A new efficient route for the formation of quinoxaline N-oxides and N,N′-dioxides using HOF·CH3CN

Article abstract of DOI:10.1021/jo0608016

HOF·CH₃CN, a very efficient oxygen-transfer agent made readily from fluorine and aqueous acetonitrile, was reacted with various quinoxaline derivatives to give the corresponding mono N-oxides and especially the N,N′-dioxides in very good yields under mild conditions and short reaction times.

Full text of DOI:10.1021/jo0608016

A New Efficient Route for the Formation of Quinoxaline N-Oxides
and N,N′-Dioxides Using HOF‚CH3CN
Mira Carmeli and Shlomo Rozen*
School of Chemistry, Tel-AViV UniVersity, Tel-AViV 69978, Israel
rozens@post.tau.ac.il
ReceiVed April 16, 2006
HOF‚CH3CN, a very efficient oxygen-transfer agent made readily from fluorine and aqueous acetonitrile,
was reacted with various quinoxaline derivatives to give the corresponding mono N-oxides and especially
the N,N′-dioxides in very good yields under mild conditions and short reaction times.
SCHEME 1. Antibacterial Quinoxaline N,N′-Dioxides
Introduction
Quinoxaline N,N′-dioxides and phenazine 5,10-N,N′-dioxides
have found application as antibacterial agents, possessing a wide
range of activities.1 Methyl 3-[(2-quinoxalinyl)methylene]car-
bazate N,N′-dioxide (1), for example, known commercially as
Carbadox,2 has proven to be an efficient antibacterial and
growth-promoting material. Another example is a series of
penicillin derivatives of quinoxaline N,N′-dioxide carboxylic
acids such as 2,3 which exhibit exceptional activity against
Salmonella and Proteus species (Scheme 1).4
One of the most common methods for generating some
quinoxaline mono- and N,N′-dioxides consists in the total
synthesis based on the reaction of carbonyl derivatives with
benzofurazan oxide.5 An alternative, more general, and poten-
tially easier pathway should be the direct transfer of oxygen to
the nitrogen atoms of quinoxaline derivatives in order to obtain
the corresponding mono N-oxides and particularly quinoxaline
N,N′-dioxides, which are hardly reported in the literature. The
main reason for this “anomaly” is the second nitrogen of the
ring system, which makes the diazines less reactive than pyridine
toward electrophilic substitutions, including electrophilic oxygen
transfer. Electron-withdrawing substituents, such as halogens,
reduce the basicity of the ring nitrogens even further. Conse-
quently, while it was difficult to make haloquinoxaline mono-
N-oxides, it was practically impossible to form haloquinoxaline
N,N′-dioxides by the orthodox oxygen transfer agents.6 The
reason for this failure is the first step that produces the mono-
N-oxide moiety, which in turn deactivates almost completely
the second nitrogen atom toward any further electrophilic attack.
We report here a general method for direct oxidation of
quinoxaline derivatives using the HOF‚CH3CN complex. This
oxygen transfer agent stands alone in its ability to oxidize azides
and vicinal diamines into the corresponding nitro7 and dinitro8
derivatives. It was employed in the oxidation of CdN-containing
compounds,9 forming 1,10-phenanthroline N,N′-dioxide deriva-
(1) Haddadin, M. J.; Issidorides, C. H. Heterocycles 1976, 4, 767.
(2) (a) Chas Pfizer and Co. U.S. Patent 1968, 3 371 090. (b) Research
Corp. U.S. Patent 1968, 3 398 141. (c) Haddadin, M. J.; Kattan, A. M. A.;
Issidorides, C. H. J. Org. Chem. 1985, 50, 129.
(3) Edwards, M. L.; Bambury, R. E.; Ritter, H. W. J. Med. Chem. 1976,
19, 330.
(4) Kim, H. K. U.S. Patent 1972, 3 644 363.
