Benzoxazinophenoxazines: Neutral and charged species

Article abstract of DOI:10.1016/S0040-4039(01)01575-1

A new electron-donor, benzoxazinophenoxazine 1, was synthesized. It showed a low oxidation potential of +0.21～+0.23 V versus SCE. The neutral molecule, its radical cation salts, and charge transfer complexes were synthesized and characterized.

Full text of DOI:10.1016/S0040-4039(01)01575-1

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 42 (2001) 7591–7594
Benzoxazinophenoxazines: neutral and charged species
Toshihiro Okamoto,a Masatoshi Kozaki,a Yoshiro Yamashitab,† and Keiji Okadaa,*
aDepartment of Chemistry, Graduate School of Science, Osaka City University, Sugimoto, Sumiyoshi-ku, Osaka 558-8585, Japan
bInstitute for Molecular Science, Myodaiji, Okazaki 444-8585, Japan
Received 30 May 2001; accepted 24 August 2001
Abstract—A new electron-donor, benzoxazinophenoxazine 1, was synthesized. It showed a low oxidation potential of +0.21ꢀ
+0.23 V versus SCE. The neutral molecule, its radical cation salts, and charge transfer complexes were synthesized and
characterized. © 2001 Elsevier Science Ltd. All rights reserved.
The synthesis and characterization of electron-donating
compounds are important not only in their chemistries
but also in their material sciences including their solid-
state properties, e.g. conductivity, magnetism, and elec-
troluminescence. To date, many studies have focussed
on 6p- or 7p-electronic systems in five-membered hete-
rocyclic rings (e.g. thiophenes1 or TTFs1–3). Although
8p-electronic systems in six-membered heterocyclic
rings (e.g. phenoxazines and phenothiazines) are well
known as good electron-donors, studies of such com-
pounds have revealed little.4–6
The benzoxazino- and benzothiazino-condensed phe-
noxazines and phenothiazines, 1 and 2, are interesting
compounds expected to have superior electron-donat-
ing ability. The sulfur-compound 2 has recently been
synthesized by two groups: Mu¨llen’s group7 used
nitrene-induced cyclization (Cadogan cyclization8) and
Silberg’s group used a more convenient sulfur-bridging
method9,10. Electrochemical studies of the neutral 2 and
spectral studies (UV–vis and EPR) of the radical
cations, charge transfer complexes, and deprotonated
radical species have been reported, but the solid-state
properties of these species have not been clarified. In
contrast to 2, the oxygen analogue 1, which is expected
to be a better electron-donor, has not been described in
the literature. Obviously, 1 cannot be synthesized by
Silberg’s method because of the lack of a suitable
oxygen-bridging method. We report the synthesis of 1,
Scheme 1. Reagents and conditions: (a) o-bromophenol (2 equiv.), NaH/DMSO, 90°C, 90 min; (b) SnCl2·2H2O/EtOH, 70°C, 40
min; (c) Ac2O, 0°C, 15 min; (d) Cu, K2CO3/nitrobenzene, 180–185°C, 50 min; (e) KOH (20 equiv.)/EtOH, rt, 25 min; (f) n-BuLi,
Me2SO4/THF-toluene (1:2).
Keywords: benzoxazinophenoxazine; charge transfer complex; radical cation; conductivity.
* Corresponding author. Tel.: +81-6605-2568; fax: +81-6690-2709; e-mail: okadak@sci.osaka-cu.ac.jp
† Present address: Department of Electronic Chemistry, Interdisciplinary Graduate School of Science and Engineering, Tokyo Institute of
Technology, 4259 Nagatsuta-cho, Midori-ku, Yokohama 226-8502, Japan.
0040-4039/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(01)01575-1

