6-Endo-trig and 5-exo-trig selective aryl radical cyclisations of N-(o-bromobetenzyl) enamides

Article abstract of DOI:10.1039/b003790g

Bu₃SnH-mediated aryl radical cyclisation of enamide 9 proceeded in a 6-endo-trig manner to give exclusively tetrahydroisoquinoline derivative 12, whereas enamide 10b having a (Z)-phenylthio group at the terminus of the N-vinylic bond gave exclu

Full text of DOI:10.1039/b003790g

6-Endo-trig and 5-exo-trig selective aryl radical cyclisations of
N-(o-bromobenzyl) enamides
Hiroyuki Ishibashi,* Issei Kato, Yoshifumi Takeda, Momoyo Kogure and Osamu Tamura
Faculty of Pharmaceutical Sciences, Kanazawa University, Takara-machi, Kanazawa 920-0934, Japan.
E-mail: isibasi@dbs.p.kanazawa-u.ac.jp
Received (in Cambridge, UK) 11th May 2000, Accepted 28th June 2000
Bu3SnH-mediated aryl radical cyclisation of enamide 9
proceeded in a 6-endo-trig manner to give exclusively
tetrahydroisoquinoline derivative 12, whereas enamide 10b
having a (Z)-phenylthio group at the terminus of the N-
vinylic bond gave exclusively the 5-exo-trig cyclisation
product 16.
When a mixture of Bu3SnH (2.2 eq.) and azobis(cyclohex-
ancarbonitrile) (ACN) (0.4 eq.) in toluene was added slowly to
a boiling solution of 9 in toluene over a period of 3.5 h, the
6-endo cyclisation product 125 was obtained in 68% yield,
along with the simple reduction product 14 (16% yield)
(Scheme 2). Similarly, the N-formyl congener 11 gave 13 and
15 in 43 and 19% yields, respectively. On the other hand,
treatment of the sulfur substituted (E)-isomer 10a with
Bu3SnH–ACN afforded the 5-exo cyclisation product 16 in 40%
yield, along with dihydroisoquinoline derivative 17 in 41%
yield. The corresponding (Z)-isomer 10b gave 16 as a sole
product in 75% yield.
The exclusive formation of the 6-endo cyclisation product 12
from 9 is of great interest in view of previous work on radical
cyclisations of the related compounds 1, 3 and 5, which gave the
5-exo cyclisation products 2, 4 and 6, respectively. Formation of
4 (from 3) and 12 (from 9) can be rationalized by an attack of
Bu3SnH on the primary radical 18 and on the secondary radical
21, respectively. Since enamide 3 gave no 6-endo cyclisation
product via the secondary radical 19, formation of 12 from 9
could not be explained by assuming that the nitrogen-
substituted secondary radical 21 might be more stable than the
primary radical 20.
Aryl radical cyclisations are now widely used in organic
synthesis for the construction of fused aromatic compounds. A
5-exo-trig cyclisation is generally preferred over a 6-endo-trig
ring closure in those systems having an alkenic bond at the
5-position relative to the aryl radical centre. For example, aryl
bromides 1 (X = CH2, O or NCOR), upon treatment with
Bu3SnH in the presence of AIBN, gave almost exclusively the
5-exo cyclisation products 2.1 This was also the case for the
One possible explanation for the formation of 12 from 9 may
involve a consecutive 5-exo cyclisation and neophyl-like†
rearrangement of the resulting radical 20. The possibility,
cyclisations of enamide 3 and acryloylanilide 5, which gave
only the five-membered lactams 42 and 6,3,4 respectively.
Herein we wish to report that N-(o-bromobenzyl) enamide 9
undergoes aryl radical cyclisation in a 6-endo-trig manner to
give exclusively tetrahydroisoquinoline derivative 12, and that
the mode of cyclisation can be shifted to a 5-exo-trig manner by
introducing a (Z)-phenylthio group at the terminus of the N-
vinylic bond.
Scheme 2 Reagents and conditions: i, Bu3SnH, ACN, toluene, reflux; ii,
Bu3SnH, Et3B, toluene, rt.
The requisite radical precursors 9, 10a and 10b were prepared
as shown in Scheme 1.
Scheme 1 Reagents and conditions: i, EtCOCl, Et3N, CH2Cl2, rt, 90%; ii,
MCPBA, CH2Cl2, 0 °C, 93%; iii, xylene, NaHCO3, reflux, 81%; iv,
(CF2CO)2O, CH2Cl2, rt; v, toluene, reflux, 46% for 10a, 13% for 10b (based
on 8).
DOI: 10.1039/b003790g
Chem. Commun., 2000, 1527–1528
This journal is © The Royal Society of Chemistry 2000
1527

