 3-8 mM, [TPC] ) 10
mM, and 60 °C for 1-2 h. b The reaction was carried out at room
temperature. c Other products (16d, 14d, and 15d) were also detected or
isolated for the amine bases.
) 1.1-1.3). The reaction in methanol containing NaOMe (0.1
M) was too fast at 60 °C, and the rate constants for 1d
determined at 25 °C were kH ) 6.9 × 10-4 s-1 and kD ) 3.0 ×
10-4 s-1 (kH/kD ) 2.3).
Table 7. Pseudo-First-Order Rate Constants (kH and kD) for the
Reactions of Iodonium Salts 1 and 1-d3 with Tetrabutylammonium
Acetate or Piperidine in Chloroform at 60 °C
Discussion
[base] (M)
103kH (s-1
)
103kD (s-1
)
kH/kD
Mechanism for the ipso and cine Substitution by Acetate.
Solvolysis of cyclohex-1-enyl(phenyl)iodonium tetrafluoroborate
(1) in unbuffered alcohols and acetic acid gives solely the ipso
substitution products via an SN1-type mechanism with the
cyclohexenyl cation as an intermediate (eq 3).10 The reaction
with bromide ion in chloroform also afforded the ipso bro-
mocyclohexene (Table 4, runs 1 and 6) probably via a ligand
coupling mechanism.8,9 The ipso bromo product was reported
to form via ligand coupling in the presence of copper(I) salt,
though the reaction was occurring as the copper ligands.40
In contrast, the reaction of 4-substituted 1 with tetrabutyl-
ammonium acetate in chloroform gave the cine substitution
product 4 in addition to the ipso acetate 3 (eq 6), with the ratio
of 3/4 depending on the 4-substituent (Table 1). The same
mixture of regioisomeric substitution products was also obtained
from the 5-substituted isomeric substrate 2 (eq 7). The largest
selectivity in favor of 4 was observed with the tert-butyl
substrates 1d and 2d; 3d/4d ) 25/75-20/80 (runs 8-13). The
formation of ipso and cine substitution products and the regio-
convergent product ratio obtained from the isomeric substrates
suggest the intervention of a common intermediate and are best
interpreted by the elimination-addition (EA) mechanism with
a cyclohexyne intermediate as illustrated in Scheme 3. A similar
EA mechanism for the aromatic nucleophilic substitution via a
benzyne intermediate is well-known, though under strongly basic
conditions. In accord with the EA mechanism, the deuterium
incorporation at the vinylic carbon of the acetate products 3
and 4 was observed in the reaction of 1 with acetate in the
presence of methanol-d in chloroform (eq 11). Similar results
were also obtained in the acetate reaction of the deuterated
substrate 1-d3 under normal conditions; the protium was
incorporated at the vinylic carbon of the products, 3 and 4 (eq
12).
(a) reaction of 1a with acetatea
0.001
0.010
0.025
0.050
0.10
10.4
10.5
10.9
9.59
9.65
4.05
3.48
3.76
3.56
3.37
2.6
3.0
2.9
2.7
2.9
(b) reaction of 1d with acetatea
0.001
0.010
0.050
0.10
4.62
4.44
4.73
4.97
2.03
1.81
1.98
1.86
2.3
2.5
2.4
2.7
(c) reaction of 1d with piperidine
0.010
0.050
0.10
0.57
0.81
0.98
1.64
0.50
0.64
0.90
1.52
1.14
1.27
1.09
1.08
0.50
a The ionic strength of the solution was maintained at 0.10 M with
tetrabutylammonium perchlorate.
as colorless crystals from the base (t-BuOK) reaction of 1a in
the presence of Pt(PPh)3 in THF at 0 °C (eq 18).
