Structures, properties, and applications of soluble polyazulene and azulene containing copolymers 4

Chapter 7
Novel polyradicals stabilized by the vertical and
horizontal delocalization of the electrons
Introduction
Organic radicals are usually known as unstable, transient intermediates in organic
reactions. However, some organic radicals, such as radical crystals of galvinoxyl and
stericly hindered nitroxyl, are stable enough to remain unreact for several months at room
temperature.
1
In the past decade, stable organic radicals, based on crystals of small
radicals (e.g., nitroxides) or charge transfer salts, have been prepared.
2-4
Yet, despite
extensive research, the experimental structures of only four types of stable organic
radicals have been reported: (1) triarylmethyl radical, R1; (2) a per-pyridinium
substituted allyl radical, R2; (3) cyclopentadienyl radical, R3; and (4)
dodecamethylcarba-closo-dodecaboranyl radical (CB
11
-Me
12
•). From analysis of these
stable radical structures, we can conclude that all of these radicals are strongly stabilized
thermodynamically by effective delocalization of the unpaired electron, and kinetically
by the bulky substituents.
5,6

XC
3
P+ P+
P+

P++P
P+ = 4-Me
N-pyridine
2
R1
R2
R3

Another strategy for preparing stable radicals is to use polymeric building blocks to
construct a molecular based radical; polymers bearing a number of free radical groups are
called polyradicals.
7
In the past decade, there has been a very strong interest in the

220

synthesis and properties of conjugated polyradicals, based on the hope that such systems
can eventually be utilized as components of designable magnetic materials. Numerous π-
conjugated polymer substitutes with pendant radicals have been synthesized and
characterized.
8-11
For example, Rajca et al. have created an elegant body of work on
backbone-conjugated and dendritic polyradicals, based on triarylmethyl radicals.
12,13

Dougherty and colleagues have also worked on the synthesis and characterization of
doped plaronic conjugated materials with ferromagnetic (FM) and high spin-coupling.
14,15

Various researchers have synthesized conjugated polymers employing pendant

polynitroxide, polynitronylnitroxide, and polyphenoxyl types of spin-bearing units.
16-18

However, most of the polyradicals are designed with the radical spin sites as pendant
groups. It is often difficult in finding a suitable radical center that can be conjugated and
linked to the polymer backbone.
In Chapter 3, when we studied the electronic properties of our copolymers, we found that
they formed highly stable cation radicals, either by iodine doping or TFA protonation. In
the solid state, a protonated copolymer film showed no significant change in the EPR
signal, even after 2 weeks. Furthermore, the stability was also investigated in detail using
the nitrogen and oxygen permeation test. The high stability was attributed to the stability
of the azulenium ion and the delocalization of the electrons along the conjugated polymer
backbone.
These results tell us that azulene is possibly a suitable radical center that can be linked
into the conjugated polymers to form new polyradical systems. This is because azulene
displays many of the criteria essential for the formation of a stable cation radical. For
one, it has an asymmetric charge distribution. It has a tendency to stabilize cations, as
well as anions, due to its remarkable polarizability.
19
Secondly, azulene and their

221

alkylated derivatives are quite unique in that the ongoing disruption of aromaticity from
the neutral to the charged state is counterbalanced by the gain in resonance energy upon
formation of azulenylium carbocation, a 6π-electron aromatic tropylium analogue (Figure
7-1).
20
Thus, azulene appears to be an extremely novel and versatile system with regard
to radical centers.

Figure 7-1. Resonance forms of azulenium carbocation.
When the azulene was inserted into the polymer backbone via 1,3-position, and after
oxidation or protonation, the azulenium cation formed at the seven-membered ring and
the radicals were formed on the five-membered ring that is linked to the conjugated
polymer backbone. Thus the un-paired electron can be mediated in a vertical by the
aromatic tropylium cation; what is more, the un-paired electrons can also be delocalized
along the conjugated polymer backbone. This concept is illustrated clearly in Scheme 7-
1.
Conjugated Units
Radical Center
Radicals
Direction of
Delocalization
of Electrons

Scheme 7-1. Illustration of the design of the stable polyradicals.
Based on this analysis, we designed a conjugated polymer system containing azulene in
the polymer backbone, and used a 1,3-conjugation (Figure 7-2) to demonstrate the

222

concept. The aims of the present study were to prepare a highly stable polyradical system
and study their interesting properties.
R
R
R
R
R
R

R
R
R
R
R
R
Ox
[H
+
]
R = C
8
H
17
; OC
8
H
17

Figure 7-2. The design of the high stable polyradical system based on the conjugated copolymers
containing azulene moiety.

