Superiority of branched side chains in spontaneous nanowire formation: Exemplified by poly(3-2-methylbutylthiophene) for high-performance solar cells

Article abstract of DOI:10.1002/smll.201002196

One-dimensional nanostructures containing heterojunctions by conjugated polymers, such as nanowires, are expected to greatly facilitate efficient charge transfer in bulk-heterojunction (BHJ) solar cells. Thus, a combined theoretical and experimental approach is pursued to explore spontaneous nanowire formation. A dissipative particle dynamics simulation is first performed to study the morphologies formed by rodlike polymers with various side-chain structures. The results surprisingly predict that conjugated polymers with branched side chains are well suited to form thermodynamically stable nanowires. Proof of this concept is provided via the design and synthesis of a branched polymer of regioregular poly(3-2-methylbutylthiophene) (P3MBT), which successfully demonstrates highly dense nanowire formation free from any stringent conditions and stratagies. In BHJ solar cells fabricated using a blend of P3MBT and [6,6]-phenyl-C71-butyric acid methyl ester (PC₇₁BM), P3MBT polymers are self-organized into highly crystalline nanowires with a d₁₀₀ spacing of 13.30 A. The hole mobility of the P3MBT:PC₇₁BM (1:0.5 by weight) blend film reaches 3.83 A× 10^-4 cm ² V^-1 s^-1, and the maximum incident photon-to-current efficiency reaches 68%. The results unambiguously prove the spontaneous formation of nanowires using solution-processable conjugated polymers with branched alkyl side chains in BHJ solar cells.

Full text of DOI:10.1002/smll.201002196

full papers
Solar Cells
Superiority of Branched Side Chains in Spontaneous
Nanowire Formation: Exemplified by Poly(3-2-
methylbutylthiophene) for High-Performance Solar Cells
Hsieh-Chih Chen, I-Che Wu, Jui-Hsiang Hung, Fu-Je Chen, I-Wen P. Chen,
Yung-Kang Peng, Chao-Sung Lin, Chun-Hsien Chen, Yu-Jane Sheng,
Heng-Kwong Tsao,* and Pi-Tai Chou*
O ne-dimensional nanostructures containing heterojunctions by conjugated polymers,
such as nanowires, are expected to greatly facilitate efficient charge transfer in bulk-
heterojunction (BHJ) solar cells. Thus, a combined theoretical and experimental
approach is pursued to explore spontaneous nanowire formation. A dissipative
particle dynamics simulation is first performed to study the morphologies formed
by rodlike polymers with various side-chain structures. The results surprisingly
predict that conjugated polymers with branched side chains are well suited to form
thermodynamically stable nanowires. Proof of this concept is provided via the design
and synthesis of a branched polymer of regioregular poly(3-2-methylbutylthiophene)
(P3MBT), which successfully demonstrates highly dense nanowire formation free
from any stringent conditions and stratagies. In BHJ solar cells fabricated using a
blend of P3MBT and [6,6]-phenyl-C71-butyric acid methyl ester (PC71BM), P3MBT
polymers are self-organized into highly crystalline nanowires with a d100 spacing
of 13.30 Å. The hole mobility of the P3MBT:PC71BM (1:0.5 by weight) blend
film reaches 3.83 × 10−4 cm2 V−1 s−1 , and the maximum incident photon-to-current
efficiency reaches 68%. The results unambiguously prove the spontaneous formation
of nanowires using solution-processable conjugated polymers with branched alkyl
side chains in BHJ solar cells.