Kinetics of the Reaction. The acetate reactions of 1a and
1d and their deuterated analogues 1-d3 in chloroform were
followed at 60 °C and the ionic strength of 0.10 (Bu4NClO4)
by monitoring the decrease in absorbance at 280 nm due to the
iodonium salt. The reaction followed reasonably well the
pseudo-first-order kinetics, and the observed first-order rate
constants are given in Table 7 as averages of at least three runs
within (5%. The unsubstituted idodonium salt 1a is about 2
times more reactive than the tert-butyl derivative 1d, and the
observed rate constants are practically independent of the
concentration of acetate. The rate constant (kD) for the reaction
of 1-d3 is smaller than that of 1 (kH) by 2.3-3.0 times.
Trapping experiments confirmed the formation of the cyclo-
hexyne intermediate during the acetate reaction (as well as other
basic reactions). The reactions of 1 or 2 in the presence of
The reaction of 1d with piperidine (0.01-0.1 M) in chloro-
form was slower, but 1d and 1d-d3 showed similar rates (kH/kD
(40) Ochiai, M.; Sumi, K.; Takaoka, Y.; Kunishima, M.; Nagao, Y.; Shiro, M.;
Fujita, E. Tetrahedron 1988, 44, 4095-4112.
9
J. AM. CHEM. SOC. VOL. 126, NO. 24, 2004 7553

A R T I C L E S
Fujita et al.
trapping agents, R-pyrone and tetraphenylcyclopentadienone
(TPC), gave effectively the adducts 21 and 22 of cyclohexyne
(eqs 16 and 17), respectively. Trapping of cyclohexyne was also
observed with bromide. The reaction of 1d in the presence of
both acetate and bromide in chloroform gave the cine bromide
substitution product 20d in addition to the ipso bromide 19d as
well as the acetate products (eq 15, Table 4). This is in contrast
to the selective ipso substitution of the bromide reaction in the
absence of acetate ion. The preferential formation of 20d over
19d was observed from both 1d and 2d in the presence of
varying concentrations of acetate ion. This regioselectivity is
in accord with that of the acetate substitution, 3d < 4d.41
Intermediacy of the cyclohexyne is now established, but the
mechanism of formation is another problem. Rates for the
reactions of 1a and 1d and their deuterated counterparts 1-d3
with acetate were determined in chloroform at 60 °C (Table 7).
The reactions of the deuterated substrates are significantly slower
than the corresponding protium substrates, kH/kD ) 2.3-3.0.
The observed primary kinetic isotope effects are consistent with
a mechanism involving the rate-determining deprotonation at
C2, i.e., the E2 mechanism for the formation of the cyclohexyne
intermediate.
aqueous treatment of the reaction mixture obtained in chloroform
(eq 13). These products must come from hydrolysis of the
primary products of isomeric enamines, hydrolysis of which is
known to be very fast.42 The ipso and cine substitutions by the
amine must occur via the EA mechanism with the cyclohexyne
intermediate. However, the formation of various byproducts is
apparent in this reaction (eq 13, Table 2). The byproducts, 14d,
15d, and 16d, are typical of those derivable from a cyclohexenyl
cation intermediate and were found in the thermal decomposition
of the iodonium salts43 and during the solvolysis.10 The Friedel-
Crafts recombination product, cyclohexenyliodobenzene 16d,
is considered to be derived from the contact cyclohexenyl
cation-iodobenzene pair.10 These results implicate that the EA
mechanism of the piperidine reaction proceeds via the E1
elimination.
The kinetic isotope effects observed are in fact very small in
the reaction with piperidine (Table 7). The fraction of the EA
route is found to be over 50% when [piperidine] ) 0.05-0.5
M (Table 2), and the expected overall kinetic isotope effect kH/
kD would be greater than 1.5 provided that the kH/kD for the E2
reaction of 3 is operative (Table 7). The observed values (1.08-
1.27) are differentiated from this value, although the present
kinetic measurements are not perfectly precise due to the side
reactions and the isotopic purity of the substrate 1-d3 (90% D).
It can be concluded that the EA pathway of the piperidine
reaction in chloroform does not show appreciable primary
kinetic isotope effect, and the elimination proceeds mainly via
the E1 mechanism as implied from the product analysis.