Results and Discussion
Monomer synthesis and characterization of the cation radical
To understand better of the formation of the cation radicals, the model compound, 1,3-
diphenylazulene, was first prepared. 1,3-Diphenyl-azulene was synthesized following the
general Grignard reaction (Scheme 7-1). One equiv of 1,3-dibromoazulene was reacted
with 2.5 equiv. of Grignard reagent of bromobenzene in anhydrous ether, yielding the
green oil product, which solidified after long-term storage.
Br

Br MgBr
i
ii/iii
Br

Scheme 7-1. The synthesis of 1,3-diphenyl-azulene.


223

The monomer 1,3-diphenylazulene was then characterized by NMR and FT-IR, then the
formation of the azulenium cation radical was confirmed by respective HNMR and EPR
spectrum.
(a)
(b)
Figure 7-3. HNMR spectrum of 1,3-diphenyl-azulene (a) before and (b) after protonation with
trifluoroacetic acid.
Figure 7-3 is the HNMR spectra of 1,3-diphenyl-azulene before and after the addition of
deuterated TFA. Before the addition of TFA, peaks at 8.58 (d), 8.16 (s), and 7.17 (t) were
attributed to the protons, H4,8, H2, and H5,7 on the azulene ring. The H6 peak of azulene
was overlapped by multiple peaks of the benzene ring. The pattern of these peaks
indicated the formation as expected symmetrically structured compound. After addition
of TFA, an important feature noted was the formation of new peaks in the low field. As
shown in Figure 7-3b, two new peaks appeared at 9.08 ppm and 8.92 ppm, which were
attributed to the protons 4,8 and 5,7 on the seven-membered ring of the azulenium

224

cation.
22,23

We also found that the compound was partially protonated, as can be seen
from the peaks at 8.7 ppm and 7.2 ppm that were attributed to the H4,8 and H5,7 of the
neutral azulene. From from the integration, about 30% protonation was found. This
conclusion was confirmed by further EPR and UV-vis studies. Additionally, the HNMR
spectrum of the protnated 1,3-diphenyl-azulene also suggests that the protonation mainly
occurred on the azulene ring, as there was little change in the peaks belonging to the
benzene ring (Figure 7-3b).
The UV-vis spectra of 1,3-diphenyl-azulene showed an maximum absorption wavelength
(λ
max
) of 387 nm (in chloroform), suggesting a developed π-conjugation between the
azulene and benzene rings. Upon addition of TFA to the solution, two new peaks appear
around 466 nm and 353 nm suggesting the formation of the azulenium cations.
24
On
further protonation, the new bands had a gradual increase, accompanied with a decrement
of the bands at 300 nm. The color of 1,3-diphenyl-azulene solution changed from
yellowish-green to brown. Comparing the absorption intensity changes of the new peaks
with the absorption band at 300 nm, we can conclude that the compound is partially
protonated. This is in agreement with what we observed in the HNMR study.
400 600
0
1
2
3
TFA concentration
40%
30%
20%
10%

5%
1%
Absorption
Wavelength (nm)
(a)

225

400 600 800
0
1
2
3
4
2 days
2mins
Time
2 mins
10 mins
30 mins
2 days
Absorption
Wavelength (nm)
(b)
Figure 7-4. Continuous change in UV-vis spectrum of 1,3-diphenylazulene (a), with different
TFA concentration; (b) in 10% TFA concentration with different time.

Figure 7-4a also displays that the protonation can be saturated with 20% TFA solution.
This result may suggest that 1,3-diphenyl-azulene is easily protonated by TFA, even at
low TFA concentrations. At higher TFA concentrations, there is little change of the

spectrum. However, when we follow the UV-vis spectrum change of 1,3-diphenyl-
azulene in 10% TFA concentration at different storage times, we found a kinetic behavior
with the protonation process. With an increase in storage time of 10% TFA/1,3-diphenyl-
azulene solution, an increasing amount of 1,3-diphenyl-azulene was protonated.
However, the spectrum changed faster in the first 30 mins, whereas the protonation rate
slowedafter 30 mins. It is also interesting to note that after 2 days of storage at room
temperature, there is no other change of the spectrum except in the increase of intensity
of the longest wavelength absorption. This result indicates that the azulenium cation
radical is very stable.

