Approach to spirocyclohexadienimines and corresponding dienones through radical ipso cyclization onto aromatic azides

Article abstract of DOI:10.1002/anie.200804333

(Chemical Equation Presented) Ipso easy: Cyclization of aryl radicals at the ipso-position of p-azido-substituted benzamides or 2-phenylacetamides results in effective production of spirocyclohexadieniminyl radicals through prompt elimination of molecular nitrogen by transient azidocyclohexadienyl radicals. The resultant iminyl radicals can be exploited for the synthesis of oxindoles or quinolones bearing spirocyclohexadienimine/ cyclohexadienone rings (see picture, TTMSS=tris(trimethylsilyl)silane).

Full text of DOI:10.1002/anie.200804333

Angewandte
Chemie
DOI: 10.1002/anie.200804333
Synthetic Methods
Approach to Spirocyclohexadienimines and Corresponding Dienones
through Radical ipso Cyclization onto Aromatic Azides**
Tommaso Lanza, Rino Leardini, Matteo Minozzi,* Daniele Nanni,* Piero Spagnolo,* and
Giuseppe Zanardi
Spirocyclohexadienones, such as 1 and 3, are pivotal inter-
mediates in the synthesis of important biologically active
compounds (Figure 1). In particular, reduction and depro-
Figure 1. Spirooxindoles (1, 2) and spirohydroquinone (3) intermedi-
ates in the preparation of biologically active compounds.
Scheme 1. Synthesis of spirooxindoles and spirodihydroquinolones by
five- and six-membered ipso-cyclization of aryl radicals. TTMSS=tris-
(trimethylsilyl)silane; TBS=tert-butyldimethylsilyl; Trt=triphenyl-
methyl.
tection of 1 provides spirooxindole 2, an intermediate in the
synthesis of the vasopressin inhibitor SR121463A,[1–4] whereas
spirodihydroquinolone 3 is a key intermediate in the prepa-
ration of azagalanthamine[5a] and other indole alkaloids.[5b]
In a recent study, Curran and co-workers devised an
elegant radical procedure for the synthesis of both spiroox-
indoles and spirodihydroquinolones, entailing five- and six-
membered ipso cyclization of aryl radicals arising from
deiodination of p-alkoxy-substituted anilides 4 and 5: the
resulting spirocyclohexadienyl radicals afford dienones 6 and
TBS), which in turn required the rather longwinded prepa-
ration of TBS-protected 4-hydroxybenzoyl and, especially, 2-
(4-hydroxyphenyl)acetyl chloride (four steps). Moreover, the
production of spirooxindole 6 (from benzanilides 4) was
significantly limited by the concomitant generation of phen-
anthridinone 8, arising from cyclization at the aromatic
ortho-position.
In previous studies, we discovered that a-azidoalkyl
radicals can form iminyl radicals by extrusion of dinitrogen.[7]
In principle, suitable azidocyclohexadienyl radicals, which
would arise from radical ipso attack at the para position of
aromatic azides, might behave similarly to afford cyclohexa-
dieniminyl radicals. We therefore considered that spirocyclo-
hexadienyl radicals, analogous to those devised by Curran,
but bearing a p-azido rather than a p-alkoxy substituent,
should be very attractive candidates for testing our assump-
tion. Indeed, spirocyclic azidocyclohexadienyl radicals, such
as 13 and 14 (Scheme 2), might afford the corresponding
iminyl radicals 15 and 16, which, under reductive conditions,
would yield cyclohexadienimines 17 and 18 in an efficient
fashion. Since those imines should undergo easy hydrolytic
conversion into dienones, we envisioned that our study might
eventually lead to an alternative synthetic route to spirocy-
clohexadienones 6 and 7, based on the primary radical
reaction of p-azido-substituted 2-iodoanilides 9 and 10 with
tris(trimethylsilyl)silane (TTMSS) and Et3B.
À
7 upon b cleavage of the O R’ bond with elimination of the
alkyl/silyl R’ radical (Scheme 1).[6] The effectiveness of this
method is, however, somewhat limited by the efficiency of the
b-elimination process, which was especially effective only
when a highly stable species, such as a triphenylmethyl (trityl,
Trt) radical, could be eliminated. Furthermore, the construc-
tion of the trityloxy precursors 4 and 5 (R’ = Trt) was not
trivial, as it entailed deprotection and tritylation of prelimi-
nary silyloxy anilides 4 and 5 (R’ = tert-butyldimethylsilyl,
[*] Dr. T. Lanza, Prof. R. Leardini, Dr. M. Minozzi, Prof. D. Nanni,
Prof. P. Spagnolo, Prof. G. Zanardi
Dipartimento di Chimica Organica “A. Mangini”
Universitꢀ di Bologna
Viale Risorgimento 4, I-40136 Bologna (Italy)
Fax: (+39)051-2093654
E-mail: minus@ms.fci.unibo.it
Aromatic azides are widely known to undergo smooth
radical attack at the azido moiety to give aminyl radicals[8]
but, to our knowledge, their potential use as alternative
precursors of cyclohexadieniminyl radicals is to date totally
unprecedented.
nanni@ms.fci.unibo.it
spagnolo@ms.fci.unibo.it
[**] We gratefully acknowledge financial support from MIUR (2006–
2007 Funds for “Extending the Application of Organic Free Radicals
to New Synthetic Methods”).
The use of azidoanilides 9 and 10 as radical cyclization
precursors was also encouraged by our expectation that such
compounds should be more easily accessible than the oxygen-
Supporting information for this article (experimental details) is
available on the WWW under http://dx.doi.org/10.1002/anie.
200804333.
Angew. Chem. Int. Ed. 2008, 47, 9439 –9442
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
9439