The iodocyclohexene product, 5 from 1 or 6 from 2, formed
from the cleavage of the aromatic carbon-iodine bond is also
apparent at higher concentrations of piperidine (Table 2, runs
4 and 5). This product was found also in the acetate reaction,
especially in more polar aprotic solvents (Table 1, runs 10, 11,
and 19; Table 2, runs 6 and 7), and with the phenyl-substituted
iodonium salts, 1e and 2e (Table 1, runs 14-17). However,
acetoxybenzene was not detected as its counterpart product. The
product observed was benzene (or mesitylene from 1d′), a
reduction product. The possibility of reduction product formation
via benzyne intermediate can be eliminated: a similar reduction
was observed in both cases of 1d and 1d′, although 1d′ cannot
give the benzyne type intermediate. The pair of iodocyclohexene
and benzene is evidently derived from homolysis of the C-I
bond, and the reaction may involve electron transfer from the
nucleophile to the iodonium ion.44,45
The observed rate was unexpectedly independent of the
concentration of acetate base (Table 7). Yields of the products
were also not dependent on the concentration of acetate. These
observations may be explained by a mechanism involving the
initial equilibrium formation of the adduct (λ3-iodane) followed
by an intramolecular elimination within the adduct (eq 19). The
formation constant of the adduct must be large in chloroform.
A similar mechanism of â-elimination was proposed for the
reaction of alk-1-enyl(phenyl)iodonium salts with chloride and
bromide (Scheme 1)5,6 and also that with carboxylates.7 In the
latter case, a stronger base-like acetate preferentially induces
R-elimination, but a less basic carboxylate like trifluoroacetate
can induce the intramolecular-type â-elimination.7 In the present
case, the acetate can participate in the intramolecular â-elimina-
tion within the λ3-iodane due to the lack of R-hydrogen, but
the competing bimolecular elimination cannot be excluded. Such
an intramolecular elimination may contribute to efficient forma-
tion of cyclohexyne during the reaction with acetate and fluoride
ions in chloroform. In any case, the deprotonation is involved
in the rate-determining step, and the mechanism can be taken
as an E2 elimination.
Reaction in Methanol. The reaction of 1 in methanol usually
gives simply the ipso 1-methoxycyclohexene 17,10 but in the
presence of a strong base, sodium methoxide, the products
include those of both ipso and cine substitution (eq 14 and Table
3). The isomeric iodonium salts 1d and 2d gave a similar
mixture of the products 17d and 18d. Again, the intermediary
formation of cyclohexyne is strongly suggested, and it was
Reaction with Piperidine. The reaction of 1d with piperidine
provides the two isomeric cyclohexanones 11d and 12d on
(41) Effects of added methanol on the relative amounts of bromide and acetate
products (Table 4, runs 4, 5, 8, and 10) may be worth mentioning here,
since the acetate reactions summarized in Table 1 are not affected by added
methanol. A small amount of methanol increases the yields of bromide
products 19d and 20d, while reducing the acetate products 3d and 4d,
keeping the relative preference for the regioisomer (20d > 19d and 4d >
3d). The results may be rationalized from the effect of methanol on the
relative nucleophilicity of acetate and bromide toward the cyclohexyne. A
more effective hydrogen-bonding solvation of acetate reduces more
effectively its reactivity than that of bromide. Alternatively, the increased
yields of the bromide products could be ascribed to the reversibility of the
bromide addition to the cyclohexyne: The interception of the â-bromovinyl
anion with methanol would increase the contribution from the bromide
addition without affecting the irreversible acetate addition.
(42) Maas, W.; Janssen, M. J.; Stamhuis, E. J.; Wynberg, H. J. Org. Chem.
1967, 32, 1111-1115.
(43) Okuyama, T.; Fujita, M.; Gronheid, R.; Lodder, G. Tetrahedron Lett. 2000,
41, 5125-5129.
(44) (a) Tanner, D. D.; Reed, D. W.; Setiloane, B. P. J. Am. Chem. Soc. 1982,
104, 3917-3923. (b) Grushin, V. V.; Demkina, I. I.; Tolstaya, T. P. J.
