Modification of a donor-acceptor photovoltaic polymer by integration of optoelectronic moieties into its side chains

Article abstract of DOI:10.1016/j.polymer.2015.01.008

In this study, a strategy to modify photovoltaic properties of a known material by integrating certain optoelectronic moieties in its side chains has been described. Thus, a plenty of single and dendritic carbazole units were introduced into the side chains of poly(2,7-(9,9-dialkyl-fluorene)-alt-5,5′-(4,7-di-2-thienyl-2,1,3-benzothiadiazole)) (PFDTBT), a famous donor-acceptor alternative conjugated polymer, to see what and how they can change the latter optoelectronic properties. It was found that such modifications not only increase the polymer light-harvesting capabilities in the UV region, but also enhance hole mobility in the pure film state. Furthermore, complicated photophysical and photochemical processes, including energy transfer, electron transfer and site-isolation effect, were observed to take place between carbazole units and the PFDTBT conjugated backbone. These factors work comprehensively and finally improve the polymer photovoltaic properties when modified with single carbazole units, but deteriorate when modified with dendritic carbazole units.

Full text of DOI:10.1016/j.polymer.2015.01.008

Polymer 59 (2015) 57e66
Contents lists available at ScienceDirect
Polymer
journal homepage: www.elsevier.com/locate/polymer
Modification of a donor-acceptor photovoltaic polymer by integration
of optoelectronic moieties into its side chains
Lin-Feng Yu, Cong-Wu Ge, Jin-Tu Wang, Xuan Xiang, Wei-Shi Li*
CAS Key Laboratory of Synthetic and Self-Assembly Chemistry for Organic Functional Molecules, Shanghai Institute of Organic Chemistry, Chinese Academy
of Sciences, Shanghai 200032, China
a r t i c l e i n f o
a b s t r a c t
Article history:
In this study, a strategy to modify photovoltaic properties of a known material by integrating certain
optoelectronic moieties in its side chains has been described. Thus, a plenty of single and dendritic
carbazole units were introduced into the side chains of poly(2,7-(9,9-dialkyl-fluorene)-alt-5,50-(4,7-di-2-
thienyl-2,1,3-benzothiadiazole)) (PFDTBT), a famous donor-acceptor alternative conjugated polymer, to
see what and how they can change the latter optoelectronic properties. It was found that such modifi-
cations not only increase the polymer light-harvesting capabilities in the UV region, but also enhance
hole mobility in the pure film state. Furthermore, complicated photophysical and photochemical pro-
cesses, including energy transfer, electron transfer and site-isolation effect, were observed to take place
between carbazole units and the PFDTBT conjugated backbone. These factors work comprehensively and
finally improve the polymer photovoltaic properties when modified with single carbazole units, but
deteriorate when modified with dendritic carbazole units.
Received 4 November 2014
Received in revised form
30 December 2014
Accepted 3 January 2015
Available online 10 January 2015
Keywords:
Organic optoelectronic materials
Photovoltaics
Poly(2,7-(9,9-dialkyl-fluorene)-alt-5,50-(4,7-
di-2-thienyl-2,1,3-benzothiadiazole))
© 2015 Elsevier Ltd. All rights reserved.
1. Introduction
modification with optoelectronic moieties on the side chains of
conjugated polymers. Up to date, only few examples have been
Polymer solar cells (PSCs) have attracted significant attention in
terms of light-weight, solution-processable, low cost, and adapt-
able to flexible and large area devices. In the past decade, PSCs have
achieved a substantial progress [1]. Power conversion efficiency
(PCE) for single-junction devices has improved from 3 to 5% before
2005 [2] to nowadays over 9% [3], while that of tandem cells has
surpassed 10% [4]. Such remarkable progress makes PSCs as a more
attractive and realistic solar electric generation technology. In
addition to the improvement in device fabrication technologies, the
progressive innovation of active materials plays one of the most
important roles to push the field forward [1,5]. Looking back the
history of photovoltaic polymers, one may conclude three main
streams in the new material design strategies: (1) invention of new
conjugated building blocks [6], (2) newly choice and organization
reported [9,10]. For examples, certain mesogens were integrated
into the side chain of conjugated polymers for tuning their packing
structures in solid state [9]. Hole or electron-transporting units
have been appended in the side chains of conjugated polymers for
improving their light-emitting performances [10].
Recently, we endeavored to improve photovoltaic properties of
conjugated polymers by integration of optoelectronic moieties into
their side chains. Here, we report the work on poly(2,7-(9,9-dialkyl-
fluorene)-alt-5,50-(4,7-di-2-thienyl-2,1,3-benzothiadiazole))
(PFDTBT) (Chart 1). PFDTBT is a famous donor-acceptor (D-A)
conjugated polymer, firstly reported by Andersson et al., in 2003
[11] and laters by others [12]. The alternative electron-rich and
-deficient structure in its backbone endows the polymer a broad
and intense light absorption spectrum in the range of 450e700 nm.
Moreover, PFDTBT possesses a deep highest occupied molecular
orbital (HOMO) and has been demonstrated to give an impressing
large open-circuit voltage (VOC) (around 1 V) with its bulk hetero-
junction cells in conjunction with [6,6]-phenyl-C61-butyric acid
methyl ester (PC61BM). Although the above mentioned properties
of PFDTBT are better than poly(3-hexylthiophene) (P3HT), its
photovoltaic efficiency still lags behind P3HT [2]. One of possible
reasons is coming from its low hole mobility [13]. Since carbazole
of known p-conjugated moieties for conjugated polymer backbone
[7], and (3) aliphatic side-chain engineering in order to tune the
solubility and interchain interactions among polymer chains [8].
However, little attention has been paid on the non-conjugated
* Corresponding author. Tel./fax: þ86 21 5492 5381.
E-mail address: liws@mail.sioc.ac.cn (W.-S. Li).
http://dx.doi.org/10.1016/j.polymer.2015.01.008
0032-3861/© 2015 Elsevier Ltd. All rights reserved.

58
L.-F. Yu et al. / Polymer 59 (2015) 57e66
Chart 1. Chemical structures of PFDTBT-C12, PFDTBT-Cz1 and PFDTBT-Cz3.
derivatives usually have good hole-transport properties and been
widely used as hole transport materials in organic light-emitting
diodes [14], we propose here to modify PFDTBT with carbazole
units in its side chains. Thus, two polymers, PFDTBT-Cz1 and
PFDTBT-Cz3, having single or dendritic carbazole units respectively
(Chart 1), were designed and synthesized. We anticipated that a
large number of carbazole units surrounding the conjugated
backbone would form their own network even in the amorphous
state and serve as an additional hole-transport channel besides the
conjugated backbone pathway, and thus improve charge trans-
portation in polymer films. The amorphous materials with good
charge-transport properties would be greatly welcomed for PSC
real application, since they provide a solution to the poor stability
and reproducibility of PSC devices owing to the variation and
evolution of the microstructure of the active layer based on crys-
talline materials.
polydispersity index (PDI) were determined by gel permeation
chromatography using monodispersed polystyrenes as standard
(Table 1). The data shows a large difference among three polymers.