226

The formation of the azulenium cation radical was further confirmed by EPR
experiments. A solution of 1,3-diphenyl-azulene in 10% TFA solution (in chloroform)
reveals a symmetrical singlet absorption with a g factor (2.0020) near to that of free
electrons, having a line width of 3.2 G. The EPR signal intensity change of the compound
in 10% TFA solution with time is similar to that which we observed in the UV-vis spectra
study. The stability of the cation radical was measured by testing the EPR absorption
intensity change after leaving the 1,3-diphenyl-azulene solution in 10% TFA, at
atmosphere for 2 days. The EPR spectrum showed an increase in the intensity of the EPR
signal, without obvious g value change. These results are in good agreement with the
UV-vis spectra study.
Polymers synthesis and characterization
The strategy for synthesis the desired polymers is outlined in Scheme 7-2. 1,4-Dibromo-
2,5-di-n-octylbenzene and 1,4-dibromo-2,5-n-octyloxybenzene were used as starting
materials and synthesized according to a reference procedure.
25
Here, 1,4-dibromo-2,5-
di-n-octylbenzene and 1,4-dibromo-2,5-n-octyloxybenzene were reacted with excess n-
butylthium at a low temperature, subsequently quenched with trimethyl borate, and then

hydrolyzed with hydrochloric acid to produce 2,5-dioctyloxybenzene-1,4-bis(boronic
acid), or 2,5-dioctylbenzene-1,4-bis(boronic acid).
26
The purity of the di-acid was
checked by HNMR spectroscopy and CNMR spectroscopy.
R
R
Br Br
i
ii
R
R
(HO)
2
BB(OH)
2
Br
Br
iii
iv
n
R
R
PAzBzC8 R = C
8
H
17
PAzBzOC
8
R = OC

8
H
17

Scheme 7-2. The synthesis of PAzBzC8 and PAzBzOC8 by Suzuki coupling.

227


Synthesis of the compolymers (Az-Bz) was carried out by a Suzuki coupling reaction,
with Pd(0) catalysis.
27-29
The polycondensation of equimolar quantities of the diboronic
acid with 1,3-dibromoazulene was performed in a biphasic medium (toluene/aqueous
sodium carbonate) solution at 80
0
C,

with catalytic amounts of tetrakis-
(triphenylphosphine)palladium added. The polymers were precipitated by pouring into
the organic solution into methanol. Purification was carried out by dissolveing the
polymer in chloroform and re-precipitating in methanol.
GPC analysis showed a number-average molecular weight of 6,900 and 8,100 for
PAzBzC8 and PAzBzOC8 respectively. These corresponded to a chain lengths of about
32-34 aromatic rings, which is in agreement with published degrees of polymerization
when using diboronic acid in the Suzuki polycondensation.
30
The low molecular weight
may be due to the impurities within the diboronic acids that easily condense
spontaneously to boroxines to varying degrees.

26


Figure 7-5.
1
HNMR spectrum of polymer PAzBzOC8.

These polymers were characterized by FT-IR, NMR, and elemental analyses. A
representative
1
HNMR spectrum of the polymer PAzBzOC8 is depicted in Figure 7-5.
The chemical shifts of the azulene protons were manifested at δ 8.52, 8.35, 8.10, 7.62

228

ppm which are associated, respectively, with the protons of H4,8, H2, and H6 on the
azulene. Phenyl protons appeared at δ 7.29 and 7.18 ppm. The remaining resonance at δ
3.91 ppm and 0.88-1.60 ppm corresponded to the n-octyloxy pendant chains. Due to the
presence of both the head-to-head (HH) and head-to-tail (HT) units, the resonance peaks
depicted both broad and multisignal response. Similar behavior has been reported in other
copolymers systems.
31

The FT-IR spectra of the polymers depicted strong C-H stretching (2920 and 2850 cm
-1
)
of the alkyl, or alkoxyl, side group with weak stretching of azulene and benzene at 3018
cm
-1
. Peaks at 1570, 1200, and 736 cm

-1
also confirmed the presence of the azulene
moiety and benzene ring in our copolymers. Polymer PAzBzOC8, doped with iodine and
protonared with TFA, is also illustrated in Figure 7-6. The increase in absorption in the
region of 1400-700 cm
-1
indicated the formation of cation radicals.
4000 3000 2000 1000
(c)
(b)
(a)
Relative Intensity
-1
Wavenumber (cm )

Figure 7-6. FT-IR spectrum of (a) neutral PAzBzOC8, (b), iodine doped PAzBzOC8, and (c),
TFA protonated PAzBzOC8.

Thermal properties.

