Unexpected reduction of ethyl 3-phenylquinoxaline-2- carboxylate 1,4-di-N-oxide derivatives by amines.

The unexpected tendency of amines and functionalized hydrazines to reduce ethyl 3-phenylquinoxaline-2-carboxylate 1,4-di-N-oxide (1) to afford a quinoxaline 1c and mono-oxide quinoxalines 1a and 1b is described. The experimental conditions were standardized to the use of two equivalents of amine in ethanol under reflux for two hours,with the aim of studying the distinct reductive profiles of the amines and the chemoselectivity of the process. With the exception of hydrazine hydrate, which reduced compound 1 to a 3-phenyl-2-quinoxalinecarbohydrazide derivative, the amines only acted as reducing agents.

Introduction

Quinoxaline and quinoxaline 1,4-di-N-oxide are heterocycles that are often used in the synthesis of biologically active compounds [1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31]. The former is described as a bioisoster of the quinoline, naphthyl, benzothienyl and other aromatic rings [32], and it can be found in the structure of anti-bacterial [1], anti-tuberculosis [5,6,7,9,10,11,13,14,16,22], anti-cancer [8,18,20,28], anti-malarial [12,24,26,29,30,31], anti-Chagas [15,25] and anti-inflammatory [19,27] drug candidates. The widespread activity of quinoxaline 1,4-di-N-oxides can be associated with the generation of free radicals [33]. In our continuing efforts to find quinoxaline-1,4-di-N-oxide derivatives with anti-malarial activity [12,23,24,26,29,30,31], the conversion of compound 1 into hydrazides 2 and amide derivatives 3 was carried out by means of hydrazinolysis and aminolysis reactions, in which unexpected results were obtained. In this work we describe the reducing profiles of amine derivatives when they act upon the ethyl 3-phenylquinoxaline-2-carboxylate 1,4-di-N-oxide derivative 1.

Results and Discussion

In a continuing effort to synthesize new anti-malarial drug candidates, we wished to substitute the carboethoxy moiety present in lead-compound 1, with a functionalized hydrazide subunit 2 (Scheme 1). For this purpose compound 1 was first treated with phenylhydrazine in the presence of ethanol under reflux for two hours. After workup, the crude oil was analyzed by 1H-NMR, which revealed the presence of four different carboethoxy group signals. This crude oil was purified by silica gel column chromatography and the separated products analyzed by 1H-NMR, IR, mass spectra and elemental analyses. From these analyses, it was observed that the reaction with phenylhydrazine had failed to produce the functionalized hydrazide derivative 2, but rather this reaction surprisingly gave a mixture of quinoxaline 1c, quinoxalines N-4 monoxide 1a, N-1 monoxide 1b and 1,4-di-N-oxide 1 (Table 1). While the reducing activity of hydrazine hydrate [34,35] is well known, the aforementioned information was not as clear for phenylhydrazine. In an attempt to determine if the reduction of 1 by phenylhydrazine could be influenced by the electronic profile of quinoxaline-ester 1, the derivatives 4-7, attached to electron-withdrawing and electron-donating groups, were treated with phenylhydrazine in ethanol under reflux for two hours; the results are given in Table 1 (entries 2, 3, 4 and 5). These results showed no selective reduction for substrates 5-7, with the exception of compound 4 (4’-nitro, σp = 0.81) that could be selectively reduced to quinoxaline 4c (entry 2). In this particular case, different reducing profiles were observed for phenylhydrazine and hydrazine hydrate because, with the latter no chemoselectivity was observed, and consequently, the nitro group was also reduced and the hydrazinolysis product was formed (data not shown). Intrigued by these results, the possibility that other amines could act as reducing agents towards the quinoxaline 1,4-di-N-oxide system was then studied. Derivative 1, used as template, was treated with different amines (two equivalents) in the presence of ethanol under reflux for two hours (Table 1). The results clearly demonstrated the ability of hydroxylamine (entry 9), methylamine (entry 10), ethanolamine (entry 11), methyl hydrazine (entry 6), and 2,4-dinitrophenylhydrazine (entry 7) to reduce compound 1 to the mono-oxide derivatives 1a and 1b and to quinoxaline 1c, while amines such as triethylamine (entry 12) and aniline (entry 13), were unable to reduce compound 1, as no significant difference was noted when compared to the experiment carried out in the absence of amine (entry 14).

Scheme 1

Design of new quinoxaline 1,4-di-N-oxide derivatives.

Table 1

Reduction of ethyl 3-phenylquinoxaline-2-carboxylate 1,4-di-N-oxide (1).