Communications
Scheme 3. Route to indolones by five-membered radical spirocycliza-
tion, and concurrent formation of phenathridinone by-products.
a R=H; b R=Me; c R=Br. [a] Yield not determined, owing to unsat-
isfactory purification; [b] yield based on starting azides 9a–c.
Scheme 2. Radical approach to cyclohexadienimines from p-azido-sub-
stituted 2-iodoanilides.
the filtrate gave the respective phenanthridinone by-products
19b and c (Scheme 3).
substituted counterparts 4 and 5. Furthermore, previous
results from this group suggested that tris(trimethylsilyl)silyl
radicals, rather than adding to the azido moiety, would effect
selective iodine abstraction from the azidoiodides 9 and 10 to
give aryl radicals 11 and 12.[7a,8a,9] Finally, the azido function-
ality is known to effect stronger radical stabilization than an
alkoxy group[10] and, in the case of 9, this effect should favor
the five-membered ring-closure ratio of aryl radical 11, hence
disfavoring the competing formation of the phenanthridone.
Herein we report the highly rewarding results provided by
our preliminary studies, using azides 9a–c and 10a–c
(Scheme 2), and the congener 21 (Scheme 5). As anticipated,
compounds 9a–c and 10a–c were prepared in good yields (65–
85%) by coupling the appropriate N-methyl-2-iodoaniline
with 4-azidobenzoyl or 2-(4-azidophenyl)acetyl chloride.
These acyl chlorides were, in turn, readily prepared through
conversion of the corresponding commercially available
amino acids into azides.
Analogous radical reactions of acetanilides 10a–c also led
to satisfactory results, giving rise to the precipitation of large
amounts of quinolone imine hydroiodides 18Aa–Ac, which
were again isolated by filtration. In each case, column
chromatography of the filtrate furnished modest amounts of
the respective 2-hydroxy-substituted acetanilides 20a–c,
which conceivably arose from the 1,5-H-transfer reaction of
aryl radicals 12a–c, followed by trapping of the translocated
benzyl radicals by oxygen (Scheme 4).[7a,11]
All azides were reacted by adopting experimental con-
ditions strictly comparable with those previously employed
for alkoxyanilide substrates 4 and 5.[6] After 16 h, the reaction
of benzamide 9a gave rise to extensive precipitation of an
orange solid, which was isolated by filtration and found to be
sparingly soluble in the common nonpolar solvents, and more
soluble in polar solvents, such as acetone, methanol, and, to a
lesser extent, water. The spectroscopic data were consistent
with the structure of imine 17a, in the form of its hydroiodide
salt 17Aa (Scheme 3). This salt (70% yield) presumably arose
from the initial formation of 17a through subsequent reaction
with hydrogen iodide, possibly produced upon hydrolysis of
tris(trimethylsilyl)iodosilane, and/or direct reaction with the
iodosilane followed by fast hydrolysis of the ensuing N-
silyliminium iodide. Column chromatography of the filtrate
remaining after removal of 17Aa afforded the minor product
azidophenanthridinone 19a (30% yield), that is, the product
of the competing six-membered cyclization of aryl radical
11a. In a similar fashion, direct filtration of the reaction
mixture from the radical reactions of benzanilides 9b and c
furnished comparable amounts of the analogous crude imine
hydroiodides 17Ab and Ac. Chromatographic separation of
Scheme 4. Route to quinolones by six-membered radical spirocycliza-
tion, and concurrent formation of 2-hydroxy-substituted acetanilide by-
products. a R=H; b R=Me; c R=Br. [a] Yield not determined, owing
to unsatisfactory purification; [b] yield based on starting azides 10a–c.
The above overall results confirmed our original assump-
tions that aryl radicals 11 and 12 could undergo both five- and
six-membered spirocyclization reactions onto the internal
aromatic azide, and that the resulting intermediate spirocyclic
azidocyclohexadienyl radicals 13 and 14 could extrude
9440 www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 9439 –9442

