A new and efficient synthetic approach to dichlorofluoflavines. Study of the stability of isomeric fluoflavines by HF and B3LYP procedures

Article abstract of DOI:10.1016/j.tet.2009.01.059

The first method for the synthesis of 7,10-dichloro-5,12-dihydroquinoxalino[2,3-b]quinoxalines is reported. Treatment of 3,3,6,6-tetrachloro-1,2-cyclohexanedione with diaminomaleonitrile leads to 5,8-dichloro-2,3-dicyanoquinoxaline in near quantitative yi

Full text of DOI:10.1016/j.tet.2009.01.059

Tetrahedron 65 (2009) 2254–2259
Contents lists available at ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
A new and efficient synthetic approach to dichlorofluoflavines. Study of the
stability of isomeric fluoflavines by HF and B3LYP procedures
,
a
a
a
b,
*
Antonio Guirado a , Jose I. Lopez-Sanchez , Alfredo Cerezo , Delia Bautista , Jesus Galvez
*
´
´
´
´
´
a
´
´
Departamento de Qu´ımica Organica, Facultad de Quımica, Universidad de Murcia, Campus de Espinardo, 30071 Murcia, Apartado 4021, Spain
b Departamento de Qu´ımica F´ısica, Facultad de Qu´ımica, Universidad de Murcia, Campus de Espinardo, 30071 Murcia, Apartado 4021, Spain
a r t i c l e i n f o
a b s t r a c t
Article history:
The first method for the synthesis of 7,10-dichloro-5,12-dihydroquinoxalino[2,3-b]quinoxalines is re-
ported. Treatment of 3,3,6,6-tetrachloro-1,2-cyclohexanedione with diaminomaleonitrile leads to 5,8-
dichloro-2,3-dicyanoquinoxaline in near quantitative yields. This compound has been found to be an
excellent synthetic equivalent of unavailable 2,3,5,8-tetrachloroquinoxaline. It reacts with o-phenyl-
enediamines providing the corresponding dichlorofluoflavines in fair to high yields. These compounds
have been identified by NMR spectroscopy and X-ray crystallography. Molecular structures, chemical
Received 9 December 2008
Received in revised form 9 January 2009
Accepted 12 January 2009
Available online 20 January 2009
Keywords:
Fluoflavines
Quinoxalines
1,2-Phenylenediamines
Dehydrochlorination
Aromatization
hardness,
HF and B3LYP density functional theory methods.
D
Gs298-values, and relative stabilities of all possible isomeric products have been calculated by
Ó 2009 Elsevier Ltd. All rights reserved.
1. Introduction
a sure criterion to assume the actual position of both N–H groups
in the parent compound.
The first reference on fluoflavines is Hinsberg and Pollack, who
gave the name fluoflavine to the dihydroquinoxalino[2,3-b]qui-
noxaline formed in only 5% yield by reaction of 2,3-dichloro-
quinoxaline with o-phenylenediamine.1 They formulated its
Over the years different features of the chemistry of fluoflavines,
such as synthesis,1,2a,4 oxidation,1,4d,5 electrophilic attack,2b,4i,5a and
electrochemical2b,6 and photochemical4k,7 reactions have received
considerable attention. Fluoflavines have been described as colo-
rants,4c,e,8 substances with anthelmintic activity,4i and as practical
components in confering peculiar properties to polymers4e,9 and
for other uses.4a,g,m,5a,10 Recently described electronic properties of
metal complexes containing this four N-donor system also evi-
dence a notable interest in fluoflavines.11
It is clear that the preferred strategy to obtain fluoflavines is
based on 2,3-dichloroquinoxaline chemistry.1,4b,c,i Some related
preparations starting from 2,3-diamino-, 2,3-dihydroxy- or 2,3-
diphenoxyquinoxaline instead of 2,3-dichloroquinoxaline have also
been described.2a,4b,e,j It should be mentioned that certain less
general reactions yielding fluoflavines or fluoflavine derivatives,
but without participation of quinoxaline intermediates, have also
been reported.4d,f,h,k,l A survey of the literature shows that there is
effective methodology to build the parent compound of the family.
However, access to functionalized fluoflavines may be unfeasible or
severely limited by unavailability of the required quinoxaline or
1,2-benzenediamine intermediates. This is just the case for the
hitherto unknown 7,10-dichlorofluoflavines here reported. The
number of chlorinated fluoflavines described is very scanty.4c,i,m,8,12
However, considering that the synthetic usefulness of chloro-
phenazines in affording phenazine derivatives is well docu-
mented,13 chlorofluoflavines are of interest in themselves, but also
structure as 5,12-dihydro
1 but later it was controversially
assigned to be either compound 1 or its 5,11-dihydro isomer2,3
2
(Fig. 1). This is understandable considering that NMR studies are
particularly difficult, due to the high insolubility of this product,
which does not provide fully conclusive spectra if the sample is
solved in acidic media. Certain N-derivatized fluoflavines were
found to be somewhat more soluble than fluoflavine itself.
