First synthesis of 2-phosphonylated quinoxaline 1,4-dioxides: An extension to the Beirut reaction

Article abstract of DOI:10.1016/j.tetlet.2010.07.148

An extension of the Beirut reaction for the preparation of the first members of the 2-phosphonylated quinoxaline 1,4-dioxide series is described. Contrary to their carboxylated equivalents, preparation of these new compounds could not be achieved under basic conditions but required the use of powdered molecular sieves. Good and reproducible yields were obtained only when the initial suspension in THF was transformed into a pasty film by slow evaporation of ca. 90% of the initial solvent volume.

Full text of DOI:10.1016/j.tetlet.2010.07.148

Tetrahedron Letters 51 (2010) 5516–5520
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
First synthesis of 2-phosphonylated quinoxaline 1,4-dioxides: an extension
to the Beirut reaction
⇑
Samir Dahbi, Ebtissem Methnani, Philippe Bisseret
Laboratoire de Chimie Organique et Bioorganique, ENSCMu, 3, rue Alfred Werner, 68093 Mulhouse Cedex, France
a r t i c l e i n f o
a b s t r a c t
Article history:
An extension of the Beirut reaction for the preparation of the first members of the 2-phosphonylated
quinoxaline 1,4-dioxide series is described. Contrary to their carboxylated equivalents, preparation of
these new compounds could not be achieved under basic conditions but required the use of powdered
molecular sieves. Good and reproducible yields were obtained only when the initial suspension in THF
was transformed into a pasty film by slow evaporation of ca. 90% of the initial solvent volume.
Ó 2010 Elsevier Ltd. All rights reserved.
Received 17 June 2010
Revised 26 July 2010
Accepted 26 July 2010
Available online 3 August 2010
Keywords:
Quinoxaline
Quinoxaline 1,4-dioxide
Phosphonate
Beirut reaction
Molecular sieves
1. Introduction
dimethyl phosphonate, freshly prepared from dimethylphosphite
by the action of NaH in THF, in four chloroquinoxalines 5–8. For
Among heteroaryls, quinoxaline 1,4-dioxides corresponding to
general scaffold 1 are endowed with many biological properties
that in particular largely encompass by their diversity the proper-
ties of the non-oxgenated quinoxalines.1 Indeed, already recog-
nized in the 1940’s for their antibacterial activities,2 many
quinoxaline dioxide derivatives have gained much attention for
their antimycobacterial, anticandida, antiprotozoal (especially
antimalaria) as well as anticancer properties.1,3 Most bioactive
members of this family possess either an oxo or a sulfonyl group
in position 2 as in the anticandida phenylsulfone 2,3a the antituber-
cular benzylcarboxylate, and benzylcarboxamide 33b,c or the re-
cently described tumor-specific propenoyl derivative 43d (Fig. 1).
Considering the importance of phosphonate derivatives in
medicinal chemistry as bioisosters of carboxylates and sulfones,4
it is surprising that no 2-phosphonylated quinoxaline 1,4-dioxides
have been reported so far. To our knowledge, as indicated in Figure
2, only a handful of phosphonates derived from quinoxalines have
been described so far.5
that purpose, chloroquinoxalines 6–8 were prepared simulta-
neously in excellent overall yield by oxidizing commercial chloro-
quinoxaline 5 using Caro’s acid in concentrated sulfuric acid
(Scheme 1).6 Under these conditions, we were pleased to isolate
the recently described dioxide derivative 8,7 albeit in low yield, be-
sides the expected pair of chloro monoxides 6 and 7.
This series of experiments is summarized in Scheme 2. After
exposure to an excess of sodium dimethyl phosphonate at room
temperature, quinoxalines 5, 7, and 8 did not yield significant
amounts of phosphonate derivatives. The parent quinoxaline 5
was recovered entirely whereas oxides 7 and 8 were simply con-
verted into 5 and 6, respectively, following a deoxygenation reac-
tion similar to the action of trimethylphosphite or phosphorus
trichloride on quinoxaline oxide derivatives, as already reported.8
The experiment with chloroquinoxaline monoxide 6 was more
gratifying, yielding dimethyl(quinoxalin-2-yl)phosphonate 10 in
good yield along with trace amounts of monoxide 9. Using a lower
temperature and/or only one equivalent of NaP(O)(OMe)2 the reac-
tion was very sluggish, giving in particular no better yield in phos-
phonate 9.9 Utilization of lithium dimethyl phosphonate, prepared
by deprotonation (LiHMDS) of dimethylphosphite,10 or dimethyl-
phosphite in the presence of K2CO3, Et3N, or CsF in place of sodium
dimethyl phosphonate also proved disappointing, giving at best a
low yield of 10.
