Radical cyclizations to quinolone and isoquinolone systems under oxidative and reductive conditions

Article abstract of DOI:10.1016/j.tetlet.2004.01.123

Radical cyclizations to quinolone and isoquinolone systems under Fenton-type and n-Bu₃SnH-mediated conditions are described. For N-iodoalkylquinolones, ca. 3:1 mixtures of oxidative cyclization products at C-2, and unexpectedly at C-8, were obtained under both conditions. Five- or six-membered oxidative cyclization products were obtained from N-iodoalkylisoquinolones under Fenton-type conditions, whereas n-Bu ₃SnH-mediated reactions gave products of reductive cyclization in the five, six, and seven-membered series.

Full text of DOI:10.1016/j.tetlet.2004.01.123

Tetrahedron
Letters
Tetrahedron Letters 45 (2004) 2855–2858
Radical cyclizations to quinolone and isoquinolone systems under
oxidative and reductive conditions
Yazmin M. Osornio,a,* Luis D. Miranda,a Raymundo Cruz-Almanzaa,z and
Joseph M. Muchowskib,*
a
ꢀ
ꢀ
ꢀ
Instituto de Quımica, Universidad Nacional Autonoma de Mexico, Circuito Exterior, Ciudad Universitaria,
Coyoacꢀan 04510 Mꢀexico, DF, Mexico
bChemistry, Roche Palo Alto, 3431 Hillview Avenue, Palo Alto, CA 94304-1320, USA
Received 18 December 2003; revised 23 January 2004; accepted 26 January 2004
Abstract—Radical cyclizations to quinolone and isoquinolone systems under Fenton-type and n-Bu3SnH-mediated conditions are
described. For N-iodoalkylquinolones, ca. 3:1 mixtures of oxidative cyclization products at C-2, and unexpectedly at C-8, were
obtained under both conditions. Five- or six-membered oxidative cyclization products were obtained from N-iodoalkylisoquinol-
ones under Fenton-type conditions, whereas n-Bu3SnH-mediated reactions gave products of reductive cyclization in the five, six, and
seven-membered series.
ꢀ 2004 Elsevier Ltd. All rights reserved.
Heterocyclic moieties are an exceedingly frequently
encountered structural feature of pharmacologically
important compounds, and new synthetic routes to such
substances, including radical based methods,1 have
received much attention. The intramolecular addition of
carbon-centered radicals to an heteroaromatic nucleus
followed by an in situ oxidative restoration of the dou-
ble bond (oxidative radical cyclization), is a process,
which has emerged as a powerful synthetic tool within
this area.2–4 Several methods have been devised to effect
such radical cyclizations, with n-Bu3SnH-mediated
reactions being the most explored.2 The highly toxic
nature of the tin reagent, and the difficulties associated
with the removal of its reaction products, have led
synthetic organic chemists to find alternative methods to
accomplish such reactions.3;4 Some time ago, an efficient
tin-free oxidative radical cyclization of (x-iodoalkyl)-
indoles and pyrroles using Fenton-type conditions was
described,5a and more recently5e it was used in a tandem
radical addition–oxidative cyclization sequence based
on N-(2-iodoethyl)indoles and pyrroles. Implicit in this
methodology is the resolution of the problem of the
oxidative re-aromatization step, which arises in tin
hydride mediated reactions.5 The importance of the
benzoindolizidine and benzoquinolizidine alkaloids6 led
us to consider the extension of this process to the radical
cyclization of primary and secondary N-iodoalkyl-
quinolones and isoquinolones. In this article we present
some of our recent results on the cyclization of these
compounds using the oxidative Fenton-type and the
reductive n-Bu3SnH-mediated conditions.
The radical precursors 2a–d were synthesized in two
steps by alkylation of the N-unsubstituted quinolones
1a–c with the appropiate a,x-dibromoalkane and
subsequent halogen exchange with an excess of sodium
iodide in acetonitrile (Scheme 1). The quinolones 1b
(R ¼ CO2Et) and 1c (R ¼ CN) were prepared by a
thermal Gould–Jacobs cyclization of the corresponding
anilinomethylenemalonates.7 Preliminary attempts to
effect the desired radical cyclization were carried out by
dropwise addition of 30% hydrogen peroxide (10 equiv)
to a sonicated8 solution of 2a in DMSO containing
3 equiv of heptahydrated ferrous sulfate. The mildly
exothermic reaction was easily controlled at ca. 40 ꢁC b y
the rate of the peroxide addition (ꢀ0.5 h). It was
observed that after complete addition of hydrogen per-
oxide, some starting material remained and addition of
further quantities of peroxide and/or Fe(II) failed to
effect its consumption. Consequently, the desired prod-
uct 3a was isolated in a very low yield (Table 1, entry 1)
Keywords: Cyclization; Radicals; Isoquinolone; Quinolone; Fenton-
type conditions.
