Radical cleavage of a β-hydroxy azide: A reversal of regioselectivity in the oxidative fragmentation of hydroindoles

Article abstract of DOI:10.1016/S0040-4039(00)00706-1

Treatment of hydroindole 17 with PhI(OAc)₂/I₂ leads to oxidative fragmentation of the β-hydroxy azide moiety and results in the exclusive formation of keto nitrile 18 via radical cleavage of the lateral ring C-C bond. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0040-4039(00)00706-1

Tetrahedron Letters 41 (2000) 4723±4727
Radical cleavage of a b-hydroxy azide: a reversal of
regioselectivity in the oxidative fragmentation of hydroindoles
Peter Wipf* and David A. Mareska
Department of Chemistry, University of Pittsburgh, Pittsburgh, Pennsylvania 15260, USA
Received 17 March 2000; accepted 26 April 2000
Abstract
Treatment of hydroindole 17 with PhI(OAc)2/I2 leads to oxidative fragmentation of the b-hydroxy azide
moiety and results in the exclusive formation of keto nitrile 18 via radical cleavage of the lateral ring C±C
bond. # 2000 Elsevier Science Ltd. All rights reserved.
Keywords: regioselective oxidative fragmentation; hydroxy azide; keto nitrile.
The synthesis of medium-sized rings may be accomplished by the oxidative cleavage of alkoxy
radicals attached to fused bicyclic systems.1,2 Recently, we developed new methodology for the
ring expansion of 4-hydroxyhydroindoles such as 1 to azononanes in an approach toward the
total synthesis of tuberostemonone 6 (Scheme 1, 1±5).3 Presumably, this conversion proceeds
through a nitrogen-stabilized radical intermediate 3. Evidence suggests that fragmentations of this
type are reversible processes and that ®nal product formation is in¯uenced by both the stability of
the radical intermediates and their relative rates of trapping.2
As part of our e€orts to synthesize several of the biologically interesting and structurally
challenging Stemona and Croomia alkaloids4 from hydroindole 7,5 we envisioned directing the
oxidative cleavage to a€ord products arising from C(3a)±C(4) bond fragmentation (Scheme 2).
Such a cleavage a€ords an alkyl substituted pyrrolidinone that may be useful for the preparation
of alkaloids such as croomine6,7 and parvistemonine.8 However, careful examination of the
reaction mixture from oxidative fragmentation of 1 and related compounds3 revealed no evidence
of C(3a)±C(4) bond cleavage. Even after the addition of a radical stabilizing vinyl substituent at
C(4) in substrate 10, only complex product mixtures were obtained (Fig. 1). Increasing the steric
bulk about the ring nitrogen (11) or protonating at this position (12) to reduce the stability of a
radical analogous to 3 was similarly unsuccessful in yielding C(3a)±C(4) scission products.
Recent studies by Suarez have shown that a radical a to an azide group is trapped by the
release of nitrogen and nitrile formation.9 From these results, we hypothesized that an azide
* Corresponding author.
0040-4039/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(00)00706-1

4724
Scheme 1.
Scheme 2.
Figure 1.
appropriately placed within a hydroindole framework may direct ®nal product formation to arise
from C(3a)±C(4) cleavage. b-Hydroxy azides suitable for probing this alternative fragmentation
strategy were obtained from the conjugate addition of sodium azide to hydroindole 7 which is
readily available from l-tyrosine via an oxidative cyclization process (Scheme 3).5a After ®ve
days, a separable mixture of isomers 13 and 14 as well as a signi®cant amount of recovered
starting material were obtained.10 Attempts to optimize this conversion via Lewis acid mediated
conjugate additions with diethylaluminum azide under anhydrous conditions failed to deliver any
desired product.11,12 Upon exposure to oxidative fragmentation conditions, both ketoazides 13
and 14 failed to a€ord any fragmentation product and were instead converted back to enone 7
along with decomposition products. As the presence of the carbonyl appeared to allow for facile
elimination, the major isomer 13 was reduced under Luche conditions and protected as the silyl
ether 17.13
The oxidative fragmentation of 17 was attempted under a variety of conditions (Eq. (1)) and
Table 1).14 Treatment of 17 with iodobenzene diacetate and iodine a€orded the desired ketonitrile
18 with no evidence for the formation of the transannular fragmented product 19 (entry 1). In the