We disclose herein our efforts to synthesize 2-phosphonylated
quinoxaline 1,4-dioxides.
2. Results and discussion
In order to prepare 2-phosphonylated quinoxaline 1,4-dioxides,
we first tested the nucleophilic substitutions of chlorine by sodium
We then examined the effect of replacing Clꢀ by other nucleo-
fuges: after having checked that 2-fluoroquinoxaline monoxide
was far too unstable for being used,11 we performed a few experi-
ments starting from 2-iodoquinoxaline monoxide, obtained in 85%
⇑
Corresponding author.
E-mail address: philippe.bisseret@uha.fr (P. Bisseret).
0040-4039/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2010.07.148

S. Dahbi et al. / Tetrahedron Letters 51 (2010) 5516–5520
5517
O
N
O
N
O
N
O
N
+
+
+
8
5
+
+
1
Cl
2
3
7
6
X
N
O
Cl
N
O
N
O
S
O
N
O
4
+
+
+
O
O
O
4
(X = O or NH)
3
2
1
Figure 1. The quinoxaline 1,4-dioxide scaffold 1 and examples of biologically relevant 2-substituted quinoxaline 1,4-dioxides.
O
HN
O
P
NH2
CO2H
OH
OH
OEt
OEt
OEt
OEt
N
N
R
Cl
Cl
N
N
N
N
CF3
N
N
N
N
OMe
OMe
OEt
OEt
OEt
OEt
P
O
EtO
EtO
P
P
O
P
O
P
O
O
(R = CH3, Ph,
NH2, NHtBu)
Figure 2. Quinoxaline-derived phosphonates described in the litterature.5
O
N
O
+
+
N
N
N
N
N
H2SO4, K2S2O8
20 ºC, 18 h
+
Cl
+
N
+
Cl
N
O
Cl
+
Cl
O
66%
17%
10%
8
5
6
7
Scheme 1. Synthesis of chloroquinoxaline oxides.
O
N
N
NaP(O)(OCH3)2
(2 equiv)
+
NaP(O)(OCH3)2
(2 equiv)
no reaction
THF, 20 ºC, 4 h
N
Cl
Cl
THF, 20 ºC, 18 h
N
Cl
Cl
85%
5
7
O
N
N
N
N
N
+
+
NaP(O)(OCH3)2
NaP(O)(OCH3)2
(2 equiv)
(2 equiv)
O
+
O
N
O
P
P
N
O
+
THF, 20 ºC, 4 h
80%
THF, 20 ºC, 4 h
OMe
+
OMe
OMe
N
O
OMe
60%
5%
10
9
6
8
Scheme 2. Action of sodium dimethyl phosphonate on chloroquinoxalines.
yield from 6 by the action of sodium iodide in the presence of HI.12
Action of NaP(O)(OMe)2 resulted mainly in deiodination, leading to
quinoxaline monoxide, whereas pallado-catalyzed Hirao cross-
coupling trials were disappointing, offering at best a sluggish con-
version into phosphonates 9 and 10.13
ide moiety,unfortunately no dioxide derivatives were obtained. We
decided to investigate at this stage a modification of the Beirut reac-
tion. This reaction which involves the condensation of a b-ketoester
on benzofuroxan under basic conditions has already been frequently
used in the synthesis of quinoxaline dioxide derivatives.14
We then investigated oxidizing reactions on phosphonate 10. As
presented in Table 1, by using conditions previously selected for
the preparation of the chloroquinoxalines 6–8, we could prepare
9 in 30% yield, while recovering the starting phosphonate also in
30% yield (entry 3). Forcing the experimental conditions, as well
as using only a slight excess of potassium persulfate proved unsat-
isfactory (entries 4 and 6) and adding methanol in the medium, in
the hope of esterifying any phosphonic acid derivatives that could
have been formed, yielded only a complex mixture (entry 7). Using
mCPBA (entry 1) or better freshly prepared peracetic acid (entry 2)
enabled us to prepare the isomeric monoxide 11 in very good yield.