(5) (a) Issidorides, C. H.; Haddadin, M. J. J. Org. Chem. 1966, 31, 4067.
(b) Kluge, A. F.; Maddox, M. L.; Lewis, G. S. J. Org. Chem. 1980, 45,
1909.
(6) Katritzky, A. R.; Lagowski, J. M. Chemistry of the Heterocyclic
N-Oxides; Academic Press: New York, 1971; Chapter 2. (b) Cheeseman,
G. W. H.; Werstiuk, E. S. G. AdV. Heterocycl. Chem. 1972, 14, 99.
(7) Rozen, S.; Carmeli, M. J. Am. Chem. Soc. 2003, 125, 8118-8119.
(8) Golan, E.; Rozen, S. J. Org. Chem. 2003, 68, 9170-9172.
(9) Carmeli, M.; Rozen, S. Tetrahedron Lett. 2006, 47, 763.
10.1021/jo0608016 CCC: $33.50 © 2006 American Chemical Society
Published on Web 06/29/2006
J. Org. Chem. 2006, 71, 5761-5765
5761

Carmeli and Rozen
SCHEME 3. Oxidation of 2-Chloroquinoxaline Using
SCHEME 2. Oxygen Transfer to Some Quinoxaline
Derivatives Using HOF‚CH3CN
Various Popular Oxidation Reagents
reaction mixture did not change the outcome. It seems that this
result is related to the modest steric hindrance in an indirect
way. Since the oxidation at the N-4 position is slower than that
of N-1, the small amount of protons originating from the HF
(formed simultaneously along the HOF) compete with the
electrophilic oxygen and protonate the hindered nitrogen atom
N-4. This protonation does not allow, of course, any further
electrophilic oxygen transfer at that position. However, since
the separation of 4d and 5d (after the appropriate workup that
regenerates the free base) is easy, the latter can be brought once
again in contact with HOF‚CH3CN forming 4d in quantitative
yield, but in 32% conversion. This process can be repeated until
the original 40% yield of 4d could be increased to higher than
90%.
Steric hindrance at the 2 and 3 positions did not cause any
problem. Oxidation of 2,3-diphenylquinoxaline (3e) using 4
equiv of the oxidation agent for a few seconds resulted in a
35% yield of the 2,3-diphenylquinoxaline N-oxide (5e) and 65%
yield of 2,3-diphenylquinoxaline N,N′-dioxide (4e). For com-
parison, oxidation of 3e with either m-CPBA for 18 h or
hydrogen peroxide in acetic acid resulted in 0.6% yield of the
N,N′-dioxide (4e).17,18 It should be noted that here again,
separating 4e from 5e makes it possible to repeat the reaction
with 5e resulting in 4e in higher than 95% yield (75%
conversion). This brings the potential overall yield of 4e to
nearly quantitative.
Benzylic halogens, which are excellent synthetic entries, are
also suitable substrates. Less than 1 min was needed for an
excess of HOF‚CH3CN to transfer two oxygen atoms to 2,3-
bis(bromoethyl)quinoxaline (3f) forming the corresponding 2,3-
bis(bromoethyl)quinoxaline N,N′-dioxide (4f) in 90% yield
(Scheme 2). To the best of our knowledge, no direct oxidation
of 3f has been achieved before.
Using HOF‚CH3CN for oxidation of 2-chloroquinoxaline (6)
produced the unknown 2-chloroquinoxaline N,N′-dioxide (9) in
60% yield along with 30% of 2-chloroquinoxaline N-oxide (8),
whose regioselectivity was controlled by the relative basicity
of the quinoxaline nitrogens. Scheme 3 compares reactions of
several peroxy reagents which are capable of transferring an
oxygen atom to 6 (forming either 7 or 8)19 with HOF‚CH3CN.
