A solid phase traceless synthesis of quinoxalinones

Article abstract of DOI:10.1016/S0040-4039(00)00297-5

A solid-phase traceless synthesis of quinoxalinones in three combinatorial steps is reported. An aldehyde functionalized polystyrene resin was reductively alkylated by amino acid methyl esters, and then the resin bound secondary amines were reacted with o

Full text of DOI:10.1016/S0040-4039(00)00297-5

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 41 (2000) 2835–2838
A solid phase traceless synthesis of quinoxalinones
∗
Viktor Krch nˇ ák, Lajos Szabo and Josef Vágner
SIDDCO, 9040 South Rita Road, Tucson, AZ 85747, USA
Received 22 December 1999; revised 15 February 2000; accepted 16 February 2000
Abstract
A solid-phase traceless synthesis of quinoxalinones in three combinatorial steps is reported. An aldehyde
functionalized polystyrene resin was reductively alkylated by amino acid methyl esters, and then the resin
bound secondary amines were reacted with o-fluoronitrobenzenes. The resulting o-nitroanilines were reduced by
tin chloride to the dianilines, which spontaneously cyclized. The amide nitrogen of the dihydroquinoxalinones
obtained was alkylated by alkyl halides. After cleavage from the resin, the dihydroquinoxalinones were air oxidized
to quinoxalinones. © 2000 Elsevier Science Ltd. All rights reserved.
1
We have recently described a solid-phase traceless synthesis of benzimidazoles. Our concept of
traceless synthesis of heteroaromatic compounds was based on the acid liability of N-arylbenzylamines.
An aldehyde linker was derivatized to form a resin bound N-alkyl-N-arylbenzylamine, allowing acidolytic
cleavage of the carbon–nitrogen bond, and leaving the nitrogen with only a hydrogen atom on the
target structure. In the synthesis of benzimidazoles we constructed a linear precursor on solid support
followed by off resin cyclization. In this letter we describe a traceless synthesis of quinoxalinones
based on the same concept. The combinatorial solid-phase synthesis of quinoxalinones has already
2
been described. However, these syntheses used linkers that leave a carboxylate functional group on
all target compounds. As reported here, complete tracelessness was achieved by off resin air oxidation of
dihydroquinoxalinones 6 to quinoxalinones 7, leaving no residual linkage functionality at all.
The traceless synthesis of quinoxalinones followed the route shown in Scheme 1. The synthesis
was developed on (4-(4-formyl-3-methoxyphenoxy)butyryl) AM resin (Novabiochem, Laufelfingen,
Switzerland). The aldehyde resin 1 was reductively alkylated with an amino acid ester using the standard
3
1
protocol developed by others and us. In order to evaluate the yield and purity of this initial step, we
reacted the secondary resin-bound amine 2 with Fmoc-chloroformate (Fmoc-Cl). The resin-bound Fmoc
protected amino acid ester was then cleaved by acid (TFA, gaseous HCl or HF) and analyzed.
The second combinatorial step involved nucleophilic fluorine displacement using o-fluoronitrobenzene
with the polymer-supported secondary amine 2. The purity and yield of the product o-nitroaniline
was evaluated using reaction with Fmoc-Cl and cleavage by TFA. In order to optimize the nucleo-
philic displacement, we tested a less sterically demanding linker (4-(4-formylphenoxy)butyric acid),
∗
Corresponding author. Tel: 520-663-4001; fax: 520-663-0795; e-mail: krchnak@siddco.com (V. Krch nˇ ák)
0
040-4039/00/$ - see front matter © 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(00)00297-5
tetl 16577