Although this first set of experiments enabled us to prepare, for
the first time, two phosphonates attached to a quinoxaline monox-
Utilization of Ca(OH)2 or NaOEt as a base, as often recommended
for the Beirut reaction, unfortunately led to an unexploitable com-
plex mixture starting with dimethyl 2-oxopropylphosphonate. By
using either cesium carbonate or cesium fluoride as a base, we could
obtain the 2-methylquinoxaline dioxide 12 in good yield probably
via
a transient phosphonylated alcoholate, as indicated in
Scheme 3. No traces of phosphonylated quinoxalines were detected
under these conditions.
In the hope of trapping by protonation the transient alkoxide,
thus avoiding phosphonate elimination, we replaced the basic salt
and instead we used molecular sieves 3 Å.15 We were pleased to
isolate under these conditions the desired phosphonate 13a as a
yellow solid, along with trace amounts of 12 and the dimethyl-

5518
S. Dahbi et al. / Tetrahedron Letters 51 (2010) 5516–5520
Table 1
Oxidation experiments on dimethyl(quinoxalin-2-yl)phosphonate
O
N+
N
N
conditions
O
P
O
P
O
P
+
N+
O
OMe
OMe
OMe
OMe
OMe
N
N
OMe
10
9
11
Entry
Conditions
mCPBA/CHCl3/40 °C/3 d
AcOH/Ac2O/H2O2 (35%) 50 °C/20 h
H2SO4/K2S2O8 (3 equiv) 20 °C/16 h
H2SO4/K2S2O8 (5 equiv) 20 °C/16 h
H2SO4/K2S2O8 (3 equiv) 4 °C/3 d
10a (%)
9b (%)
11b (%)
1
2
3
4
5
6
7
30
12
30
nd
60
44
nd
ndc
2
30
10
15
6
30
86
nd
nd
nd
nd
nd
H2SO4/K2S2O8 (1.5 equiv) 20 °C/48 h
MeOH/H2SO4/K2S2O8 (5 equiv)/20 °C/16 h
nd
a
b
c
Recovered starting material.
Yield of isolated product.
Not detected.
Cs2CO3, THF
60 ºC, 2 d
60%
of a small amount of the dephosphorylated compound 12. Interest-
ingly, no traces of 14 were detected (entry 2). Completing the evap-
oration of the solvent to dryness resulted in a poorer yield in
phosphonate (entry 3). The best result was obtained in THF after
three days at 30 °C (entry 4), minimizing in particular the forma-
tion of the side product 12. We then tested a few solvents (entries
5–10) but could not reach the efficiency of experiments in THF. In
methanol, the reaction was particularly fast but gave 12 as a major
product which probably arose from the decomposition of phospho-
nate 13a by nucleophilic attack of MeOH on the phosphonate
moiety.16
O
N+
O
N+
O
O
P
OMe
OMe
N+
O
+
O
or
CsF, THF
60 ºC, 4 d
40%
N
12
O
O
N+
N+
O
O
OMe
OMe
N+
O
OMe
OMe
N+
O
P
O
P
O
We then investigated the reactivity of a few b-ketophospho-
nates under the optimized previous conditions (Table 3).
In most cases, the reaction proceeded very well, giving good
yields of quinoxaline dioxide derivatives.9,17 In the case of R = H,
a complex mixture was obtained after one day with no formation
of quinoxaline derivatives whereas for R = OCH3, the starting prod-
ucts were entirely recovered. In the case of R = Ph, it was better to
stop the reaction after 4 h, thus enabling the recovery of half of the
starting phosphonate: a longer reaction time did not afford a better
yield in the quinoxaline derivative 13g but resulted in lower yields
of recovery of the starting phosphonate.18
In summary, we have prepared the first quinoxaline dioxide-de-
rived phosphonates. This offers stimulating perspectives in medic-
inal chemistry, taking into account the variety of biological
properties attributed to quinoxaline dioxide derivatives. Although
our attempts to prepare them by oxidation of quinoxaline or quin-
oxaline monoxide precursors failed, we developed with success an
extension of the Beirut reaction by using molecular sieves instead
of the much more classically used basic conditions. In order to
reach reproducible and good yields in phosphonylated quinoxaline
dioxides, the reacting medium had to be confined to a pasty film
obtained by slow evaporation of 90% of the initial suspension into
THF. We are currently exploring the chemistry of these new
derivatives.