tives which eluded chemists for so many decades,10 transferring
oxygen atoms to sulfides, including electron-depleted ones,11
to thiophenes,12 polythiophenes,13 and much more.14
Results and Discussion
Direct oxidation of quinoxaline with H2O2 is reported in the
literature, but the yields never exceeded 50% and long reaction
times were required.15 It took less than 2 min for the HOF‚
CH3CN complex to convert quinoxaline (3a) to quinoxaline
N,N′-dioxide (4a) in 95% yield (Scheme 2). Although HOF‚
CH3CN is known to react even with aromatic rings,16 its reaction
with both nitrogen atoms is faster and no difficulties were
encountered with the more electron-rich aromatic ring of
2-methylquinoxaline (3b). This compound was transformed in
2 min to 2-methylquinoxaline N,N′-dioxide (4b) in 90% yield.
It is noteworthy that compound 4b has been previously obtained
via the benzofurazane route by a multistep synthesis eventually
resulting in less than 20% yield.5b Increased substitution on the
quinoxaline skeleton was also tolerated, and when 2,3-dimeth-
ylquinoxaline (3c) was reacted with 2 equiv of HOF‚CH3CN,
2,3-dimethylquinoxaline N,N′-dioxide (4c) was obtained after
3 min in 95% yield.
It was of interest to see whether the HOF‚CH3CN could
overcome steric hindrances, and 5-methylquinioxaline (3d)
served as a test case. Using an excess of the oxidation agent
provided a mixture of only 5-methylquinoxaline 1-N-oxide (5d)
in 50% yield and 5-methylquinioxaline N,N′-dioxide (4d) in
40% yield. Adding more than 5 equiv of the reagent to the
(10) (a) Rozen, S.; Dayan, S. Angew. Chem., Int. Ed. 1999, 38, 3471.
(b) Carmeli, M.; Rozen, S. J. Org. Chem. 2005, 70, 2131.
(11) (a) Rozen, S.; Bareket, Y. J. Org. Chem. 1997, 62, 1457. (b) Toyota,
A.; Ono, Y.; Chiba, J.; Sugihara, T.; Kaneko, C. Chem. Pharm. Bull. 1996,
44, 703.
(12) Rozen, S.; Bareket, Y. J. Chem. Soc., Chem. Commun. 1994, 1959.
(13) Amir. E.; Rozen, S. Angew. Chem., Int. Ed. 2005, 44, 7374.
(14) Rozen, S. Eur. J. Org. Chem. 2005, 2433.
(15) (a) Cooper, S. M.; Heaney, H.; Newbold, J. A.; Sanderson, W. R.
Synlett 1990, 9, 533. (b) Kobayashi, Y.; Kumadaki. I.; Sato, H.; Sekine,
Y.; Hara, T. Chem. Pharm. Bull. 1974, 22, 2097.
(17) Nasielski, J.; Heilporn, S.; Nasielski-hinkens, R.; Geerts-Evrard, F.
Tetrahedron 1987, 43, 4338.
(18) Landouist, J. K.; Stacey, G. J. J. Chem. Soc. 1953, 282.
(19) Mixan, C. E.; Pews, R. G. J. Org. Chem. 1977, 42, 1869.
(16) Kol, M.; Rozen, S. J. Org. Chem. 1993, 58, 1593.
5762 J. Org. Chem., Vol. 71, No. 15, 2006

Formation of Quinoxaline N-Oxides and N,N′-Dioxides
SCHEME 4. Oxidation of 2,3-Dichloroquinoxaline Using
Various Oxidation Agents
SCHEME 6. Oxidation of Phenazine Using HOF‚CH3CN
SCHEME 7. Oxidation of 6,7-Dimethyl-2,3-di(2-pyridyl)-
quinoxaline
SCHEME 5. Oxidation of Tetrachloroquinoxaline
SCHEME 8. Synthesis of Quinoxaline Di-N-oxide with
[18]O Isotope
None of the former was able to produce the N,N′-dioxide 9. As
in the cases of 5d and 5e, adding large excess of the hypo-
fluorous acid complex to 6 did not increase the 9:8 ratio
apparently because both steric and electronic effects prioritize
the protonation vs the oxygenation on N-1. During the workup,
however, the N-1 in 8 regained its basicity, and since the
separation between 9 and 8 is easy, the mono N-oxide 8 could
be subjected once again to a very short reaction with HOF‚
CH3CN increasing the overall yield of 9 to higher than 80%.