Spectroscopic studies of these compounds were made.2b In this
case, however, the site where an N-derivatization occurs is not
H
N
H
N
N
N
N
N
H
N
H
N
1
2
Figure 1. The structures of isomeric fluoflavines.
* Corresponding authors. Tel.: þ34 968367490; fax: þ34 968364148.
E-mail address: anguir@um.es (A. Guirado).
0040-4020/$ – see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2009.01.059

A. Guirado et al. / Tetrahedron 65 (2009) 2254–2259
2255
to investigate their potential utility as intermediates to provide
further fluoflavine derivatives.
of a single product was observed. It was isolated and identified as
5,8-dichloro-2,3-dicyanoquinoxaline 7, a novel quinoxaline de-
rivative. The yield of this one-pot preparative process was almost
quantitative. In the hope of obtaining fluoflavines from 8 we
attempted a reaction with 1,2-phenylenediamine in hot pyridine.
The formation of a strongly fluorescent product was perceptible
after a short time, whereas the total conversion of the starting
materials to this product occurred after 6 h. It was isolated as
a highly pure yellow powder whose spectral and elemental anal-
yses revealed a dichlorofluoflavine formed in high yield, which was
identified as 7,10-dichloro-5,12-dihydroquinoxalino[2,3-b]qui-
noxaline 8a.
An approach to 7,10-dichlorofluoflavines 8 by the Hinsberg and
Pollack route (Scheme 1) appears to be unfeasible. Preparation of
2,3,5,8-tetrachloroquinoxaline 4 following the standard route de-
scribed for 2,3-dichloroquinoxaline but involving 3,6-dichloro-1,2-
phenylenediamine 3 instead of 1,2-phenylenediamine appears to
be highly conflictive since 3 is a rare and expensive compound of
extremely difficult availability.14 Its most efficient preparation has
been reported in only 15% yield through the synthesis of ben-
zo[c][1,2,5]selenadiazole and subsequent chlorination and re-
duction reactions.14a
Recently we developed a new, highly efficient methodology for
the synthesis of chlorophenazines15 starting from 3,3,6,6-tetra-
chlorocyclohexanedione 5, which is a cheap and readily available
intermediate that can be quantitatively generated by simple chlo-
rination of commercial trans-cyclohexanediol. Compound 5 reacts
with 1,2-phenylenediamines exhibiting the behavior of an excellent
synthetic equivalent of unavailable 3,6-dichloro-1,2-benzoquinone.
It also provided an exclusive approach to polychloro-1,2-epoxy-
cyclopentane-1-carboxylic acids and their alkyl esters.16 Given the
interest of these results, we continued working to expand the
classes of specifically chlorinated heterocyclic compounds by
starting from this profitable reagent. The first outcome of this
project is the synthesis of 7,10-dichloro-5,12-dihydroquinox-
alino[2,3-b]quinoxalines 8 by an expeditious approach that implies
the first preparation of 5,8-dichloro-2,3-dicyanoquinoxaline 7 and
its use as a key synthetic intermediate.
Reactions with further 1,2-benzenediamines were also carried
out under a similar protocol leading to the corresponding dichlor-
ofluoflavines 8 in fair to high yields. These compounds were also
highly fluorescent. These results evidenced that inexpensive and
easily available 5,8-dichloro-2,3-dicyanoquinoxaline 7 behaves as
an excellent synthetic equivalent of unavailable 2,3,5,8-tetra-
chloroquinoxaline 4, providing access to this previously unattain-
able class of fluoflavine derivatives.
Given that the prepared dichlorofluoflavines exhibit a modest
but sufficient solubility in usual spectroscopy solvents, our purpose
was to realize a fully reliable characterization of the products
formed by a structural study involving a conjunction of tools such
as 1D and 2D NMR spectroscopy, single crystal X-ray crystallogra-
phy, and HF and B3LYP density functional theory methods.
Figure 2 shows the three possible molecular arrangements for
any dichlorofluoflavine originated by reaction between 5,8-
dichloro-2,3-dicyanoquinoxaline
7 and 1,2-phenylenediamine.