PFDTBT-Cz1 has a number-average molecular weight (Mn) of
21,882 and a PDI of 17.91. The large PDI indicates the complexity for
the Stille coupling polymerization. In contrast, PFDTBT-Cz3 has a
Mn of 7842 with a PDI of 1.28. The small molecular weight of
PFDTBT-Cz3 reflects that the large size of dendritic carbazole units
may hamper the proceeding of the polymerization. The reference
polymer PFDTBT-C12 shows a moderate molecular weight and
polydispersity (Mn ¼ 10,057, PDI ¼ 2.10). Since molecular weight
and its polydispersity have been demonstrated to have great
impact on the photovoltaic performance [15], the so prepared three
polymers with such large molecular weight differences could not
be directly used to investigate the effect of carbazole and dendritic
carbazole units. Therefore, in order to exclude the influences of the
molecular weight and its polydispersity, preparative size-exclusion
chromatography (SEC) was applied on PFDTBT-Cz1 and PFDTBT-
C12. The fractions having molecular weight similar to that of
2. Results and discussion
PFDTBT-Cz3 were collected (for PFDTBT-Cz1: Mn
¼
7950,
2.1. Synthesis and characterization
PDI ¼ 1.24; for PFDTBT-C12: Mn ¼ 7504, PDI ¼ 1.99) and used for
the following property investigation. By means of thermogravi-
metric analysis (TGA), the thermal decomposition temperature at
5% weight loss was determined to be 407, 399, and 346 ꢀC for
PFDTBT-C12, PFDTBT-Cz1, and PFDTBT-Cz3 respectively (Fig. 1 and
Table 1), indicating the integration of carbazole units slightly
decrease thermal stability of the polymers.
The syntheses of single and dendritic carbazole-appended
PFDTBT (PFDTBT-Cz1 and PFDTBT-Cz3) are outlined in Scheme 1.
The preparation of PFDTBT-Cz1 started from the attachment of two
carbazole units into the side chains of 2,7-dibromofluorene
monomer (F-Cz1) via two steps from 9H-carbazole (Cz), and
finished by Stille coupling polymerization of F-Cz1 with 4,7-bis(5-
trimethylstannyl-thienyl-2-yl-2,1,3-benzothiadiazole with a yield
of 48%. On the other side, the synthesis of PFDTBT-Cz3 is a little
more complicated. Compound Cz3, an important dendritic carba-
zole precursor, was first prepared via Ullmann condensation reac-
tion of acetyl-protected 3,6-diiodo carbazole (I-Cz-Amide) with
9H-carbazole and followed by deprotection. Afterward, Cz3 was
subsequently reacted with 1,12-dibromododecane and 9H-2,7-
dibromofluorene, affording dendritic carbazole-appended 2,7-
dibromofluorene monomer (F-Cz3). Stille coupling polymeriza-
tion of this monomer with 4,7-bis(5-trimethylstannyl-thienyl-2-yl-
2,1,3-benzothiadiazole finally produced PFDTBT-Cz3 in a yield of
53%. In order to check the effect of the introduced single and
dendritic carbazole units on their properties, PFDTBT without
carbazole units in its side chains (PFDTBT-C12, Chart 1) was also
prepared by similar Stille coupling polymerization and used as a
reference.
2.2. Optical properties
In CHCl3 solution at room temperature, polymer PFDTBT-C12
exhibited two main absorption bands centered at 542, and 388 nm,
respectively (Fig. 2a, Table 2). These peaks originate from the
conjugated backbone of PFDTBT. Similarly, polymer PFDTBT-Cz1
showed these two absorption bands at 539 and 386 nm, a slightly
blue-shifted from that of PFDTBT-C12. However, in case of PFDTBT-
Cz3, the blue shifts became larger. The two bands appeared at 523
and 372 nm, 19 and 16 nm blue-shifted from those of PFDTBT-C12,
respectively. Since the molar molecular weight of the repeating
unit of PFDTBT-Cz3 is larger than those of PFDTBT-C12 and
PFDTBT-Cz1, the actual polymerization degree (the average num-
ber of repeating units in one polymer chain) of PFDTBT-Cz3 was
small. This would be one of the possible reasons for the above blue-
shifted observation with PFDTBT-Cz3. Another possible reason may
be attributed by the large carbazole dendritic substituents, which
All the synthesized polymers have good solubility in CHCl3,
chlorobenzene, and dichlorobenzene. Their molecular weight and

L.-F. Yu et al. / Polymer 59 (2015) 57e66
59
Scheme 1. Syntheses of PFDTBT-Cz1 and PFDTBT-Cz3.
would exert a large steric hindrance on the PFDTBT backbone and
induced a most twisted configuration, and thus disfavor to
respectively (Fig. 3b, Table 2). Only 1 and 3 nm red shifts were
observed as compared with those in solution, suggesting that
carbazole units almost stay in isolated state even in films. Calcu-
lated from the absorption onsets of the film spectra (PFDTBT-C12:
680 nm, PFDTBT-Cz1: 673 nm, and PFDTBT-Cz3: 665 nm), the
optical band gap of these three polymers is 1.82, 1.84, and 1.86 eV,
respectively (Table 2).
In CHCl3 solution, all three polymers fluoresced when excited at
their absorption peak-tops of the conjugated backbones (Fig. 3a,
Table 2). PFDTBT-Cz1 showed an emission band around 630 nm,
similar to PFDTBT-C12 (628 nm). Whereas, PFDTBT-Cz3 displayed
a fluorescent band around 620 nm, 8e10 nm-blue shifted from
those of PFDTBT-C12 and PFDTBT-Cz1. When normalized the
absorbance of the excitation wavelength to be 0.1, the area intensity
of the emission band increased following the order of 13,800
(PFDTBT-Cz1), 24,440 (PFDTBT-C12), and 42,128 (PFDTBT-Cz3).
This result means the fluorescence from PFDTBT conjugated back-
bone enhances when the introduction of dendritic carbazole units
(Cz3), but quenches when attaching single carbazole units (Cz1)
into its side chains. Since carbazole derivatives can generally
behave as electron donating quencher, the fluorescence titration
experiments were carried out. As shown in Fig. 3b, upon addition of
Cz1-Br into the solution of PFDTBT-C12, its fluorescence was
p
-
conjugation. Anyway, as compared with those of PFDTBT-C12 and
PFDTBT-Cz1, the narrower absorption spectrum of PFDTBT-Cz3,
would be one of important factors that suggest it is not a good
material for photovoltaic applications. In addition to the above two
bands, PFDTBT-Cz1 exhibited two other peaks around 348 and
334 nm (Fig. 2a, Table 2), consistent with those of compound Cz1-
Br. Therefore, these two peaks can be assigned to the absorption of
the periphery carbazole units. In case of PFDTBT-Cz3, three peaks
observed around 343, 327, and 317 nm originate from the periphery
dendritic carbazole units. Since the number of carbazole units is
large in PFDTBT-Cz3, the absorption bands of dendritic carbazole
units becomes more intense than those of conjugated backbone.