Prof. Y.-J. Sheng
Dr. H.-C. Chen, Dr. I.-C. Wu, Dr. I.-W. P. Chen, Y.-K. Peng, Prof. C.-H. Chen,
Prof. P.-T. Chou
Department of Chemistry
National Taiwan University
Taipei, 106, Taiwan
Department of Chemical Engineering
Institute of Polymer Science and Engineering
National Taiwan University
Taipei, 106, Taiwan
E-mail: chop@ntu.edu.tw
Prof. H.-K. Tsao
Department of Chemical and Materials Engineering
Department of Physics
National Central University, Jhongli, 320, Taiwan
E-mail: hktsao@cc.ncu.edu.tw
J.-H. Hung
Institute of Polymer Science and Engineering
National Taiwan University
Taipei, 106, Taiwan
Dr. F.-J. Chen, Prof. C.-S. Lin
Department of Materials Science and Engineering
National Taiwan University
Taipei, 106, Taiwan
DOI: 10.1002/smll.201002196
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1098 wileyonlinelibrary.com
small 2011, 7, No. 8, 1098–1107

Superiority of Branched Side Chains in Spontaneous Nanowire Formation
than systematic design. However, recent advancements in
polymer simulations have enabled predictive studies of the
dependence of morphology upon side-chain types. To explore
the mesoscale morphology, a coarse-grained model of conju-
gated polymers must be employed. The dissipative particle
dynamics (DPD) method, proposed by Hoogerbrugge and
Koelman,[23] is a genuine mesoscale approach for studying
the behavior of complex fluids, including micelles, polymeric
systems, liquid crystals, and colloidal suspensions.[24–27] DPD
incorporates the idea of coarse-grained beads, which are clus-
ters of several atoms or molecules that have similar proper-
ties.The interaction potential between any pair of DPD beads
is thus soft, allowing the system to evolve much faster. There-
fore, DPD can treat a wider range of length and time scales
compared to atomistic simulations such as classical molecular
dynamics. Herein, conjugated polymers were modeled as rod-
like chains with various side-chain structures, and DPD was
performed to investigate the influence of the side-chain type
on the morphology formed by polymers.
1. Introduction
Polymer solar cells (PSCs) based on solution-processable
conjugated polymers are a promising candidate for the
production of a low-cost renewable energy source.[1] Such
PSCs are light-weight and mechanically flexible, and the
properties of the photoactive layer can be straightforwardly
fine-tuned by device architecture and chemical composition.
An outstanding configuration of PSCs should be credited to
the bulk heterojunction (BHJ), which is based on the inter-
penetration of electron donor (D) and acceptor (A) phases
and has been extensively exploited in the active layer.[2–16]
In blends of separate D and A phases, the characteristic sizes
of D and A domains play a key role in the improvement of
photovoltaic performance. To facilitate efficient charge trans-
port prior to charge recombination, an ideal domain size is
close to the exciton diffusion length (i.e., ∼10–20 nm).[17–19]
Another attractive approach is to provide a direct pathway
for charge transport through 1D nanostructures, such as
nanowires[20,21] and nanofibers.[22] In general, more efficient
transport of charge carriers is critical for creating more effi-
cient solar cells.
2. Results and Discussion
BHJ devices require sensitive control over domain
morphology on the nanoscale, which is largely affected by
the choice of materials, solvents, and the D:A weight ratio.
A prototypical polymer/fullerene system is the blend of
regioregular poly(3-hexylthiophene) with [6,6]-phenyl-C61-
butyric acid methyl ester (P3HT:PC61BM). Linear alkyl side
2.1. The Influence of Side-Chain Bulkiness on the
Morphological Structure of Polymer Aggregates
Conjugated polymers with three types of side-chain struc-
chains are attached to increase the solubility of conjugated tures were considered, as illustrated in Figure 1a. Polymer
polymers in common solvents. In addition, the performance architectures labeled A to C are represented by a rod-
of P3HT:PC61BM blend solar cells has been progressively like backbone with linear, double, and branched structures,
improved by various annealing protocols, and the mechanism respectively. Note that the rods are depicted in light gray, and
of enhanced performance is attributed to the improvement of the side chains are in black. The influence of side-chain bulki-
the two-phase nanoscale morphology frozen in the metastable ness on the morphological structure of polymer aggregates
state associated with the phase separation of the blend. Evi- was thus systematically studied. As a result, we found an
dently, the formation of 1D nanostructures greatly facilitates interesting and perhaps previously unrecognized morpholog-
charge carrier mobility by the provision of a superhighway ical transition. Figure 1a shows the overall morphology from
for charge carriers to travel along to the electrode. Although the top, the cross-sectional view and isosurface of the domain
nanofibers and nanowires have been prepared from regioreg- formed by rods, and the aggregates of backbones for the three
ular P3HT and poly(3-butylthiophene) (P3BT), respectively, molecular architectures considered. For the linear type, the