4725
Scheme 3.
absence of iodine, starting material was recovered quantitatively (entry 2). Performing the reaction
at 5ꢀC (entry 3) signi®cantly retarded the formation of 18. The use of iodobenzene bis(tri¯uoro-
acetate) did not e€ect the desired reaction and instead desilylated the starting material in almost
quantitative fashion (entry 4). Lead tetraacetate also failed to yield any desired 18 (entry 5).
ꢁ1†
Table 1
Oxidative fragmentation of hydroxy azide 17
The formation of 18 may be explained by an oxidative fragmentation as proposed by Suarez
(Scheme 4).9 Transannular and lateral C±C bond cleavage intermediates 21 and 22 may both be
accessible from alkoxy radical 20. Rapid and irreversible oxidation of 22 to 23 followed by elim-
ination of N2 provides ketonitrile 18. An alternative, ionic mechanism involving oxidation of
azide prior to C±C bond cleavage cannot be excluded, however.

4726
Scheme 4.
In summary, we have been able to reverse the high selectivity for ring expansion previously
observed in the oxidative fragmentation of hydroindoles 1 by incorporating an azide substituent
at C(4) and taking advantage of the irreversible nitrogen elimination of a-azido radicals. We
expect that this strategy will be generally useful in directing radical-based fragmentation processes
and that the selective access toward azanonanes 5 and pyrrolidinones 18 from hydroxy indoles
will further enhance the use of tyrosine oxidation products as sca€olds for natural product
synthesis.
Acknowledgements
This work was supported by the National Institutes of Health (AI/GM-33506). We thank Mr.
Wenjie Li for attempting the oxidative fragmentation of 10.
References
1. (a) Dowd, P.; Zhang, W. Chem. Rev. 1993, 93, 2091. (b) Suginome, H.; Kondoh, T.; Gogonea, C.; Singh, V.;
Goto, E.; Osawa, E. J. Chem. Soc., Perkin Trans. 1 1995, 69. (c) Mihailovic, M. Lj.; Lorenc, Lj.; Pavlovic, V.
Tetrahedron 1977, 33, 441. (d) Arencibia, M. T.; Freire, R.; Perales, A.; Rodrõguez, M. S.; Suarez, E. J. Chem.
Soc., Perkin Trans. 1 1991, 3349. (e) Boto, A.; Hernandez, R.; Velaquez, S. M.; Suarez, E. J. Org. Chem. 1998, 63,
4697. (f) Saicic, R. N. Tetrahedron Lett. 1997, 38, 295.
2. (a) Macdonald, T. L.; O'Dell, D. E. J. Org. Chem. 1981, 46, 1501. (b) Beckwith, A. L. J.; Kazlauskas, R.; Syner-
Lyons, M. R. J. Org. Chem. 1983, 48, 4718. (c) O'Dell, D. E.; Loper, J. T.; Macdonald, T. L. J. Org. Chem. 1988,
53, 5225. (d) Hernandez, R.; Melian, D.; Prange, T.; Suarez, E. Heterocycles 1995, 41, 439.
3. Wipf, P.; Li, W. J. Org. Chem. 1999, 64, 4576.
4. Pilli, R. A.; de Oliveira, M. C. F. Nat. Prod. Rep. 2000, 17, 117 and references cited therein.
5. (a) Wipf, P.; Kim, Y. Tetrahedron Lett. 1992, 33, 5477. (b) Wipf, P.; Kim, Y.; Goldstein, D. M. J. Am. Chem. Soc.
1995, 117, 11106. (c) Goldstein, D. M.; Wipf, P. Tetrahedron Lett. 1996, 37, 739.
6. Noro, T.; Fukushima, S.; Ueno, A.; Litaka, Y.; Sakai, Y. Chem. Pharm. Bull. 1979, 27, 1495.
7. Martin, S. F.; Barr, K. J.; Smith, D. W.; Bur, S. K. J. Am. Chem. Soc. 1999, 121, 6990.
8. Lin, W.-H.; Yin, B.-P.; Tang, Z. J.; Xu, R. S.; Zhong, Q.-X. Acta Chim. Sin. 1990, 48, 811.
9. Hernandez, R.; Leon, E. I.; Moreno, P.; Suarez, E. J. Org. Chem. 1997, 62, 8974.
10. The relative con®guration of 13 and 14 was deduced from their J4,5 values and HMQC, HMBC, and NOE
techniques. 4-(R)-Azido-3a-hydroxy-6-oxo-octahydro-indole-1,2-dicarboxylic acid 1-benzyl ester 2-methyl ester
(13). A solution of hydroindole 7 (1.0 g, 2.9 mmol) in 8 mL of MeOH, 4 mL of H2O, and 6 mL of glacial AcOH

