Access to 4-substituted 3,4-dihydroquinolin-2(1H)-ones by an unusual radical cyclisation of a secondary amide

Article abstract of DOI:10.1016/j.tetlet.2005.09.006

A novel route to 3,4-dihydroquinolin-2(1H)-ones involving two radical addition steps is reported, starting from readily accessible xanthates and N-aryl-3-butenamides.

Full text of DOI:10.1016/j.tetlet.2005.09.006

Tetrahedron
Letters
Tetrahedron Letters 46 (2005) 7503–7506
Access to 4-substituted 3,4-dihydroquinolin-2(1H)-ones by
an unusual radical cyclisation of a secondary amide
*
´
Gregori Binot and Samir Z. Zard
`
´
Laboratoire de Synthese Organique associe au CNRS, Ecole Polytechnique, F-91128 Palaiseau, France
Received 13 July 2005; revised 22 August 2005; accepted 2 September 2005
Available online 19 September 2005
Abstract—A novel route to 3,4-dihydroquinolin-2(1H)-ones involving two radical addition steps is reported, starting from readily
accessible xanthates and N-aryl-3-butenamides.
Ó 2005 Elsevier Ltd. All rights reserved.
Although quinolin-2(1H)-one skeletons are rarely found
in naturally occurring substances, these heterocycles
possess varied and powerful biological properties.1 For
example, compound A is a strong inhibitor of HIV-1
reverse transcriptase, whereas B (carteolol) is a b-adren-
ergic blocking agent (Fig. 1).2
tical industry to deal with heavy metal residues espe-
cially associated with organotin chemistry, and also
because of the relative difficulty of building six-mem-
bered rings by radical cyclisation onto an aromatic
nucleus. The tin-free xanthate radical chemistry we have
developed7 formally gives a longer lifetime to the inter-
mediate radicals, allowing them to undergo difficult
transformations. This technology provides a useful tool
for the construction of the desired dihydroquinolinone
skeleton under mild conditions.
Development of new synthetic pathways to these com-
pounds has thus recently attracted some attention from
organic chemists, broadening the scope of substrates
classically obtained by Friedla¨nder-type methods or Fri-
edel–Crafts cyclisations.3 These new routes include, for
instance, multi-component reactions,4 palladium chem-
istry or rhodium catalysis.5 In contrast, radical chemis-
try has not been considered in this context.6 This is, in
part, a consequence of the reluctance of the pharmaceu-
In the course of our studies centred around the prepara-
tion of new xanthates by conjugate addition of xanthic
acid to various electrophiles,8 xanthate 1 was prepared
in poor yield (15%) by reaction of the corresponding
N-crotonanilide with potassium O-ethyl xanthate in a
mixture of trifluoroacetic acid (TFA) and dichlorometh-
ane (DCM). However, to our surprise, refluxing 1 in
chlorobenzene or 1,2-dichloroethane (DCE) in the pres-
ence of a stoichiometric amount of lauroyl peroxide led
to dihydroquinolinone 2 in 57% yield (Scheme 1). This
result was particularly unexpected in view of the trans-
conformational preference of most secondary amides9
and because of related studies,10a which seemed to indi-
cate that substitution of the free position on the amide
nitrogen was required to achieve radical cyclisation in
synthetically useful yields. Only in the synthesis of homo-
phthalimides could we perform the ring closure in the
absence of a substituent on the nitrogen but this
required a reaction temperature of 160 °C.10b
Figure 1.
Keywords: 3,4-Dihydroquinolin-2(1H)-ones; Xanthates; Radical
cyclisation.
To circumvent the lack of reactivity of the parent anilide
towards conjugate addition, an electron-withdrawing
isobutoxycarbonyl group was placed on the nitrogen
*
Corresponding author. Tel.: +33 (0)169334872; fax: +33
(0)169333851; e-mail: zard@poly.polytechnique.fr
0040-4039/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2005.09.006