229

The thermal stability of our polymers was determined using TGA in a nitrogen
atmosphere. The copolymers depicted very good thermal stability with an onset of
degradation at about 305
0
C and 412
0
C for PAzBzC8 and PAzBzOC8, respectively.
The fist degradation step of PAzBzC8 occurred at 305

0
C,

with only a 2.8% weight loss;
this is possibly due to some left over oligomers. The second degradation began at 409
0
C,
(similar to the degradation onset of PAzBzOC8), and ended at 496
0
C. The second
degradation was attributed to the cleavage of the pendant group, as calculated from the
weight change. Polymer PAzBzOC8 displayed only one-step degradation that began at a
temperature of 412
0
C, and finished at 474
0
C, which was also attributed to the cleavage of
the side chains. From these observations, we concluded a high stability for our
conjugated polymer backbones. The char yield of the polymers in nitrogen atmosphere
was keep at about 45-50%, even with temperature increases of up to 1000
0
C.
0400800
8
12
16
PAzBzOC8
PAzBzC8
Weight (%)
0

Temperature ( C)

Figure 7-7. TGA curves of polymer PAzBzOC8 and PAzBzC8.

Electrochemical Properties
The p-doping characteristics of our polymers were investigated by using cyclic
voltammetry. The onset of the p-doping process (oxidation) of PAzBzC8 occurred at a

230

potential of ca.1.0 V (vs. Ag/AgCl). The anodic current increased quickly with an anodic
peak appearing at ca. 1.2 V, and the corresponding reversible reduction peak occurred at
ca. 0.98 V. For polymer PAzBzOC8, the on-set of oxidation occur at ca. 0.82 V (vs
Ag/AgCl) with an anodic peak observed at 1.06 V. Its corresponding reversible reduction
peak appeared at ca. 0.75 V, as a small peak (Figure 7-8). Concurrently, the p-doping
process of our polymers was accompanied by an obvious color change from yellowish-
green, in the neutral film, to dark brown in the p-doped polymeric films. Comparing the
p-doping process of PAzBzC8 and PAzBzOC8, we found a significant negative shift of
the oxidation potential for PAzBzOC8. This is may be due to the electron-donating
properties of the oxygen in the alkoxy side group, which decreased the oxidation
potential of the resulting conjugated polymers.
0 500 1000 1500
PAzBzC8
PAzBzOC8
0.5 mA
Cathodic
Anodic
Potentials (mV) vs Ag/AgCl

Figure 7-8. Cyclic voltammograms of PAzBzC8 and PAzBzOC8, measure in 0.1 M TBAHP

(tetrabutylammonium hexafluoro-phosphate) solution of acetonitrile with a scan rate of 80 mV/s.

Electronic spectroscopy study
The solution-phase UV-vis absorption of our polymers was recorded at room temperature
in chloroform solution (Figure 7-9). PAzBzC8 and PAzBzOC8 exhibited the longest

231

wavelength absorption of up to 375 nm and 390 nm, respectively. For PAzBzC8, two
main absorption peaks appeared in the UV region. One strong absorption appeared at 300
nm that was attributed to the π-π* transition of the monomeric units in the polymers, with
a shoulder appearing at approximately 375 nm. The latter was attributed to the π-π*
transition of the conjugated polymer backbone. While for polymer PAzBzOC8, three
absorption bands were found in the UV region. One main absorption band appeared at
297 nm and the other two appeared as shoulders at 340 nm and 390 nm. Comparing the
longest wavelength absorption of PAzBzOC8 with that of PAzBzC8, PAzBzOC8
showed a bathochromic shift of more than 15 nm. This may suggest that the electron-
donating alkoxy side group may mediate the conjugation of the polymer backbone, this
was also observed in the cyclic voltammagram study.
300 400 500 600
0
1
2
PAzBzOC8
PAzBzC8
Absorption
Wavelength (nm)

Figure 7-9. Comparison of the UV-vis spectrum of PAzBzC8 and PAzBzOC8 in chloroform
solution.

The UV-vis measurements also suggest the formation of a stable radical cationic species
in solution. As shown in Figure 7-10, when TFA was added to the polymer solution, a
new peak appeared in the visible range at 654 nm and 620 nm for PAzBzC8O and
PAzBzOC8, respectively. The new absorption was attributed to the formation of

232

azulenium cation radicals as we previously mentioned in the monomers study.
24
On
further protonation, the new band in the visible region had gradual increase, accompanied
by a decrease of the absorption in UV region. The color of the polymer solutions changed
to yellowish-brown and purple for PAzBzC8O and PAzBzOC8, respectively. As can be
seen in Figure 7-10, we found that the polymers are sensitive to TFA concentration. The
polymer absorption changed greatly at low TFA concentrations, and showed little change
when the TFA concentration was more than 10%. Similar behavior was found in the
PAzBzC8 polymer solution.
Stability of the azulenium cation radicals in solution was also investigated by UV-vis
spectrum analysis. A solution of PAzBzOC8 and 10% TFA (in chloroform) was left in
atmospheric conditions for 2 days, only an increase in intensity of the longest wavelength
absorption was seen. The solution was further heated at 60
0
C for 8 hours and then
adjusted to its original volume; the UV-vis spectrum showed no significant changes. All
these results indicated that the azulenium cation radicals are stable enough in solution to
endure oxygen exposure from the atmosphere.
400 600 800
0
1
2

TFA concentration
20%
10%
5%
0%
Absorption
Wavelength (nm)

Figure 7-10. UV-vix spectrum change of PAzBzOC8 in different TFA concentration solution.