Scheme 3. First tentatives of Beirut reaction with dimethyl 2-oxopropylphos-
phonate.
phosphate 14 resulting from the incorporation by an as yet un-
known mechanism of one molecule of THF (Scheme 4).
On repeating this experiment, we encountered difficulties to get
reproducible yields (see Scheme 4). We examined several reaction
conditions including varying the temperature, the nature of the
molecular sieves (3–5 Å, in powder or in beads, used as such or
preactivated) as well as using acetonitrile as a solvent. None of
these conditions improved the reproducibility of the reaction and
the yield of 13a. The reaction mixture being heterogenous, we sus-
pected that stirring could be responsible at least for the poor repro-
ducibility of these experiments. We then performed trials without
stirring, confining the reacting medium to a pasty film after evap-
oration of ca. 90% of the initial solvent volume containing the ben-
zofuroxan, dimethyl 2-oxopropylphosphonate, and molecular
sieves in powder. As shown in Table 2, these changes proved very
fruitful. We first investigated experiments in a film of THF at 50 °C
during one day, checking before all the decisive role of the molec-
ular sieves (entry 1). In the presence of molecular sieves, a very
good yield of phosphonate 13a was observed with the formation
O
N+
O
O
O
N+
N+
N+
O
O
P
MS 3Å, THF
60 °C, 2 d
OMe
OMe
+
O
+
+
OMe
OMe
OMe
OMe
N+
O
P
O
O
N+
O
N
O
N
P
O
(5-20%)
14
(10%)
12
13a (20-40%)
Scheme 4. First successful Beirut reaction with dimethyl 2-oxopropylphosphonate.

S. Dahbi et al. / Tetrahedron Letters 51 (2010) 5516–5520
5519
Table 2
Optimization studies for the synthesis of phosphonate 13a
O
O
O
N+
N+
N+
O
O
P
conditionsa
OMe
OMe
O
+
OMe
+
N+
O
P
O
N+
O
N
OMe
13a
12
Entry
Solvent
T (°C)/time (days)
Conversionb (%)
Yieldb (%) 13a/12
1c
2
THF
THF
THF
THF
50/1
50/1
50/1
30/3
30/2
40/1
30/2
50/1
50/1
0
0/0
83
68
83
45
37
84.5
82
47
77/6
58/10
79/4
36/9
6/31
2.5/82
68/14
38/9
3d
4
5
6
7
9
CH3CN
CH3CN/H2O (95:5, v/v)
CH3OH
Toluene
1,4-Dioxane
10
a
b
c
Conditions: benzofuroxan (0.65 mmol), phosphonate (0.3 mmol), preactivated molecular sieves (0.3 g); solvent (2 mL) is added and concentrated to ca. 0.2 mL.
Based on NMR analysis of the crude mixture.
No molecular sieves were added.
d
The medium was evaporated to dryness.
4. For a recent review, see: Palacios, F.; Alonso, C.; De Los Santos, J. M. Chem. Rev.
Table 3
2005, 105, 899–932.
Syntheses of 3-substituted-1,4-dioxoquinoxaline-2-dialkylphosphonatesa
5. (a) Aparicio, D.; Attanasi, O. A.; Filippone, P.; Ignacio, R.; Lillini, S.; Mantellini,
F.; Palacios, F.; De Ios Santos, J. M. J. Org. Chem. 2006, 71, 5897–5905; (b) Adam,
M. S. S.; Kindermann, M. K.; Köckerling, M.; Heinicke, J. W. Eur. J. Org. Chem.
2009, 27, 4655–4665; (c) Teng, M.; Johnson, M. D.; Thomas, C.; Kiel, D.; Lakis, J.
N.; Kercher, T.; Aytes, S.; Kostrowicki, J.; Bhumralkar, D.; Truesdale, L.; May, J.;
Sidelman, U.; Kodra, J. T.; Jorgensen, A. S.; Olesen, P. H.; Cornelis de Jong, J.;
Madsen, P.; Behrens, C.; Pettersson, I.; Knudsen, L. B.; Holst, J. J.; Lau, J. Biorg.