Dihalogenated quinoxalines are notoriously difficult to oxidize
by direct oxidation, and the yields in the few successful cases
are usually lower than 5%.20 Oxidation of the electron-poor 2,3-
dichloroquinoxaline (10) was attempted by several oxidation
agents,19 but once again only the HOF‚CH3CN complex was
able to produce the desired N,N′-dioxide 12 in 45% yield and,
as in the previous examples, after deprotonation of 11 and a
second short reaction with HOF‚CH3CN, the overall yield of
12 could be increased to 65% (Scheme 4).
Polyhalogenated diazines were practically impossible to be
converted to N,N′-dioxides. Systems such as peroxydichloro-
maleic acid,21 90% H2O2,22 or 60% H2O223 were found to form
exclusively the corresponding mono-N-oxides. Oxidation of
2,3,6,7-tetrachloroquinoxaline (13) was difficult even for the
HOF‚CH3CN complex. It took a large excess of the oxidizing
agent and 40 min to produce the two unknown derivatives
2,3,6,7-tetrachloroquinoxaline N-oxide (14) in 55% yield and
2,3,6,7-tetrachloroquinoxaline N,N′-dioxide in 40% yield (15)
(Scheme 5).
oxide with various polluting heavy metal catalysts.15b,24 The
reaction with HOF‚CH3CN proceeds smoothly without any
presence of catalysts and in 90% yield (Scheme 6).
Oxidation of 6,7-dimethyl-2,3-di(2-pyridyl)quinoxaline 18
presented an additional challenge for the HOF‚CH3CN complex.
This compound has never been fully oxidized previously.
Reacting 18 at room temperature with an excess of 6 molar
equiv of the reagent produced the 6,7-dimethyl-2,3-di(2-pyridyl)-
quinoxaline N,N′,N′′,N′′′-tetraoxide (19) in 80% yield after only
a few seconds (Scheme 7).
It was of interest to determine the oxygenation order of all
of the nitrogen atoms in 18. To answer this question, this
compound was reacted with 1 equiv of HOF‚CH3CN. It was
easy to conclude that one of the pyridine rings was first
oxygenated forming 20, followed by the oxygenation of the
nitrogen atom in the second pyridine ring, 21, as another molar
equivalent of the reagent was added. Continuing addition of
another 1.5 equiv of the acetonitrile complex of the hypofluorous
acid resulted in another oxygen transfer, this time to one of the
quinoxaline’s nitrogens forming 22, and finally, the last 2.5
equiv of the oxidizing agent afforded the tetra-N-oxide (19).
In many cases, especially when dealing with potential drugs,
it is essential to have information on the metabolism of the
studied compound. One of the techniques toward this end is to
label the relevant molecule, or part of it, with some uncommon
isotopes. One of the advantages of the HOF‚CH3CN complex
is that its electrophilic oxygen originates from water, which is
the best source for all oxygen isotopes. We passed fluorine
through a solution of acetonitrile and H218O, obtained as H18-
Phenazine (16) was also reacted with the HOF‚CH3CN
complex. The corresponding phenazine 5,10-N,N′-dioxide (17)
was obtained in a few seconds using a 2-fold excess of the oxi-
dizing agent. There are some reported methods for obtaining
17 using tert-amyl hydroperoxide, peracids, or hydrogen per-
(20) Palamidessi, G.; Bernardi, L.; Farmaco, A. L. Ed. Sci. 1966, 21,
805.
(21) Pollak, A.; Zupan, M.; Sket, B. Synthesis 1973, 495.