233


EPR spectroscopy studies
In the EPR spectrum of PAzBzOC8 in 10% TFA solution, only a single broad signal was
observed at a g-value of 2.0017, as shown in Figure 7-11a. The EPR peak-to-peak line
width (∆H
pp
) of the polyradical solid was 4.1 G for PAzBzOC8 and 3.7 G for PAzBzC8.
In solid state, the cation radicals can be obtained by iodine doping or by TFA
protonation. In fact, because of the formation of azulenium cation radicals, PAzBzOC8
and PAzBzC8 showed good solubility in TFA, despite both of them contained the long
non-polar side chain. Figure 7-11b shows the EPR signal of PAzBzOC8 doped with
iodine. As we have previously discussed, the iodine doping induced a broad EPR signal
that may have been due to the spin-spin interaction in the polyradical system. An EPR
signal at 2.0011 (g-value), with a peak-to-peak width (∆H
pp
) of 12.4 G, was found for
PAzBzOC8 when doped with iodine. However, a relatively narrow peak was found for
the polyradical system when the polymer was protonated with TFA. Figure 7-11c is the

EPR spectrum of PAzBzOC8 when protonated with TFA. As we can see, it gave an EPR
signal at a g value of 2.0021, with a peak-to-peak width of 1.9 G.
(a)


(b)


234

(c)





Figure 7-11. EPR spectra of polymer PAzBzOC8 in (a) 10% TFA solution, (b), solid state doped
with iodine, (c), solid state protonated with TFA vapor.

The spin concentration of polyradicals reached a value of ca. 0.45 spin/unit (based on the
azulene) upon iodine oxidation (for 4 hours) and 0.28 spin/unit upon TFA protonation
(using DPPH as standard). The polyradical solid was much more stable, even under
atmospheric conditions. Their EPR intensities showed no significant change after storage
in atmospheric conditions for 2 months. The stability of these polyradicals was also
investigated by the nitrogen/oxygen permeation measurements, as we have discussed in
Chapter 3. Similar results were obtained in these measurements. Stability of the
copolymer polyradical system, was determined with EPR measurements by heating the
polyradical system at 80
0
C. Higher stability was seen for the copoly(azulene-benzene)

system, which may be due to both the stability and low reactivity of benzene moiety.
After heating the polyradicals, produced by protonation with TFA vapor at 80
0
C under
atmospheric conditions, there was no significant change of the EPR intensities (even after
2 days).


235

Conclusions
A novel and stable stable polyradical system, stabilized via the vertical delocalization and
horizontal conjugation, was designed and prepared. In this portion of research,
PAzBzOC8 and PAzBzC8 were synthesized to produce the polyradical systems and to
demonstrate our new concept. These copolymers were synthesized via a Suzuki coupling
reaction and characterized by HNMR, CNMR, and FT-IR spectrum analysis. Both
polymers showed high thermal stability and oxidation potentials.
Upon doping with iodine or protonation with TFA, both polymers could be, as expected,
converted into the stable polyradicals, which was confirmed by UV-vis, FT-IR, and EPR
experiments. UV-vis spectra study showed that both polymers were easily protonated to
the polyradical systems at low TFA concentrations. UV-vis spectra also revealed the high
stability of our polyradicals in solution. Polymer solution protonated with 10% TFA (in
chloroform) was left in atmospheric conditions for 2 days; only an increase in the
intensity of the longest absorption was seen.
Further stability tests in the solid state of the polyradical systems were carried out using
EPR measurements. EPR intensities of the polyradical systems showed no significant
change, even after storage in atmospheric conditions for 2 months.
To gain insight into the formation of azulenium cation radicals, a model compound
(1,3-diphenyl-azulene) was prepared by a general Grignard coupling reaction. This
compound was characterized by HNMR, CNMR and FT-IR. Formation of the radical

cation was confirmed and investigated by HNMR, UV-vis, and EPR measurements.