4727
was cooled to 0ꢀC and treated with NaN3 (1.9 g, 29 mmol). The mixture was stirred at rt for 5 days, diluted with
EtOAc and washed with H2O and brine. The aqueous layer was extracted with EtOAc. The organic layers were
dried (MgSO4), ®ltered, and chromatographed on SiO2 (1.5:1, EtOAc:hexanes) to a€ord 125 mg (13%) of 7, 600
mg (53%) of 13 and 250 mg (22%) of 14 as waxy solids: compound 13: [ꢀ]D ^42.3 (c 1.0, CHCl3); IR (neat) 3434
(br), 2108, 1708 cm^1; 1H NMR (500 MHz, CDCl3) ꢁ 7.39±7.29 (m, 5H), 5.23±5.19 (m, 1H), 5.09 and 4.99 (2 d, 1H,
J=12.1 Hz), 4.91 and 4.51 (2 bs, 1H), 4.51±4.45 (m, 1H), 4.30 and 4.18 (2 dd, 1H, J=7.3 Hz), 4.14 and 4.08 (2 dd,
1H, J=5.3, 11.9 Hz), 3.88 and 3.58 (2 s, 3H), 3.04 and 2.89 (2 dd, 1H, J=6.4, 15.4 Hz), 2.77±2.71 (m, 1H), 2.69±
2.58 (m, 1H), 2.42±2.32 and 2.21±2.17 (m, 1H), 2.42±2.32 (m, 1H), 2.21±2.17 (m, 1H); 13C NMR (125 MHz,
CDCl3) ꢁ 204.5, 203.9, 175.0, 174.5, 154.4, 154.0, 135.6, 128.5, 128.4, 128.3, 128.2, 127.9, 81.4, 80.5, 67.8, 67.5,
64.1, 63.5, 61.7, 57.7, 57.6, 53.2, 52.8, 44.1, 43.0, 42.2, 41.9, 35.6, 34.3; HRMS m/z calcd for C18H20N4O6:
388.1383, found: 388.1368.
11. Rawal, V. H.; Zhong, H. M. Tetrahedron Lett. 1994, 35, 4947.
12. Chung, B. Y.; Park, Y. S.; Cho, I. S.; Hyun, B. C. Bull. Korean Chem. Soc. 1988, 9, 269.
13. Attempted conversion of the minor isomer 14 under analogous conditions failed to produce the expected diol and
instead returned a complex mixture of products.
14. 5-(R)-[2-(R)-tert-Butyl-dimethyl-silanyloxy)-3-cyano-propyl]-4-oxo-pyrrolidine-1,2-dicarboxylic acid 1-benzyl ester
2-methyl ester (18). Iodobenzene diacetate (45 mg, 0.14 mmol) and iodine (35 mg, 0.14 mmol) were added
sequentially at rt to a solution of hydroxy azide 17 (27 mg, 0.054 mmol) in 1.2 mL of CH2Cl2. The purple solution
was stirred for 3 h, diluted with EtOAc and washed with saturated aqueous NaHCO3 and 10% aqueous Na2S2O3.
The aqueous phase was extracted with EtOAc. The combined organic layers were dried (MgSO4), ®ltered, and
chromatographed on SiO2 (1:3, EtOAc:hexanes) to a€ord 15 mg (59%) of 18 as an oil and 3 mg (ꢂ10%) of 17
contaminated with an unknown by-product. Compound 18: [ꢀ]D ^45.0 (c 1.0, CHCl3); IR (neat) 2249, 1758, 1712
cm^1; 1H NMR (500 MHz, CDCl3) ꢁ 7.38±7.34 (m, 5H), 5.31±5.22 (m, 1H), 5.16±5.09 (m, 1H), 4.72 (d, 1H, J=6.2
Hz), 4.28±4.11 (m, 2H), 3.76 and 3.53 (2 s, 3H), 2.97 (dd, 1H, J=6.2, 11.2 Hz), 2.60±2.45 (m, 2H), 2.50±2.30 (m,
2H), 2.00±1.90 and 1.80±1.72 (2 m, 1H), 0.91 and 0.85 (2 s, 9H), 0.14±0.01 (m, 6H); 13C NMR (125 MHz, CDCl3)
ꢁ 209.1, 208.6, 172.2, 155.5, 154.7, 135.6, 128.6, 128.4, 117.2, 116.7, 68.1, 67.6, 65.1, 63.9, 59.3, 58.3, 55.9, 55.8,
52.7, 52.6, 39.7, 39.5, 39.4, 39.1, 26.7, 26.4, 25.6, 17.9, ^4.6, ^4.8; HRMS m/z calcd for C24H34N2O6Si: 474.2186,
found: 474.2181.