In film state, polymer PFDTBT-C12 showed an absorption band
centered at 567 nm for its conjugated backbone, 25 nm red-shifted
from that in solution (Fig. 2b, Table 2). This peak appeared at 560
and 544 nm for PFDTBT-Cz1 and PFDTBT-Cz3, respectively, about
21 nm red-shifted from their solution ones. These observations
indicate the introduction of carbazole units does not have great
impact on the interchain interactions among the conjugated
backbone in solid state, and further imply that the original hole
transportation channel among polymer backbone would be
reserved even after attaching a large dendritic side chains. On the
other hand, the absorption band of carbazole units in both PFDTBT-
Cz1 and PFDTBT-Cz3 films showed up at 349 and 346 nm,
gradually quenched with
a
SterneVolmer constant (KSV
) of
1.26 ꢁ 103
M
ꢂ1, indicating Cz1-Br is really a quencher for the
fluorescence of PFDTBT. The fact suggests the occurrence of photo-

60
L.-F. Yu et al. / Polymer 59 (2015) 57e66
Table 1
would be a main reason. Indeed, upon excitation at carbazole units
at 348 nm, all polymers showed a fluorescence peak in the range of
550e750 nm (Fig. 4b), which is corresponding to the emission band
of PFDTBT backbone. The area intensity of this band was 12,007,
12,002, and 30,535 for PFDTBT-C12, PFDTBT-Cz1, and PFDTBT-Cz3,
respectively, under the same polymer concentration (1.0 ꢁ 10ꢂ6 M).
Although the area intensity of PFDTBT-Cz1 was almost the same as
that of PFDTBT-C12, the occurrence of energy transfer from
carbazole units to the PFDTBT backbone can be deduced from the
above mentioned fluorescence quench fact of PFDTBT-Cz1 as
referred to PFDTBT-C12. In the case of PFDTBT-Cz3, the fluores-
cence enhancement ratio is 2.54 (30,535/12,007), larger than that
when excited at PFDTBT backbone with 542 nm light (42,128/
24,440 ¼ 1.72). This analysis clearly confirms that energy transfer
took place from dendritic carbazole units to the PFDTBT backbone.
The existing of energy transfer indicates that carbazole units could
behave as light-absorption antenna to enhance light acquisition in
the UV region, which may be favorable for their photovoltaic
properties.
Molecular weight and thermal stabilities of PFDTBT-C12, PFDTBT-Cz1, and PFDTBT-
Cz3.
SEC fractiona
Td (ꢀC)
b
Polymer
So-prepared
Mn (KDa)
PDI
2.10
Mn (kDa)
PDI
PFDTBT-C12
PFDTBT-Cz1
PFDTBT-Cz3
10.1
21.9
7.84c
7.50
7.95
e
1.99
1.24
e
407
399
346
17.9
1.28c
a
After treated by preparative size-exclusion chromatography (SEC) and used in
the study.
b
c
Decomposition temperature at 5% weight loss.
PFDTBT-Cz3 was not subjected to preparative size-exclusion chromatography
and used directly in the study.
induced electron transfer between the PFDTBT backbone and the
side carbazole units in PFDTBT-Cz1. The much larger quenching
extent for PFDTBT-Cz1 than the titration experiment implies a
more efficient photo-induced electron transfer when carbazole
units are covalently attached. Although the fluorescence titration
experiments (Fig. 3c) also showed that compound Cz3-Br can
quench the fluorescence of PFDTBT-C12 with
a
KSV of
2.3. Electrochemical properties
1.76 ꢁ 103
M
ꢂ1, the fluorescence of PFDTBT-Cz3 still enhanced as
compared with that of PFDTBT-C12. This fluorescence enhance-
ment phenomenon has been often observed in dendrimer chem-
istry when optoelectronic functionalities are encapsulated in a
large dendritic wedge [16]. In such cases, the dendrimer encapsu-
lation allows optoelectronic functionalities isolated from each
other and then avoids collision fluorescence quench. In the case of
PFDTBT-Cz3, a lot number of dendritic carbazole units may impose
two conflictive effects on the PFDTBT backbone: fluorescence
quench by means of photo-induced electron transfer, and fluores-
cence enhancement by means of dendritic encapsulation effect.
Since fluorescence quench effect is weak, PFDTBT-Cz3 finally ex-
hibits fluorescence enhancement.
The electrochemical properties of the prepared polymers were
investigated by means of cyclic voltammetry (Fig. 5). The polymer
film samples were prepared by casting their CHCl3 solutions onto a
glassy-carbon working electrode and measured in deoxygenated
MeCN containing 0.1 M Bu4NPF6 supporting electrolyte at room
temperature with a Pt plate counter electrode and a Ag/AgNO3
reference electrode at a scan rate of 100 mV sꢂ1. The onset oxidation
potential (Eox) was determined to be 0.65 V for PFDTBT-C12, and
0.63 V for both PFDTBT-Cz1 and PFDTBT-Cz3 (Table 2). Under the
same conditions, the redox potential of ferrocenium (Fcþ)/ferro-
cene (Fc) was 0.13 V. Since the absolute energy level of Fcþ/Fc redox
Carbazole derivatives are good emitters in general. In this study,
we also found that Cz1-Br and Cz3-Br showed very intense fluo-
rescence with bands around 368 nm and 390 nm, respectively,
upon excitation at 348 nm (Fig. 4a). However, these bands were
almost quenched when they were attached to PFDTBT side chains.
Since no spectral position change was observed for the absorption
bands of carbazole units of PFDTBT-Cz1 and PFDTBT-Cz3 in solu-
tion as compared with those of Cz1-Br and Cz3-Br respectively
(Fig. 2a), the fluorescence quench would not cause by the chro-
mophoric packing effect among carbazole units. This implies that
energy transfer between carbazole units and PFDTBT backbone
Fig. 2. . The UVevis absorption spectra of PFDTBT-C12, PFDTBT-Cz1, and PFDTBT-Cz3
(a) in CHCl3 solution and (b) in film state at room temperature. The UVevis absorption
spectra of compounds Cz1-Br and Cz3-Br in CHCl3 solution are also displayed for
comparison.
Fig. 1. TGA of PFDTBT-C12, PFDTBT-Cz1, and PFDTBT-Cz3 under N2 atmosphere with a
heat rate of 10 ꢀC minꢂ1
.

L.-F. Yu et al. / Polymer 59 (2015) 57e66
61
Table 2
Optical and electrochemical properties of PFDTBT-C12, PFDTBT-Cz1, and PFDTBT-Cz3.
a
b
c
d
e
f
g
Polymer
labs, soln (nm)
lFL, soln (nm)
labs, film (nm)
labs, onset (nm)
Eg, opt (eV)
Eox, onset (V)
EHOMO (eV)
ELUMOh (eV)
PFDTBT-C12
542 (3.28),
388 (2.90)
539 (3.07),
386 (2.76),
348 (2.00),
334 (1.63)
523 (2.24),
372 (2.62),
343 (3.20),
327 (3.25),
317 (3.37)
628
630
567,
394
560,
394,
349,
336
544,
386,
346
680
673
1.82
1.84
0.65
0.63
ꢂ5.32
ꢂ5.30
ꢂ3.50
ꢂ3.46
PFDTBT-Cz1
PFDTBT-Cz3
620
665
1.86
0.63
ꢂ5.30
ꢂ3.44
a
Absorption bands in CHCl3 solution at room temperature, their molar absorption coefficients are shown in parentheses with a unit of 104 Mꢂ1 cmꢂ1
Fluorescence bands in CHCl3 solution at room temperature when excited at their absorption peak-top.
Absorption bands in film state at room temperature.
.
b
c
d
e
f
Absorption onset in film state.
Optical band gap, calculated by Eg, opt ¼ 1240/labs, onset
.
Oxidation onset measured by CV.
g
h
EHOMO ¼ -e (Eox, onset þ 4.8 - 0.13).