For the sulfur substituted (E)-isomer 10a, the two conformers
syn-10a and anti-10a can be considered (Fig. 1). As in the case
of anti-9, there is a severe steric repulsion between the EtCO
and CNC groups in anti-10a, and the cyclisation might therefore
proceed via the conformer syn-10a in a 6-endo manner to give
17 through an elimination of a benzenethiyl radical from the
resulting intermediate radical A (Fig. 2). The major product of
the reaction of 10a, however, is the 5-exo cyclisation product
16. This is probably because the sulfur atom of the intermediate
radical B can strongly stabilise the neighboring radical
centre.9
On the other hand, both conformers syn-10b and anti-10b for
the (Z)-isomer 10b have a more severe steric constraint between
the o-BrC6H4CH2 and SPh groups for the former and between
the COEt and SPh groups for the latter (Fig. 1), and hence the
CaNCb bond and amide nitrogen might not be conjugated in
enamide 10b.10 If the CaNCb bond is almost perpendicular to the
amide bond, as depicted in C (Fig. 2), the resulting radical can
attack the more proximate Ca-position to give exclusively the
observed 5-exo cyclisation product 16.11
Fig. 1 Ar = o-BrC6H4.
Fig. 2
Notes and references
† The IUPAC name for neophyl is 2-methyl-2-phenylpropane.
however, could be ruled out by the following work to
simultaneously examine the effects of various Bu3SnH concen-
trations, addition times and reaction temperatures.6 Thus,
treatment of 9 with 4 eq. of Bu3SnH (not using the slow addition
technique) in the presence of triethylborane in toluene at rt for
16 h also gave the 6-endo cyclisation product 12 in 51% yield,
along with the reduction product 14 (23%). The most plausible
explanation for the results with 3 and 9, therefore, may be
derived from the consideration of the rotation of enamide.7 Two
conformers can be considered for both radical precursors, i.e.
syn-3 and anti-3 for 3 and syn-9 and anti-9 for 9. In the
1 For X = CH2, see: A. N. Abeywickrema, A. L. J. Beckwith and S.
Gerba, J. Org. Chem., 1987, 52, 4072; For X = O, see: S.-K. Chung and
F.-F. Chung, Tetrahedron Lett., 1979, 2473; H. Togo and O. Kikuchi,
Tetrahedron Lett., 1988, 29, 4133; For X = NCOR, see: J. P. Dittami
and H. Ramanathan, Tetrahedron Lett., 1988, 29, 45; Y. Özlü, D. E.
Cladingboel and P. J. Parsons, Tetrahedron, 1994, 50, 2183.
2 H. Ishibashi, K. Ohata, M. Niihara, T. Sato and M. Ikeda, J. Chem. Soc.,
Perkin Trans. 1, 2000, 547.
3 K. Jones, M. Thompson and C. Wright, J. Chem. Soc., Chem. Commun.,
1986, 115; K. Jones and J. M. D. Storey, Tetrahedron Lett., 1993, 34,
7797.
4 A limited example of 6-endo selective cyclisation has been reported for
the palladium-mediated reaction of N-acryloyl-7-bromoindoline. See:
J. W. Dankwardt and L. A. Flippin, J. Org. Chem., 1995, 60, 2312.
5 C. Aubert, C. Huard-Perrio and M.-C. Lasne, J. Chem. Soc., Perkin
Trans. 1, 1997, 2837.
6 Careful examinations on the effects of varying Bu3SnH concentration,
addition time and reaction temperature, have frequently shown that