7504
G. Binot, S. Z. Zard / Tetrahedron Letters 46 (2005) 7503–7506
Scheme 1.
of the amide. In this case, the Michael addition pro-
ceeded smoothly to give xanthate 3 in excellent yield.
Unfortunately, and very surprisingly, the radical cyclisa-
tion of the latter was more sluggish, leading to dihydro-
quinolinone 4 in only 25% isolated yield.
This is the first instance to our knowledge where a sec-
ondary amide cyclises more efficiently than a substituted
imide. The unexpected lack of reactivity in the present
case may arise from the geometry of the system which
puts the radical centre too far from the aromatic ring,
slowing down the process and thus allowing side reac-
tions to occur. Selective removal of the isobutoxycar-
bonyl group in the presence of the xanthate proved
difficult.11 Another strategy outlined in Scheme 2 was
therefore considered, relying on the xanthate transfer
method. An efficient access to cyclisation precursors
could indeed be provided by means of the powerful
intermolecular radical addition of xanthates to unacti-
vated olefins. Hence readily accessible, unprotected,
butenanilides 5a–d12 underwent the radical addition
of xanthates 6a–c to give the desired adducts 7a–e in
very good isolated yields (Scheme 3). These adducts
were then refluxed in chlorobenzene in the presence of
stoichiometric amounts of lauroyl peroxide, added por-
tionwise, to give the desired 4-substituted dihydroquin-
olinones 8a–e, which were isolated in moderate yield
(28–56%) along with 9–13% of the reduced noncyclised
material (corresponding to 9 in Scheme 2).
Scheme 3.
The radical cyclisation appears to be rather insensitive
to the electronic nature of the aromatic ring. Formation
of dihydroquinolones 8b and 8d from adducts 7b and 7d,
containing, respectively, an electron poor and an elec-
tron rich aromatic ring, took place in roughly the same
yield. This fact is particularly interesting in the case of
8b. The presence of two strongly electron-withdrawing
trifluoromethyl groups essentially eliminates the possi-
bility of relying on the classical Friedel–Crafts-type of
reaction to access such a structure. Our method thus
allows the easy preparation of fluorinated compounds,
which are known to be of some importance in medicinal
chemistry as well as in agrochemistry.13
In the light of these encouraging preliminary results, the
two-step sequence was simplified since the intermediate
xanthate adducts could be directly engaged in the cycli-
sation step without purification. For this purpose, the
1,2-dichloroethane used in the first step was simply
replaced by chlorobenzene at the completion of the
intermolecular radical addition, as indicated by the dis-
appearance of the starting anilide 5. The cyclisation was
then performed by portionwise addition of a stoichio-
metric amount of lauroyl peroxide to furnish the desired
dihydroquinolones 8f–j in useful overall yields (36–51%;
Scheme 2.