(22) Chivers, G. E.; Suschitzky, H. Chem. Commun. 1971, 28.
(23) Kyriacou, D. J. Heterocycl. Chem. 1971, 8, 697.
(24) Tolstikov, G. A.; Jemilev, U. M.; Jurjev, V. P.; Gershanov, F. B.;
Rafikov, S. R. Tetrahedron. Lett. 1971, 30, 2807.
J. Org. Chem, Vol. 71, No. 15, 2006 5763

Carmeli and Rozen
8.36 (1 H, d, J ) 1.8 Hz) 8.44 (1 H, dd, J1 ) 4.6 Hz, J2 ) 0.8 Hz)
8.68 (1 H, d, J ) 1.8 Hz); 13C NMR 19.7, 118.7, 130.9, 131.8,
133.8, 139.7, 140.8, 146.4, 147.2; MS (CI) m/z ) 377.1 (M + 1)+.
Anal. Calcd for C9H8N2O: C, 67.49; H, 5.03; N, 17.49. Found:
C, 67.20; H, 5.19; N, 17.34.
OF‚CH3CN, which was then reacted with 3a. The product’s
HRMS (CI) m/z ) 167.06193 (M + 1) (calcd for C8H6N218O2,
167.059242) clearly showed that both oxygen atoms of 4a-[18]-
O are the 18O isotope (Scheme 8).
The N,N′-dioxide 4d was obtained in 40% yield as a yellow
solid: mp ) 174-177 °C; UV-vis (CHCl3) λmax 240 (ꢀ ) 1.5 ×
104), 375 (ꢀ ) 0.9 × 104), 390 (ꢀ ) 1 × 104); 1H NMR 3.10 (3 H,
s), 7.55 (1 H, d, J ) 3.8 Hz), 7.69 (1 H, dd, J1 ) 3.8 Hz, J2 ) 3.8
Hz), 8.11 (1 H, d, J ) 2.8 Hz) 8.16 (1 H, d, J ) 2.8 Hz) 8.50 (1
H, d, J ) 3.8 Hz); 13C NMR 25.6, 120.7, 131.9, 133.4, 133.5, 136.7,
136.9, 139.9, 142; HRMS (CI) m/z ) calcd 177.066403, found
177.066494 (M + 1)+. Anal. Calcd for C9H8N2O2: C, 61.36; H,
4.58; N, 15.90. Found: C, 61.16; H, 4.85; N, 15.49. Compound
4d could also be prepared by a similar procedure reacting 14 equiv
of HOF‚CH3CN with 5d forming 4d in quantitative yield, but with
32% conversion raising the yield of this N,N′-dioxide to nearly 60%.
2,3-Diphenylquinoxaline N-oxide (5e) and 2,3-diphenylqui-
noxaline N,N′-dioxide (4e) were obtained from 3e (0.6 g, 2.1
mmol), as described above, using 4 equiv of the oxidizing agent.
The mixture was separated and 5e obtained in 35% yield as a white
solid: mp ) 191 °C; UV-vis (CHCl3) λmax 249 (ꢀ ) 2.6 × 104),
Conclusion
HOF‚CH3CN is unique in its ability to oxidize quinoxaline
derivatives, including sterically hindered and electron-depleted
ones to the corresponding mono-N-oxides and N,N′-dioxides
and if needed to transfer the former to the latter. Expanding
this family of compounds and simplifying their synthesis may
provide an easy tool for synthesis of new antibiotics and other
antibacterial agents. Considering the commercial availability of
premixed fluorine/nitrogen mixtures and the technical ease of
the reaction, chemists should be encouraged to take advantage
of the unique synthetic abilities of HOF‚CH3CN complex as a
powerful oxygen transfer agent in organic chemistry.
Experimental Section
1
270 (ꢀ ) 2.5 × 104), 334 (ꢀ ) 0.9 × 104); H NMR 7.41-7.22
1H NMR and 13C NMR were obtained at 400 and 50 MHz,
respectively, with CDCl3 as a solvent and Me4Si as an internal
standard. MS spectra were measured under CI, EI, or FAB
conditions. UV spectra were recorded in CHCl3.