Chem. Soc., Perkin Trans. 2 1992, 505-511. (c) Ochiai, M.; Shu, T.;
Nagaoka, T.; Kitagawa, Y. J. Org. Chem. 1997, 62, 2130-2138. (d) Chen,
D.-W.; Ochiai, M. J. Org. Chem. 1999, 64, 6804-6814.
(45) One of the referees pointed out the possibility of formation of cyclohexyne
via the E2 reaction of 5. This possibility is excluded since the cyclohexyne
products always accompany the formation of iodobenzene and 5 is much
less reactive toward elimination than 1.
9
7554 J. AM. CHEM. SOC. VOL. 126, NO. 24, 2004

Cyclohexynes from Iodonium Salts
A R T I C L E S
Scheme 4
actually trapped by TPC under these conditions (Table 6). The
observed kinetic isotope effect of kH/kD ) 2.3 for the methoxide
reaction suggests the E2 mechanism of the elimination.
In contrast, acetate cannot effect the cine-substitution at all
in methanol. Only the normal SN1-type methanolysis to give
the ipso methoxy product occurs. Piperidine and cyanide gave
the cine methoxy product as well as the ipso methoxide, but
the ratio is heavily biased to the side of the ipso product. The
ratios of 17d/18d obtained from 1d and 2d are significantly
different, that is, a poor convergence of regioselctivity was
observed. The cine product must come from the cyclohexyne
intermediate (via the EA mechanism), but a considerable part
of the reaction occurs via a competing ipso route, probably the
SN1 solvolysis through the cyclohexenyl cation intermediate.
The mechanism of the elimination step in the EA pathway can
be E1 or E2, but due to a small fraction of the elimination route,
the kinetic isotope effect cannot clearly differentiate these
mechanisms.
Scheme 5
The reaction of 1d with alkoxide in trifluoroethanol and
hexafluoro-2-propanol (HFIP) was examined to see if the
elimination reaction can lead to formation of the cine-substitu-
tion product, but no sign of the cine substitution was found.
The alkoxides in these acidic alcohols are not basic enough to
induce the elimination to afford highly strained cyclohexyne.
This is in contrast to the efficient cycloheptyne formation via
the E1-type mechanism with alkoxide in HFIP,29 where the
cycloheptenyl cation is formed from a cyclohexylidenemethyl
derivative via participation and is deprotonated.
stronger base than piperidine in chloroform. This must be a main
reason the former can induce the E2 reaction of the cyclohex-
enyliodonium salt while the latter can only work on the
cyclohexenyl cation intermediate, which was allowed to form
in the poorly basic medium, to lead to the E1 reaction. The
intramolecularity of the acetate reaction may be another reason
for the ease of the E2-type elimination.
Trapping Cyclohexyne by Platinum. While cyclohexyne
is labile and exists only in matrixes at 77 K,24a group 10
transition metals are known to form stable complexes with
distorted alkynes.39 In fact, when the iodonium salt 1a was
treated with Pt(PPh3)3 in the presence of t-BuOK in THF at 0
°C, the cyclohexyne complex 23 of platinum(0) was obtained
in 69% yield as colorless crystals (eq 18). This can be taken as
another experiment for trapping cyclohexyne. However, for the
synthesis of the related platinum(0) cyclopentyne complex from
1,2-dibromocyclopentene, the reduction by Na/Hg to give
cyclopentyne was reported to occur after coordination of the
olefin to platinum, as shown in Scheme 4.39a It was also
suggested that free cyclohexyne is not involved in the analogous
synthesis of 23 from 1,2-dibromocyclohexene.39a
Trapping of the Cyclohexyne. Efficiency of formation of
cyclohexyne was compared by trapping experiments with TPC
under various conditions with changing bases and solvents. As
summarized in Table 6, acetate and fluoride are the most
effective for formation of cyclohexyne in chloroform and yields
of isolated adducts 22 are more than 80% for various substrates
1 (and 2d) with different alkyl and phenyl groups, practically
quantitative in many cases. The basic reaction seems to be less
effective in more polar solvents such as THF and acetonitrile,
but it still provides sufficiently good access to cyclohexyne
(>75% yield) under mild conditions. Strong bases such as
methoxide in methanol and tert-butoxide in THF afforded
satisfactory yields of the adducts of cyclohexynes at lower
temperature.