First, we focused on determining the relative stabilities of the non-
chlorinated isomers (1, 2) as well as the dichlorinated ones (8a, 8a0,
8a00) by computational methodology. With this purpose, full geo-
metry optimizations were performed at the B3LYP/6-311þþG(d,p)
level of theory. It was found that 5,12-dihydroquinoxalino[2,3-
b]quinoxaline 1 is more stable than 5,11-dihydroquinoxalino[2,3-
2. Results and discussion
3,3,6,6-Tetrachloro-1,2-cyclohexanedione 5 was treated with
diaminomaleonitrile17 6 in the presence of pyridine. The formation
b]quinoxaline 2 with relative values of
D
Gs298¼0 and ꢀ1.93 kcal/
Cl
H2N
H2N
Cl
O
C
O
C
NH2
NH2
N
N
mol, respectively. In the case of dichlorofluoflavines, the lowest
stability corresponds to 1,4-dichloro-5,11-dihydroquinoxalino[2,3-
b]quinoxaline 8a00, the most stable isomer is 1,4-dichloro-5,12-
dihydroquinoxalino[2,3-b]quinoxaline 8a0, whereas stability of
7,10-dichloro-5,12-dihydroquinoxalino[2,3-b]quinoxaline 8a is
Cl
Cl
1) RO
OR
2) PCl5
2 HCl
Cl
3
Cl
4
Cl
Cl
H
somewhat lower than that of 8a0 (the relative
compounds are 0, ꢀ2.76, and ꢀ0.77 kcal/mol). Since it has been
reported that in most of the cases the chemical hardness ( ) is
a good indicator for predicting the most stable isomer,18 the
-values for these isomers were also computed at the B3LYP/6-
D
Gs298 values for these
N
N
N
h
N
H
8a
h
Cl
Cl Cl
311þþG(d,p) level by using the equation
N
N
O
O
CN
(80%)
H2N
H2N
CN
2 HCN
2 H2O
I ꢀ A
2 HCl
(95%)
H2N
h
¼
(1)
CN
CN
2
H2N
Cl Cl
5
Cl
Cl
where I is the ionization potential and A the electron affinity of the
molecule. To carry out calculations for the positive and negative
ions of each isomer, which are necessary to determine the values of
I and A in Eq. 1, the respective geometries of the neutral molecules
were used. Alternatively, I and A can be approximated in terms of
the eigenvalues corresponding to the HOMO and LUMO orbitals,
6
7
H
Cl
H
N
N
Me
Me
N
N
Me
N
N
N
H
N
H
Cl
Cl
Cl
8b (85%)
8c (87%)
Cl
H
Ac
Cl
A
H
N
C
N
H
Cl
H
N
H
N
Cl
N
N
Cl
Cl
N
N
N
B
N
N
N
N
D
N
N
N
N
N
N
H
N
H
H
Cl
Ac
Cl
Cl
Cl
Cl
8a
8a'
8a''
8d (60%)
9 (95%)
Scheme 1.
Figure 2. The structurs of isomeric dichlorofluoflavines.

2256
A. Guirado et al. / Tetrahedron 65 (2009) 2254–2259
Table 1
i.e., I¼ꢀEHOMO, A¼ꢀELUMO. The calculated values of
h for these
Selected bond lengths and bond angles of crystal and calculated structures of 9 and
their differences
isomers were 3.23 eV (8a), 3.40 eV (8a0), and 3.12 eV (8a00). These
values confirm that the most energetically stable isomer (8a0) is
associated with the maximum hardness, while 8a00 exhibiting the
Bond
Crystal
structure (Å)
Calculated
D
(Å)
Calculated
D(Å)
structure (Å)a
structure (Å)b
minimum hardness has also the largest value of
D
Gs298. This last
structure was clearly discarded by 13C NMR since only 7 resonance
signals were recorded instead of 14. However, these spectra were
not able to distinguish between structures 8a and 8a0. This problem
was resolved by 15N HSQC and 15N HMBC experiments, which were
found to be in good agreement with arrangement 8a. Thus, 15N
HMBC spectroscopy showed that protonated nitrogens (N5–N12,
117 ppm) correlate with hydrogens H4–H1, whereas non-pro-
tonated nitrogens (N6–N11, 239 ppm) correlate with hydrogens
H8–H9. This led to the conclusion that the dihydropyrazine system
is located in the ring C, whereas pyrazine arrangement pertains to
ring B. The 15N HSQC spectrum was consistent with this assumption
since protons H5–H12 correlated with N5–N12.