the formation of stable 1D nanostructures remains elusive equilibrium morphology is a multishell, football-like aggre-
because the thermodynamics and kinetics of blend phase sep- gate. For the double type, the rod domain of the aggregate
aration cannot be controlled due to the numerous variables is formed by sticklike regions. When side chains further grow
involved.[21,22] It is difficult to gain improvement in a system- to a branched pattern, a bundle of distinct strings is formed
atic manner using current methods, which involve control- in the cylindrical aggregate, as is clearly shown in the cross-
ling the film morphology of the polymer/fullerene blend by section. This result can be accounted for by the bulkiness of
annealing protocols and may cause poor stability and dura- the side chain. In the absence of solvophilic side chains, the
bility of BHJ PSCs.As a consequence, spontaneous formation solvophobic backbones cannot dissolve in the selective sol-
of 1D nanostructures (i.e., without resorting to any strin- vent. However, the presence of side chains aids in the dis-
gent requirements) should be the most attractive approach. solution of the rodlike polymers. Nonetheless, the polymers
Comparative studies of regioregular poly(3-akylthiophene) still form aggregates, similar to micelles formed by surfactant
indicate a large variation in morphology and mobility upon molecules.[28,29] Due to the incompatibility between rods and
varying the length of the alkyl side chain.[7,15] It is therefore side chains, they tend to aggregate with themselves and form
anticipated that the film morphology will be influenced by rod and side-chain domains, respectively. In the rod domain,
the architecture of the side chain; the nanowires may eventu- the steric interactions among rods lead to alignment of the
ally be a thermally favorable state if the side-chain structures rods to maximize the translational entropy, similar to liquid
are appropriately designed.
crystals, as clearly demonstrated in Figure 1b.
Thus far, efforts to obtain 1D nanostructures that enhance
The morphology of the aggregates was then determined
hole mobility have proceeded largely by trial and error rather by the free energy competition between the interfacial
small 2011, 7, No. 8, 1098–1107
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
wileyonlinelibrary.com 1099

H. -C. Chen et al.
full papers
Figure 1. a) Architectures of the rodlike polymers with various types of side chains (A–C). From the top of each column, morphologies of the polymer
aggregate in the simulation cell, cross-sectional view (left)/isosurface of the rod domain (right), and the aggregate of the backbones for three types
of side chains. For the aggregate morphologies, the side chains are in black and the rods are in light gray. b) For the rodlike polymer with branched
side chains, from left to right, the enlarged nanowire morphology, microscopic arrangement of rods, and the isosurface of the rod domain. The
figure also shows a snapshot of a unit string and the pattern of the strings within a wire-like structure with side chains removed.
energy associated with unfavorable interactions (between the side-chain domains tends to be as smooth and as minimal
rod and side-chain domains) and the free-volume entropy as possible to minimize the interfacial energy increment, as
associated with side chains. The linear side chains are less shown in the isosurface of Figure 1a for the linear type. When
bulky, and their demand for free volume is thus easier to the side chain becomes bulky, the entropic penalty increases.
satisfy. The reduction of interfacial energy is dominant and As a result, the isosurface of the rod domain becomes rough,
can be fulfilled by the formation of a continuous shell of and the stick-like regions appear, as depicted in Figure 1a for
rod domain. Moreover, the interface between the rod and the double type. When the side chains become very bulky, as
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1100 wileyonlinelibrary.com
small 2011, 7, No. 8, 1098–1107

Superiority of Branched Side Chains in Spontaneous Nanowire Formation
in the branched type, the entropic factor becomes dominant,
and the interfacial energy must be sacrificed to provide more
free volume. Consequently, a nanowire formed by a bundle
of dispersed strings appears, which have much more inter-
facial area than the shell. Detailed arrangement of the rods
belonging to a string within a nanowire is further presented
in Figure 1b to illustrate its microscopic structure.
Br
Br
NBS
THF/AcOH
S
S
Mg
NidpppCl2
3
Note that polymer chains in conventional knowledge are
expected to be perpendicular to the ribbons as in previous
experiments for P3HT.[30–32] However, our simulation results
indicate a different scenario; our polymer chains (P3MBT;
poly(3-2-methylbutylthiophene)) are clearly parallel to the
long axis of the nanowire. It should be noted that nanorib-
bons formed by P3HTs are precipitated out into polycrystal-
line material by a cooling process while P3MBT nanowires
are spontaneously formed in solution through a self-assembly
process at ambient temperature. The formation mechanisms
are fundamentally different.