ELUMO ¼ EHOMO þ Eg, opt
.
coupling is 4.8 eV below vacuum [17], the HOMO energy level of the
polymers can be derived by the following equation:
configuration of glass/ITO/PEDOT:PSS/polymer film/Au. According
to MotteGurney law, SCLC theory can be described as
EHOMO ¼ ꢂeðEox þ 4:8 ꢂ 0:13ÞeV
9
8
ðVa ꢂ VbiÞ2
J ¼ ε0εrm
;
d3
The results were ꢂ5.32 eV for PFDTBT-C12 and ꢂ5.30 eV for
both PFDTBT-Cz1 and PFDTBT-Cz3. Following the same method,
the onset oxidation potential for the both reference compounds,
Cz1-Br and Cz3-Br, was determined to be 0.63 V, giving their
HOMO level of ꢂ5.30 eV. This value is the same as that of PFDTBT-
Cz1 and PFDTBT-Cz3, and no substantially different from that of
PFDTBT-C12, implying that hole can freely transfer between
carbazole units and the PFDTBT backbone. From the HOMO energy
level and optical band gap, the lowest unoccupied molecular
orbital (LUMO) of PFDTBT-C12, PFDTBT-Cz1 and PFDTBT-Cz3 are
calculated to be ꢂ3.50, ꢂ3.46, and ꢂ3.44 eV, respectively
(Table 2).
where
J is current density, ε0 is permittivity of vacuum
(8.85 ꢁ 10ꢂ12 F mꢂ1), εr is relative permittivity of the material and
generally taken as 3 for organic polymers,
m is mobility, Va is
applied voltage, Vbi is built-in voltage (0.1 V in this work by
taking 5.1 eV for the work function of Au and 5.0 eV for work
function of PEDOT:PSS), and d is the thickness of the active film
[18]. For each material, at least five devices with film thickness in
the range of 50e120 nm were checked and averaged for the final
results. Fig. 6 displays the current density-voltage data of the
check devices at SCLC region and their fitting curves. By this
method, the hole mobility of PFDTBT-C12 film was determined to
be (4.5 0.4) ꢁ 10ꢂ7 cm2 Vꢂ1 sꢂ1, while those of PFDTBT-Cz1 and
2.4. Hole mobility
PFDTBT-Cz3 films to be (1.5
0.2)
ꢁ
10ꢂ6 and
(2.8 0.3) ꢁ 10ꢂ6 cm2
V
ꢂ1 sꢂ1 respectively. The data obviously
In order to check the effect of the introduced single and den-
dritic carbazole side chains on the polymer charge transportation
properties, the hole mobility of the polymer films were measured
by space-charge-limited current (SCLC) method with a device
show that PFDTBT-Cz1 film possessed a higher hole mobility
than PFDTBT-C12, while PFDTBT-Cz3 film further improved the
value slightly. This result confirms that the introduction of
carbazole units indeed improves charge transportation
Fig. 3. (a) Fluorescence spectra of PFDTBT-C12 (excited at 542 nm), PFDTBT-Cz1 (excited at 542 nm), and PFDTBT-Cz3 (excited at 522 nm) in CHCl3 solution at room temperature
normalized by the absorbance at the excitation wavelength to be 0.1, and fluorescence spectral change of PFDTBT-C12 (1 ꢁ10ꢂ5 M) in CHCl3 at room temperature excited at 542 nm
upon titration with (b) Cz1-Br, and (c) Cz3-Br. Insets show their SterneVolmer plots.

62
L.-F. Yu et al. / Polymer 59 (2015) 57e66
Fig. 4. (a) Fluorescence spectra of PFDTBT-C12, PFDTBT-Cz1, PFDTBT-CZ3, and compound Cz1-Br and Cz3-Br in CHCl3 solution at room temperature upon excitation at 348 nm. The
spectra were normalized by the concentration of PFDTBT-C12, PFDTBT-Cz1, PFDTBT-Cz3 to be 1.0 ꢁ 10ꢂ6 M, while that of Cz1-Br and Cz3-Br to be 2.0 ꢁ 10ꢂ6 M. (b) The enlarged
region from 550 to 750 nm of the spectra shown in Fig. 4a. The peaks around 700 nm are the frequency-doubled harmonic peaks of the excitation light.
properties of the polymer, which would be favor for their
photovoltaic applications.
However, the EQE values in this range were all around 15%,
resulting in small JSC and FF values. As compared with that of
PFDTBT-C12-based device, the spectrum of PFDTBT-Cz1ebased
cell clearly had larger EQE values in the range of 300e450 nm,
which is corresponding to the absorption band of carbazole units.
This suggests that the introduction of a large number of carbazole
units enhances the light absorption in the UV region and then
contributes more photocurrent. As a contrast, the introduction of
dendritic carbazole unit (Cz3) did not enhance but decreased EQE
values in the whole spectral range (less than 7%), resulting in the
worst device performance for the PFDTBT-Cz3-based cell. From
their EQE data and the AM 1.5 G solar spectrum, the JSC value were
calculated to be 2.4, 2.97, and 0.6 mA cmꢂ2 for the devices based on
PFDTBT-C12, PFDTBT-Cz1 and PFDTBT-Cz3, respectively. The error
between the measured and the calculated JSC values is less than 5%.
As mentioned in the above, the property comparison among the
three polymers indicate that the integration of dendritic carbazole
units does increase the hole mobility and also enable a much
intense light-absorption band in the UV region. However, PFDTBT-
Cz3 exhibited so inferior photovoltaic performance to PFDTBT-C12
and PFDTBT-Cz1. In order to know the reason, we firstly measured
the SCLC hole and electron mobilities for three polymer/PC61BM
blend films, which compositions were the same as the active layers
of their PV devices. The device configuration for hole mobility
measurement was the same as that for pure polymer films, while
that for electron mobility changed to Al/blend film/Al. As shown in
Table 3, we found that there was no big difference in hole and
electron mobilities among three polymer/PC61BM blend films. All
2.5. Photovoltaic properties
The photovoltaic properties of the polymers were studied with
bulk heterojunction OPV cells having a structure of glass/ITO/
PEDOT:PSS/active layer/Ca/Al. The active layer was casted from a
1,2-dichlorobenzene solution and composed of the blend of the
checked polymer and PC61BM with an optimized weight ratio of
1:3. Upon irradiation by an AM 1.5 G solar simulator with a light
density of 100 mW cmꢂ2, the cell based on the blend of PFDTBT-
Cz1/PC61BM showed the best performance, having a VOC of 0.84 V, a
short-circuit current (JSC) of 3.03 mA cmꢂ2, a fill factor (FF) of 39.3%,
giving a PCE of 1.00% (Fig. 7a, Table 3). Under the same conditions,
the PFDTBT-C12-based device exhibited a VOC of 0.77 V, a JSC of
2.48 mA cmꢂ2, a FF value of 32.5%, then affording a PCE of 0.62%.
While, the device based on PFDTBT-Cz3 displayed the smallest
values of VOC (0.54 V) and JSC (0.63 mA cmꢂ2). Combined with its FF
value (38.4%), the efficiency of the PFDTBT-Cz3-based cell was only
0.13%.
The observed device performance difference can be explained
by their external quantum efficiency (EQE) spectra (Fig. 7b). The
device based on PFDTBT-C12 showed photovoltaic response in the
range of 300e700 nm, consistent with its absorption spectrum.
these blend samples had
a hole mobility in the range of
6.3e8.1 ꢁ10ꢂ5 cm2 Vꢂ1 sꢂ1 and an electron mobility in the range of
4.0e6.6 ꢁ 10ꢂ4 cm2
V
ꢂ1 sꢂ1. Obviously, charge transportation
should not be a cause for their photovoltaic performance differ-
ences. Secondly, the charge separation efficiency between donor
polymer and PC61BM was checked by fluorescence quenching
experiment. As illustrated in Fig. 8, all three blend films displayed
an extinguished fluorescence spectrum as compared with their
pure films. However, their quenching extents were quite different.