236

References
1 Coppinger, G. M. J. Am Chem. Soc. 1957, 79, 501.
2 Kinoshita, M.; in Handbook of Organic Conductive Molecules and Polymers, Vol. 1 (Ed.:
Nalva, H. S.) Wiley, New York, 1997, chapter 15, 781.
3 Buchanenko, A. L. Russ. Chem. Rev. 1990, 59, 307.
4 Crayston, J. A.; Devine, J. N. Walton, J. C. Tetrahedron 2000, 56, 7829.
5 Dimagno, S.; Waterman, K.; Sperr, D.; Streitweser, A. Angew. Chem. Int. Ed. Engl. 1991, 30,
4679.
6 Sizzmann, H.; Boese, R. Angew. Chem. Int. Ed. Angl. 1987, 26, 971.
7 Rajca, A. Chem. Rev. 1994, 94, 871.
8 Yamaguchi, K.; Toyoda, Y.; Fueno, T. Synth. Met. 1987, 47, 297.
9 Lahti, P. M.; Ichimura, A. S. J. Org. Chem. 1991, 56, 3030.
10 Nishide, H.; Kaneko, T.; Nii, T.; Katoh, K.; Tsuchida, E.; Yamaguchi, K. J. Am. Chem. Soc.
1995, 117, 548.
11 Takahashi, M.; Nakazawa, T.; Tsuchida, E.; Nishide, H. Macromolecules 1999, 32, 6383.
12 Rajca, A.; Utamapanya, S. J. Am. Chem. Soc. 1993, 115, 2396.
13 Rajca, S.; Rajca, A. J. Am. Chem. Soc. 1995, 117, 9172.
14 Kaisaki, D. A.; Chang, W.; Dougherty, D. A. J. Am. Chem. Soc. 1991, 113, 2764.
15 West Jr., A. P.; Silverman, S. K.; Dougherty, D. A. J. Am. Chem. Soc. 1996, 118, 1452.
16 Fujii, A.; Ishida, T.; Koga, N.; Iwamura, H.; Macromolecules 1991, 24, 1077.
17 Miura, Y.; Inui, K.; Markromol. Chem. 1992, 193, 2137.
18 Abdelkader, M.; Drench, W.; Mwijer, E. W. Chem. Mater. 1991, 3, 598.
19 Ito, S.; Nomura, A.; Morita, N.; Kabuto, C.; Kobayashi, H.; Maehima, S.; Fujimori, K.;
Yasunami, M. J. Org. Chem. 2002, 67, 7295.

20 Oudar, J. L.; Chemla, S. S. Chem. Phys. 1977, 66, 2664.
21 Hendrickx, E.; Clays, K.; Dehu, C.; Brédas, J. L. J. Am. Chem. Soc. 1995, 117, 3547.

237

22 Oda, M.; Kajioka, T.; Uchiyama, T.; Nagara, K.; Okujima, T.; Ito, S.; Morita, N.; Sato, T.;
Miyatake, R.; Kuroda, S. Tetrahedron 1999, 55, 6081.
23 Farrell, T.; Manning, A. R.; Mitchell, G.; Heck, J.; Meyer-Fridrichsen, T.; Malessa, M.;
Wittenburg, C.; Prosenc, M. H.; Cunningham, D.; McArdle, P. Eur. J. Inorg. Chem. 2002,
1677.
24 Ito, S.; Nomura, A.; Morita, N.; Kabuto, C.; Kobayashi, H.; Maejima, S.; Fujimori, K.;
Yasunami, M. J. Org. Chem. 2002, 67, 7295.
25 Rehahn, M.; Schlüter, A D.; Feast, W. J. Synthesis, 1988, 386.
26 Todd, M. H.; Balasubramanian, S.; Abell, C. Tetradron Lett. 1997, 38, 6781.
27 Miyaura, N.; Yanagi, T.; Suzuki, A. Synth. Commun. 1981, 11, 513.
28 McCarthy, T. F.; Witteler, H.; Pakula, T.; Wegner, G. Macromolecules 1995, 28, 8350.
29 Remmers, M.; Schulze, M.; Wegner, G. Macromol. Rapid Commun. 1996, 17, 239.
30 Scherf, U.; Müllen, K. Macromolecules 1992, 25, 3546.
31 Yamamoto, T.; Zhou, Z. H.; Kanbara, T. and et al. J. Am. Chem. Soc. 1996, 118, 10389.

238

Chapter 8
Conjugation control by changing the main backbone
conjugation type or by side aromatic substituent
Introduction
π-conjugated oligomers and polymers have been actively investigated for a variety of
electronic and optoelectronic applications.
1-3
A key feature of these novel semiconductors
materials is that manipulation of the chemical structure allows control of the electronic
and optical band gap.
4
In recent years, there has been considerable interest in design and
synthesis of the novel conjugation systems that the band gap can be controlled. For
example, the low-band gap materials received much attention because these materials
may afford intrinsically conducting polymers and their incorporation as the active layer in
polymer photovoltaic cell for light absorption and charge generation.
5-8
At the same time,
much efforts have also been done on adjustment of HOMO and LUMO energy level of
the polymers by introducing phenylene ring as side pendant to improve the
photoluminescence (PL) quantum yield.
9,10
In this contribution, We would like to carry
out an systemically study of the band-gap control by changing the main backbone
electronic structure or by introduction of the phenyl group to the side position based on

the azulene containing conjugated polymers.
First, we designed the low band gap polymers by changing the main backbone electronic
structure. One successful approach to low band-gap polymers is insertion of alternation
electron-donating and electron-accepting moieties along the polymers backbone.
11,12
The
alternating electron donor-acceptor arrangement causes an interaction of the highest
occupied molecular orbital of the donor and the lowest occupied molecular orbital of the