General Procedure for Working with Fluorine. This element
is a strong oxidant and very corrosive material. It should be used
only with an appropriate vacuum line.25 For the occasional user,
however, various premixed mixtures of F2 in inert gases are
commercially available, simplifying the process. If elementary
precautions are taken, work with fluorine is relatively simple and
we had no bad experiences working with it.
General Procedure for Producing HOF‚CH3CN. Mixtures of
10-20% F2 with nitrogen were used in this work. The gas mixture
was prepared in a secondary container before the reaction was
started. It was then passed at a rate of about 400 mL per minute
through a cold (-15 °C) mixture of 100 mL of CH3CN and 10 mL
of H2O in regular glass reactor. The development of the oxidizing
power was monitored by reacting aliquots with an acidic aqueous
solution of KI. The liberated iodine was then titrated with
thiosulfate. Typical concentrations of the oxidizing reagent were
around 0.4-0.6 mol/L.
General Procedure for Working with HOF‚CH3CN. A
quinoxaline derivative (usually 0.4-0.6 g) was dissolved in about
30 mL of CHCl3, and the mixture was cooled to 0 °C. The oxidizing
agent was then added in one portion to the reaction vessel (1 equiv
means 1 molar equiv per each nitrogen). The excess of HOF‚CH3-
CN was quenched with saturated sodium bicarbonate, the solution
extracted with CHCl3, the organic layer dried over MgSO4, and
the solvent evaporated. The crude product was usually purified by
vacuum flash chromatography using silica gel 60-H and increasing
portions of EtOAc in PE as an eluent. Similar chromatography was
used for separation of the mono- and N,N′-dioxides when obtained
as a mixture. The spectral and physical properties of the known
products thus obtained were compared with those reported in the
literature. In every case, excellent agreement was obtained. All
known compounds were referenced throughout this work. Data for
the new compounds, or for those not well defined in the literature,
is given below.
(10 H, m), 7.74 (1 H, ddd, J1 ) 8 Hz, J2 ) 7.1 Hz, J3 ) 1.6 Hz),
7.84 (1 H, ddd, J1 ) 8 Hz, J2 ) 7.1 Hz, J3 ) 1.6 Hz), 8.2 (1 H,
ddd, J1 ) 8 Hz, J2 ) 1.6 Hz, J3 ) 0.8 Hz), 8.65 (1 H, ddd, J1 )
8, Hz J2 ) 1.6 Hz, J3 ) 0.8 Hz); 13C NMR 119.3, 127.9, 128.2,
128.9, 129.2, 129.5, 129.9, 130, 130.6, 131.4, 135.9, 137.9, 140,
143.8, 156.2; HRMS (CI) m/z ) calcd 299.118438, found 299.117583
(M + 1)+.
The N,N′-dioxide 4e was obtained in 65% yield as a yellow
solid: mp ) 212-215 °C; UV-vis (CHCl3) λmax 240 (ꢀ ) 2.2 ×
1
104), 290 (ꢀ ) 1.9 × 104), 395 (ꢀ ) 1.0 × 104); H NMR 7.31-
7.21 (10 H, m), 7.9 (2 H, dd, J1 ) 6.7 Hz, J2 ) 3.2 Hz), 8.7 (2 H,
dd, J1 ) 6.7 Hz, J2 ) 3.2 Hz); 13C NMR 119.1, 122.4, 126.6, 127.9,
128.7, 129.8, 130.1, 131, 132, 133.5, 137.8, 142.7; MS (FAB) m/z
) 315 (M + 1)+. Anal. Calcd for C20H14N2O2: C, 76.42; H, 4.49;
N, 8.91. Found: C, 76.63; H, 4.58; N, 8.74. Compound 4e could
also be prepared by a similar procedure from 5e in quantitative
yield, but with 75% conversion using 12 equiv of the reagent.