For PFDTBT-C12 and PFDTBT-Cz1, both quenching extents were
over 90%. But for PFDTBT-Cz3, the quenching extent was only 80%,
suggesting a much less efficient charge separation process between
PFDTBT-Cz3 and PC61BM. Again, this may be related to site
encapsulation effect that dendritic carbazole side units impose on
the central PFDTBT core. Such effect may not only prevent the
collision fluorescence quenching but also hamper electron transfer
process between PFDTBT backbone and outside units. The latter
would reduce the charge separation efficiency between PFDTBT
Fig. 5. CV curves of PFDTBT-C12, PFDTBT-Cz1, and PFDTBT-Cz3 films on glass carbon
working electrode in CH3CN containing 0.1 M Bu4NPF6 at a scan rate of 0.1 V minꢂ1 at
room temperature, Pt disk and Ag/AgNO3 electrode as counter and reference elec-
trodes, respectively.

L.-F. Yu et al. / Polymer 59 (2015) 57e66
63
units impose on the central PFDTBT backbone may hamper the
efficient charge separation process between PFDTBT backbone and
PC61BM, and thus deteriorate its photovoltaic performance.
Furthermore, the large dendritic substitutes also brought PFDTBT-
Cz3 a narrow absorption spectrum and a poor film morphology
when blending with PC61BM, which would be important factors for
its poor photovoltaic performance. Although the present work did
not give a very positive example, the side chain modification with
optoelectronic moieties would be one of important strategies to
develop novel new organic optoelectronic materials, and would
gain more and more attention in the future.
4. Experimental
Fig. 6. Current density-voltage characteristics and their fitting curves at SCLC region of
hole-only devices based on PFDTBT-C12, PFDTBT-Cz1, and PFDTBT-Cz3 films.
4.1. Measurements and characterization
1H NMR spectra were recorded on a Varain Mercury 300 MHz
spectrometer. The chemical shifts were determined using tetra-
methyl silane (TMS) as an internal standard. MALDI-TOF mass
spectroscopy was carried out on a Shimadzu Biotech Axima Per-
formance Mass Spectrometer using dithranol or CHCA as a matrix.
Molecular weights and polydispersity of the polymers were
determined by a Waters 1515 HPLC with a Waters 2489 UV detector
using monodispersed polystyrenes as the standard. The optical
absorption and emission spectra were recorded on a Hitachi U-3310
UVevis spectrometer and a Hitachi F-4600 fluorescence spec-
trometer, respectively. Cyclic voltammetry (CV) was performed on a
CHI 660C instrument with a standard three-electrode cell using a
glassy carbon as working electrode, a platinum wire as counter
electrode, and Ag/AgNO3 as reference electrode. The samples were
first casted on glassy carbon to form a film, then measured in
CH3CN in the presence of 0.1 M Bu4NPF6 with a scan rate of
0.1 V minꢂ1. Thermogravimetric analysis (TGA) was performed by a
TGA Q500 instrument under N2 atmosphere with a heat rate of
10 ꢀC minꢂ1. Atomic force microscopy (AFM) was performed on a
Veeco Nanoscope IIIa multimode apparatus by a tapping mode with
a silicon tip.
backbone and PC61BM, and thus finally lead to poor photovoltaic
performance. Thirdly, the morphology of the active layers for the
photovoltaic devices were checked by atomic force microscopy
(AFM, Fig. 9). Of interest, a number of short fiber-like objects
appeared in the blend film of PFDTBT-C12/PC61BM (Fig. 9a). In the
case of PFDTBT-Cz1/PC61BM, a large number of small dot-like ag-
gregates appeared (Fig. 9b). However, the morphology of PFDTBT-
Cz3/PC61BM turned to bad with the appearance of many big
aggregated lump-like structure (Fig. 9c). Obviously, such
morphology is not good for efficient charge separation between the
donor polymer and PC61BM and would contribute the partial
reason for its poor photovoltaic performance.
3. Conclusions
Efficient light harvesting, charge separation and transportation
are three main basic requirements for high performance OPV ma-
terials. In this study, we have demonstrated that the integration of
carbazole and dendritic carbazole units into the side chains of
PFDTBT polymers have great impact on their basic optical proper-
ties, hole mobility and photovoltaic performance. The results
showed that such structural modifications endow the polymers
certain complicated photophysical and photochemical communi-
cations, including energy transfer, electron transfer and dendrimer
site isolation effect, between carbazole side units and the PFDTBT
backbone. It was found that the integration of carbazole or den-
dritic carbazole units does not only enhance light absorption in the
UV region, but also improve the hole mobility for the pure polymer
films. However, the site isolation effect that dendritic carbazole
4.2. Solar cell device fabrication and characterization
The solar cell prototype devices were fabricated with a struc-
tural configuration of ITO glass/PEDOT:PSS/active layer/Ca/Al. First,
ITO-coated glass substrates with sheet resistance ꢃ10
U
sqꢂ1 were
cleaned by ultrasonic treatment in deionized water, acetone, and
isopropyl alcohol, successively, and then subjected to UV/Ozone
cleaning for 20 min. Second, poly(3,4-ethylenedioxythiophene)-
Fig. 7. (a) Current density-voltage curves of polymer PV cells based on the blend of PFDTBT-C12, PFDTBT-Cz1, and PFDTBT-Cz3 with PC61BM (1:3, w/w) under AM 1.5 G illu-
mination with a density 100 mW cmꢂ2. (b) EQE spectra of polymer PV cells based on the blend of PFDTBT-C12, PFDTBT-Cz1, and PFDTBT-Cz3 with PC61BM (1:3, w/w).

64
L.-F. Yu et al. / Polymer 59 (2015) 57e66
Table 3
mixture was poured into water and extracted with ethyl ether. The
organic phase was washed with water, dried over MgSO4, and
evaporated to dryness. The residue was subjected to silica column
chromatography using chloroform/hexane (6:1, v/v) as an eluant,
allowing to separate compound Cz1-Br (5.00 g) as white solid in a
Device parameters of polymer PV cells based on the blend of PFDTBT-C12, PFDTBT-
Cz1, and PFDTBT-Cz3 with PC61BM (1:3, w/w).
Polymer
VOC
(V)
JSC
FF
(%)
PCE mh (10ꢂ5 cm2 me (10ꢂ4 cm2
(%)
Vꢂ1 sꢂ1 Vꢂ1 sꢂ1
(mA cmꢂ2
)
)
)
yield of 40%. 1H NMR (300 MHz, CDCl3,
d
): 8.10 (d, J ¼ 8 Hz, 2H), 7.44
PFDTBT-C12 0.77 2.48
PFDTBT-Cz1 0.84 3.03
PFDTBT-Cz3 0.54 0.63
32.5 0.62 6.3 0.2
39.3 1.00 6.6 0.3
38.4 0.13 8.1 0.4
6.5 0.1
4.0 0.2
6.6 0.1
(m, 4H), 7.22 (m, 2H), 4.30 (t, J ¼ 8 Hz, J ¼ 2H), 3.40 (t, J ¼ 7 Hz, 2H),
1.84 (m, 4H), 1.24 (m, 16H).