* Corresponding authors. Tel.: +52-55-56-22-4440; fax: +52-55-56-16-
2217; e-mail: osornio_yaz@correo.unam.mx
z
Deceased.
0040-4039/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2004.01.123

2856
Y. M. Osornio et al. / Tetrahedron Letters 45 (2004) 2855–2858
O
O
O
N
O
O
R
R
I
R
R
R
path a
i, ii
N
N
N
N
H
I
n
2
6
5
FeIII
2a n = 1, R = H
2b n = 1, R = CO2Et (69%)
2c n = 1, R = CN (64%)
2d n = 2, R = CO2Et (75%)
(57%)
1a R=H
1b R= CO2Et
1c R=CN
Me
path b
FeII
O
N
O
O
FeII FeIII
R
R
R
O
O
R
R
H
N
N
+
H
see
table 1
N
N
9
7
8
n
n
-H
-H
3a-d
4a-d
4
3
Scheme 1.
Reagents
and
conditions:
(i)
BrCH2CH2(CH2)nBr; (ii) NaI, CH3CN, reflux, 24 h.
NaH,
DMF,
Scheme 2. Proposed mechanism for the oxidative radical cyclizations.
adding small incremental amounts of AIBN until
1.1 equiv had been reached. At this stage, the starting
material had largely disappeared, and the above men-
tioned tricyclic products were formed in remarkably
high yields (entries 6–8). That products of oxidative
radical cyclization can form using n-Bu3SnH is no
longer unusual.2 The large amounts of AIBN, which are
required, however, for the reaction to proceed efficiently
is consistent with this reagent functioning as the oxidant
of the intermediate radical species 6 and 8 (Scheme 2).10
Table 1. Radical cyclization
Entry Substrate
n
R
Cond.
Product,
yield (%)
1
2
3
4
5
6
7
8
2a
2b
2c
2d
2b
2b
2c
2d
1
1
1
2
1
1
1
2
H
A
A
A
A
B
3a (5)
4a (0)
CO2Et
CN
3b (48) 4b (15)
3c (56) 4c (23)
3d (53) 4d (0)
3b (26) 4b (9)
3b (60) 4b (24)
3c (65) 4c (22)
3d (90) 4d (0)
CO2Et
CO2Et
CO2Et
CN
C
C
C
CO2Et
To extend the study to alkyl radical cyclizations in the
isoquinolone series, the required halides 14b–e were
prepared following the same protocol as for compounds
2a–d, but the secondary halo derivative 14a had to be
prepared by the reaction sequence shown in Scheme 3.
When the Fenton-type conditions were applied to
halides 14a–c, the oxidative radical cyclization products
benzo[1,2-b]indolizidinones 15a and 15b, and the benzo-
[1,2-b]quinolizidinone 15c were produced in modest
yields (Scheme 4), and considerable amounts of the
starting materials were recovered. Neither the primary
alkyl iodide 14d nor the secondary alkyl iodide 14e gave
the anticipated products. The failure to observe the
formation of the seven-membered product 14d is not
surprising, but the negative result with 14e was unan-
ticipated, especially since oxidative radical annulation of
14a to 15a took place reasonably well.
Conditions: (A) FeSO4Æ7H2O, H2O2, DMSO, ))))); (B) nBu3SnH
(1.2 equiv), AIBN (0.4 equiv), benzene, reflux; (C) nBu3SnH
(1.2 equiv), AIBN (1.1 equiv), benzene, reflux.
along with considerable starting material (ꢀ20%). In the
light of this result, we turned our attention to the more
electrophilic quinolones 2b–d (Scheme 1) wherein the
strongly electron attracting CO2Et (rm ¼ 0:37) and CN
(rm ¼ 0:56) groups were expected to decrease the
SOMO–LUMO energy difference at C-3 and favor
radical attack at this site relative to that found for 2a.