To gain mechanistic insights into the formation of 23 from
1a, the reaction of 1a with Pt(PPh3)3 was examined in the
absence of the base, t-BuOK. From this reaction, the olefin
complex of platinum shown by 24 was not isolated, which is
In contrast, amines are not good bases for this reaction. Low
yields of the cyclohexyne adducts using amine bases are due
to the formation of various byproducts as discussed for the
piperidine reaction. The contrasting tendency in efficiency of
the cyclohexyne formation with ionic and neutral bases in
chloroform can be rationalized by different mechanisms of
elimination: E2 for the former and E1 for the latter bases. This
contrasting conclusion on the mechanism for the elimination
step induced by acetate and piperidine is rather unexpected on
the basis of general knowledge of basicity: pKa of acetic acid
) 4.76 and pKa of dialkylammonium ion ) ca. 10 but in
aqueous solution. In aprotic solvents, the basicity of an ionic
base increases enormously, while that of a neutral base remains
essentially unchanged, e.g., pKa’s in DMSO of acetic acid and
triethylammonium ion are 12.6 and 9.0, respectively.46 These
effects would be still greater in chloroform, and acetate is a
analogous to the preformed cyclopentene complex in the reaction
of Scheme 4. Instead, the cationic cyclohexenyl platinum(II)
complex 25 was produced in 85% yield as colorless crystals
(Scheme 5). Complex 25 is stable toward oxygen and moisture
in solid, and the combustion analysis and NMR spectra are
consistent with the formulation. The molecular structure of 25
was determined by X-ray crystallographic analysis, in which
the 1-cyclohexenyl ligand is situated approximately perpen-
(46) Maskill, H. The Physical Basis of Organic Chemistry; Oxford University
Press: Oxford, 1985; p 185.
9
J. AM. CHEM. SOC. VOL. 126, NO. 24, 2004 7555

A R T I C L E S
Fujita et al.
Table 8. Regioselectivity of Nucleophilic Addition to Substituted
Scheme 6
Cyclohexynes
product ratio, 3/4a
substituent
σIb
reaction of 1
reaction of 2
H
4-Me
4-Et
4-t-Bu
4-Ph
4-CN
3-Me
0.00
-0.04
-0.05
-0.07
0.10
0.56
-0.04
50/50
46/54
39/61
28/72
57/43
50/50
39/61
21/79
49/51
81/19c
42/58d
37/63e
a The product ratio obtained from the acetate reaction at [acetate] ) 0.10
M (Table 1). b Inductive substituent constants.48 c The product ratio of 3f/
4f obtained from a 61:39 mixture of 1f and 2f. d The product ratio of 8b/
9b obtained from 7b. e The product ratio of 8b/9b obtained from 10d.
the 1,2-diene was found during the acetate reaction of the
iodonium salt in chloroform: 1-acetoxy-3-alkylcyclohexene 8
was not formed from 2. The concerted E2 reaction requires the
coplanarity of the orbitals involved in the bondings of both the
electrofuge and the nucleofuge, which strongly favors formation
of cycloalkyne over 1,2-diene from the cycloalkenyl substrate
due to the restricted rotation of the skeletal C-C bonds. In
contrast, the ElcB mechanism involving a vinylic anion
intermediate may be at work in the elimination of 1-halocy-
clohexene due to the lower nucleofugality of the halide and use
of the strong base. In this reaction, the deprotonation would
provide the more stable carbanion, allylic anion, leading to the
1,2-diene. The E2 reaction of the iodonium salt should be more
E1-like rather than E1cB-like due to the nucleofugality.
Regioselectivity of Nucleophilic Addition to Cyclohexyne.