CH3MgBr
Br
Br
NidpppCl2
S
1
n
S
2
Scheme
1. Synthetic
strategy
for
regioregular
poly(3-2-
methylbutylthiophene) (1). NBS, THF, AcOH, dppp represent
N-bromosuccinimide, tetrahydrofuran, acetic acid, and 1,3-bis
(diphenylphosphino)propane.
In previous works,[30–32] P3HTs were first dissolved in their
marginal solvents at high temperature.They were cooled down
to room temperature to be deposited from their solutions and
then crystallized dynamically into polycrystalline material
upon solidification. In a crystallization process, it is known
that chains tend to fold together to form ordered regions
called lamellae. Therefore, it is anticipated that nanoribbons
are the result of the layer-by-layer stacking of the lamellas.
As a consequence, the width of the nanoribbon grows with
concentration. In stark contrast, in our experiments, the
nanowires are formed spontaneously in the selective solvents.
Similar to micelle formation, these nanowires are thermody-
namically stable, self-assembled structures. As a result, the
polymer backbones prefer to orient parallel to the nanowire
in order to increase the translational entropy and reduce the
interfacial area between solvophobic backbones and solvents.
The diameters of the nanowires are essentially independent
of the solution concentration, which is notably different from
the nanoribbons formed by P3HT[32] (vide infra).
The above theoretical standpoints defy conventional
wisdom regarding, for example, polythiophenes with linear
alkyl side chains, for which the formation of 1D nanowires
can occur under specific conditions (e.g., slow sedimentation
in a selected solvent and at a specified temperature).[21,22,30–32]
Bearing this controversy in mind, we then carried out the
synthesis of a conjugated polymer with branched side chains,
namely, regioregular P3MBT (Scheme 1), via a facile syn-
thetic route (see Supporting Information, SI). Figure 2a
depicts the structures of P3MBT together with a sketch of
the energy levels for P3MBT and [6,6]-phenyl-C71-butyric
acid methyl ester (PC71BM).
2.2. Morphology and Structure
Figure 3a,e show the morphological features and phase-
mode images, respectively, of pristine P3MBT by tapping-
mode atomic force microscopy (TM-AFM). The P3MBT
film exhibits an extremely homogeneous nanowire structure,
which is indicative of a relatively extended conformation
Figure 2. a) Chemical structures and energy level diagram of P3MBT
and PC71BM. The highest occupied molecular orbital (HOMO) levels
are taken from cyclic voltammetry (CV) measurements while the lowest
unoccupied molecular orbital (LUMO) levels were calculated from
UV–vis absorption spectroscopy. b) A schematic representation of
of the polymer aggregate. Figure 4a reveals transmission P3MBT:PC71BM morphology after annealing.
small 2011, 7, No. 8, 1098–1107
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
wileyonlinelibrary.com 1101

H. -C. Chen et al.
full papers
Figure 3. Morphology characterization of a–d) TM-AFM topography and e–h) phase images of P3MBT:PC71BM blend films with various compositions:
a,e) 1:0; b,f) 1:0.5; c,g) 1:1; and d,h) 1:2. The area is 1.5 μm × 1.5 μm.
Figure 4. TEM images in combination with selected-area electron diffraction (SAED) patterns and enlarged images of P3MBT:PC71BM blend films
with different P3MBT:PC71BM compositions: a) 1:0, b) 1:0.5, c) 1:1, and d) 1:2.
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1102 wileyonlinelibrary.com
small 2011, 7, No. 8, 1098–1107

Superiority of Branched Side Chains in Spontaneous Nanowire Formation
electron microscopy (TEM) images and
the corresponding selected-area electron
diffraction (SAED) pattern for pristine
P3MBT; the latter was used to iden-
tify the internal order of the polymer
films. As shown by the SAED patterns,
a large fraction of the film is occupied
by nanowires, and these nanowire fea-
tures are ascribed to the P3MBT crystal-
lites originating from very tight stacks of
several polymer chains. Figure 4a also
reveals that the nanowire-based P3MBT
film possesses an elongated shape with
an average diameter of 12.5 nm, which
is consistent with the TM-AFM obser-
vation. Note that the P3MBT nanowires
are spontaneously formed in common
solvents, such as ortho-dichlorobenzene
(ODCB) or chloroform (see Figure S3
in the SI), simply by solvent evaporation
in an open system at room temperature.
