Synthesis, structure, and electron-donating ability of 2,2′:6′, 2″-dioxatriphenylamine and its sulfur analogue

Article abstract of DOI:10.1246/cl.2004.1174

2,2′:6′,2″-Dioxa- and dithia-triphenylamines 4a and 4b were prepared. Both compounds had a C₂-like structure. The sulfur compound 4b had a more twisted structure than 4a; the dihedral angle between two edge phenyl rings was 43° for 4a, and 62°

Full text of DOI:10.1246/cl.2004.1174

1174
Chemistry Letters Vol.33, No.9 (2004)
Synthesis, Structure, and Electron-Donating Ability of 2,20:60,200-Dioxatriphenylamine
and Its Sulfur Analogue
Masato Kuratsu, Masatoshi Kozaki, and Keiji Okadaꢀ
Department of Chemistry, Graduate School of Science, Osaka City University,
Sugimoto, Sumiyoshi-ku, Osaka 558-8585
(Received June 14, 2004; CL-040679)
2,20:60,200-Dioxa- and dithia-triphenylamines 4a and 4b were
prepared. Both compounds had a C2-like structure. The sulfur
compound 4b had a more twisted structure than 4a; the dihedral
angle between two edge phenyl rings was 43ꢁ for 4a, and 62ꢁ for
4b. The oxygen compound 4a had a reversible oxidation wave at
þ0:81 V vs SCE, whereas 4b had an irreversible peak at
þ1:14 V.
NO2
Br
NO2
Br
a
b
Cl
Cl
X
X
X = O, 65%
X = S, 72%
X = O, 80%
X = S, 80%
5a,b
Br
NH2
Br
c (for 6a)
X
X
S
4a
34%
6a,b
d
Some p-substituted triphenylamine radical cations are stable
species in a propeller-like structure in solid state.1 However, the
unsubstituted triphenylamine shows an irreversible oxidation
peak at þ0:98 V vs SCE in its cyclic voltammogram,2 indicating
that the radical cation has much poorer stability. Flattening the
structure into a planar form would increase the stability and low-
er the oxidation potential of triphenylamine. Hellwinkel and
Melan prepared interesting compounds, 1 and 2;3a,b the latter un-
derwent smooth oxidation to give well-resolvable EPR spectrum
of the radical cation.3c They also described tri(heteroatom)-
bridged triphenylamines 3a, 3b as interesting unknown species
of this class.3a Although these compounds have also been refer-
red to in a patent,4 the synthetic routes have not so far been re-
ported. The related trioxatriphenylphosphine was recently pre-
pared and shown to be a bowl-type structure exhibiting a pyro-
electric property by Krebs and co-workers.5
73%
(for 6b)
I
NH2
I
e
S
4b
86%
7b
Scheme 1. (a) o-Bromophenol or o-bromothiophenol, NaH/
DMSO, heat. (b) H2NNH2ꢃH2O, Pd/C, p-bromophenol/EtOH,
reflux. (c) NaOt-Bu, Pd(dba)2, P(t-Bu)3/toluene, reflux. (d) KI–
Cu/HMPA, 160 ꢁC. (e) K2CO3–Cu/o-dichlorobenzene, reflux.
considered a stepwise route using a linear intermediate 6 as a
key intermediate (Scheme 1). Aromatic nucleophilic substitution
reaction of 2,6-dichloronitrobenzene with o-bromophenol (5
equiv.) or o-bromothiophenol (2 equiv.) using dimsyl anion as
a base gave 5a (65%, 130 ꢁC for 1 day) or 5b (72%, 90 ꢁC for
2 h). The reduction of 5a, 5b was carried out by heating with hy-
drazine monohydrate (30 equiv.) in ethanol in the presence of p-
bromophenol (10 equiv.) and a Pd/C catalyst for 2 h, giving 6a,
6b in good yields. The presence of p-bromophenol was essential
to avoid a competing reduction of aromatic bromide. Intramolec-
ular cyclization of 6a was achieved using Pd(0)-mediated cross-
coupling reaction conditions [NaOt-Bu–Pd(dba)2–P(t-Bu)3 in
toluene, reflux 2 h],7 affording the desired 4a8 in 34% yield. Ap-
plication of similar conditions to 6b produced 4b in less than 3%
yield. Then, halogen exchange (Br to I) was examined (KI–CuI
in HMPA, 160 ꢁC for 16 h), giving crude iodide 7b (73% yield).