Cl(1)–C(4)
N(1)–C(5)
N(1)–C(6)
N(2)–C(6)
N(2)–C(7)
N(3)–C(l2)
N(2)–C(17)
N(3)–C(15)
O(1)–C(17)
1.729 (2)
1.366 (3)
1.302 (3)
1.412 (3)
1.439 (3)
1.427 (3)
1.416 (3)
1.401 (3)
1.209 (3)
1.747
1.362
1.305
1.405
1.439
1.439
1.421
1.421
1.215
0.018
0.004
1.732
1.361
1.273
1.397
1.433
1.433
1.399
1.399
1.191
ꢀ0.003
0.005
0.029
0.015
0.006
ꢀ0.006
0.017
ꢀ0.003
0.007
0.000
ꢀ0.012
ꢀ0.005
ꢀ0.020
ꢀ0.006
0.002
0.018
Bond
Crystal
Calculated
D
(ꢁ)
Calculated
D
(ꢁ)
structure (ꢁ) structure (ꢁ)a
structure (ꢁ)b
C(5)–C(4)–Cl(1)
C(3)–C(4)–Cl(1)
C(14)–C(1)–Cl(2)
C(2)–C(1)–Cl(2)
C(6)–N(1)–C(5)
N(1)–C(6)–N(2)
N(4)–C(13)–N(3)
C(6)–N(2)–C(7)
119.40 (16)
120.07 (16)
118.53 (16)
120.67 (16)
116.89 (17)
121.36 (18)
120.45 (18)
114.45 (17)
119.85
119.75
119.85
119.75
118.13
121.24
121.24
114.40
114.40
ꢀ0.45 120.05
0.32 119.81
ꢀ1.32 120.05
0.92 119.81
ꢀ0.65
0.26
ꢀ1.52
Unfortunately, single crystals suitable for an X-ray diffraction
analysis could not be obtained despite intensive effort. Instead, we
were successful in an attempt to analyze a diacetylated derivative.
The molecular structure found corresponds to 9, which shows an
angular geometry with an interplanar angle of 138.3ꢁ (Fig. 3). Se-
lected intramolecular distances and bond angles for this crystal
structure are given in Table 1. The optimized geometry of 9 calcu-
lated at the HF/6-31G(d) and B3LYP/6-31G(d) levels of theory were
in excellent agreement with that found by X-ray analysis and so the
computed B3LYP/6-31G(d) interplanar angle was 140.4ꢁ. The cal-
culated bond lengths and bond angles are also included in Table 1
for comparison. A perspective of the crystal packing showing hy-
drogen interactions is shown in Figure 4. As far as we know, this is
the first time that a crystallographic molecular structure of
a member of the fluoflavines family is reported. Since the optimized
geometries of 8a, 8a0, and 8a00 were found to be fully planar
molecules at all levels of computation, it is worthy of note that
formation of product 9 involves a remarkable geometrical change
attributable to the loss of conjugation of the nitrogen lone electron
pair with both adjacent aromatic systems motivated by a high de-
localization at the carbonyl oxygen. It should also be noted that in
agreement with the observed in the case of compound 8a, an 15N
HMBC spectrum of 9 clearly showed correlation between acylated
nitrogens and H1–H4 protons, whereas nitrogens integrated in the
aromatic system correlated with H8–H9 protons. Therefore, any
possibility of tautomerism disturbing the initial position of the
dihydro system with respect to that where diacetylation occurs
could be discarded.
0.86
ꢀ1.24
118.48
ꢀ1.59
ꢀ1.10
ꢀ2.01
0.52
0.12 122.46
ꢀ0.79 122.46
0.05 113.93
ꢀ0.84 113.93
C(13)–N(3)–C(12) 113.56 (17)
ꢀ0.37
a
Geometry optimized at the B3LYP/6-31G(d) level of theory.
Geometry optimized at the HF/6-31G(d) level of theory.
b
instead of the more stable 8a0 (the relative values of
D
Gs298 are
ꢀ0.77 and ꢀ2.76 kcal/mol, and the corresponding hardness 3.23
and 3.40 eV, see above). It is apparent from Scheme 1 that the
compound initially formed would be 8a. To clarify why this isomer
does not undergo transformation to 8a0, the transition states for the
intramolecular transformation 8a/8a00/8a0 have been calculated.
The activation energies for this pathway were found to be ex-
tremely high, 51.1 and 49.7 kcal/mol at the B3LYP/6-31G(d) level
(the corresponding values when correlation energy is not taken
into account are even higher, 68.7 and 65.3 kcal/mol at the HF/6-
31G(d) level).
In conclusion, a proficient and general method for the synthesis
of 7,10-dichlorofluoflavines on the basis of 3,3,6,6-tetrachloro-1,2-
cyclohexanedione chemistry is reported. Good yields and easy
availability of starting materials are valuable, noteworthy advan-
tages of the method, which allows a privileged access to previously
unattainable products.