Angewandte
Chemie
molecular nitrogen in a highly efficient fashion to afford
cyclohexadieniminyl radicals 15 and 16, and thence the
corresponding imines 17 and 18. The yields of the latter
compounds were significantly higher than those of cyclo-
hexadienones 6 and 7, arising from silyloxy- and trityloxy-
substituted amides 4 and 5 (R’ = TBS or Trt).[6] This fact
validated our additional hypothesis that the azido substituent
promotes aryl radical spirocyclization more effectively than
an alkoxy group. These interesting results might pave the way
to important future applications of aromatic azides for the
construction of uncommon spirocyclic cyclohexadieni-
mines.[12,13]
Gratifyingly, upon brief heating at 608C in methanol/
water in the presence of a few drops of hydrochloric acid, 17A
and 18A afforded the respective spirocyclohexadienones 6
and 7, in satisfactory yields based on the original azides 9 and
10 (Scheme 3 and 4).
We finally attempted the synthesis of spiroquinolone 3, a
valuable intermediate in the preparation of indole alkaloids.
Coupling of N-methyl-2-iodo-6-methoxyaniline with 2-(4-
azidophenyl)acetyl chloride gave azidoacetamide 21 in high
yield (82%). Treatment of 21 with TTMSS/Et3B led to imine
hydroiodide 22 (80%), which was hydrolyzed to afford the
target quinolone 3 in 54% overall yield (Scheme 5). Com-
Experimental Section
All azides (0.5 mmol) were treated in either benzene or toluene
(8 mL) with TTMSS (1.2 equiv) and Et3B (1.2 equiv) and the
resulting mixtures were stirred at RT under air until total consump-
tion of the starting material. Filtration of the reaction mixtures
afforded crude imine hydroiodides, which were then hydrolyzed,
whereas chromatography of the filtrates yielded the by-products
(phenanthridones or hydroxyanilides).
Received: September 2, 2008
Published online: October 29, 2008
Keywords: azides · cyclization · radical reactions ·
.
reactive intermediates · spiro compounds
[1] a) J. Halasz, B. Podanyi, A. Santa-Csutor, Z. Bocskei, K. Simon,
M. Hanusz, I. Hermecz, J. Mol. Struct. 2003, 644, 187; b) I.
Hermecz, A. Santa-Csutor, C. Gꢁnczi, G. Heja, E. Csikos, K.
Simon, A. Smelko-Esek, B. Podanyi, Pure Appl. Chem. 2001, 73,
1401.
[2] a) A. Di Malta, L. Foulon, G. Garcia, D. Nisato, R. Roux, C.
Serradeil-Le Gal, G. Valette, J. Wagon, US 5849780, 1998
[Chem. Abstr. 1999, 130, 66390]; b) N. A. Jonsson, P. Moses,
SE 366309, 1974 [Chem. Abstr. 1975, 83, 79076]; c) N. A.
Jonsson, P. Moses, Acta Chem. Scand. Ser. B 1974, 28, 225.
[3] a) H. Venkatesan, M. C. Davis, Y. Altas, J. P. Snyder, D. C.
Liotta, J. Org. Chem. 2001, 66, 3653; b) C. Gonczi, E. Csikos, I.
Hermecz, G. Heja, A. Illar, L. Nagy, A. Santa-Csutor, K. Simon,
A. Smelko-Esek, T. Szomor, PTC Int. Appl. WO0105760, 2001
[Chem. Abstr. 2001, 134, 115851]; c) L. Foulon, G. Garcia, C.
Serradeil-Le Gal, G. Valette, WO9715556A1, 1997 [Chem.
Abstr. 1997, 127, 5010]; d) R. S. Johnos, T. O. Lovett, T. S.
Steven, J. Chem. Soc. C 1970, 796.
[4] E. M. Beccalli, F. Clerici, M. L. Gelmi, Tetrahedron 2003, 59,
4615.
Scheme 5. Synthesis of spirodihydroquinolone 3. [a] Yield based on
starting azide 21.
[5] a) C. Thal, C. Guillou, J. C. Beunard, E. Gras, P. Potier,
Fr 2826005A1, 2002 [Chem. Abstr. 2003, 138, 39446]; b) J.
Pereira, M. Barlier, C. Guillou, Org. Lett. 2007, 9, 3101.
[6] F. Gonzalez-Lopez de Turiso, D. P. Curran, Org. Lett. 2005, 7,
151.
pound 3 had previously been prepared in moderate yield
through a long-winded, expensive multistep sequence based
on intramolecular Heck cyclization of a protected 2-(4-
oxocyclohexenyl)acetanilide[5a] and, subsequently, in 40%
yield, through direct radical cyclization of the TBSO analogue
of azide 21.[6]
In conclusion, we have shown that aromatic azides can
provide an unprecedented straightforward synthetic entry to
cyclohexadieniminyl radicals that can be suitably exploited in
the valuable production of indolones and quinolones bearing
spirocyclohexadienimine/spirocyclohexadienone rings. As far
as the dienones are concerned, our new synthetic protocol is
evidently superior to both Curranꢀs radical procedure[6] and
other reported nonradical methods.[4,5] Our procedure, in fact,
while giving rise to comparable or even better yields of those
dienones, employs very easy to prepare azido precursors,
entails very simple workup,[14] and results in highly minimized
atom waste. Our present discovery could allow novel, highly
appealing applications of radical chemistry in the synthesis of
biologically active compounds.
[7] a) G. Bencivenni, T. Lanza, R. Leardini, M. Minozzi, D. Nanni, P.
Spagnolo, G. Zanardi, J. Org. Chem. 2008, 73, 4721; b) P. C.
Montevecchi, M. L. Navacchia, P. Spagnolo, J. Org. Chem. 1997,
62, 5846.
[8] Selected examples: a) L. Benati, G. Bencivenni, R. Leardini, M.
Minozzi, D. Nanni, R. Scialpi, P. Spagnolo, G. Zanardi, J. Org.
Chem. 2006, 71, 5822; b) L. Benati, G. Bencivenni, R. Leardini,
M. Minozzi, D. Nanni, R. Scialpi, P. Spagnolo, G. Zanardi, J. Org.
Chem. 2006, 71, 434; c) L. Benati, G. Bencivenni, R. Leardini, M.
Minozzi, D. Nanni, R. Scialpi, P. Spagnolo, S. Strazzari, G.
Zanardi, C. Rizzoli, Org. Lett. 2006, 8, 2499; d) P. C. Montevec-
chi, M. L. Navacchia, P. Spagnolo, Eur. J. Org. Chem. 1998, 1219;
e) S. Kim, G. H. Joe, J. Y. Do, J. Am. Chem. Soc. 1993, 115, 3328;
f) S. Kim, J. Y. Do, J. Chem. Soc. Chem. Commun. 1995, 1607;
g) S. Kim, K. M. Yeon, K. S. Yoon, Tetrahedron Lett. 1997, 38,
3919; h) B. Patro, J. A. Murphy, Org. Lett. 2000, 2, 3599; i) for a
pioneering work about the addition of radicals to azides, see:
B. P. Roberts, J. N. Winter, J. Chem. Soc. Perkin Trans. 2 1979,
1353.
[9] a) L. Benati, R. Leardini, M. Minozzi, D. Nanni, P. Spagnolo, S.
Strazzari, G. Zanardi, Org. Lett. 2002, 4, 3079; b) M. Kizil, B.
Patro, O. Callaghan, J. A. Murphy, M. B. Hursthouse, D. Hibbs,
J. Org. Chem. 1999, 64, 7856; c) S. Kim, G. H. Joe, J. Y. Do, J.
Am. Chem. Soc. 1994, 116, 5521.
Angew. Chem. Int. Ed. 2008, 47, 9439 –9442
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
9441