The cross-coupling reaction of 7b under the above conditions or
the Ullmann reaction conditions (K2CO3–Cu in o-dichloroben-
zene, reflux 5 h) gave 4b8 in good yields.
Me Me
O
N
N
Me
Me
Me
Me
O
O
2
1
X
N
N
X
X
X
X
3, a: X = O, b: X = S
4, a: X = O, b: X = S
The structures of these compounds were determined by X-
ray structure analysis.9 Figure 1 shows the ORTEP drawings
of these compounds (left for 4a, right for 4b). Both compounds
were in a pseudo-C2-like structure. Two similar but independent
structures were found in a unit cell for 4b. One of them is shown
in the figure. The selected bond lengths and angles around the
heteroatoms are shown in the caption of Figure 1. The C–O,
In the course of our study to develop new electronic and
magnetic materials,6 we have been interested in planar triphenyl-
amine radical cations 3ꢂþ as spin-building blocks of magnetic
materials. In order to develop a general synthetic path for these
compounds and also to obtain insight into their electron-donat-
ing ability, we have investigated the lesser bridged triphenyl-
amines, 2,20:60,200-dioxatriphenylamine 4a and its sulfur ana-
logue 4b. In this paper, we report the synthesis, structure, and ox-
idation potential of 4a, 4b.
ꢀ
C–S, and C–N bond lengths were in the region of 1.36–1.42 A
ꢀ
ꢀ
for C–O, 1.69–1.82 A for C–S, 1.33–1.47 A for C–N bonds. Be-
cause of the longer C–S bond compared to the C–O bond, the di-
hedral angle between two edge phenyl rings was larger in the
For the synthesis of heteroatom-bridged triarylamines, we
Copyright Ó 2004 The Chemical Society of Japan

Chemistry Letters Vol.33, No.9 (2004)
1175
R. I. Walter, J. Am. Chem. Soc., 99, 6910 (1977). c) R. P.
Scaringe and S. L. Reynolds, Acta Crystallogr., Sect. A,
A37, C202 (1981).
A
B
γ
a
A
γ
B
b
a
b
f
2
3
E. T. Seo, R. F. Nelsen, J. M. Fritsch, L. S. Marcoux, O. W.
Leedy, and R. N. Adams, J. Am. Chem. Soc., 89, 3498
(1966).
a) D. Hellwinkel and M. Melan, Chem. Ber., 104, 1001
(1971). b) D. Hellwinkel and M. Melan, Chem. Ber., 107,
616 (1974). c) F. A. Neugebauer, S. Bamberger, and W. R.
Groh, Chem. Ber., 108, 2406 (1975).
H. Takizawa, Jpn. Kokai Tokkyo Koho JP 11339868 (1999).
F. C. Krebs, P. S. Larsen, J. L. Larsen, C. S. Jacobsen, C.
Boutton, and N. Thorup, J. Am. Chem. Soc., 119, 1208
(1997).
a) S. Hiraoka, T Okamoto, M. Kozaki, D. Shiomi, K. Sato, T.
Takui, and K. Okada, J. Am. Chem. Soc., 126, 58 (2004).
b) T. Okamoto, E. Terada, M. Kozaki, M. Uchida, S.
Kikukawa, and K. Okada, Org. Lett., 5, 373 (2003). c) T.
Okamoto, M. Kozaki, Y. Yamashita, and K. Okada,
Tetrahedron Lett., 42, 7591 (2001).