Med. Chem. Lett. 2007, 17, 5472–5478; (d) Jirkovsky, I. L.; Baudy, R. B.;
Greenblatt, L. P. U.S. Patent 51,18,675, 1992; (e) Acklin, P.; Allgeier, H.;
Auberson, Y. P.; Bischoff, S.; Ofner, S.; Sauer, D.; Schmutz, M. Bioorg. Med. Chem.
Lett. 1998, 8, 493–498.
O
N
O
+
+
O
O
P
R
P
OR'
OR'
MS 3Å
N
+
OR'
O
O
+
R
THF, 30ºC
N
N
O
OR'
15a-g
13a-g
b-Ketoester
R
R0
Time (day)
Yieldb (%)
6. Mixan, C. E.; Pews, R. G. J. Org. Chem. 1977, 42, 1869–1871.
7. Carmeli, M.; Rozen, S. J. Org. Chem. 2006, 71, 5761–5765.
8. (a) Dirlam, J. P.; McFarland, J. W. J. Org. Chem. 1977, 42, 1360–1364; (b) El-
Abadelah, M. M.; Sabri, S. S.; Tashtoush, H. I. Tetrahedron 1979, 35, 2571–2576;
(c) El-Abadelah, M. M.; Anani, A. A.; Khan, Z. H.; Hassan, A. M.; Katrib, A. J.
Heterocycl. Chem. 1980, 17, 1671.
15a
15b
15c
15d
15e
15f
CH3
CH3
C2H5
C5H11
H
CH3
3
3
3
5
1
1
0.25
79 (70)
85 (75)
73 (63)
68 (43)
0c
C2H5
C2H5
CH3
C2H5
CH3
OCH3
Ph
0d
9. Selected analytical data:
15g
CH3
45 (42)e
2-(Dimethyl)phosphonyl-quinoxaline 10:
1H NMR (CDCl3, 300 MHz) d 9.26 (s, 1H), 8.16 (dd, 1H, J = 2.1 and 7.8 Hz), 8.07
(dd, 1H, J = 2.1 and 7.8 Hz), 7.8 (m, 2H), 3.93 (d, 6H, J = 11.1 Hz); 13C NMR
a
Conditions: see Table 2.
Yield based on NMR analysis of the crude mixture (isolated yield after
chromatography).
b
(CDCl3, 75 MHz)
d 146.5 (d, J = 222.7 Hz), 145.9 (d, J = 27.7 Hz), 142.8 (d,
J = 2.5 Hz), 142.0 (d, J = 21.3 Hz), 132.0, 130.7, 129.9, 129.3, 53.7 (d, J = 6.1 Hz);
31P NMR (CDCl3, 162.1 MHz) d 12.0; HRMS calcd for C10H12N2O3P, 239.0580,
found 239.0588.
c
No traces of 13e and only 5% of 15e.
Compound 15f was completely recovered.
Half of the starting phosphonate 15g was recovered; longer reaction times
d
2-(Dimethyl)phosphonyl-quinoxaline 1-N-oxide 9:
e
1H NMR (CDCl3, 300 MHz) d 9.06 (d, 1H, J = 2.4 Hz), 8.58 (dd, 1H, J = 1.2 and
8.6 Hz), 8.17 (dd, 1H, J = 1.2 and 8.6 Hz), 7.91 (td, 1H, J = 1.4 and 6.9 Hz), 7.79
(td, 1H, J = 1.4 and 6.9 Hz), 4.05 (d, 6H, J = 11.7 Hz); 13C NMR (CDCl3, 75 MHz) d
149.2 (d, J = 11.9 Hz), 146.6, 137.2 (d, J = 2.8 Hz), 133.1, 131.1 (d, J = 210.6 Hz),
130.7, 130.2, 118.9, 54.6 (d, J = 5.6 Hz); 31P NMR (CDCl3, 162.1 MHz) d 8.6;
HRMS calcd for C10H12N2O4P, 255.0529, found 255.0532.
resulted in no improvement in the yield in 13g.