This result unambiguously supports the
prediction based on the DPD approach
Figure 5. GIXRD diffraction patterns of P3MBT:PC71BM blend films with various
compositions.
(vide supra).
Further fabrication of the solar cell
requires blending P3MBT with an acceptor
(e.g., PC71BM). It was thus imperative to examine whether to the higher concentration of PC71BM. The surface rough-
the spontaneous formation of nanowire takes place in the ness of this blend is 12.3 nm, which is significantly greater than
presence of PC71BM. Indeed, nanowires in P3MBT:PC71BM that of the other blends (∼5–8 nm) as shown in Figure 3a–c. In
blend films with various compositions (by weight) were addition, it is noteworthy that in Figure 3h, P3MBT nanowires
also spontaneously formed, as shown in panels b–d and are surrounded by large PC71BM clusters with an average size
f–h of Figure 3. For the sake of clarity in the following dis- of 53.2 nm (Figure 4d). Evidently, the presence of PC71BM
cussion, a cartoon-like 3D schematic representation of the changes the morphology of P3MBT nanowires, which varies
P3MBT:PC71BM morphology is depicted in Figure 2b. The with the amount of PC71BM added, and may accordingly alter
TM-AFM images (Figure 3b,f) show that the composite film the crystallite size and absorption spectrum, which are pivotal
of the P3MBT:PC71BM (1:0.5) blend has a relatively smooth to photovoltaic performance (vide infra).
surface, displaying densely packed nanowires. It suggests a rel-
The SAED patterns associated with the P3MBT:PC71BM
atively good mixing of the two components, which would not blends of various compositions show sharp diffraction rings,
inhibit charge transfer. The TEM image in Figure 4b indicates implying crystalline formation of P3MBT. The crystallinity
that the nanowires have a lateral dimension ranging from and crystalline size of nanowires can be further investigated
10 to 17 nm, which is approximately the same as the exciton through their diffraction patterns via grazing-incidence X-ray
diffusion length, and may effectively assist charge transport diffraction (GIXRD) measurements. Figure 5 shows that all of
prior to recombination.[17–19] The PC71BM molecules tend to the blend films contain ordered structures. A strong (100) dif-
homogeneously fill the spaces in the P3MBT network[33–35] fraction peak at 2θ = 6.64° with a d100 spacing of 13.30 Å cor-
(as depicted by the TEM image in Figure 4b) and form an responds to an ordered and self-organized lamellar structure,
amorphous matrix, providing a continuous pathway for elec- with an interlayer spacing between two P3MBT backbones
tron transport. Moreover, the enlarged image of the 1:0.5 separated by the branched alkyl side chains.The peaks at 2θ =
blend (insert of Figure 4b) clearly suggests a lack of phase 13.28° and 19.70° correspond to the (200) and (300) reflec-
separation; this may be attributed to the fact that phase sep- tions, respectively. The amorphous hump typically observed
aration may only take place in tiny domains that are very near 20° appears to be very weak. These results reaffirm that
small compared to the film thickness, resulting in no observ- the spontaneously assembled P3MBT nanowires are highly
able contrast. In comparison, the TM-AFM images of the crystalline. All of the P3MBT and PC71BM peaks and the cor-
1:1 blend (Figure 3c,g) reveal larger nanowire diameters (c.f. responding d spacing values are listed in Table S2 in the SI.
1:0.5 blend). Furthermore, beadlike clusters can be found spo- Furthermore, the crystal size of the nanowires (t) can be esti-
radically, suggesting the formation of PC71BM nanocrystals mated through the full width at half maximum (FWHM) of
embedded in the amorphous PC71BM matrix.The TEM image the GIXRD diffraction peaks with the Scherrer formula:[36]
shown in Figure 4c reveals the growth of large PC71BM crys-
tallite domains with an average size of 31.6 nm. In Figure 3d,
the surface of the 1:2 blend becomes increasingly uneven due
8
t = 0.9 ×
(1)
B × cos2B
small 2011, 7, No. 8, 1098–1107
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
wileyonlinelibrary.com 1103