2-Chloroquinoxaline N,N′-dioxide (9) was prepared from 6 (0.4
g, 2.43 mmol), as described above, using 2 equiv of the oxidizing
agent resulting in the 65% yield of a yellow solid: mp ) 198-
199 °C; UV-vis (CHCl3) λmax 240 (ꢀ ) 1.3 × 104), 266 (ꢀ ) 1.8
1
× 104), 380 (ꢀ ) 0.98 × 104), 390 (ꢀ ) 0.99 × 104); H NMR
7.87-7.95 (2 H, m), 8.47 (1 H, s), 8.58 (1 H, d, J ) 4.2 Hz), 8.65
(1 H, d, J ) 4.2 Hz); 13C NMR 122.4, 122.5, 132.5, 133.8, 134.8,
136, 139.4, 139.9; MS (CI) m/z ) 197 (M + 1)+. Anal. Calcd for
C8H5N2O2Cl: C, 48.88; H, 2.56; N, 14.25; Cl, 18.03. Found: C,
48.54; H, 2.79; N, 13.94; Cl, 18.23.
2,3-Dichloroquinoxaline mono-N-oxide (11) and 2,3-dichlo-
roquinoxaline N,N′-dioxide (12) were prepared from 10 (0.5 g,
2.51 mmol), as described above, using 3 equiv of the oxidizing
agent. The mixture was separated by chromatography to give 11
in 50% yield as a white solid: mp ) 140-142 °C; UV-vis (CHCl3)
1
λmax 250 (ꢀ ) 3.5 × 104), 325 (ꢀ ) 0.7 × 104); H NMR 7.75-
7.90 (2 H, m), 8.06 (1 H, ddd, J1 ) 7.5 Hz, J2 ) 1.6 Hz, J3 ) 1.2
Hz), 8.52 (1 H, ddd, J1 ) 7.5 Hz, J2 ) 1.6 Hz, J3 ) 1.2 Hz); 13C
NMR 118.9, 129.2, 130.7, 132.2, 136.9, 140.5, 146.4; MS (CI) m/z
) 232 (M + 1)+. Anal. Calcd for C8H4N2Cl2O: C, 44.68; H, 1.87;
N, 13.03; Cl, 32.97. Found: C, 44.50; H, 2.01; N, 12.85; Cl, 33.30.
The second component proved to be 12 obtained in 50% yield as
a yellow solid: mp ) 230-233 °C; UV-vis (CHCl3) λmax 240 (ꢀ
) 2.2 × 104), 270 (ꢀ ) 3.5 × 104), 370 (ꢀ ) 1.3 × 104), 385 (ꢀ )
1.5 × 104); 1H NMR 7.92 (2 H, dd, J1 ) 3.3 Hz, J2 ) 1.7 Hz) 8.65
(2 H, dd, J1 ) 3.2 Hz, J2 ) 1.8 Hz); 13C NMR 120.3, 132.4, 135.4,
136.2; MS (CI) m/z ) 231 (M)+. Anal. Calcd for C8H4N2Cl2O2:
C, 41.59; H, 1.75; N, 12.13; Cl, 30.69. Found: C, 41.43; H, 1.86;
N, 11.95; Cl, 30.21. Compound 12 could also be prepared by a
5-Methylquinoxaline N-oxide (5d) and 5-methylquinoxaline
N,N′-dioxide (4d) were obtained from 3d (0.62 g, 4.3 mmol), as
described above, using 2.5 equiv of the oxidizing agent. The mixture
was separated, and 5d was obtained in 50% yield as a white solid:
mp ) 123-126 °C; UV-vis (CHCl3) λmax 240 (ꢀ ) 3.8 × 104),
330 (ꢀ ) 0.8 × 104); 1H NMR 2.80 (3 H, s), 7.62-7.68 (2 H, m),
(25) Dayan, S.; Kol, M.; Rozen, S. Synthesis 1999, 1427.