2
836
Scheme 1. Traceless synthesis of quinoxalinones. Reagents: (i) amino acid ester, NaB(AcO)
3
H, DMF/AcOH; (ii)
o-fluoronitrobenzene, DMSO, 75°C, 1–3 days, see Table 1; (iii) SnCl2·2H
2
O, NMP, rt, 2 h; (iv) 2-t-butyl-
imino-2-diethylamino-1,3-dimethylperhydro-1,3,2-diazaphosphorine (BEMP), alkyl halide, DMF, rt, 2 h; (v) TFA or gaseous
HCl or gaseous HF, rt, 2 h; (vi) air oxidation, MeOH, rt, overnight
several solid supports (polystyrene resin, Tentagel, Chiron’s crowns), and various solvents (DMF, N-
methylpyrrolidinone (NMP), DMSO, THF, alcohols, ionic liquid). We did not observe any substantial
differences on the rate of nucleophilic substitution with solid support or linker. The best conversion
was observed in DMSO at 75 to 90°C, with premature cleavage from the resin observed at higher
temperatures. Typical results are given in Table 1.
Table 1
Reaction of resin-bound amino acid esters with 2-fluoronitrobenzenes
The percent conversion varies with the α-amino acid side-chain and the additional substituents on the
o-fluoronitrobenzenes. Reactivity is affected by the reduced nucleophilicity of an amino group alpha to
the electron-withdrawing carboxylic ester as well as by steric factors. The yields with amino acids having
bulkier side-chains (Phe, Ile) or o-fluoronitrobenzenes with no additional strong electron withdrawing
substituents often was less than 10%. Because neither TFA nor HF cleaves the unarylated amino acid
ester 2 from the solid support, incomplete reaction reduces the overall yield but does not affect the purity
of the final product.
The nitro group of the resin-bound o-nitroaniline 3 was reduced by 2 M 99.99% grade tin(II)
chloride dihydrate solution in NMP degassed by bubbling argon, as recommended by Morales et
2
b
al. The reduction was complete in 2 h and the reduced intermediate spontaneously cyclized to
dihydroquinoxalinone 4. During the 2 h reduction of several dinitro substituted fluorobenzenes (2,4-
dinitrofluorobenzene and 1,5-difluoro-2,4-dinitrobenzene) only the nitro group ortho to the amino
substituent was reduced.
Because the number of commercially available o-nitrofluorobenzenes is limited, the diversity of
quinoxalinones was extended as shown in Scheme 2. We substituted the remaining fluorine or chlorine of
the resin-bound fluorodinitroaniline 8a (from 1,5-difluoro-2,4-dinitrobenzene) or dichloronitroaniline 8b

2
837
(from 1,2-dichloro-4-fluoro-5-nitrobenzene) by amines to give 9a or 9b, respectively. As expected, the
replacement of the chlorine of 8b was substantially slower than the fluorine of 8a and required elevated
temperature and longer reaction time.
Scheme 2. Extending the diversity of quinoxalinones. Reagents: (i) amine, DMF, rt, overnight; (ii) amine, DMSO, 75°C,
overnight; (iii) SnCl2·2H
2
O, NMP, rt, 2 h; (iv) alkyl halide, BEMP, DMF, rt, 2 h; (v) TFA or gaseous HF or HCl, rt, 2 h;
vi) air oxidation, MeOH, rt, overnight
(
4
Dihydroquinoxalinones 6 were cleaved from the resin by TFA or gaseous reagents HCl or HF. If
TFA was used, it was evaporated in a stream of nitrogen, and the product extracted into MeOH. LC/MS
1
analysis showed a molecular ion corresponding to the oxidized product quinoxalinone 7. The H NMR
6
spectrum was also consistent with a quinoxalinone structure. The oxidation of a dihydroquinoxalinone
to quinoxalinone has already been described by Lee et al.,2 however, the dihydroquinoxalinone was the
a
major product. In order to isolate the dihydroquinoxalinone, the product was cleaved by gaseous HCl and
1
extracted directly into DMSO-d under a nitrogen atmosphere. The H NMR spectrum was consistent
with the dihydroquinoxalinone.
6
5
The last combinatorial step involved alkylation of the amide nitrogen of 4 by an electrophile. Sev-
eral bases were tested and 2-t-butylimino-2-diethylamino-1,3-dimethylperhydro-1,3,2-diazaphosphorine
(
BEMP) in DMF provided the cleanest product, while KHMDS left some starting material (<10%). We
5
have synthesized numerous target quinoxalinones in manually operated Domino Blocks. Structures of
representative examples are shown in Fig. 1. The purity of the products, evaluated by analytical HPLC,
was consistently high and ranged from 84 to 99% (Table 2). Products were characterized by LC/MS
data (Table 2) and NMR.
7
8
5
In summary, we have developed a traceless solid-phase synthesis of quinoxalinones with three
combinatorial steps. The synthetic concept, based on the use of an acid labile N-arylbenzylamine
linker, allows for cleavage of a carbon–heteroatom bond followed by air oxidation of initially formed
dihydroquinoxalinones, yielding target compounds with no trace of the linker.
Acknowledgements
The authors wish to thank Mike Kobel for analytical assistance with the mass spectral measurements.