Indeed, when 2b and 2c were subjected to the above
conditions, not only were the expected pyrrolo[1,2-
a]quinolines 3b and 3c formed in quite acceptable yields
(entries 2 and 3), significant amounts of the unexpected
benzo[ij]quinolines 4b and 4c, derived from cyclization
at C-8, were obtained as well. The observed preference
for cyclization at C-2 (path a, Scheme 2) over C-8 (path
b) is no doubt a consequence of the greater energetic
cost of breaking resonance in the benzenoid ring. In
contrast, under the same conditions 2d gave the tricyclic
compound 3d as the sole product (entry 4). The failure
to observe any of the seven-membered product 4d
stemming from cyclization at C-8 may well be due to an
unfavorable entropic effect.9
H
N
N
O
O
O
X
O
N
i
ii
93%
97%
10
11
12 R=OH
13 R=OSO2CH3
14a R=Br
The same pattern of product formation was observed
when 2b–d were reacted with n-Bu3SnH/AIBN in benz-
ene solution at reflux temperature, but low product
yields and incomplete consumption of the starting
materials were observed using catalytic amounts of
AIBN (e.g., entry 5). This situation was rectified by
iii, 91%
iv, 89%
Scheme 3. Reagents and conditions: (i) methyl vinyl ketone, C6H6,
120 ꢁC; (ii) NaBH4, CH3OH; (iii) MeSO2Cl, Et3N; (iv) LiBr, THF,
0 ꢁC.

Y. M. Osornio et al. / Tetrahedron Letters 45 (2004) 2855–2858
2857
Acknowledgements
X
R
R
i
N
N
We thank DGAPA (PAPIIT-IN205901) for generous
ꢀ
n
n
O
O
~
finantial support. Also we thank R. Patino, J. Perez, L.
Velasco, H. Rios, N. Zavala, E. Huerta and A. Pena, for
14a n=2, R=CH3, X=Br
14b n=1, R=H, X=Br
14c n=2, R=H, X=I
14d n=3, R=H, X=I
14e n=2, R=CH3, X=I
15a n=2, R=CH3 41% (59%)
15b n=1, R=H, 28% (41%)
15c n=2, R=H, 36% (73%)
~
technical support.
15d n=3, R=H,
0%
15e n=2, R=CH3, 0%
References and notes
ii
R
R
1. Bowman, W. R.; Bridge, C. F.; Brookes, P. J. Chem. Soc.,
Perkin Trans. 1 2000, 1–14.
+
N
N
n
n
2. (a) Studer, A. In Radicals in Organic Synthesis; Renaud,
P., Sibi, M., Eds.; Wiley VCH: Weinheim, 2001; Vol. 2, pp
62–76; (b) Murphy, J. A.; Sherburn, M. S. Tetrahedron
1991, 47, 4077; (c) Suzuki, F.; Kuroda, K. J. Heterocycl.
Chem. 1993, 30, 811; (d) Antonio, Y.; de la Cruz, E.;
Galezzi, E.; Guzman, A.; Bray, B. L.; Greenhouse, R.;
Kurz, L. J.; Lustig, D. A.; Maddox, M. L.; Muchowski, J.
M. Can. J. Chem. 1994, 72, 15; (e) Tim, C. T.; Jones, K.;
Wilkinson, J. Tetrahedron Lett. 1995, 36, 6743; (f) Osaki,
S.; Mitoh, H.; Ohmori, H. Chem. Pharm. Bull. 1996, 44,
2020; (g) Dobbs, P. A.; Jones, K.; Veal, K. T. Tetrahedron
1997, 53, 8287; (h) Aldabbagh, F.; Bowman, W. R.;
Mann, E. Tetrahedron Lett. 1997, 38, 7937; (i) Hagan, D.
O
O
16b n=1, R=H,
16c n=2, R=H, 77%
16d n=3, R=H, 51%
16e n=2, R=CH3, 0%
88%
17b n=1, R=H,
17c n=2, R=H,
17d n=3, R=H,
7%
16%
38%
17e n=2, R=CH3, 46%
Scheme 4. Reagents and conditions: (i) FeSO4Æ7H2O, H2O2, DMSO,
)))); (ii) nBu3SnH (1.2 equiv), AIBN (0.4 equiv), benzene, reflux.
When compounds 14b–d were subjected to standard
n-Bu3SnH/AIBN reaction conditions, the reductive
cyclization products 16b–d were obtained in good yields
with even the seven-membered compound 16d being
formed quite efficiently (Scheme 4).11 Significant
amounts of the reductively dehalogenated compounds
17b–d were formed in each case, the amount thereof
increasing in parallel with the increasing ring size of the
cyclization products 16b–d. These reactions were all ef-
fected using catalytic AIBN; even large concentrations
of this reagent did not divert the reactions away from
the observed reductive cyclization products. Finally,
both of the secondary halides 14a and 14e failed to give
reductive cyclization products. Only the reductive
dehalogenation product 17e was obtained from 14e,
while 14a underwent cyclization to the isoquinolinium
salt 18, the structure of which was established by X-ray
crystallography (Fig. 1, Br omitted for simplification
purposes). Interestingly, 14a did undergo cyclization
under Fenton-type conditions, presumably at least in
part, because of the lower reaction temperature.