4.3.3. F-Cz1
poly(styrenesulfonate) (PEDOT:PSS, Heraeus Clevios P VP. Al 4083)
was spin-coated from an aqueous solution and heated at 150 ꢀC for
10 min, giving a thickness of about 30 nm. The active layer was
composed of the checked polymer and PC61BM (Lumitec LT-8905)
in a weight ratio of 1:3, and prepared by spin-coating from a 1,2-
dichlorobenzene solution (40 mg/mL) at a speed of 3000 rpm for
30 s in a nitrogen-filled glovebox. Afterward, the substrates were
transferred into an evaporator and pumped down to 2 ꢁ 10ꢂ6 mbar.
Subsequently, the cathode of the device, consisting of 10 nm Ca and
100 nm Al, was thermally deposited onto the active layer. Finally,
the devices were annealed at 110 ꢀC for 10 min. The active area of
the device was 7 mm2 for each cell. Current density-voltage (J-V)
characteristics were measured in the dark and under AM 1.5 G
illumination by a solar simulator (Oriel 94043A, 450 W). The in-
tensity was adjusted to be 100 mW cmꢂ2 under the calibration with
a NREL-certified standard silicon cell (Orial reference cell 91150).
External quantum efficiency (EQE) was detected with a 75 W Xe
lamp, Oriel monochromator 74125, optical chopper, lock-in
amplifier and a NREL-calibrated crystalline silicon cell. All mea-
surements were performed in a N2-filled glovebox at room
temperature.
The mixture of Cz1-Br (3.03 g, 7.28 mmol), 2,7-Dibromo-9H-
fluorene (0.78 g, 2.40 mmol), tetrabutyl ammonium iodide (0.09 g,
0.24 mmol), toluene (12 mL), and NaOH aqueous solution (50%,
12 mL) were stirred vigorously at 70 ꢀC for 12 h. After the reaction
mixture was cooled down to room temperature, the organic phase
was separated, washed with water, dried over MgSO4 and evapo-
rated to dryness. The residue was subjected to silica column chro-
matography using chloroform/hexane (1:5, v/v) as an elunent,
allowing to separate compound F-Cz1 (2.38 g) as white solid in a
yield of 100%. 1H NMR (300 MHz, CDCl3,
7.47e7.36 (m, 18H), 7.21 (t, J ¼ 8 Hz, 4H) 4.25 (t, J ¼ 8 Hz, 4H), 1.84
(m, 8H), 1.23e0.55 (m, 36H).
d
): 8.10 (d, J ¼ 8 Hz, 4H),
4.3.4. I-Cz
To a mixture of carbazole (16.72 g, 0.1 mol) and KI (21.58 g,
0.13 mol) was added with 280 mL glacial acetic acid. After heating
to reflux to dissolve all the solids in glacial acetic acid, the mixture
was cooled down and added with ground potassium iodate (32.1 g,
0.15 mol). Then, the reaction mixture was heated again to reflux for
1 h and afterward cooled down to room temperature. A large
amount of compound I-Cz was precipitated during the reaction.
The crude product was separated by filtration and washed with
water, affording pure I-Cz (35.24 g) as gray solid in yield of 84%. 1H
4.3. Materials
NMR (300 MHz, CDCl3,
2H), 7.21 (d, J ¼ 8 Hz, 2H).
d
): 8.31 (s, 2H), 8.10 (s, 1H), 7.68 (d, J ¼ 8 Hz,
Unless indicated, all commercial reagents were used as received.
Reaction solvents were dehydrated following common methods,
tetrahydrofuran (THF), ether, and toluene refluxed over a mixture
of Na and benzophenone, while chlorobenzene dried over CaH2
under argon, and freshly distilled prior to use.
4.3.5. I-Cz-Amide
To a mixture of Compound I-Cz (48.60 g, 116 mmol) and acetic
anhydride (188.5 mL), was added with boron trifluoride diethyl
etherate (2.23 mL). Then, the reaction mixture was refluxed for 1 h
and cooled to room temperature. During this procedure, a large
amount of product I-Cz-Amide was precipitated. After separated by
filtration, the crude product was sequently washed with water and
alcohol, affording pure I-Cz-Amide (43.61 g) as brown solid in a
4.3.1. 4,7-Di(2-trimethylstannylthiophen-5-yl)-2,1,3-
benzothiadiazole
To
a
solution of 2,2,6,6-tetramethylpiperidine (4.22 g,
29.9 mmol) in THF (85 mL), n-butyllithium hexane solution (1.6 M,
18.69 mL, 29.9 mmol) was added at a temperature of ꢂ78 ꢀC. After
stirring for 1 h at ꢂ78 ꢀC, the reaction mixture was warmed up to
room temperature, and then cooled again to ꢂ78 ꢀC, followed with
the dropwise addition of a solution of 4,7-di(2-thiophene)-2,1,3-
benzothiadiazol (3.45 g, 11.5 mmol) in THF (35 mL). After stirring
at ꢂ78 ꢀC for 2 h, Me3SnCl hexane solution (1.0 M, 29.9 mL,
29.9 mmol) was added and the reaction mixture was warmed up to
room temperature. After stirred for 12 h, the reaction was quenched
with deionized water and extracted with diethyl ether. The organic
phase was washed with deionized water several times, dried over
MgSO4, and evaporated to dryness. The residue was subjected to
yield of 82%. 1H NMR (300 MHz, CDCl3,
d): 8.22 (s, 2H), 7.94 (d,
J ¼ 9 Hz, 2H), 7.76 (d, J ¼ 9 Hz, 2H), 2.84 (s, 3H).
4.3.6. Cz3-Amide
To a mixture of I-Cz-Amide (0.46 g, 1 mmol), carbazole (0.37 g,
2.2 mmol), and copper powder (0.43 g, 3 mmol) was added with
N,N-dimethylacetamide (DMAc, 7 mL). The mixture was stirred at
170 ꢀC for 36 h. After cooling to room temperature, the mixture was
filtered through Celite and then added with water to precipitate the
product. After filtration, the filtrate cake was subjected to silica
column chromatography using chloroform/hexane (2:1, v/v) as an
eluant, allowing to separate compound Cz3-Amide (0.21 g) as gray
recrystallization
trimethylstannylthiophen-5-yl)-2,1,3-benzothiadiazole (6.38 g) as
orange crystals with a yield of 89%. 1H NMR (300 MHz, CDCl3,
):
from
acetone,
affording
4,7-Di(2-
solid in a yield of 38%. 1H NMR (300 MHz, CDCl3,
2H), 8.15 (m, 6H), 7.74 (m, 2H), 7.40 (m, 8H), 7.29 (m, 4H), 3.04 (s,
3H, CH3).
d): 8.51 (d, J ¼ 8 Hz,
d
8.18 (d, J ¼ 4 Hz, 2H), 7.87 (s, 2H), 7.30 (d, J ¼ 4 Hz, 2H), 0.43 (s, 18H).