The regioselectivity of the acetate reaction in the formation of
ipso and cine substitution products should primarily reflect the
regioselectivity of nucleophilic addition to the intermediate
cyclohexyne. The conformity of the product ratios obtained from
the regioisomeric iodonium substrates 1 and 2, which should
provide a common cyclohexyne, is in fact satisfactory. This
convergency of regioisomeric products becomes loose due to
the accompanying reactions at the ipso position, the SN1-type
reaction and ligand coupling. Although contributions from the
ipso reactions seem to be greater in the reactions of piperidine,
methoxide, and bromide than in the acetate reaction, the
similarity of observed product ratios with different nucleophiles
substantiates that the intrinsic regioselectivity of cyclohexyne
toward nucleophiles can be assessed from the acetate reaction.
The regioselectivities for various substituted cyclohexynes are
mainly derived from the data given in Table 1 and are
summarized in Table 8. The magnitudes of 3/4 range from 0.3
to 4 for 4-alkyl, phenyl, and cyano substitutions.
dicular to the coordination plane consisting of platinum and
phosphorus atoms (Figure S3).
If t-BuOK deprotonates 25 at the vinylic position, this
complex could be regarded as a potential intermediate leading
to the cyclohexyne complex 23. However, it turns out not to be
the case. Treatment of 25 with t-BuOK in THF at 0 °C resulted
in a complex mixture containing a small amount of Pt(PPh3)3.
In the mixture, the cyclohexyne complex 23 was not discernible
as a product. Therefore, it is likely that Pt(PPh3)3 trapped
transient cyclohexyne generated from the deprotonation of 1a
by t-BuOK. It is interesting to note that iodobenzene formed as
a side product does not hamper the synthesis of the cyclohexyne
complex 23. The oxidative addition reaction of Pt(PPh3)3 with
iodobenzene would be fast, probably more facile than coordina-
tion of the olefin portion of 1a. This observation also indicates
that the highly reactive free-cyclohexyne indeed reacted with
Pt(PPh3)3.
Some Notes on the Mechanism. The elimination step of the
EA mechanism for the reaction of the cyclohexenyliodonium
salts is always subject to competition with other reactions
(Scheme 6). The conceivable first steps include the E2 (with a
base), ligand coupling (with a nucleophile), and spontaneous
heterolysis (to lead to E1 and SN1) as well as the cleavage of
the aromatic C-I bond.47 Strong bases are good for the E2
pathway (as observed with acetate and fluoride in chloroform
and with alkoxides). If the spontaneous heterolysis to give the
cyclohexenyl cation is allowed in poorly basic media, the
selection of the E1 and SN1 routes occurs on the cationic
intermediate depending on the basicity/nucleophilicity of the
reagents working in the reaction medium.
The regioselectivity of nucleophilic addition to substituted
benzynes has been extensively studied, and the inductive effect
of the substituent is considered to be responsible for the
selectivity.13a,35 The selectivity, log(4/3), is plotted against the
inductive substituent constant σI48 in Figure 1. Points for 4-CN,
4-Ph, H, and 4-Me seem to conform to the electronic effects,
but those for 4-Et and 4-t-Bu greatly deviate upward from the
correlation. The alkyl groups have similar electronic effects σI,
and the position of substitution is separated by two bonds from
the reaction site. Moreover, a bulky tert-butyl group prefers a
nucleophilic attack at the closer carbon of the triple bond to
The possibility of formation of cyclohexa-1,2-diene cannot
be neglected in the elimination step of the reaction of the
iodonium salt. Formation of the 1,2-diene has usually been
observed during the elimination of 1-halocyclohexene with
strong bases,30,31 and moreover cyclohexa-1,2-diene is more
stable than cyclohexyne. Nonetheless, no sign of formation of
(47) (a) The iodonium substrate is in a fast equilibrium with the λ3-iodane in
the presence of nucleophile. Although the reactivity of the vinylic group is
greater in the iodonium form than the iodane form, the latter is often the
main reactive species, and thus the apparent substrate reactivity varies with
the concentration and the identity of the nucleophile. The Michael addition
of cyanide to 1 has recently been discovered as an additional reaction
pathway to lead to the carbene intermediate, which undergoes 1,2-hydrogen
shift to give allylic cyanide.47b (b) Fujita, M.; Kim, W. H.; Okuyama, T.