ꢀ
J.; Gimenez-Arnau, E.; Schwalbe, C. H.; Stevens, M. F. G.
J. Chem. Soc., Perkin Trans. 1 1997, 2739; (j) Moody, C.
J.; Norton, C. L. J. Chem. Soc., Perkin Trans. 1 1997,
2639; (k) Rosa, A. M.; Lobo, A. M.; Branco, P. S.;
Prabhakar, S.; Pereira, A. M. D. L. Tetrahedron 1997, 53,
269; (l) Harrowen, D.; Nunn, M. I. T. Tetrahedron Lett.
1998, 39, 5875; (m) Aldabbagh, F.; Bowman, W. R.;
Mann, E.; Slawin, A. M. Z. Tetrahedron 1999, 55, 8111;
(n) Nadin, A.; Harrison, T. Tetrahedron Lett. 1999, 4073;
ꢀ
ꢀ
(o) Marco-Contelles, J.; Rodrıguez-Fernandez, M. Tetra-
hedron Lett. 2000, 41, 381; (p) Bowmann, W. R.; Mann, E.
J. Chem. Soc., Perkin Trans. 1 2000, 2991; (q) Allin, S. M.;
Barton, W. R. S.; Bowman, W. R.; McInally, T. Tetra-
hedron Lett. 2002, 43, 4191.
3. (a) Gagosz, F.; Zard, S. Z. Org. Lett. 2002, 4, 4345; (b)
Axon, J.; Boiteau, L.; Boivin, J.; Forbes, J. E.; Zard, S. Z.
Tetrahedron Lett. 1994, 35, 1719; (c) Liard, A.; Quiclet-
Sire, B.; Saicic, R. N.; Zard, S. Z. Tetrahedron Lett. 1997,
38, 1759; (d) Cholleton, N.; Zard, S. Z. Tetrahedron Lett.
1998, 39, 7295; (e) Ly, T. M.; Quiclet-Sire, B.; Sortais, B.;
Zard, S. Z. Tetrahedron Lett. 1999, 40, 2533.
4. (a) Mohan, R.; Kates, S. A.; Dombroski, M. A.; Snider, B.
B. Tetrahedron Lett. 1987, 28, 845; (b) Snider, B. B.;
Buckman, B. O. Tetrahedron 1989, 45, 6969, and refer-
ences cited therein; (c) Aidhen, I. S.; Narasimhan, N. S.
Tetrahedron Lett. 1989, 30, 5323; (d) Artis, D. R.; Cho,
I.-S.; Muchoswki, J. M. Can. J. Chem. 1992, 70, 1838.
5. (a) Artis, D. R.; Cho, I.-S.; Figueroa, S. J.; Muchoswki, J.
M. J. Org. Chem. 1994, 59, 2456; (b) Bertilsson, B.-M.;
Gustafsson, B.; Kuhn, I.; Torsell, K. Acta Chem. Scand.
1970, 24, 3590; (c) Minisci, F.; Vismara, E.; Fontana, F. J.
Org. Chem. 1992, 57, 6817; (d) Bacciochi, E.; Muraglia, E.;
Sleiter, G. J. Org. Chem. 1992, 57, 6817; (e) Miranda, L.
In closing, radical cyclizations to quinolone and isoqui-
nolone systems under both Fenton-type and n-Bu3SnH-
mediated conditions are described. When successful,
N-haloalkylquinolones gave products of oxidative
cyclization under both conditions, whereas N-halo-
alkylisoquinolones afforded oxidative cyclization prod-
ucts under Fenton-type conditions and reductive
cyclization products under n-Bu3SnH/AIBN mediated
conditions.
ꢀ
D.; Cruz-Almanza, R.; Pavon, M. ARKIVOC 2002, 15.
6. For extensive lead references to benzoindolizidine and
benzoquinolizidine alkaloids, see: The Alkaloids, A Spe-
cialist Periodical Report; The Royal Society of Chemistry:
London; Vol. 1-13.
7. Gould, R. G., Jr.; Jacobs, W. A. J. Am. Chem. Soc. 1939,
61, 2890–2895.
8. Branson Models B-2200R-B ultrasonic bath type cleaners
were used.
Figure 1. ORTEP drawing of the isoquinolinium salt 18.