Entry	Reducing Agent	W	n=n1=1(1, 4-7)	n=0, n1=1(1b, 4-7a)	n=1, n1=0(1c, 4-7b)	n=n1=0(1a, 4-7c)
1	PhNHNH2	H	18.1%	30.1%	28.4%	23.4%
2	PhNHNH2	NO2	16.7%	0%	0%	83.3%
3	PhNHNH2	CF3	22.6%	30,0%	23.7%	23.7%
4	PhNHNH2	OCH3	25.7%	28.6%	17.6%	28.1%
5	PhNHNH2	CH3	15.1%	31.8%	21.9%	31.2%
6	H3CNHNH2	H	43.8%	21.8%	22.5%	11.9%
7	2,4-diNO2PhNHNH2	H	57.4%	8.9%	24.1%	9.6%
8	N2H4.H2O	H	0%	0%	0%	100%*
9	NH2OH.H2O	H	22.6%	8.1%	33.8%	35.5%
10	NH2CH3.H2O	H	33.8%	8.0%	49.7%	8.5%
11	NH2(CH2)2OH	H	0%	34.2%	36.0%	29.8%
12	Et3N	H	80.8%	0%	19.2%	0%
13	PhNH2	H	81.2%	0%	18.8%	0%
14	EtOH	H	85.5%	0%	14.5%	0%
15	P(OCH3)3	H	0%	100%	0%	0%
16	Na2S2O4	H	0%	0%	0%	100%

*formation of 3-phenyl-2-quinoxalinecarbohydrazide; absence of chemioselectivity.

The yields were determined by 1H-NMR analysis of total crude product mixtures obtained after the work-up of each reaction. The methyl (OCH2CH3) resonances for the four compounds 1, 1a, 1b and 1c, were observed at different chemical shifts, and thus, integration of the methyl region (OCH2CH3) allowed relative molar percentages to be readily ascertained. The unequivocal structural assignments of the N-4 oxide 1a, N-1 oxide 1b and quinoxaline 1c derivatives were accomplished by 1H-NMR, IR, mass spectra and elemental analyses. In an attempt to specifically identify mono-oxide quinoxaline derivatives (1a versus 1b), a selective mono-deoxygenation of compound 1 was performed (entry 14) using trimethylphosphite (entry 15), as previously described by Kluge and coworkers [36]. By this methodology, it was only possible to obtain the N-4 oxide derivative 1a, which was characterized by 1H-NMR; the data was used for direct comparison with the N-oxides obtained from the reaction with the amines. In a similar way, the quinoxaline di-N-oxide 1 was totally reduced to quinoxaline 1c, using Na2S2O4 (entry 16) in a mixture of methanol and water [37,24]; the chemical shift of compound 1c was measured by 1H-NMR and compared with the products obtained from amine reductions. Compounds 1a, 1b and 1c were evaluated for their ability to inhibit P. falciparum (chloroquine sensitive), and were shown to be inactive as anti-malarial agents (data not shown).

Conclusions

In summary, this work clearly demonstrates a tendency of amines and functionalized hydrazines to reduce ethyl 3-phenylquinoxaline-2-carboxylate 1,4-di-N-oxide (1) to a quinoxaline 1c and mono-oxides quinoxalines 1a and 1b. Although the starting material 1 was recovered in most of the reactions (entries 1-7, 9-10, 12-14), suggesting that a longer time reaction could be necessary, the experimental conditions were standardized in two hours, using two equivalents of amine, with the aim of studying the distinct reductive profile of the amine used and the chemoselectivity of the process. With the exception of hydrazine hydrate, a well-known reducing agent, which reduced compound 1 to a 3-phenyl-2-quinoxalinecarbohydrazide derivative, the amines were unable to act as nucleophiles and acted exclusively as reducing agents. Compounds 1a, 1b and 1c were inactive as antimalarial agents.

Experimental

General

All of the synthesized compounds were chemically characterized by thin layer chromatography (TLC), infrared (IR), nuclear magnetic resonance (1H-NMR), mass spectra (MS) and elemental microanalysis (CHN). Alugram SIL G/UV254 (0.2 mm layer, Macherey-Nagel GmbH & Co. KG. Düren, Germany) was used for TLC and Silica gel 60 (0.040-0.063 mm, Merck) for Flash Column Chromatography. 1H-NMR spectra were recorded on a Bruker 400 Ultrashield instrument (400 MHz), using TMS as the internal standard with CDCl3 as the solvent; the chemical shifts are reported in ppm (δ) and coupling constants (J) values are given in Hertz (Hz). Signal multiplicities are represented by: s (singlet), d (doublet), t (triplet), q (quadruplet), dd (double doublet) and m (multiplet). IR spectra were recorded on a Nicolet Nexus FTIR (Thermo, Madison, USA) in KBr pellets. Mass spectra were measured on a MSD/DS 5973N G2577A mass spectrometer (Agilent Technologies, Waldbronn, Germany) with a direct insertion probe (DIP). The ionization method was electron impact (EI, 70 eV). Elemental microanalyses were obtained on a Leco CHN-900 Elemental Analyzer (Leco, Tres Cantos, Spain) from vacuum-dried samples. The analytical results for C, H, and N, were within ± 0.4 of the theoretical values. Chemicals were purchased from Panreac Química S.A. (Barcelona, Spain), Sigma-Aldrich Química, S.A. (Alcobendas, Spain), Acros Organics (Janssen Pharmaceuticalaan, Geel, Belgium) and Lancaster (Bischheim-Strasbourg, France).