Finally, we will briefly discuss that although 8a is not the most
stable isomer, this was the only compound isolated and identified,
3. Experimental
3.1. General
NMR spectra were recorded on Bruker Unity 200, 300 or
400 MHz spectrometers. Samples of compounds 8 were prepared
by dissolving 6–10 mg of each product in 0.6–0.8 mL (CD3)2SO. The
1H and 13C NMR spectra are relative to residual (CD2H)S(O)CD3 at
d
2.49 ppm and to (CD3)2SO at d N
39.5 ppm, respectively. 1H–15
NMR heteronuclear correlation spectra were recorded using stan-
dard techniques. A sample of 15N-formamide referenced to anhy-
drous liquid ammonia was used as an external standard for 15N
chemical shifts. In the HMBC experiment of 8a the spectral width
was set to 2.3 kHz in F2 and to 18.2 kHz in F1, the number of
transients was 32 and 1024 time increments were acquired. In the
HSQC experiment the spectral width was set to 2.4 kHz in F2 and to
18.3 kHz in F1, the number of transients was 32 and 2048 time
increments were acquired. Sample of 9 was prepared by dissolving
30 mg in 0.7 mL CDCl3 and referenced to internal TMS at
for 1H NMR and to CDCl3 at 77.1 ppm for 13C NMR. In the HMBC
experiment the spectral width was set to 3.6 kHz in F2 and to
d 0.00 ppm
d
Figure 3. ORTEP of 9, with thermal ellipsoids shown at 50% probability.

A. Guirado et al. / Tetrahedron 65 (2009) 2254–2259
2257
Figure 4. A perspective of the crystal packing of 9, showing the hydrogen interactions.
16.2 kHz in F1, the number of transients was 48 and 2048 time
increments were acquired.
crystallized from chloroform giving brown prisms mp 265–267 ꢁC
(dec). Yield 95%. (Found: C 48.36; H 0.80; N 22.28. C10H2Cl2N4 re-
All computations have been performed with the Spartan’06
package program.19 First, the most stable conformers were de-
termined by using the MMFF molecular mechanics method.20 Next,
these conformers were used as input for ab initio molecular orbital
and density functional theory calculations of geometry optimiza-
tions at the Hartree–Fock and B3LYP levels of theory with the 6-
31G(d) and 6-311þþG(d,p) base sets. There were no significant
differences between the results obtained with the different base
sets. Frequency calculations were performed at the same level of
theory as the geometry optimizations to characterize the stationary
points as local minima (equilibrium structures). Activation energies
were obtained as the difference between the total molecular energy
of the transition state and that of the ground state.
quires: C 48.23; H 0.81; N 22.50.) 1H NMR
d (DMSO-d6, 200 MHz):
8.39 (s, 2H); 13C NMR
d (DMSO-d6, 200 MHz): 114.13 (C), 131.55 (C),
132.58 (C), 134.63 (CH), 138.22 (C); MS m/z (%): 250 (Mþþ2, 70), 248
(Mþ, 100), 196 (6), 146 (34), 144 (50), 109 (20), 74 (17); IR (Nujol):
1586, 1540, 1314, 1302, 1209, 1196, 1123, 947, 854, 699, 608 cmꢀ1
.
3.3. Preparation of fluoflavines (8)
A mixture of 5,8-dichloro-2,3-dicyanoquinoxaline 7 (1.2 mmol),
the appropriate o-phenylenediamine (3.6 mmol), and pyridine
(20 mL) was refluxed for 6 h (16 h in the case of product 8d). After
cooling, the reaction mixture was poured into 200 mL of cold brine
and the solid precipitate formed was collected by filtration and
purified by column chromatography (silica gel/ethyl acetate–hex-
ane 2:1).
3.2. Preparation of 5,8-dichloro-2,3-dicyanoquinoxaline (7)
A solution of 3,3,6,6-tetrachlorocyclohexanedione 5 (20 mmol)
in dimethylformamide (20 mL) was added dropwise to a warmed
solution (60 ꢁC) of diaminomaleonitrile 6 (24 mmol) in dime-
thylformamide (80 mL). The solution was stirred at this tempera-
ture for 8 h. Pyridine (40 mmol) was added and the reaction
mixture was stirred at 80 ꢁC for 2 h and cooled to room tempera-
ture. Addition of cold brine (500 mL) provoked the formation of
a brown solid precipitate that was separated by filtration and
3.4. 7,10-Dichloro-5,12-dihydroquinoxalino[2,3-b]
quinoxaline (8a)
Yield 80%; yellow powder (ethyl acetate–hexane); mp 256–
259 ꢁC (dec). (Found: C 55.83; H 2.63; N 18.22. C14H8Cl2N4 requires:
C 55.47; H 2.66; N 18.48.) 1H NMR
d
(DMSO-d6, 300 MHz): 6.56–
6.63 (m, 4H), 7.05 (s, 2H), 10.46 (br s, 2H); 13C NMR
(DMSO-d6,
400 MHz): 113.86 (CH), 122.24 (CH), 123.93 (CH), 125.78 (C), 129.02
d

2258
A. Guirado et al. / Tetrahedron 65 (2009) 2254–2259
(C), 137.06 (CCl), 144.87 (C); MS m/z (%): 304 (Mþþ2, 70), 302 (Mþ,
100), 232 (Mþꢀ2 Cl, 5), 231 (55), 151 (3), 102 (4); IR (Nujol): 3395,
refinement. The structures were refined anisotropically against F2
(program SHELXL-97, G.M. Sheldrick, University of Go¨ttingen). The
methyl group was refined as a rigid group, other H with a riding
model. This structure is refined in Cc. The final wR2 value was
0.0842 for all reflections, 510 parameters and 2 restraints, with R1
1622, 1587, 1539, 1175, 970, 800, 745 cmꢀ1
.