G. Binot, S. Z. Zard / Tetrahedron Letters 46 (2005) 7503–7506
7505
References and notes
1. Jones, G. In Comprehensive Heterocyclic Chemistry II;
Katritzky, A. R., Rees, C. W., Scriven, E. F. V., Eds.;
Pergamon Press: Oxford, 1996; p 167.
2. Patel, M.; McHugh, R. J.; Cordova, B. C.; Klabe, R. M.;
Bacheler, L. T.; Erickson-Viitanen, S.; Rodgers, J. D.
Bioorg. Med. Chem. Lett. 2001, 11, 1943.
3. (a) Jones, G. In Comprehensive Heterocyclic Chemistry;
Boulton, A. J., McKillop, A., Eds.; Pergamon Press:
Oxford, 1984; Vol. 2, Chapter 8; (b) Li, K.; Foresee, L. N.;
Tunge, J. A. J. Org. Chem. 2005, 70, 2881; (c) Kadnikov,
D. V.; Larock, R. C. J. Org. Chem. 2004, 69, 6772, and
references cited therein.
4. Marcaccini, S.; Pepino, R.; Pozo, M. C.; Basurto, S.;
´
Garcıa-Valverde, M.; Torroba, T. Tetrahedron Lett. 2004,
45, 3999.
5. (a) Manley, P. J.; Bilodeau, M. T. Org. Lett. 2004, 6, 2433;
(b) Fujita, K.; Takahashi, Y.; Owaki, M.; Yamamoto, K.;
Yamaguchi, R. Org. Lett. 2004, 6, 2785.
6. For an important exception, see: (a) Wu, Y. L.; Chuang,
C. P.; Lin, P. Y. Tetrahedron 2000, 56, 6209; For recent
examples of ring closures onto aromatic rings leading to
six-membered rings, see: (b) Harrowven, D. C.; Wood-
cock, T.; Howes, P. D. Angew. Chem., Int. Ed. 2005, 44,
3899; (c) Bhowmik, D. R.; Venkateswaran, R. V. Tetra-
hedron Lett. 1999, 40, 7431, and references cited therein.
7. (a) Zard, S. Z. Angew. Chem., Int. Ed. 1997, 36, 672; (b)
Zard, S. Z. In Radicals in Organic Synthesis; Renaud, P.,
Sibi, M. P., Eds.; Wiley-VCH, 2001; pp 90–108.
8. Binot, G.; Quiclet-Sire, B.; Saleh, T.; Zard, S. Z. Synlett
2003, 382.
9. (a) Yamasaki, R.; Tanatani, A.; Azumaya, I.; Saito, S.;
Yamaguchi, K.; Kagechika, H. Org. Lett. 2003, 5, 1265,
and references cited therein; (b) Musa, O. M.; Horner, J.;
Newcomb, M. J. Org. Chem. 1999, 64, 1022; (c) Stork, G.;
Mah, R. Heterocycles 1989, 28, 723.
Scheme 4. Reagents and conditions: (a) lauroyl peroxide, DCE, reflux,
lauroyl peroxide, PhCl, reflux.
10. (a) Cholleton, N.; Zard, S. Z. Tetrahedron Lett. 1998, 39,
7295; (b) Quiclet-Sire, B.; Zard, S. Z. Chem. Commun.
2002, 2306.
11. After some experimentation, the best protecting group for
the transformation was found to be a methoxycarbonyl
this corresponds to an average of 60–70% yield per step)
as shown in Scheme 4.14 Once again the main side prod-
uct found in the crude mixture (5–16%) is the reduced
uncyclised intermediate adduct corresponding to 9 in
Scheme 2. The slightly higher yields observed when elec-
tron-withdrawing groups are present on the aromatic
ring can be ascribed to the moderate nucleophilic char-
acter of the cyclising radical.
group, giving xanthate
2 in 43% overall yield on
the protection/conjugate addition/deprotection/sequence,
starting from the crotonanilide.
12. Prepared by reaction of commercial anilines and vinylace-
tyl chloride in dichloromethane at rt.
13. For a recent review on the use of fluorine in the life science
industry, see: Maienfisch, P.; Hall, R. G. Chimia 2004, 58,
93.
In conclusion, we have developed a convergent and
short access to 4-substituted 3,4-dihydroquinolin-
2(1H)-ones. These deceptively simple-looking com-
pounds would be quite tedious and difficult to make
by classical methods. The starting materials and
reagents in the present route are cheap and readily avail-
able; the reaction conditions are mild and the process is
compatible with a broad spectrum of functional groups,
both on the aromatic ring and the xanthate, allowing
infinite variations on the structures. But perhaps most
importantly, we have documented the first and probably
quite rare instance where the radical cyclisation of a sec-
ondary amide is more efficient than that of the corre-
sponding fully substituted imide analogue.
14. The preparation of 8h is representative: a solution of
amide 5d (370 mg, 2.07 mmol) and xanthate 6b (300 mg,
1.36 mmol) in 1,2-dichloroethane (3 mL) was refluxed for
15 min under a slow stream of argon before DLP was
added (27 mg) from the top of the condenser. Portions of
DLP (16 mg) were added every 90 min until complete
consumption of the starting xanthate. The solvent was
evaporated and replaced by chlorobenzene (15 mL). The
mixture was then refluxed for 15 min under argon before
adding portions of DLP (110 mg) every 30 min until
complete disappearance of the intermediate (TLC moni-
toring). Concentration under reduced pressure afforded a
yellow residue, which was submitted to flash silica gel
chromatography eluting with petroleum ether/EtOAc (10–
40%) to give 8h (138 mg, 36%) as a white solid, mp 156–
157 °C. IR (CCl4, cmÀ1) mmax: 3200, 1706, 1682. 1H NMR
(400 MHz, CDCl3): d 1.13 (9H, s), 1.75–1.91 (2H, m),
2.48–2.54 (3H, m), 2.74 (1H, dd, 2J = 16.4 Hz,
3J = 6.0 Hz), 2.95–3.00 (1H, m), 6.82–6.90 (3H, m), 9.74
Acknowledgements
One of us (G.B.) gratefully acknowledges generous
financial support from the DGA.

7506
G. Binot, S. Z. Zard / Tetrahedron Letters 46 (2005) 7503–7506
1
(1H, br s). 13C NMR (100 MHz, CDCl3): d 26.4, 27.4,
(d, JCF = 241 Hz), 171.4, 215.1. MS (CI, m/z) 295
2
33.0, 35.3, 35.6, 44.1, 114.5 (d, JCF = 23 Hz), 114.7 (d,
[MNH4]+, 278 [M]+. Anal. Calcd for C16H20FNO2: C,
69.29; H, 7.27. Found: C, 69.06; H, 7.24.
3
2JCF = 23 Hz), 117.1, 128.6 (d, JCF = 7 Hz), 132.6, 158.6