2858
Y. M. Osornio et al. / Tetrahedron Letters 45 (2004) 2855–2858
9. Kaoudi, T.; Quiclet-Sire, B.; Seguin, S.; Zard, S. Z. Angew.
1H); 13C (75 MHz, CDCl3) d ppm 23.6, 24.8, 33.2, 34.7,
43.6, 55.1, 126.8, 127.5, 128.4, 130.2, 131.6, 136.6, 165.1;
HRMS (FAB+): calcd for C13H16NO: 202.1232, found:
202.1239. 16d as a white solid, mp 77–78 ꢁC, IR (KBr,
cmꢁ1): 2957, 1646, 1627, 1597; 1H NMR (300 MHz,
CDCl3) d ppm 1.53–1.75 (m, 8H), 2.78 (dd, J ¼ 4:5,
15.6 Hz, 1H), 2.98 (ddd, J ¼ 4:5, 9.2, 14.0 Hz, 1H), 3.24
(dd, J ¼ 5:6, 15.6 Hz, 1H), 3.74–3.82 (m, 1H), 4.53 (dt,
J ¼ 5:6, 14.0 Hz, 1H), 7.14 (d, J ¼ 7:5 Hz, 1H), 7.29–7.42
(m, 2H), 8.06 (d, J ¼ 7:5 Hz, 1H); 13C (75 MHz, CDCl3) d
ppm 26.1, 26.4, 28.0, 33.6, 34.1, 46.1, 57.2, 126.8, 127.5,
128.4, 129.1, 131.4, 136.6, 163.7; HRMS (EI+): calcd for
C14H17NO: 215.1310, found: 215.1324. 18 as a white solid,
mp 120–122 ꢁC; IR (KBr, cmꢁ1): 2984; 1648, 1583; 1H
NMR (300 MHz, CDCl3) d ppm 1.73 (d, J ¼ 6:6 Hz, 3H),
2.24–2.38 (m, 1H), 2.70–2.78 (m, 1H), 5.06 (ddd, J ¼ 2:7,
5.1, 14.1 Hz, 1H), 5.29 (ddd, J ¼ 5:1, 11.5, 14.1 Hz, 1H),
5.43–5.54 (m, 1H), 7.69 (d, J ¼ 7:0 Hz, 1H), 7.76–7.81 (m,
1H), 7.96–7.98 (m, 2H), 8.34–8.39 (m, 2H); 13C (75 MHz,
CDCl3) d ppm 20.3, 27.0, 50.3, 78.1, 117.3, 119.6,
125.5, 127.4, 129.8, 132.3, 135.7, 137.7, 158.1; HRMS
(FAB+) calcd for C13H15NOBr: 280.0337, found:
280.0342.
Chem., Int. Ed. 2000, 39, 731–733, These authors highlight
the difficulty of construction of seven-membered rings
adjoining aromatic rings by radical chemistry.
10. (a) Curran, D. P.; Liu, H. J. Chem. Soc., Perkin Trans. 1
1994, 1337; (b) Josien, H.; Ko, S.-B.; Bom, D.; Curran, D.
P. Chem. Eur. J. 1998, 4, 70, and references cited therein.
11. Typical experimental procedure: n-Bu3SnH-mediated con-
ditions: A solution of AIBN (0.24 mmol, 0.4 equiv),
Bu3SnH (0.72 mmol, 1.2 equiv) in benzene (3.5 mL, 5 mL/
mmol) was added dropwise (syringe pump) to a degassed
solution of the halide (0.6 mmol, 1 equiv) in refluxing
benzene (0.02 M) over 6 h. The reaction mixture was then
cooled and the solvent removed under reduced pressure.
The residue was partitioned between hexane (10 mL) and
acetonitrile (5 mL). The polar layer was washed with
hexane (4 · 10). The solvent was evaporated and the
residue was purified by column chromatography on silica
gel (hexane/EtOAc). Selected spectral data: 16c as a white
solid, mp 95–97 ꢁC (lit.12 92–93 ꢁC), IR (KBr, cmꢁ1): 2943,
1635, 1604, 1581; 1H NMR (300 MHz, CDCl3) d ppm
1.45–1.56 (m, 3H), 1.77–1.87 (m, 3H), 2.65–2.74 (m, 1H),
2.83 (dd, J ¼ 9:5, 16.0 Hz, 1H), 3.06 (dd, J ¼ 5:5, 16.0 Hz,
1H), 3.53–3.59 (m, 1H), 4.70–4.73 (m, 1H), 7.31 (d,
J ¼ 7:5 Hz, 1H), 7.38–7.41 (m, 2H), 8.11 (d, J ¼ 7:5 Hz,
12. Couture, A.; Deniau, E.; Grandclaudon, P.; Lebrun, S.
Tetrahedron Lett. 1996, 37, 7749.