Chem. Lett. 2003, 32, 382-383.
(48) Hine, J. Structural Effects on Equilibria in Organic Chemistry; John Wiley
& Sons: New York, 1975; Chapter 3.
9
7556 J. AM. CHEM. SOC. VOL. 126, NO. 24, 2004

Cyclohexynes from Iodonium Salts
A R T I C L E S
Figure 2. Plot of logarithms of ratios of the acetate products 4/3 obtained
from 1 (filled circle) and 2 (open circle) against the differences in LUMO
populations, f2 - f1. In the case of 3-methylcyclohexyne (3-Me), the starting
substrate were 7b and 10b in place of 1b and 2b, respectively, and the
product ratio of 9b/8b is employed for 4b/3b.
Figure 1. Plot of logarithms of ratios of the acetate products 4/3 obtained
from 1 (filled circle) and 2 (open circle) against the inductive parameter
σI.
Table 9. LUMO Populationsa and Bond Anglesb at the Acetylenic
Carbons of Substituted Cyclohexynes Optimized by
B3LYP/6-31G(d)
substituent
f1
f2
f2 − f1
θ1 (deg) θ2 (deg) θ2 − θ1 (deg)
H
4-Me
4-Et
4-t-Bu
4-Ph
4-CN
3-Me
0.40037 0.40037
0.39413 0.40146
0.39426 0.40301
0.38741 0.40724
0.00000 131.7 131.7
0.00733 130.9 132.0
0.00875 130.6 132.2
0.01983 129.5 133.5
0.0
1.1
1.6
4.0
0.9
-2.5
0.7
4.0
0.37927 0.37332 -0.00595 131.2 132.1
0.40728 0.37523 -0.03205 133.4 130.9
0.39543 0.39884
0.00341 131.6 132.3
0.01850 129.5 133.5
H(4-t-Bu)c 0.39100 0.40950
a f1 and f2 are the LUMO populations at C1 and C2 of the substituted
cyclohexyne, respectively. b θ1 and θ2 are the bond angles at C1 and C2 of
the substituted cyclohexyne, respectively. c A model calculation on the
optimized cyclohexyne ring for 4-tert-butylcyclohexyne, but the tert-butyl
group is replaced with hydrogen.
give 4d. What controls the regioselectivity of nucleophilic
addition to the substituted cyclohexynes?
Figure 3. Plot of the LUMO populations (f2 - f1) vs the bond angles (θ2
- θ1) for 4-substituted (filled circle) and 3-substituted cyclohexynes (open
circle). The numbering refers to Tables S3 and S5.
Theoretical calculations were carried out on various substi-
tuted cyclohexynes at the level of B3LYP/6-31G(d). The
calculations on the parent cyclohexyne show that the bond
angles of acetylenic carbons are strongly deformed to about 130°
in accord with previous calculations,19h and the LUMO is
considerably lowered and mainly located at the acetylenic
carbons developing in the plane of cyclohexyne ring. The
consequence is a facile reactivity of cyclohexyne toward a
nucleophile. The electronic populations of LUMO at the
acetylenic carbons calculated are summarized in Table 9. All
the 4-alkylcyclohexynes have a higher LUMO population at C2
(f2) than at C1 (f 1), and the tert-butyl derivative has the largest
difference (f2 - f1) among those calculated. This conforms to
the product distribution of nucleophilic addition to the cyclo-
hexyne in favor of the attack at C2 (formation of 4). The
differences between the LUMO populations at C2 and C1 (f2
- f1) are correlated with the logarithms of the product ratios
4/3 for 4-substituted cyclohexynes and the equivalent values
for the 3-methyl derivative in Figure 2.49 It shows a good
correlation between the regioselectivity and the LUMO popula-
tions. This correlation can be extended to the 4-phenyl and
4-cyano derivatives.
the LUMO? The bond angles at the acetylenic carbons of the
optimized structure are given in the last columns of Table 9.