7592
T. Okamoto et al. / Tetrahedron Letters 42 (2001) 7591–7594
its radical cation 1+, and charge-transfer complexes
together with their properties.
Synthesis of 1,4-benzoxazino[2,3-b]phenoxazine (1a)
and the N-methyl derivative (1b) was achieved through
an intramolecular Ullmann coupling as a key step as
illustrated in Scheme 1. The nucleophilic aromatic sub-
stitution reaction of 1,5-dichloro-2,4-dinitrobenzene
with 2-bromophenolate in DMSO gave 4 with 80%
yield. Reduction of 4 followed by acetylation provided
6 in good yield. Intramolecular Ullmann coupling giv-
ing 1c proceeded with moderate yield by heating at
180–185°C for 50 min. Hydrolysis of 1c gave 1a with
good yield. The methylation through lithiation in
THF:toluene=1:2 gave 1b with 78% yield. A similar
approach was possible for the known sulfur analogue
2a, which was hardly soluble in many organic solvents;
thus, N-methylation under similar conditions leading to
2b was difficult. The compound 2b was successfully
prepared (72% yield) in DMSO using dimsyl sodium as
a base and dimethyl sulfate as a methylation reagent.
Figure 1. Cyclic volutammograms of 1b and 2b in DMF.
Because of the low oxidation potentials, the oxygen
derivatives 1 can be expected to form charge-transfer
(CT) complexes with suitable electron-acceptors. For
instance, when 1b was mixed with an equivalent
The oxidation potentials of the oxygen derivatives 1a,b
are shown in Table 1 with those of the sulfur analogues
2a,b. The oxygen derivatives 1a,b have lower oxidation
potentials than the sulfur derivatives 2a,b.‡ The first
oxidation step (E1) was reversible for the N–H and
N–Me derivatives, 1a,b and 2a,b. The second oxidation
step (E2) was irreversible for the N–H derivatives 1a
and 2a, and the N–Me oxygen derivative 1b, but
reversible for the sulfur derivative 2b (Fig. 1). The
E1-values for these ring-fused compounds are smaller
than the oxidation potential of N-methylphenoxazine
(+0.59 V under similar conditions) or N-methylpheno-
thiazine (+0.76 V under similar conditions), showing
an efficient ring-fusion effect. The E1-values of the
oxygen derivatives 1a,b are even smaller than TTF
(+0.30 V under similar conditions).
Table 1. Oxidation potential of 1a–2b and the related
compoundsa
Compd
E1
E2
1a
1b
2a
2b
+0.21b
+0.23b
+0.30b
+0.50b
+0.43c
+0.74c
+0.51c
+0.85b
a V versus S.C.E. measured in DMF in the presence of n-Bu4NClO4
(0.1 M) with the sweep rate of 50 mV/s.
b Half wave potential.
Figure 2. Molecular structure of the neutral 1b (top), the
charge transfer complex 1b-TCNQ·CH3CN (middle), and the
crystal packing structure of 1b-TCNQ·CH3CN (bottom); in
the middle figure, CH3CN solvent molecules incorporated in
c Peak potential.
,
the crystal lattice was omitted for clarity; bond length (A):
‡ The difference in oxidation potentials between the phenoxazines and
phenothiazines is especially large for the N–Me derivatives. This is
probably related to the planarity of these cation radical states. The
planar N–Me phenothiazine radical cations would experience higher
steric repulsion between the NꢀMe group and the hydrogens at
peri-positions because of the longer CꢀS bond length (compared to
the CꢀO bond length).
a,a%=1.41, 1.41; b,b%=1.40, 1.40; c,c%=1.38, 1.39; d,d%=1.38,
1.40; e,e%=1.40, 1.40; f,f%=1.38, 1.38; g,g%=1.41, 1.40, h,h%=
1.38, 1.39; i,i%=1.37, 1.36, dihedral angle (°) for the neutral
1a: plane A/B=18.3, A/C=24.4, B/C=6.1, dihedral angle (°)
for the 1a-TCNQ·CH3CN: A/B=2.2, A/C=4.2, B/C=2.1.