Communications
[10] L. Benati, G. Calestani, R. Leardini, M. Minozzi, D. Nanni, P.
[13] Aromatic azides have found some use in the production of
peculiar spirocyclic cyclohexadienimines through intramolecular
nucleophilic ipso-trapping of derived nitrenium ions, see:
a) E. K. Dolence, T. Fitz, H. Morita, D. S. Watt, Tetrahedron
Lett. 1987, 28, 43; b) R. A. Abramovitch, R. Yeyaraman, J.
Chem. Soc. Chem. Commun. 1986, 283.
Spagnolo, S. Strazzari, G. Zanardi, J. Org. Chem. 2003, 68, 3454.
[11] a) A. L. J. Beckwith, J. M. D. Storey, J. Chem. Soc. Chem.
Commun. 1995, 977; b) D. P. Curran, H. Yu, H. Liu, Tetrahedron
1994, 50, 7343; c) K. Jones, J. M. D. Storey, J. Chem. Soc. Chem.
Commun. 1992, 1766.
[12] a) I. G. C. Coutts, N. J. Culbert, M. Edwards, J. A. Hadfield,
D. R. Musto, V. H. Pavidis, D. J. Richards, J. Chem. Soc. Perkin
Trans. 1 1985, 1829; b) I. G. C. Coutts, M. Edwards, D. R. Musto,
D. J. Richards, Tetrahedron Lett. 1980, 21, 5055.
[14] Imine hydroiodides are recovered by filtration of the reaction
mixtures and can be subjected to hydrolysis without any need for
further purification by column chromatography.
9442 www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 9439 –9442