H. -C. Chen et al.
full papers
where B represents the FWHM of the XRD peak, λ is the of the nanowires. Before annealing, the absorption spectrum
wavelength of the X-rays, and θB is the Bragg diffraction angle. exhibits a maximum at 506 nm and two shoulders at 553
For the (100) reflection of the P3MBT:PC71BM films, the crys- and 602 nm, which are assigned to the crystalline π-stacking
tallite sizes are 15.20, 12.77, and 10.70 nm for the 1:0.5, 1:1, structure of P3MBT.[37,38] After thermal annealing, the max-
and 1:2 blends, respectively, and these values are consistent imum absorption peak is slightly red-shifted to 510 nm, and
with those estimated by TEM. The smaller P3MBT crystallite the shoulder bands are augmented in terms of absorption
sizes for the 1:1 and 1:2 blends imply that P3MBT interchain cross-section.[6,14] The influence of PC71BM on the crystal-
interactions are disrupted due to interfacial structure incom- linity of nanowires was also studied by examining the UV–vis
patibility between P3MBT and PC71BM nanocrystals. This spectra of the blend films, as depicted in Figure 6b. When the
effect was further examined by UV–vis absorption spectra.
PC71BM composition is low, the amorphous PC71BM matrix
only weakly perturbs the P3MBT nanowires, resulting in an
extended conjugated P3MBT structure similar to that in the
absence of PC71BM. Therefore, the absorption spectral fea-
ture remains essentially unchanged. As the PC71BM compo-
sition increases, the PC71BM nanocrystals, which are formed
in proximity to the P3MBT nanowires, grow significantly
and then interfere with the interchain stacking interactions
in the nanowires. Because the crystallinity of the nanowires
declines with increasing PC71BM concentration, the absorp-
tion peak undergoes a blue-shift from 506 to 487 nm, accom-
panied by decreases in the vibronic structures located at
553 and 602 nm. A similar behavior has also been reported
for P3HT:PC61BM and P3BT:PC61BM blends.[21,22,39] The
P3MBT:PC71BM (1:0.5) film enhances the red-light absorp-
tion due to its more organized chain packing, thus harvesting
more photons for increased solar cell performance.[6,14,34,35]
2.3. Optical Properties
All of the aforementioned samples were treated by
annealing at 150 °C for 10 min to enhance crystallinity.[6,14]
Comparisons of the absorption spectra associated with the
pristine P3MBT films before and after thermal annealing are
shown in Figure 6a and demonstrate the crystallinity changes
2.4. Photovoltaic Properties
The transport of holes in conjugated polymers is often the
rate-limiting step in PSCs. To demonstrate the advantage of
nanowire morphology and the influence of PC71BM, the hole
mobility of the blends was measured using the space-charge-
limited current (SCLC) model at low voltage.[40–42] However,
in the blend devices, the presence of Ohmic contacts at both
interfaces could lead to a SCLC that is a combination of both
hole and electron current. In order to measure SCLC of one
type of charge carrier only, the SCLC of the other carrier has
to be suppressed by a large injection barrier, resulting in a
hole-only or electron-only diode device. The device charac-
teristics were extracted by modelling the dark current under
forward bias using the SCLC expression described by the
Mott–Gurney law:
J = 9g0gr:hV2/ 8L3
(2)
where J is current density, ε0 is the permittivity of free space, εr
is the dielectric constant of the polymer, μh is the hole mobility,
V is the voltage drop across the device, L is the polymer thick-
ness, and V = Vappl – Vr – Vbi (where Vappl is the applied voltage
to the device, Vr is the voltage drop due to contact resistance
and series resistance across the electrodes, and Vbi is the built-
in voltage due to the difference in work function of the two
electrodes). The resistance of the device was measured using
a blank configuration ITO/PEDOT:PSS/Al and was found to
be ∼25 Ω (ITO: indium tin oxide; PEDOT: poly(3,4-ethylene-
dioxythiophene); PSS: poly(styrenesulfonate)). The dielectric
Figure 6. a) Normalized absorption spectra for a pristine P3MBT film
before and after thermal annealing. b) Normalized absorption spectra
for P3MBT:PC71BM blend films with different compositions. All of
the films were dried slowly in a covered Petri dish and subsequently
thermally annealed at 150 °C for 10 min.