3.5. 7,10-Dichloro-2-methyl-5,12-dihydroquinoxalino[2,3-
b]quinoxaline (8b)
0.0333 for reflections with I>2s
(I); max Dr 0.695 e/Å3, S 1.05.
The program PLATON suggests the higher symmetry C2/c as
default with ADDSYM, but ADDSYM EXACT does not find it, which
suggests ‘almost but not quite’ C2/c. We did not manage to refine it
successfully in C2/c. When all the atoms are anisotropic, with
normal anisotropic displacement parameters, and all hydrogens are
fixed the R1¼0.1698 and wR2¼0.4151 and no residual electron
density is higher than 0.5 e/Å3.
Complete crystallographic data (excluding structure factors)
have been deposited at the Cambridge Crystallographic Data Centre
under the number CCDC 711405.
Yield 85%; yellow powder (ethyl acetate–hexane); mp 215–
217 ꢁC (dec). (Found: C 57.11; H 3.24; N 17.49. C15H10Cl2N4 requires:
C 56.80; H 3.18; N 17.66.) 1H NMR
d
(DMSO-d6, 400 MHz): 2.11 (s,
3H), 6.47–6.53 (m, 3H), 7.11 (s, 2H), 10.44 (br s, 1H), 10.46 (br s, 1H);
13C NMR
(DMSO-d6, 400 MHz): 20.37 (CH3), 114.00 (CH), 114.56
d
(CH), 123.09 (CH), 124.20 (CH), 124.36 (CH), 125.89 (C), 125.98 (C),
126.86 (C), 129.11 (C), 131.80 (C), 137.27 (CCl), 137.47 (CCl), 145.37
(C), 145.41 (C); MS m/z (%): 318 (Mþþ2, 68), 316 (Mþ, 100), 281
(MþꢀCl, 10), 245 (31), 114 (23), 103 (28), 89 (41), 76 (41), 52 (41); IR
(Nujol): 3418, 1628, 1589, 1539, 1186, 975, 944, 799 cmꢀ1
.
Acknowledgements
3.6. 7,10-Dichloro-2,3-dimethyl-5,12-dihydroquinoxalino[2,3-
b]quinoxaline (8c)
We gratefully acknowledge the financial support of the Funda-
´
´
´
´
cion Seneca of the Comunidad Autonoma de la Region de Murcia
Yield 87%; yellow powder (ethyl acetate–hexane); mp 268–
270 ꢁC (dec). (Found: C 57.89; H 3.71; N 17.14. C16H12Cl2N4 requires:
(project 03035/PI/05).
C 58.02; H 3.65; N 16.92.) 1H NMR
d
(DMSO-d6, 300 MHz): 1.98 (s,
Supplementary data
6H), 6.39 (s, 2H), 7.05 (s, 2H), 10.41 (br s, 2H); 13C NMR
d
(DMSO-d6,
300 MHz): 18.73 (CH3), 115.26 (CH), 124.18 (CH), 125.86 (C), 126.84
(C), 130.23 (C), 137.44 (CCl), 145.47 (C); MS m/z (%): 332 (Mþþ2, 70),
330 (Mþ, 100), 315 (MþꢀCH3, 14), 317 (8), 259 (6), 165 (10); IR
X-ray structural data of compound 9; NMR spectra of com-
pounds 7, 8a, and 9; optimized geometries and energies for com-
pounds 8a, 8a0, 8a00, and 9, as well as transition state structures for
8a/8a00 and 8a00/8a0. Supplementary data associated with this
article can be found in the online version, at doi:10.1016/
j.tet.2009.01.059.
(Nujol): 3409, 1619, 1588, 1538, 1179, 962, 874, 798 cmꢀ1
.
3.7. 2,3,7,10-Tetrachloro-5,12-dihydroquinoxalino[2,3-
b]quinoxaline (8d)
References and notes
Yield 60%; yellow powder (ethyl acetate–hexane); mp 282–
284 ꢁC (dec). (Found: C 45.44; H 1.61; N 15.30. C14H6Cl4N4 requires:
1. Hinsberg, O.; Pollak, J. Chem. Ber. 1896, 29, 784.
2. See for example: (a) Badger, G.; Nelson, P. Aust. J. Chem. 1963, 16, 445; (b)
Armand, J.; Boulares, L.; Bellec, C.; Pinson, J. Can. J. Chem. 1982, 60, 2797.