6-endo cyclisation products are formed by an initial 5-exo cyclisation
followed by neophyl rearrangement. See: K. A. Parker, D. M. Spero and
K. C. Inman, Tetrahedron Lett., 1986, 27, 2833; A. N. Abeywickrema,
A. L. J. Beckwith and S. Gerba, J. Org. Chem., 1987, 52, 4072; K. Jones,
S. A. Brunton and R. Gosain, Tetrahedron Lett., 1999, 40, 8935. See
also ref. 2.
7 An initial conformation of a radical precursor has been suggested to play
an important role in deciding the course of cyclisation. See: D. P. Curran
and J. Tamine, J. Org. Chem., 1991, 56, 2746; O. M. Musa, J. H. Horner
and M. Newcomb, J. Org. Chem., 1999, 64, 1022; D. P. Curran, W. Liu
and C. H.-T. Chen, J. Am. Chem. Soc., 1999, 121, 11 012, and references
cited therein.
8 It has been also suggested that the vinyl groups of N-alkyl-N-
vinylcarbamates occupy anti-position to the alkoxycarbonyl groups.
See: O. Tamura, M. Hashimoto, Y. Kobayashi, T. Katoh, K. Nakatani,
M. Kamada, I. Hayakawa, T. Akiba and S. Terashima, Tetrahedron,
1994, 50, 3889; T. Akiba, O. Tamura, M. Hashimoto, Y. Kobayashi, T.
Katoh, K. Nakatani, M. Kamada, I. Hayakawa and S. Terashima,
Tetrahedron, 1994, 50, 3905.
conformers syn-3 and anti-9, severe steric repulsions between
the aroyl (o-IC6H4CO) and CNC groups and between the acyl
(EtCO) and CNC groups, respectively, are evident. The
conformers anti-3 and syn-9 therefore predominate, and the
resulting radicals attack on the more proximate Ca-position of
anti-3 and Cb-position of syn-9, to give the observed 5-exo
cyclisation product 4 and the 6-endo cyclisation product 12,
respectively. The NOE difference spectroscopy also indicated
that 9 exists only in the syn-9 form.8 Thus, irradiation of the
9 For sulfur-controlled exo selective radical cyclisations, see: H.
Ishibashi, T. Kobayashi and D. Takamasu, Synlett, 1999, 1286 and
references cited therein.
1
signals due to the N-benzylic protons [d 4.76 (3 3 2 H, s) and
2
10 This assumption was supported by IR spectral properties showing the
carbonyl band for (E)-isomer 10a in a higher frequency region (1680
cm21) compared to that (1660 cm21) for (Z)-isomer 10b.
11 It seems that the size of the substituent on the nitrogen atom of 10a,b
does not influence the conformer population. Thus, treatment of the N-
COBut congener of (E)-isomer 10a also gave nearly equal amounts of
5-exo cyclisation product (42%) and 6-endo cyclisation product (37%),
and the corresponding (Z)-isomer gave only the 5-exo cyclisation
product in 70% yield (compare to the results with 10a,b).
4.94 (3 3 2 H, s)] of 9 caused an enhancement of the signals due
1
to the Cb-proton cis to the nitrogen atom [d 4.26 (3 H, d, J 15.6)
2
and 4.29 (3 H, d, J 15.6)] and no enhancement of the signals due
2
1
to the Ca-proton [d 6.95 (3 H, dd, J 15.6 and 9.2) and 7.67 (3 H,
dd, J 15.6 and 9.2)].11 The preponderance of the syn-9
conformer over anti-9 seems to be independent of the size of the
N-acyl group, since 11 having a sterically less demanding N-
formyl group, also gave the 6-endo cyclisation product 13.
1528
Chem. Commun., 2000, 1527–1528