5764 J. Org. Chem., Vol. 71, No. 15, 2006

Formation of Quinoxaline N-Oxides and N,N′-Dioxides
similar procedure from 11 in quantitative yield, but with 40%
conversion using 8 molar equiv of the reagent bringing the overall
yield of 12 to 65%.
6,7-Dimethyl-2,3-di(2-pyridyl)quinoxaline tetra-N-oxides (19)
was prepared from 18 (0.6 g, 1.92 mmol), using altogether 6 equiv
of the oxidizing agent as described above, resulting in 80% yield
of a yellow solid which proved to be 19: mp ) 274 °C; UV-vis
(CHCl3) λmax 240 (ꢀ ) 4.9 × 104), 270 (ꢀ ) 4.3 × 104), 375 (ꢀ )
2,3,6,7-Tetrachloroquinoxaline N-oxide (14) and 2,3,6,7-
tetrachloroquinoxaline N,N′-dioxide (15) were obtained as a
mixture from 13 (0.5 g, 1.87 mmol) as described above, using 7
equiv of the oxidizing agent. After separation by chromatography,
compound 14 was obtained in 55% yield as a yellow solid: mp )
184 °C; UV-vis (CHCl3) λmax 260 (ꢀ ) 4.0 × 104), 330 (ꢀ ) 0.7
× 104), 365 (ꢀ ) 0.5 × 104); 1H NMR 8.17 (1 H, s), 8.63 (1 H, s);
13C NMR 120.2, 128.7, 130, 135.5, 136.3, 137.8, 139.2, 147.9; MS
(CI) m/z ) 285 (M + 1)+. Anal. Calcd for C8H2N2Cl4O: C, 33.84;
H, 0.71; N, 9.87; Cl, 49.95. Found: C, 34.21; H, 1.10; N, 9.78; Cl,
50.36. The second fraction proved to be the N,N′-dioxide 15
obtained in 40% yield as a yellow solid: mp ) 188-189 °C; UV-
vis (CHCl3) λmax 245 (ꢀ ) 1.4 × 104), 260 (ꢀ ) 1.2 × 104), 285 (ꢀ
1
1.3 × 104), 395 (ꢀ ) 1.4 × 104); H NMR 2.56 (6 H, s), 7.16-
7.36 (4 H, m), 7.63 (2 H, dd, J1 ) 7.5 Hz, J2 ) 1.7 Hz), 8.24 (2 H,
dd, J1 ) 6.4 Hz, J1 ) 0.4 Hz), 8.42 (2 H, s); 13C NMR 20.4, 119.8,
125.4, 127.6, 129.1, 134.8, 136.7, 139.4, 139.9, 143.9 ppm; MS
(FAB) m/z ) 377 (M + 1)+. Anal. Calcd. for C20H16N4O4: C,
63.83; H, 4.28; N, 14.89. Found: C, 63.83; H, 4.60; N, 14.51. As
described in the Results and Discussion, it was possible to observe
the formation of the mono-, di-, and trioxides 20, 21, and 22 with
a recorded NMR spectrum of 21: 2.53 (6 H, s), 7.22-7.39 (2 H,
m), 7.85-7.91 (2 H, m), 7.95 (2 H, s), 8.01-8.07 (2 H, m), 8.44-
8.47 (2 H, m).
1
) 1.4 × 104), 400 (ꢀ ) 0.3 × 104); H NMR 8.74 (2 H, s); 13C
NMR 121.7, 117.6; MS (CI) m/z ) 301 (M + 1)+. Compound 15,
which could not be obtained in analytical purity, was also prepared
from 14 by a similar procedure using 8 equiv of the reagent in
quantitative yield, but with 10% conversion only.
Acknowledgment. This work was supported by the Israel
Science Foundation.
JO0608016
J. Org. Chem, Vol. 71, No. 15, 2006 5765