3. See for example: (a) Hunig, S.; Scheutzow, D.; Schlap, H.; Quast, H. Justus Liebigs
Ann. 1972, 765, 110; (b) Akimoto, Y. Bull. Chem. Soc. Jpn. 1956, 29, 460.
4. (a) Kawai, S.; Kimura, K.; Furuhata, T. Proc. Jpn. Acad. 1954, 29, 344; (b)
Woodburn, H.; Hoffman, W. J. Org. Chem. 1958, 23, 262; (c) Riedel, G.; Deuschel,
W. Patent GB 971048, 1964; Chem. Abstr. 1965, 62, 9503; (d) Ried, W.; Scha¨fer, D.
Chem. Ber. 1970, 103, 2225; (e) Marcel, C.; Tucson, A. U.S. Patent 3,563,917, 1971;
Chem. Abstr. 1971, 74, 126521; (f) Ried, W.; Schmidt, A.; Knorr, H. Chem. Ber.
1975, 108, 538; (g) Sugimoto, A.; Kato, S.; Inoue, H.; Imoto, E. Bull. Chem. Soc.
Jpn. 1976, 49, 337; (h) Hu¨ nig, S.; Pu¨tter, H. Chem. Ber. 1977, 110, 2532; (i) Fisher,
M.; Lusi, A.; Egerton, J. J. Pharm. Sci. 1977, 66, 1349; (j) Cheeseman, G.; Cookson, R.
In Condensed Pyrazines; Weissberger, A., Taylor, E., Eds.; The Chemistry of Het-
erocyclic Compounds; JohnWiley & Sons: New York, NY, 1979; Vol. 35, p 186; (k)
Tauer, E.; Grellmann, K. Chem. Ber. 1990, 123, 1149; (l) Koutentis, P.; Rees, C. J.
Chem. Soc., Perkin Trans. 1 2000, 2601; (m) Hitoshi, M. Jpn. Pat. JP 2002265811,
2002; Chem. Abstr. 2002, 137, 234007.
5. (a) Kuhn, R.; Skrabal, P.; Fischer, P. Tetrahedron 1968, 24, 1843; (b) Kuhn, R.;
Skrabal, P. Chem. Ber. 1968, 101, 3913.
6. (a) Armand, J.; Boulares, L.; Bellec, C.; Bois, C.; Philochelevisalles, M.; Pinson, J.
Can. J. Chem. 1984, 62, 1028; (b) Armand, J.; Bellec, C.; Boulares, L.; Pinson, J.
J. Org. Chem. 1983, 48, 2847.
7. Tauer, E.; Grellmann, K.; Noltemeyer, M.; Sheldrick, G. Angew. Chem., Int. Ed.
Engl. 1989, 28, 338.
8. (a) Burke, Jr.; Humphreys, V. U.S. Patent 4,132,561, 1979; Chem. Abstr. 1979, 90,
123210; (b) Graser, F. Ger. Pat. DE 3439045 A1, 1986; Chem. Abstr. 1986, 105,
99143.
9. (a) De Schryver, F.; Marvel, C. J. Polym. Sci., Part A: Polym. Chem. 1967, 5, 545; (b)
Jadamus, H.; Schryver, F.; Winter, W.; Marvel, C. J. Polym. Sci., Part A: Polym.
Chem. 1966, 4, 2831; (c) Tadao, K.; Masana, Y.; Yoshimitsu, I.; Masayuki, O.;
Masao, H.; Tadao, T.; Muneyoshi, M. U.S. Patent 3,491,057, 1970; Chem. Abstr.
1970, 72, 91450.
10. Horner, M.; Hu¨nig, S.; Pu¨ tter, H. Electrochim. Acta 1982, 27, 205.
11. Cotton, F. A.; Li, Z.; Liu, C.; Murillo, C.; Villagra´n, D. Inorg. Chem. 2006, 45, 767.
12. (a) Kehmann, V.; Bener, C. Helv. Chim. Acta 1925, 8, 20.
13. For example: (a) Tada, M. Bull. Chem. Soc. Jpn. 1974, 47, 1803; (b) Vivian, D. L.
J. Org. Chem. 1956, 21, 565; (c) Vivian, D. L. J. Am. Chem. Soc. 1951, 73, 457; (d)
Vivian, D. L.; Belkin, M.; Hogart, R. M. J. Org. Chem. 1961, 26, 112; (e) Tada, M.
Bull. Chem. Soc. Jpn. 1978, 51, 3093; (f) Pachter, I. J.; Kloetzel, M. C. J. Am. Chem.
Soc. 1952, 74, 971; (g) Tada, M. J. Mol. Struct. 1975, 48, 363 and references cited
therein.