4.3.2. Cz1-Br
4.3.7. Cz3
To a mixture of carbazole (5.02 g, 30 mmol), tetrabutyl ammo-
nium bromide (0.39 g, 1.2 mmol) was added with 15 mL toluene,
1,12-dibromododecane (19.69 g, 60 mmol) and NaOH aqueous so-
lution (50%, 15 mL). After stirred at 70 ꢀC overnight, the reaction
To compound I-Cz-Amide (1.21 g, 2.24 mmol), was added with
THF (78.5 mL), DMSO (33.5 mL), H2O (2.24 mL), and then KOH
(1.25 g, 22.28 mmol). The mixture was refluxed for 2 h, then cooled
to room temperature, and added with water to precipitate the

L.-F. Yu et al. / Polymer 59 (2015) 57e66
65
Fig. 8. (a) Fluorescence spectra of PFDTBT-C12, PFDTBT-Cz1, and PFDTBT-Cz3 films in pure form and blended with PC61BM (1:3, w/w) upon excitation at 560 nm. The spectra were
normalized to the A560 nm to be 0.1. (b) Fluorescence quenching extents of the above three blend films as referred to their pure films, respectively.
product. After filtration, the crude product was washed with water,
yield of 30%. 1H NMR (300 MHz, CDCl3,
d): 8.20 (s, 4H), 8.15 (d,
J ¼ 8 Hz, 8H), 7.63 (s, 8H), 7.46e7.37 (m, 22H), 7.27 (m, 8H), 4.45 (t,
J ¼ 8 Hz, 4H), 2.01 (m, 4H), 1.85 (m, 4H), 1.30e0.53 (m, 36H).
MALDI-TOF MS (m/z (%)): 1651.6 (100) [Mþ].
affording pure Cz3 (1.01 g) as white solid in a yield of 91%. 1H NMR
(300 MHz, CDCl3,
7.72 (d, J ¼ 8 Hz, 2H), 7.63 (dd, J ¼ 8 Hz, 2H), 7.39 (m, 8H), 7.29
d
): 8.50 (s, 1H), 8.21 (s, 2H), 8.17 (d, J ¼ 8 Hz, 4H),
(m, 4H).
4.3.10. PFDTBT-C12
4.3.8. Cz3-Br
To a mixture of 2,7-dibromo-9,9-didodecyl-fluorene (2.64 g,
4.0 mmol) and 4,7-di(2-trimethylstannylthiophen-5-yl)-2,1,3-
benzothiadiazol (2.50 g, 4.0 mmol) was added with freshly
distilled toluene (60 mL). After degassed by freezeepumpethaw
cycle three times, the mixture was added with Pd(PPh3)4 (120 mg,
0.104 mmol) under Ar protection, and then subjected to freeze-
epumpethaw cycle again to completely get rid of oxygen. After
refluxing for 4 d, the reaction mixture was poured into methanol to
precipitate the polymer product. After filtration, the crude product
was subjected to Soxhlet extraction with methanol, acetone, and
chloroform in sequence. The chloroform fraction was collected,
evaporated off the solvent, and dried in vacuum, giving polymer
PFDTBT-C12 (620.2 mg) as purplish red solid in a yield of 19%. Mn:
10057, PDI: 2.10. 1H NMR (400 MHz, 1,2-dichlorobenzene-d4, 80 ꢀC,
To a mixture of tetrabutyl ammonium bromide (0.003 g,
0.01 mmol), and aqueous 50% sodium hydroxide (0.3 mL) was
added a solution of 3,6-di(9H-carbazol-9-yl)-9H-carbazole (0.10 g,
0.2 mmol) and 1,12-dibromododecane (0.13 g, 0.4 mmol) in toluene
(0.3 mL). The mixture was stirred at 70 ꢀC overnight and then
poured into water. The organic components were extracted with
ethyl ether. The organic phase was washed with water and dried
over MgSO4. After removal of solvent, the residue was purified by
column chromatography. The eluant was trichloromethane/hexane
(2:1, v/v) giving the product (0.065 g, yield: 44%). 1H NMR
(300 MHz, CDCl3,
d
): 8.22 (s, 2H), 8.17 (d, J ¼ 8 Hz, 4H), 7.67 (s, 4H),
7.39 (m, 8H), 7.29 (m, 4H), 4.50 (t, J ¼ 8 Hz, 2H), 3.40 (t, J ¼ 7 Hz, 2H),
2.05 (m, 2H), 1.84 (m, 2H), 1.30 (m, 16H).
d): 7.97e6.53 (m, 12H) 2.03 (m, 4H), 1.07e0.59 (m, 46H).
4.3.9. F-Cz3
The mixture of Cz3-Br (0.29 g, 0.39 mmol), 2,7-Dibromo-9H-
fluorene (0.58 g, 0.18 mmol), tetrabutyl ammonium iodide (0.02 g,
0.05 mmol), toluene (20 mL), and NaOH aqueous solution (50%,
20 mL) were stirred vigorously at 70 ꢀC for 12 h. After the reaction
mixture was cooled down to room temperature, the organic phase
was separated, washed with water, dried over MgSO4 and evapo-
rated to dryness. The residue was subjected to silica column chro-
matography using chloroform/hexane (2:3, v/v) as an elunent,
allowing to separate compound F-Cz3 (0.090 g) as white solid in a
4.3.11. PFDTBT-Cz1
To a mixture of F-Cz1 (1.02 g, 1.03 mmol) and 4,7-di(2-
trimethylstannylthiophen-5-yl)-2,1,3-benzothiadiazol (0.64 g,
1.03 mmol) was added with freshly distilled toluene (30 mL). After
degassed by freezeepumpethaw cycle three times, the mixture
was added with Pd(PPh3)4 (31 mg, 0.027 mmol) under Ar protec-
tion, and then subjected to freezeepumpethaw cycle again to
completely get rid of oxygen. After refluxing for 4 d, the reaction
mixture was poured into methanol to precipitate the polymer
Fig. 9. Topographical height AFM images (5 ꢁ 5
m
m2) of blend films of (a) PFDTBT-C12, (b) PFDTBT-Cz1, and (c) PFDTBT-Cz3 with PC61BM (1:3, w/w).

66
L.-F. Yu et al. / Polymer 59 (2015) 57e66
[5] (a) Chen J, Cao Y. Acc Chem Res 2009;42:1709e18;
(b) Zhou H, Yang L, You W. Macromolecules 2012;45:607e32;
(c) Ye HY, Li W, Li WS. Chin J Org Chem 2012;32:266e83.
[6] For examples, see (a) Mei C-Y, Liang L, Zhao F-G, Wang J-T, Yu L-F, Li Y-X, et al.
Macromolecules 2013;46:7920e31;
product. After filtration, the crude product was subjected to Soxhlet
extraction with methanol, acetone, and chloroform in sequence.
The chloroform fraction was collected, evaporated off the solvent,
and dried in vacuum, giving polymer PFDTBT-Cz1 (559.6 mg) as
purplish red solid in a yield of 48%. Mn: 21,882, PDI: 17.9. 1H NMR
(b) Zhou P, Zhang Z-G, Li Y, Chen X, Qin J. Chem Mater 2014;26:3495e501;
(c) Jiang J-M, Lin H-K, Lin Y-C, Chen H-C, Lan S-C, Chang C-K, et al. Macro-
molecules 2014;47:70e8.
(400 MHz, 1,2-dichlorobenzene-d4, 80 ꢀC,
d): 7.92e6.53 (m, 28H)
[7] For examples, see (a) Ong K-H, Lim S-L, Tan H-S, Wong H-K, Li J, Ma Z, et al.
Adv Mater 2011;23:1409e13;
3.87 (t, J ¼ 7 Hz, 4H), 2.03 (m, 8H), 1.47e0.66 (m, 36H).