Acknowledgments
3-Methyl-2-[(dimethyl)phosphonyl]-quinoxaline 1,4-di-N-oxide 13a:
1H NMR (CDCl3, 300 MHz) d 8.63 (dd, 1H, J = 1.2 and 8.6 Hz), 8.56 (dd, 1H,
J = 1.2 and 8.6 Hz), 7.90 (ddd, 1H, J = 1.4, 7.2 and 8.6 Hz), 7.83 (ddd, 1H, J = 1.4,
7.2 and 8.6 Hz), 4.03 (d, 6H, J = 11.7 Hz), 2.99 (d, 3H, J = 1.4 Hz); 13C NMR
The authors thank the CNRS for the financial support. S.D. is
grateful to the ‘Ministère de la Recherche et de l’Education Natio-
nale’ for a fellowship.
(CDCl3, 75 MHz)
d 146.1 (d, J = 19.7 Hz), 138.2 (d, J = 2.0 Hz), 136.8 (d,
J = 9.1 Hz), 133.4, 133.3 (d, J = 211.5 Hz), 131.3, 120.5 (d, J = 1.3 Hz), 120.4,
54.8 (d, J = 6.4 Hz), 15.4; 31P NMR (CDCl3, 162.1 MHz) d 7.9; HRMS calculated
for C11H13N2O5P, 284.0562, found 284.0556.
References and notes
1. For reviews, see: (a) Carta, A.; Corona, P.; Loriga, M. Curr. Med. Chem. 2005, 12,
2259–2272; (b) González, M.; Cerecetto, H.; Monge, A. Top. Heterocycl. Chem.
2007, 11, 179–211.
2. (a) McIlwain, H. J. Chem. Soc. 1943, 322–325; (b) Moore, P. R.; Evenson, A.;
Luckey, T. D.; McCoy, E.; Elvehjem, C. A.; Hart, E. B. J. Biol. Chem. 1946, 165, 437–
441.
3. (a) Carta, A.; Paglietti, G.; Nikookar, M. E. R.; Sanna, P.; Sechi, L.; Zanetti, S. Eur. J.
Med. Chem. 2002, 37, 355–366; (b) Jaso, A.; Zarranz, B.; Aldana, I.; Monge, A.
Eur. J. Med. Chem. 2003, 38, 791–800; (c) Ancizu, S.; Moreno, E.; Solano, B.;
Villar, R.; Burguete, A.; Torres, E.; Pérez-Silanes, S.; Aldana, I.; Monge, A. Bioorg.
Med. Chem 2010, 18 (ASAP); (d) Das, U.; Pati, H. N.; Panda, A. K.; De Clercq, E.;
Balzarini, J.; Molnár, J.; Baráth, Z.; Ocsovszki, I.; Kawase, M.; Zhou, L.; Sakagami,
H.; Dimmock, J. R. Bioorg. Med. Chem. 2009, 17, 3909–3915.
3-Methyl-2-[(diethyl)phosphonyl]-quinoxaline 1,4-di-N-oxide 13b:
1H NMR (CDCl3, 300 MHz) d 8.63 (dd, 1H, J = 1.2 and 8.6 Hz), 8.55 (dd, 1H,
J = 1.2 and 8.6 Hz), 7.89 (ddd, 1H, J = 1.2, 6.9 and 8.6 Hz), 7.81 (ddd, 1H, J = 1.2,
6.9 and 8.6 Hz), 4.42 (m, 4H), 3.02 (d, 3H, J = 1.4 Hz), 1.42 (td, 6H, J = 0.7 and
7.1 Hz); 13C NMR (CDCl3, 75 MHz) d 146.0 (d, J = 20.0 Hz), 138.0 (d, J = 2.0 Hz),
137.0 (d, J = 9.0 Hz), 134.1 (d, J = 209.9 Hz), 133.2, 131.2, 120.5, 120.3, 64.7 (d,
J = 6.0 Hz), 16.4 (d, J = 6.0 Hz), 15.4; 31P NMR (CDCl3, 162.1 MHz) d 4.9; HRMS
calcd for C13H17N2O5P, 312.0875, found 312.0870.