239

acceptor, and hence resulting in reduction of the band gap in these materials. Following
this strategy, a number of low-band gap materials have been synthesized.
13,14
Theoretical
calculations have confirmed the band gap of polyarylenes was a strong function of
molecular geometry.
15
Often, the quinoid isomer gave small band gap.
16
To test this idea
that quinoid geometry can substantially reduce the band gap of the conjugated polymer,
Wudl and co-workers synthesized polyisothianaphthene (PITN) and showed that it has a
very small band gap (E
g
= 1.13 eV) due to the preferred stabilization of the quinoid
resonance form by the fused benzene ring.
17
Based on the theoretical and experimental
studies, led us to conclude that azulene seems to be a good candidate for designing the

low band gap polymers because of its unique electronic structure. First, azulene contains
fused five-membered ring and seven-membered ring as a bicyclic structure and it is a
nonalternant polycyclic aromatic compound with a 10-π electron system.
18
A dipole
moment of µ ≈ 1 D (1D = 3.33 × 10
-30
cm) was observed for azulene, indicating the
contribution of the dipolar structure as shown in Figure 8-1a. Thus azulene can be
regarded as a intra-molecule electron-donating and electron-accepting system.
19,20

Secondly, azulene can be easily converted to its quinoid geometry by flash vacuum
pyrolysis or photooxidation.
21,22
However, up to now, azulene was only scarcely
employed as precursors for the synthesis of conducting polymers, and all the reported
polyazulene were synthesized by linkage of the 1,3-position of azulene. In the present
work, the design and synthesis of the novel conjugated polymers containing azulene in
the polymer backbone via coupling the 2,6-positions of the azulene will be described.
Effective conjugation of 1,3-azulene based and 2,6-azulene based copolymers will be
compared. To further understand the conjugation effect, model compounds of this series
polymers were prepared and studied.

240

(a)
R
R
R

R
Ar
Ar Ar
Ar
(b)

Figure 8-1. The resonance structure and quinoid structure of azulene.

On the other hand, the properties of conjugated polymers such as polythiophenes (PTs)
are intimately connected with the coplanarity and conjugation of the arylene ring
systems. In the straight-chain, unsubstituted oligothiophene and polythiophenes, the only
influence on the extent of the conjugation is by the chain length. By assigning suitable
side groups to the thiophene ring, one should be able to control the conjugation by either
direct conjugation of the substituent with the polymers backbone or indirect conjugation
by introducing steric restriction to the coplanarity of the chain rings or to the
intermolecular order in the solid state. This is of considerable value in design conducting
polymers for various purpose.
23
Aromatic substituents are particularly interesting as they
present a possibility to extend its conjugation to polymer backbone. For instance, 3-
phenyl-thiophene is known to polymerize to a polymer having a very good conductivity
and capable of undergoing facile doping-undoping process even in the cathodic region.
24

Whether this is due to the conjugation of the phenyl group with the oligothiophene
backbone or just the steric enhancement of favorable properties is an interesting question
which has raised some controversy. According to Kaeriyama et al.
25
the substituent
effects shown by poly(3-(4-methoxyphenyl)thiophene) support the conjugation of the

pendant aryl group with the main chain. However, Yoshino et al.
26
have deduced from
the band gap of poly(3-phenylthiophene) that the phenyl ring is not coplanar with the
thiophene rings and thus is not in conjugation with the main chain. To make clear about
this question, we designed and synthesized a series of model compounds with azulene
moiety in the center and phenyl group substituted at different side positions as shown in

241

Figure 8-2b. As azulene is a nonalternant compound, the conjugation influence can easily
be detected by the chemical shift of the azulene protons and the bond alternation. To
estimate the influence of steric and conjugation effect of the side phenyl groups, a series
model compounds and related polymers with phenyl groups substituted at different
positions were prepared and investigated.