(b) Liang L, Wang J-T, Mei C-Y, Li W-S. Polymer 2013;54:2278e84;
(c) Hendriks KH, Li W, Wienk MM, Janssen RAJ. J Am Chem Soc 2014;136:
12130e6;
(d) Lee J, Jo SB, Kim M, Kim HG, Shin J, Kim H, et al. Adv Mater 2014;26:
6706e14;
4.3.12. PFDTBT-Cz3
To a mixture of F-Cz3 (0.089 g, 0.05 mmol) and 4,7-di(2-
trimethylstannylthiophen-5-yl)-2,1,3-benzothiadiazol (0.034 g,
0.05 mmol) was added with freshly distilled toluene (5 mL). After
degassed by freezeepumpethaw cycle three times, the mixture
was added with Pd(PPh3)4 (31 mg, 0.027 mmol) under Ar pro-
tection, and then subjected to freezeepumpethaw cycle again to
completely get rid of oxygen. After refluxing for 4 d, the reaction
mixture was poured into methanol to precipitate the polymer
product. After filtration, the crude product was subjected to
Soxhlet extraction with methanol, acetone, and chloroform in
sequence. The chloroform fraction was collected, evaporated off
the solvent, and dried in vacuum, giving polymer PFDTBT-Cz3
(50.7 mg) as purplish red solid in a yield of 53%. Mn: 7842, PDI:
(e) Jiang J-M, Yang P-A, Lan S-C, Yu C-M, Wei K-H. Polymer 2013;54:155e61.
[8] For examples, see (a) Osaka I, Kakara T, Takemura N, Koganezawa T,
Takimiya K. J Am Chem Soc 2013;135:8834e7;
(b) Cabanetos C, Labban AE, Bartelt JA, Douglas JD, Mateker WR, Frechet JMJ,
et al. J Am Chem Soc 2013;135:4656e9.
[9] (a) Chen L, Chen Y, Yao K, Zhou W, Li F, Chen L, et al. Macromolecules 2009;42:
5053e61;
(b) Yao K, Chen Y, Chen L, Li F, Li X, Ren X, et al. Macromolecules 2011;44:
2698e706;
(c) Watanabe M, Tsuchiya K, Shinnai T, Kijima M. Macromolecules 2012;45:
1825e32.
[10] (a) Shu CF, Dodda R, Wu F-I, Liu MS, Jen AK-Y. Macromolecules 2003;36:
6698e703;
(b) Gu C, Fei T, Lv Y, Feng T, Xue S, Lu D, et al. Adv Mater 2010;22:2702e5;
(c) Dong W, Xue S, Lu P, Deng J, Zhao D, Gu C, et al. J Polym Sci Polym Chem
2011;49:4549e55;
1.28. 1H NMR (400 MHz, 1,2-dichlorobenzene-d4, 80 ꢀC,
d):
7.87e6.53 (m, 56H) 4.06 (s, 4H), 1.98e1.64 (m, 8H), 1.33e0.69 (m,
36H).
(d) Yuan M-C, Su M-H, Chiu M-Y, Wei K-H. J Polym Sci Polym Chem 2010;48:
1298e309.
[11] (a) Svensson M, Zhang F, Veenstra SC, Verhees WJH, Hummelen JC, Kroon JM,
et al. Adv Mater 2003;15:988e91;
€
(b) Zhang FL, Gadisa A, Inganas O, Svensson M, Andersson MR. Appl Phys Lett
Acknowledgments
2004;84:3906e8.
[12] (a) Slooff LH, Veenstra SC, Kroon JM, Moet DJD, Sweelssen J, Koetse MM. Appl
Phys Lett 2007;90:143506;
This work is supported by NFSC (No. 21474129), Science and
Technology Commission of Shanghai Municipality (No.
13JC1407000), and Chinese Academy of Sciences.
(b) Chen M-H, Hou J, Hong Z, Yang G, Sista S, Chen L-M, et al. Adv Mater
2009;21:4238e42;
(c) Calabrese A, Schimperna G, Po R, Yohannes T, Debebe SE, Tinti F, et al.
J Appl Phys 2011;110:113106.
[13] Lee WY, Cheng K-F, Wang T-F, Chen W-C, Tsai F-Y. Thin Solid Films 2010;518:
́
2119e23.
References
[14] (a) Huang F, Cheng Y-J, Zhang Y, Liu MS, Jen AK-Y. J Mater Chem 2008;18:
4495e509;
[1] For selected reviews, see (a) Cheng Y-J, Yang S-H, Hsu C-S. Chem Rev
2009;109:5868e923;
(b) Gong W-L, Zhong F, Aldred MP, Fu Q, Chen T, Huang D-K, et al. RSC Adv
2012;2:10821e8;
(b) Li G, Zhu R, Yang Y. Nat Photonics 2012;6:153e61;
(c) Heeger A. J Adv Mater 2014;26:10e28;
(d) Su Y-W, Lan S-C, Wei K-H. Mater Today 2012;15:554e62.
[2] (a) Li G, Shrotriya V, Huang JS, Yao Y, Moriarty T, Emery K, et al. Nat Mater
2005;4:864e8;
(c) Jiang H, Sun J, Zhang J. Curr Org Chem 2012;16:2014e25.
ꢀ
[15] (a) Thompson BC, Frechet JMJ. Angew Chem Int Ed 2008;47:58e77;
(b) Coffin RC, Peet J, Rogers J, Bazan GC. Nat Chem 2009;1:657e61;
(c) Ashraf RS, Schroeder BC, Bronstein HA, Huang Z, Thomas S, Kline RJ, et al.
Adv Mater 2013;25:2029e34;
(b) Ma W, Yang C, Gong X, Lee K, Heeger AJ. Adv Funct Mater 2005;15:
1617e22.
(d) Wang J, Ye H, Li H, Mei C, Ling J, Li W, et al. Chin J Chem 2013;31:1367e79.
ꢀ
[16] (a) Hecht S, Frechet JMJ. Angew Chem Int Ed 2001;40:74e91;
[3] (a) He Z, Zhong C, Su S, Xu M, Wu H, Cao Y. Nat Photonics 2012;6:591e1622;
(b) Liao S-H, Jhuo H-J, Cheng Y-S, Chen S-A. Adv Mater 2013;25:4766e71;
(c) Ye L, Zhang S, Zhao W, Yao H, Hou J. Chem Mater 2014;26:3603e5;
(d) Nguyen TL, Choi H, Ko S, Uddin MA, Walker B, Yum S, et al. Energy Environ
Sci 2014;7:3040e51.
(b) Li W-S, Jiang D-L, Aida T. Angew Chem Int Ed 2004;43:2943e7.
[17] (a) Li Y, Cao Y, Gao J, Wang D, Yu G, Heeger AJ. Synth Met 1999;99:243e8;
(b) Zhang J, Cai W, Huang F, Wang E, Zhong C, Liu S, et al. Macromoleculers
2011;44:894e901.
ꢀ
[18] (a) Goh C, Kline RJ, McGeheea MD, Kadnikova EN, Frechet JMJ. Appl Phys Lett
[4] (a) You J, Dou L, Yoshimura K, Kato T, Ohya K, Moriarty T, et al. Nat Commun
2013;4:1446;
2005;86:122110;
(b) Liang L, Wang J-T, Xiang X, Ling J, Zhao F-G, Li W-S. J Mater Chem A
(b) Chen C-C, Chang W-H, Yoshimura K, Ohya K, You J, Gao J, et al. Adv Mater
2014;26:5670e7.
2014;2:15396e405.