3-Ethyl-2-[(diethyl)phosphonyl)]-quinoxaline 1,4 di-N-oxide 13c:
1H NMR (CDCl3, 300 MHz) d 8.62 (dd, 1H, J = 1.2 and 8.6 Hz), 8.54 (dd, 1H,
J = 1.2 and 8.6 Hz), 7.89 (ddd, 1H, J = 1.4, 7.1 and 8.6 Hz), 7.81 (ddd, 1H, J = 1.4,
7.1 and 8.6 Hz), 4.43 (m, 4H), 3.58 (q, 2H, J = 7.2 Hz), 1.42 (t, 6H, J = 7.1 Hz), 1.36
(t, 3H, J = 7.2 Hz); 13C NMR (CDCl3, 75 MHz) d 150.5 (d, J = 20.1 Hz), 138.3 (d,

5520
S. Dahbi et al. / Tetrahedron Letters 51 (2010) 5516–5520
J = 2.0 Hz), 137.1 (d, J = 9.0 Hz), 133.9 (d, J = 210.9 Hz), 133.2, 131.2, 120.6 (d,
13.
A
very poor conversion (<10%) into phosphonates was observed using
J = 0.5 Hz), 120.5 (2 carbons), 64.6 (d, J = 6.0 Hz), 22.0, 16.4 (d, J = 8.0 Hz), 10.9;
31P NMR (CDCl3, 162.1 MHz) d 5.0; HRMS calcd for C14H19N2O5P, 326.1032,
found 326.1028.
3-Phenyl-2-[(diethyl)phosphonyl)]-quinoxaline 1,4 di-N-oxide 13g:
1H NMR (CDCl3, 300 MHz) d 8.63 (m, 2H), 7.93 (ddd, 1H, J = 1.6, 6.9 and 8.6 Hz),
7.88 (ddd, 1H, J = 1.6, 6.9 and 8.6 Hz), 7.53 (m, 5H), 4.22 (m, 2H), 4.99 (m, 2H),
1.10 (td, 6H, J = 0.5 and 7.2 Hz); 31P NMR (CDCl3, 162.1 MHz) d 3.9; HRMS calcd
for C18H19N2O5P, 374.1032, found 374.1024.
Pd(OAc)2/dppf, as recently precognized for the synthesis of
heteroarylphosphonates by Hirao cross-coupling: Belabassi, Y.; Alzghari, S.;
Montchamp, J.-L. J. Organomet. Chem. 2008, 693, 3171–3178. Switching dppf to
xantphos doubled the yield of the reaction.
14. Haddadin, M. J.; Issidorides, C. H. Heterocycles 1993, 35, 1503–1525.
15. Utilization of molecular sieves for the Beirut reaction has been reported by M.
Hasegawa’s group: (a) Takabatake, T.; Hasegawa, Y.; Hasegawa, M. J. Heterocycl.
Chem. 1993, 30, 1477–1479; (b) Takabatake, T.; Miyazawa, T.; Hasegawa, M. J.
Heterocycl. Chem. 1996, 33, 1057–1061; (c) Hasegawa, M.; Takabatate, T.;
Miyazawa, T. Yakugaku Zasshi 2001, 121, 379–393.
16. Subjecting phosphonate 13a to treatment in methanol in the presence of
molecular sieves under reaction conditions yielded a complete conversion into
2-methyl-quinoxaline 1,4-di-N-oxide 12.
17. We noted a good stability at room temperature of reference phosphonate 13a
in solution into THF, CHCl3, or MeOH.
18. The reaction appeared quite sluggish in the case of R = iPr or cyclopropyl: we
observed in NMR on prolonged reaction time a complex mixture containing
less than 15% of the desired phosphonate.
10. The utilization of LiHMDS has been recommended for the preparation of
lithium salts from P–H bonds, see: Belabassi, Y.; Antczak, M. I.; Tellez, J.;
Montchamp, J.-L. Tetrahedron 2008, 64, 9181–9190.
11. On treating 2-chloroquinoxaline monoxide with an excess of CsF and [18:6]
crown ether in THF at room temperature for 16 h, we obtained a crude mixture
containing about 50% of the quite unstable 2-fluoroquinoxaline monoxide.
12. Experiments run without adding HI resulted in no conversion into 2-
iodoquinoxaline monoxide. This latter compound has already been prepared,
albeit in lower yield, by thermal decomposition of the corresponding 2-
diazonium fluoroborate (Balz–Schiemann reaction).