C 45.20; H 1.63; N 15.06.) 1H NMR
d
(DMSO-d6, 400 MHz): 6.70 (s,
2H), 7.18 (s, 2H), 10.70 (br s, 2H); 13C NMR
d
(DMSO-d6, 300 MHz):
114.41 (CH), 123.37 (C), 124.46 (CH), 126.15 (C), 129.7 (C), 136.67 (C),
144.00 (C); MS m/z (%): 374 (Mþþ4, 49), 372 (Mþþ2, 100), 370 (Mþ,
79), 301 (27), 299 (38), 265 (7), 186 (18), 149 (19), 57 (23); IR
(Nujol): 3364, 1645, 1580, 1520, 1181, 952, 881, 799 cmꢀ1
.
3.8. Preparation of 5,12-diacetyl-7,10-dichloro-5,12-
dihydroquinoxalino[2,3-b]quinoxaline (9)
7,10-Dichloro-5,12-dihydroquinoxalino[2,3-b]quinoxaline
5a
(1.4 mmol) was treated with acetic anhydride (15 mL) at reflux
temperature for 16 h and the resulting reaction mixture was con-
centrated to dryness yielding a crude material that was recrystal-
lized from methanol–chloroform to give brown prisms mp
232–235 ꢁC (dec). Yield 95%. (Found: C 56.12; H 3.16; N 14.32.
C18H12Cl2N4O2 requires: C 55.83; H 3.12; N 14.47.) 1H NMR
400 MHz): 2.71 (s, 6H), 7.30–7.35 (m, 2H), 7.74 (s, 2H), 7.92–7.96 (m,
2H); 13C NMR
(CDCl3, 400 MHz): 25.13 (CH3), 125.27 (CH), 126.38
d (CDCl3,
d
(CH), 129.43 (CH), 131.03 (C), 131.08 (C), 136.47 (C), 144.38 (C),
168.71 (CO); MS m/z (%): 344 (MþꢀC2H2O, 6), 304 (53), 302
(Mþꢀ2C2H2O, 100), 266 (9), 231 (30), 102 (10); IR (Nujol): 1692,
1344, 1298, 1256, 1234, 1188, 1016, 762, 670 cmꢀ1
.
Crystal data. C18.5H12.5Cl3.5N4O2, Mr¼446.90, monoclinic, space
group
b
Cc,
a¼15.6451(8),
b¼17.5842(8),
c¼14.5831(7) Å,
¼111.283(2)ꢁ, V¼3738.3(3) Å3 at ꢀ100 K; Z¼8, Dx¼1.588 g/cm3,
F(000)¼1816,
m¼0.59/mm. Data collection. A pale yellow prism
0.30ꢂ0.29ꢂ0.26 mm was mounted in inert oil on a glass fibre and
transferred to the cold gas stream of the diffractometer (Bruker
SMART APEX CCD). Of 21,503 measured reflections, 8424 were
unique (Rint¼0.0196) and were used for all calculations. Structure

A. Guirado et al. / Tetrahedron 65 (2009) 2254–2259
2259
14. (a) Bird, C.; Cheeseman, G.; Sarsfield, A. J. Chem. Soc. 1963, 4767; (b) Pesin, V.;
Sergeev, V.; Khaletskii, A. Zh. Obshch. Khim. 1964, 34, 3028; (c) Pesin, V.; Ser-
geev, V. Khim. Geterotsikl. Soedin. 1968, 6, 1001.
16. Guirado, A.; Cerezo, A.; Lo´pez-Sa´nchez, J. I.; Sa´ez-Ayala, M.; Bautista, D. Tetra-
hedron Lett. 2006, 47, 7583.
17. A cheap commercially available reagent.
´
15. (a) Guirado, A.; Cerezo, A.; Ramırez de Arellano, C. Tetrahedron 1997, 53, 6183;
18. (a) Chemical Hardness, Structure and Bonding; Sen, K. D., Ed.; Springer: Berlin,
1993; Vol 80; (b) Ghanty, T. K.; Ghosh, S. K. J. Phys. Chem. A 2000, 104, 2975; (c)
Kolandaivel, P.; Senthilkumar, K. J. Mol. Struct. 2001, 535, 61.
19. Spartan’06 Wavefunction, Inc. Irvine, CA.
(b) Guirado, A.; Cerezo, A.; Andreu, R. Tetrahedron Lett. 2000, 41, 6579; (c)
Guirado, A.; Cerezo, A.; Andreu, R.; Lo´pez-Sa´nchez, J. I.; Bautista, D. Tetrahedron
2004, 60, 6747; (d) Guirado, A.; Cerezo, A.; Lopez-Sanchez, J. I.; Bautista, D.
Tetrahedron Lett. 2007, 48, 9137.
´
´
20. Halgren, T. A. J. Comput. Chem. 1996, 17, 730.