T. Okamoto et al. / Tetrahedron Letters 42 (2001) 7591–7594
7593
amount of TCNQ, dark-green solids were precipitated.
Recrystallization from CH3CN gave a pure CT com-
plex with a formula of 1b-TCNQ·CH3CN. The IR
spectrum showed 2189 cm−1 (KBr) as a CN stretching
vibrational absorption, indicating that the degree of
charge-transfer was 0.81.11
symmetry elements for the space group P1 (No. 2). The
molecules, TCNQ (D in Fig. 2, bottom), 1b (E), and
TCNQ (F) are almost in parallel planes and the dis-
,
tance between these planes is about 3.2 A for D,
,
E-planes, and about 3.3 A for E, F-planes. These values
are shorter than twice the van der Waals radii of
14
,
aromatic rings (1.77 A for benzene), which indicates
Fig. 2 compares the structures of neutral 1b and CT-
complex 1b-TCNQ·CH3CN as well as the crystal pack-
ing structure of 1b-TCNQ·CH3CN.§ While the neutral
1b has a shallow butterfly structure (Fig. 2, top), the
CT complex has an almost planar structure (middle).
Furthermore, the length of the CꢀC double bond (e or
good orbital overlap between the electron-donor and
-acceptor. Acetonitrile solvents (linear molecules indi-
cated in Fig. 2, bottom) are located around the charged
TCNQ and these components make a mixed stack
along the c-axis.
,
e% bond in Fig. 2, middle; 1.40 A) of the
dicyanomethylene moiety of TCNQ is an intermediate
12
,
value between those of the neutral TCNQ (1.37 A )
+
−
13
,
and K TCNQ (1.42 A ), suggesting the partial charge
transfer character in accordance with the IR spectrum.
The bond length differences for the structure of the
donor moiety between the charged and neutral species
were small. The crystal packing structure of 1b-
TCNQ·CH3CN is also shown (Fig. 2, bottom) with the
Figure 4. Solution EPR spectra of 1b-TCNQ·CH3CN (upper)
and 1b+PF6− (lower) in CH3CN.
Figure 3. UV–vis spectra of 1b-TCNQ·CH3CN (upper) and
1b+PF6− (lower) in CH3CN.
Table 2. The charge transfer complexes and radical ion
salts with physical properties
§ Crystal data of 1b; C20H16N2O2, monoclinic, Space group P21/c
,
(No.14), a=15.133(1), b=13.167(1), c=16.902(2) A, i=
Compd.a
Method
Color
Solvent
Conductivity
(S cm−1
3
114.758(6)°, V=3058.3(4) A , Z=8, Dc=1.374 g cm−3; Cu Ka
,
)
,
radiation (u=1.54178 A); 2q<130.2°; 5219 unique reflections of
which 3323 were treated as observed [I>3|(I)]; R=0.057. Rw=
0.086, Crystal data of 1b-TCNQ·CH3CN; C34H23N7O2, triclinic,
1a-TCNQ
1b-TCNQ
Mixing
Mixing
Blue
CH3CN
7.8×10−6
9.9×10−4
5.6×10−7
3.0×10−5
9.8×10−6
Dark green CH3CN
,
space group P1 (No. 2), a=12.423(3), b=15.764(4), c=8.130(2) A,
−
1b+ClO4
Electrolysis Dark green THF
Electrolysis Dark green PhCl
Electrolysis Dark green THF
3
,
h=90.84(2), i=108.47(2), k=67.98(2)°; V=1390.2(7) A , Z=2,
−
1b2+I3
Dc=1.342 g cm−3; Mo Ka radiation (u=0.71069 A; 2q<55.0°; 6394
,
1b+PF6
−
unique reflections of which 4606 were treated as observed [I>3|(I)];
R=0.053. Rw=0.087. These data have been deposited with the
Cambridge Crystallographic Data Center (CCDC 163017 for 1b,
CCDC 163016 for 1b-TCNQ·CH3CN).
a The chemical formula accords with the elemental analysis within
0.3% except for 1a-TCNQ( 0.5%).