α
d
β
β
α
δ
c
ε
d
δ
c
f
ε
g
e
e
g
C
C
4a
4b
Figure 1. Molecular structures of 4a and 4b drawn at 50% el-
lipsoid level; hydrogen atoms are eliminated for claritꢁy. Selected
4
5
ꢁ
ꢀ
bond lengths (A), angles ( ), and dihedral angles ( ) between
phenyl planes, A, B, and C. For 4a measured at 123 K: a:
1.333(6), b: 1.472(6), c: 1.406(4), d: 1.401(6), e: 1.373(6), f:
1.364(6), g: 1.420(6), ꢂ ¼ 114:6ð5Þ, ꢃ ¼ 118:1ð5Þ, ꢄ ¼
127:1ð2Þ, ꢀ ¼ 117:3ð4Þ, " ¼ 113:5ð4Þ, A/B ¼ 43:0, A/C ¼
25:5, B/C ¼ 24:7. For two independent molecules of 4b meas-
ured at 298 K: a: 1.459(10), 1.459(9), b: 1.363(10), 1.382(10), c:
1.452(8), 1.382(9), d: 1.793(8), 1.736(7), e: 1.737(8), 1.818(10),
f: 1.762(9), 1.774(9), g: 1.685(11), 1.796(10), ꢂ ¼ 120:3ð7Þ,
123.0(7), ꢃ ¼ 118:4ð7Þ, 116.8(6), ꢄ ¼ 121:3ð6Þ, 119.9(6), ꢀ ¼
99:1ð4Þ, 98.8(4), " ¼ 98:1ð4Þ, 101.6(4), A/B ¼ 62:3, 62.4,
A/C ¼ 32:4, 32.1, B/C ¼ 36:0, 36.9.
6
7
8
a) J. F. Hartwig, M. Kawatsura, S. I. Hauck, K. H.
Shaughnessy, and L. M. Alcazar-Roman, J. Org. Chem.,
64, 5575 (1999). b) T. Yamamoto, M. Nishiyama, and Y.
Koie, Tetrahedron Lett., 39, 2367 (1998). c) D. W. Old,
J. P. Wolfe, and S. L. Buchwald, J. Am. Chem. Soc., 120,
9722 (1998).
sulfur derivative 4b; A/B ¼ 43ꢁ for 4a, and 62ꢁ for 4b. As a re-
sult, the sulfur derivative 4b is more twisted than the oxygen de-
rivative 4a. This is also supported by the small C(sp2)–S–C(sp2)
bond angles (98–102ꢁ) for 4b compared to a general nonstrained
diphenylsulfide structure (ca. 105ꢁ).10
The electron-donating ability was estimated using cyclic
voltammetry.11 The oxygen derivative 4a showed a reversible
oxidation peak at E1=2 ¼ þ0:81 V vs SCE in DMF. On the other
hand, the sulfur derivative 4b showed an irreversible peak at
Ep ¼ þ1:14 V with a shoulder (þ1:04 V), indicating instability
of the radical cation. The oxidation potential of 4a was lower
than that of triphenylamine (Ep ¼ þ1:07 V under the identical
conditions) but higher than that of 10-phenylphenoxazine 8
(E1=2 ¼ þ0:75 V). The peak potential of 4b was higher than that
4a; colorless prisms, mp 165 ꢁC, MS (FAB) m=z 273 (Mþ),
1H NMR (400 MHz, in DMSO-d6): ꢀ 6.64 (d, 2H, J ¼
8:3 Hz), 6.88 (t, 1H, J ¼ 8:3 Hz), 6.96–7.08 (m, 6H), 7.37
(d, 2H, J ¼ 8:2 Hz); 13C NMR (100 MHz, in CDCl3): ꢀ
111.06, 114.60, 117.42, 120.92, 123.34, 123.54, 123.60,
129.09, 145.33, 146.98, Anal. Calcd. for C18H11NO2: C,
79.11; H, 4.06; N, 5.13; Found: C, 78.88, H, 3.98, N, 5.02.