Molecular design
To develop low band gap conducting polymers, conjugated polymers based on the 2,6-
coupling of azulene were designed as shown in Figure 8-2a. Two conjugated polymers
with different position of the side alkyl group were designed to study the effect of steric
effect on the resulting polymers’ properties. Figure 8-2b is the design of the model
compounds with different arrangement of side phenyl group. Study of these compounds
may reveal which one plays a main role in influence of the polymer conjugation, the
steric or conjugation effect.
CO
2
Et
CO
2
Et

S
S
C
10
H
21
CO
2
Et
CO
2
Et
S
S
C
10
H
21
n
n
(a)
P26A
P26B

S
S
S
S
S
S

(b)
PhS1
PhS2
PhS3

Figure 8-2. The molecular design for the bang gap control study.





242

Results and discussion
Part 1. Conjugation control by changing the main backbone
conjugation type
Model compounds synthesis and characterization
To have a better understanding of electronic geometry of the molecule and the side group
steric hindrance affect on the resulting conducting polymers, two model compounds were
designed and synthesized as shown in Scheme 8-1.
O
OH
O
Cl
CO
2
Et
CO
2
Et

NH
2
CO
2
Et
CO
2
Et
NH
2
Br
CO
2
Et
CO
2
Et
N
2
O
CO
2
Et
CO
2
Et
HO
S
CO
2

Et
CO
2
Et
TfO
S
CO
2
Et
CO
2
Et
S
S
C
4
H
9
CO
2
Et
S
Me
iii
iii
iv/v
vi vii
viii/ix
x/ix
M26A

M26B
CO
2
Et
S

Scheme 8-1. The synthesis of model compounds for the 2,6-azulene coupling polymers. Reagents
and conditions: (i), SOCl
2
/benzene, (ii), EtONa, ethyl cyanoacetate, (iii), Br
2
/CHCl
3
, (iv),
H
2
SO
4
/1,4-dioxane, (v), NaNO
2
/HCl, (vi), h
ν
, thiophene, (vii), Tf
2
O/Et
3
N/CH
2
Cl
2

, (viii), 3-butyl-
2-thenyl-tributyl stanne, (ix), LiCl/AsPh
3
, Pd(PPh
3
)
4
, (x), 2-(4-methyl-5-phenyl)-thenyl-tributyl
stanne.

As shown in Scheme 8-1, both model compounds are prepared by Stille coupling
reaction
27-29
using thenylstannanes with the azulen triflates as the coupling partners. We
chose Stille coupling reaction because this reaction is extremely versatile, proceeds under
neutral condition and can tolerate wide range of substituents on both coupling partners.
The synthetic intermediates azulen triflate was synthesized from tropolone via 6 steps.

243

Azulene compoundderivative diethyl 2-aminoazulene-1,3-dicarboxylate was constructed
by Nozoe synthesis.
30
Reaction of equiv. amount of 2-chlorocyclohepta-2,4,6-trien-1-one
ethyl cyanoacetate in ethanol, diethyl 2-aminoazulene-1,3-dicarboxylate was obtained in
high yield. Functionalized the 6-position was successful by bromination of diethyl 2-
aminoazulene-1,3-dicarboxylate in chloroform. Diazotization of diethyl 2-amino-6-
bromo-1,3-dicarboxylate by sodium nitrite in dioxane-sulfuric acid afforded the diethyl
2-diazo-1,3-dicarboxylate-6-oxo-2,6-azulene, the structure of which was confirmed by
NMR and IR spectrum compared with the reference.

31
Photochemical reaction of
thiophene and diethyl 2-diazo-1,3-dicarboxylate-6-oxo-2,6-azulene was carried out by a
medium-pressure Hg lamp under an nitrogen atmosphere. The photochemical reaction
was first carried out between diethyl 2-diazo-1,3-dicarboxylate-6-oxo-2,6-azulene with
thiophene in ethyl acetate. However, the yield is low.
32
We think this is may be due to
the radical formation in the reaction was captured by the ethyl acetate not by thiophene as
shown in the possible mechanism (Scheme 8-2).
33


N
2
O
CO
2
Et
CO
2
Et
h
ν
O
CO
2
Et
CO
2

Et
S
O
CO
2
Et
S
H
H
2
O
CO
2
Et
S
O
CO
2
Et
CO
2
Et
CO
2
Et
CO
2
Et
HO

Scheme 8-2. Mechanism of the photochemical reaction of diethyl 2-diazo-1,3-dicarboxylate-6-
oxo-2,6-azulene with thiophene.

Thus we usedthiophene as solvent for the photochemical reaction, and this procedure
produced the 2-thenyl substituted azulene in high yield. Conversion of the 2-(diethyl-1,3-
dicarboxylate-6-hydroxy-2-azulenyl)-thiophene to its triflate was successfully by
treatment with triflic anhydride in the presence triethylamine as a base.
34
The coupling

244