7594
T. Okamoto et al. / Tetrahedron Letters 42 (2001) 7591–7594
In CH3CN solution, the CT complex (1b-TCNQ·
CH3CN) did not show the charge transfer absorption
band.15 Instead, neutral and charged species were
observed (Fig. 3). The presence of the neutral species is
in accordance with the partial charge transfer character
of 1b-TCNQ·CH3CN. The spectral assignment of 1b+
was achieved by comparison with 1b+PF6− prepared
electrochemically.
References
1. Ba¨uerle, P. In Electronic Materials: The Oligomer
Approach; Mu¨llen, K.; Wegner, G., Eds.; Wiley-VCH:
Weinheim, 1998; pp. 105–197.
2. Ishiguro, T.; Yamabe, K.; Saito, G. In Organic Supercon-
ductors; Fulde, P., Ed.; Springer: Telos, 1998; pp. 125–
220.
3. Kobayashi, H.; Kobayashi, A.; Cassoux, P. Chem. Soc.
Rev. 2000, 29, 315–324.
4. Kobayashi, H. Bull. Chem. Soc. Jpn. 1973, 46, 2945–
2949.
Similarly, the EPR spectrum of the CT complex (1b-
TCNQ·CH3CN) in CH3CN solution showed mixed sig-
nals due to the TCNQ radical anion16 and 1b+.
Identification of 1b+-signals was achieved by comparing
5. Oza, A. T. Czech. J. Phys. 1983, B33, 1148–1153.
6. Kozawa, K.; Hoshizaki, T.; Uchida, T. Bull. Chem. Soc.
1991, 64, 2039–2044.
with 1b+PF6 (g=2.0034, formally nine lines with split-
−
ting of ca 4.5 G) (Fig. 4).
7. Kistenmacher, A.; Baumgarten, M.; Enkelmann, V.;
Pawlik, J.; Mu¨llen, K. J. Org. Chem. 1994, 59, 2743–
2747.
8. Cadogan, J. I. G. Synthesis 1969, 1, 11–17.
9. Silberg, I. A.; Cristea, C. Hetreocyclic Commun. 1996, 2,
117–124.
Table 2 lists the colors and formulas (determined by
elementary analysis) of the CT-complexes or the radical
ion salts prepared by mixing with TCNQ or by electrol-
ysis in the presence of suitable electrolytes. In contrast
to 1a and 1b, the sulfur analogues did not give com-
plexes with well-definable elemental composition under
similar conditions. The conductivity for the powdered
sample of 1b-TCNQ·CH3CN showed a low conductiv-
ity (3×10−7 S cm−1). However, removing the incorpo-
rated CH3CN solvent by heating under vacuum
increased the conductivity sharply. The value in Table 2
is for the powdered sample whose elemental analysis
coincided with the formula of 1b-TCNQ. The conduc-
tivity for other salts (powdered samples) is also summa-
rized in Table 2, showing that these compounds are
semiconductors with 10−3–10−7 S cm−1.
10. Jitaru, M.; Cristea, C.; Silberg, I. A. Rev. Roum. Chim.
1999, 44, 863–866.
11. Chappell, J. S.; Bloch, A. N.; Bryden, W. A.; Maxfield,
M.; Poehler, T. O.; Cowan, D. O. J. Am. Chem. Soc.
1981, 103, 2442–2443.
12. Long, R. E.; Sparks, R. A.; Trueblood, K. N. Acta
Crystallogr. 1965, 18, 932–939.
13. Konno, M.; Ishii, T.; Saito, Y. Acta Crystallogr. 1977,
B33, 763–770.
14. Bondi, A. J. Phys. Chem. 1964, 68, 441–451.
15. Mulliken, R. S.; Person, W. B. Molecular Complexes;
John Wiley & Sons: New York, 1969; pp. 1–298.
16. Rataiczak, R. D.; Jones, M. T.; Reeder, J. R. Mol. Phys.
1985, 56, 65–77.
Compounds data for 1a and 1b are presented in the
footnote.¶
¶ Compounds data: 1a; pale yellow powder, mp>300°C, MS (EI) 288
(M+), 1H NMR (300 MHz, in DMSO-d6): l 7.93 (s, 2H), 6.72–6.67
(m, 2H), 6.54–6.50 (m, 4H), 6.40 (d, 2H, J=7.5 Hz), 6.07 (s, 1H),
5.73 (s, 1H); 13C NMR (75.45 MHz, in DMSO-d6): l 142.37,
135.35, 132.26, 127.56, 123.67, 120.09, 114.94, 113.16, 103.63, 98.79.
This compound is rather unstable under aerated conditions. 1b;
colorless plate, mp=198°C, MS (FAB) 605 (MH+), 1H NMR (300
MHz, in DMSO-d6): l 6.89–6.83 (m, 2H), 6.70–6.66 (m, 6H), 6.28
(s, 1H), 6.12 (s, 1H), 3.06 (s, 6H); 13C NMR (75.45 MHz, in
DMSO-d6): l 144.45, 137.74, 134.70, 130.37, 124.04, 120.75, 114.76,
112.01, 103.58, 97.50, 31.17.