4b; pale yellow prisms, mp 202 ꢁC, MS (FAB) m=z 305
(Mþ), 1H NMR (400 MHz, in CDCl3): ꢀ 6.93–6.98 (m,
1H), 6.98–7.02 (m, 2H), 7.04 (ddd, 2H, J ¼ 7:6, 7.2,
1.4 Hz), 7.13 (ddd, 2H, J ¼ 8:1, 7.2, 1.6 Hz), 7.18 (dd, 2H,
J ¼ 8:1, 1.4Hz), 7.21 (dd, 2H, J ¼ 7:6, 1.6 Hz); 13C NMR
(100 MHz, in CDCl3) ꢀ 120.64, 124.67, 124.83, 125.65,
125.87, 127.08, 127.45, 127.92, 139.55, 142.64, Anal. Calcd.
for C18H11NS2: C, 70.79; H, 3.63; N, 4.59; Found: C, 70.62;
H, 3.52; N, 4.51.
of triphenylamine and 10-phenylphenothiazine
9
(E1=2
¼
þ0:77 V). These results strongly suggest that the planarity in
radical cation states is very important for the oxidation of these
compounds. The radical cation species 4aꢂþ, 4bꢂþ should not be
planar and 4bꢂþ should be more twisted than 4aꢂþ. This factor
explains the higher oxidation potential of 4b. The lower oxida-
tion potential of 8, compared to that of 4a, can also be rational-
ized by the planar (twisted) phenoxazine radical cation frame-
work in 8ꢂþ (4aꢂþ).
9
Crystallographic data for 4a: a cololess prism, orthorhombic,
ꢀ
ꢀ
space group Pna21 (#33), a ¼ 7:0441ð9Þ A, b ¼ 9:738ð2Þ A,
ꢀ
ꢀ 3
c ¼ 17:841ð3Þ A, V ¼ 1223:8ð3Þ A , Z ¼ 4, T ¼ 123 K,
R1ðI > 2ꢁÞ ¼ 0:050, RwðI > 0ꢁÞ ¼ 0:067, Crystallographic
data for 4b, a pale yellow prism, orthorhombic, space group
The present study establishes a general synthetic route to
heteroatom-bridged triphenylamines. The sulfur derivative 4b
turned out to be in a largely twisted structure giving a higher
and irreversible oxidation peak. Preliminary calculations for
the targets 3a and 3b using the present bond lengths and angles
as initial parameters in the geometrical optimization suggests
that the sulfur derivative 3b is in a deep bowl-shape, whereas
the oxygen derivative 3a is in a shallow bowl-shape. Oxidation
of 3a would give rise to shrinkage (ꢄ2%) of the C–N bonds,6a
leading to a planar 3aꢂþ. It seems that 3aꢂþ is an ideal spin build-
ing block for our purpose. Synthesis of 3a and 3aꢂþ is under way.
ꢀ
ꢀ
P212121 (#19), a ¼ 8:231ð2Þ A, b ¼ 17:792ð4Þ A, c ¼
ꢀ
ꢀ 3
19:180ð3Þ A, V ¼ 2808:7ð1Þ A , Z ¼ 8, T ¼ 293 K, R1ðI >
2ꢁÞ ¼ 0:069, RwðI > 0ꢁÞ ¼ 0:073. CCDC 240403 and
240404, These data can be obtained free of charge via
www.ccdc.cam.ac.uk/conts/retrieing.html (or from the
Cambridge Crystallographic Data Center, 12 Union Road,
Cambridge CB21EZ, UK; fax: (+44)1223-336-033; or
deposite@ccdc.cam.ac.uk).
10 G. D. Andreetti, J. Garbarczyk, and M. Krolikowska, Cryst.
Struct. Commun., 10, 789 (1981).
11 The cyclic voltammograms were measured in DMF in the
presence of tetrabutylammonium perchlorate (0.1 M) using
glassy carbon as a working electrode, Pt as a counter
electrode, and SCE as a reference electrode at sweep rate
of 50 mV/s.
References and Notes
a) F. A. Bell, A. Ledwith, and D. C. Sherington, J. Chem.
Soc. C, 1969, 2719. b) G. M. Brown, G. R. Freeman, and
1
Published on the web (Advance View) August 14, 2004; DOI 10.1246/cl.2004.1174