The difference in LUMO populations (f2 - f1) can be noticed
to conform to the difference between the bond angles at C2
and C1 (θ2 - θ1), as plotted in Figure 3. The points in Figure
3 include not only those for the optimized structures but also
those for local minimum structures given in Table S5. The
difference in LUMO populations does increase as that in bond
angles increases. The acetylenic carbon of the larger bond angle
has a higher LUMO population than that of the smaller bond
angle. That is, the LUMO population is strongly correlated with
the ring structure represented by the bond angles at the
acetylenic carbons. This can be further confirmed by a model
calculation on the deformed cyclohexyne ring, which has the
same structure of the ring of the optimized 4-tert-butylcyclo-
(49) Theoretical calculations also indicate the presence of other local minimum
structure(s) of a substituted cyclohexyne as well as the optimized structure
where the substituent is equatorial. The LUMO populations can be modified
by the contribution from other conformations in equilibrium at the reaction
temperature (60 °C). The modified values are given in Table S2 of
Supporting Information, but they are not much different from the unmodi-
fied values for the optimized structures. Thus, the simple, original LUMO
population is satisfactory in representing the regioselectivity of nucleophilic
addition to substituted cyclohexynes.
If the inductive electronic effect of alkyl groups is not solely
responsible for the change in LUMO populations, what controls
9
J. AM. CHEM. SOC. VOL. 126, NO. 24, 2004 7557

A R T I C L E S
Fujita et al.
hexyne but the tert-butyl group is replaced with hydrogen (the
last entry of Table 9).
negative atomic charge, but the correlation between the atomic
charge and the product distribution is relatively poor (Figure
S2). The LUMO population best represents the regioselectivity
of nucleophilic addition to substituted cyclohexynes (Figure 2).
In conclusion, the elimination-addition mechanism for
nucleophilic substitution at the vinylic carbon is established,
and the reaction can be employed as a mild method for
generation of cyclohexynes. The regioselectivity of nucleophilic
addition to the substituted cyclohexynes is explored for the first
time and is rationalized from the angle strain and the LUMO
population at the acetylenic carbons.
When an electron-withdrawing group such as CN or CF3 is
introduced at C4 of the cyclohexyne ring, both values of θ2 -
θ1 and f2 - f1 become negative. This inverse effect of the
electron-withdrawing group compared with alkyl substituents
must be attributed to the electronic effect. In contrast, the
relatively large values of θ2 - θ1 and f2 - f1 for tert-butyl-
substituted cyclohexyne compared with those for other alkyl
derivatives cannot be explained only by their electronic effects.
Steric effects of the alkyl group may be responsible for the
deformation of the cyclohexyne ring, and the resulting bond
angles control the LUMO population and in turn the regiose-
lectivity of nucleophilic addition.
The more deformed of the two acetylenic carbons has the
lower LUMO population. When the acetylenic carbon is
deformed from the linear structure, the π orbital in the plane of
the cyclohexyne ring begins to have some s character. The
increase in s character of the π orbital results in the increase in
electron density in occupied MO and conversely the decrease
in the LUMO population. The atomic charge at the acetylenic
carbon actually changes with the bond angles (Figure S1). The
more deformed of the two acetylenic carbons has the more
Acknowledgment. This work was partially supported by a
Grant-in-Aid for Scientific Research on Priority Areas (Reaction
Control of Dynamic Complexes, No. 15036260) from the
Ministry of Education, Culture, Sports, Science and Technology,
Japan.
Supporting Information Available: Experimental and com-
putational details and additional data (CIF and PDF). This
material is available free of charge via the Internet at
http://pubs.acs.org.
JA0496672
9
7558 J. AM. CHEM. SOC. VOL. 126, NO. 24, 2004