General procedure for the reduction of ethyl 3-phenylquinoxaline-2-carboxylate 1,4-di-N-oxide (

Ethyl 3-phenylquinoxaline-2-carboxylate 1,4-di-N-oxide (1, 1 mmol) was added to ethanol (10 mL) and the amine derivative (1 mmol). The mixture was refluxed for two hours, then the reactions were worked-up by adding CH2Cl2 (50 mL), followed by extraction with 10% aqueous HCl (4 x 15 mL). The organic layer was dried (Na2SO4), filtered, and evaporated to dryness. The yields were determined by 1H-NMR analysis of the total crude product mixture. The residue was later purified by silica gel column chromatography (n-hexane-ethyl acetate).

Ethyl 3-phenylquinoxaline-2-carboxylate 1,4-di-N-oxide (1). 1H-NMR: δ 1.08 (t, J= 7.2 Hz, 3H, OCH2CH3); 4.25 (t, J= 7.2 Hz, 2H, OCH2CH3); 7.53 (m, 3H, H3’-H5’); 7.61 (m, 2H, H2’ and H6’); 7.91 (m, 2H, H6 and H7); 8.65 (m, 2H, H5 and H8) ppm; 13C-NMR: δ 13.98 (OCH2CH3), 63.65 (OCH2CH3), 120.89 (C5), 121.08 (C8), 127.84 (C1’), 129.16 (C3’ and C5’), 130.15 (C2’ and C6’), 131.26 (C4’), 132.51 (C6), 132.53 (C7), 136.53 (C2), 137.73 (C10), 138.81 (C3), 140.08 (C9), 159.66 (CO2Et) ppm; IR: 2978 (ArC-H), 1746 (C=O), 1352 (N-oxide), 701 and 666 (monosubstituted phenyl) cm-1; MS: 310 (m/z, 100%), 294 (M+, 6%), 249 (M+, 51%), 221 (M+, 46%), 77 (M+, 46%); Anal. calcd. for C17H14N2O4: C, 65.80; H, 4.52; N, 9.03. Found: C, 65.65; H, 4.57; N, 8.98.

Ethyl 3-phenylquinoxaline-2-carboxylate 4-N-oxide (1a). 1H-NMR: δ 1.04 (t, J=7.2 Hz; 3H, OCH2CH3); 4.20 (q, J=7.2 Hz; 2H, OCH2CH3); 7.56 (m, 3H, H3’-H5’); 7.61 (m, 2H, H2’and H6’); 7.88 (m, 2H, H6 and H7); 8.26 (dd, J= 1.2, 8.2 Hz; 1H, H5); 8.65 (dd, J= 1.2, 7.6 Hz; 1H, H8); IR: 2981 (ArC-H), 1742 (C=O), 1359 (N-oxide), 701 and 666 (monosubstituted phenyl) cm-1; MS: 294 (m/z, 65%), 265 (M+, 13%), 249 (M+, 26%), 221 (M+, 100%), 77 (M+, 18%); Anal. calcd. for C17H14N2O3: C, 69.38; H, 4.79; N, 9.52. Found: C, 69.37; H, 4.77; N, 9.51.

Ethyl 3-phenylquinoxaline-2-carboxylate 1-N-oxide (1b). 1H-NMR: δ 1.25 (t, J=7.2 Hz; 3H, OCH2CH3); 4.42 (q, J=7.2 Hz; 2H, OCH2CH3); 7.54 (m, 3H, H3’-H5’); 7.80 (m, 3H, H2’and H6’ and H7); 7.89 (dt, J= 8.4, 7.6 Hz; 1H, H6); 8.20 (d, J= 8.4 Hz; 1H, H5); 8.61 (d, J= 8.0 Hz; 1H, H8); IR (KBr): 2979 (ArC-H), 1735 (C=O), 1363 (N-oxide), 701 and 671 (monosubstituted phenyl) cm-1; MS: 294 (m/z, 60%), 265 (M+, 9%), 249 (M+, 31%), 221 (M+, 100%), 77 (M+, 26%); Anal. calcd. for C17H14N2O3: C, 69.38; H, 4.79; N, 9.52. Found: C, 69.39; H, 4.80; N, 9.51.

Ethyl 3-phenylquinoxaline-2-carboxylate (1c). 1H-NMR: δ 1.19 (t, J=7.2 Hz; 3H, OCH2CH3); 4.34 (q, J=7.2 Hz; 2H, OCH2CH3); 7.53 (m, 3H, H3’-H5’); 7.76 (m, 2H, H2’and H6’); 7.85 (m, 2H, H6 and H7); 8.20 (d, J= 8.0 Hz; 1H, H5); 8.24 (d, J= 7.6 Hz; 1H, H8); IR: 2990 (ArC-H), 1730 (C=O), 711 and 669 (monosubstituted phenyl) cm-1; MS: 278 (m/z, 33%), 249 (M+, 26%), 234 (M+, 15%), 206 (M+, 100%), 77 (M+, 33%); Anal. calcd. for C17H14N2O2: C, 73.37; H, 5.07; N, 10.07. Found: C, 73.35; H, 5.08; N, 10.05.