2
838
Fig. 1. Structure of representative quinoxalinones
Table 2
Purity and yield of traceless quinoxalinones
References
1
2
. Smith, J.; Krch nˇ ák, V. Tetrahedron Lett. 1999, 40, 7633–7636.
. (a) Lee, J.; Murray, W. V.; Rivero, R. A. J. Org. Chem. 1997, 62, 3874–3879; (b) Morales, G. A.; Corbett, J. W.; DeGrado,
W. F. J. Org. Chem. 1998, 63, 1172–1177; (c) Zaragoza, F.; Stephensen, H. J. Org. Chem. 1999, 64, 2555–2557.
. Swayze, E. E. Tetrahedron Lett. 1997, 38, 8465–8468 and references cited therein.
3
4
5
. Kerschen, A.; Kaniszai, A.; Botros, I.; Krch nˇ ák, V. J. Comb. Chem. 1999, 1, 480–484.
1
. H NMR spectrum (500 MHz, DMSO-d
6
) of compound 7a: δ 7.59 (s, 1H), 7.62 (d, 1H), 7.92 (d, 1H), 8.32 (s, 1H); compound
7
7
3
b: δ 2.65 (s, 3H), 8.16 (s, 1H), 8.39 (s, 1H); compound 7c: δ 4.01 (s, 2H), 6.76 (d, 1H), 7.12 (d, 1H), 7.20 (s, 1H); compound
d: δ 1.25 (d, 3H), 4.04 (q, 1H), 6.87 (s, 1H), 7.09 (s, 1H); compound 7f: δ 0.78 (d, 3H), 0.95 (d, 3H), 1.22 (d, 3H), 1.24 (d,
H), 3.65 (m, 1H), 3.88 (m, 1H), 7.38 (s, 1H), 7.60 (s, 1H); compound 7g: δ 2.97 (m, 4H), 3.74 (m, 4H), 7.33 (s, 1H), 7.49
(s, 1H), 8.15 (s, 1H); compound 7h: δ 5.22 (s, 2H), 5.71 (s, 1H), 5.95 (s, 1H), 7.72 (d, 1H), 7.79 (s, 1H), 8.06 (d, 1H), 8.45
(s, 1H); compound 7i: δ 2.21 (s, 3H), 5.47 (s, 2H), 7.11 (d, 2H), 7.15 (d, 2H), 7.65 (d, 1H), 7.70 (s, 1H), 8.04 (d, 1H), 8.46
(s, 1H); compound 7j: δ 3.85 (s, 3H), 5.48 (s, 2H), 6.87 (d, 2H), 7.24 (d, 2H), 7.65 (d, 1H), 7.79 (s, 1H), 8.05 (d, 1H), 8.48
(s, 1H) ppm.
6
7
. Krch nˇ ák, V.; Pad eˇ ra, V. Bioorg. Med. Chem. Lett. 1998, 8, 3261–3264.
. Analytical gradient HPLC profile was run on a ProC18 4.6×50 mm analytical column (YMC, Wilmington, NC), gradient
0
–70% of ACN in 7 min. The purity was estimated based on analytical traces at λ=280 nm.
8
. Lower yield of compound 7g was caused by partial cleavage during the tin(II) chloride reduction. The extent of cleavage
2
increased with increased electron donating effect of substituent R and was also observed e.g., when 2-fluoronitrobenzene,
3
-fluoro-4-nitrotoluene, 2,5-difluoronitrobenzene, or 1-bromo-4-fluoro-3-nitrobenzene were used in the second combinatorial
step. More acid-stable 4-(4-formylphenoxy)butyryl resin prevented premature loss of the product.