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1104 wileyonlinelibrary.com
small 2011, 7, No. 8, 1098–1107

Superiority of Branched Side Chains in Spontaneous Nanowire Formation
Figure 7. Determination of the hole mobility from the dark current
densities for P3MBT:PC71BM diodes with various compositions
(symbols). The solid lines are fits of the data points. The applied voltage
was corrected for the built-in voltage (Vbi) arising from the work function
difference between the contacts and the voltage loss (Vr) due to contact
resistance and series resistance across the electrodes. The thickness of
the films is indicated in the plots.
constant was assumed to be 3 in our analysis, which is a typ-
ical value for conjugated polymers.
Figure 7 shows the experimental dark current densities
(JD) of P3MBT:PC71BM blends, which were measured in
hole-only diode devices. When the applied voltage is greater
than Vbi, JD throughout all devices scales quadratically with
voltage (trap-free region), indicating the feasibility of space-
charge-limited (SCL) transport. In this range, SCL follows
the SCLC model and its value allows for a direct determi-
nation of a constant mobility. When voltage is too high or
too low, serious deviation takes place, and the result may no
longer obey the SCLC model.[41] Applying the Mott–Gurney
law for SCLC, the hole mobility values were calculated to be
3.83 × 10−4 , 1.82 × 10−4 , and 7.95 × 10−5 cm2 V−1 s−1 for the
1:0.5, 1:1, and 1:2 of P3MBT:PC71BM blends, respectively. As
a consequence, P3MBT nanowires provide short pathways to
the electrodes and thus offer efficient hole transport, supe-
rior to typical phase-separated blends. Although all of the
P3MBT:PC71BM blend films possessed an electrically bicon-
Figure 8. a) IPCE spectra of optimized devices made from different
ratios of the P3MBT:PC71BM blend illuminated by monochromatic light.
b) Current density versus voltage (J–V) characteristics of optimized
devices made from different blend ratios of P3MBT:PC71BM measured
under AM1.5 white light illumination at 100 mW cm−2 . For the 1:0.5
device (•): Voc = 0.72 V, Jsc = –9.15 mA cm−2 , FF = 0.63, and PCE = 4.15%.
For the 1:1 device ( ): Voc = 0.65 V, Jsc = –7.98 mA cm−2 , FF = 0.57, and
᭿
PCE = 2.96%. For the 1:2 device ( ): Voc = 0.63 V, Jsc = –7.06 mA cm−2
FF = 0.54, and PCE = 2.40%.
,
᭡
tinuous morphology, the 1:0.5 blend displayed a significantly efficiency (IPCE) is closely related to the current of the
higher hole mobility, the value of which is comparable to the minority carrier.According to the PSC devices fabricated with
reported values of highly efficient solar cell polymers.[3,16] various P3MBT:PC71BM blend ratios, very broad panchro-
This result is attributed to the higher crystallinity of the matic spectra over the entire excitation spectral range were
nanowires, as manifested by the crystalline size and red-shift obtained from IPCE measurements, as shown in Figure 8a.
of the absorption peak. As a result, the film with a blend The spectrum of the 1:0.5 active layer exhibited a maximum
ratio of 1:0.5 is capable of attaining a better balance between IPCE of ∼68% in the 460–510-nm range and nearly 50% in
transport of electrons and holes across the device, which is the 388–570-nm range. Moreover, this result also demonstrates
advantageous to the fill factor[43,44] and power-conversion that the photon wavelength for the current generation extends
efficiency of PSCs owing to the accumulated SCLC charges, to ∼750 nm, as expected from the absorption spectrum of the
and recombination processes were reduced by the increased active layer (Figure 6b). Regardless of the blend ratios, the
hole mobility and enhanced charge collection efficiency.
The spontaneous formation of P3MBT nanowires facili- exceeds 45% for more than half of the visible region.
tates efficient hole transport, while the presence of PC71BM
All of the photovoltaic parameters for P3MBT:PC71BM
data for these devices highlight the important fact that IPCE
may affect the nanowire crystallinity and the solar cell nanowire-based PSCs are summarized in Table 1. A s
performance. For example, the incident photon-to-current anticipated, a PSC with a blend ratio of 1:0.5 exhibits the
small 2011, 7, No. 8, 1098–1107
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
wileyonlinelibrary.com 1105

H. -C. Chen et al.
full papers
T a b l e 1 . Optimized performance of P3MBT:PC71BM nanowire-based PSCs.
P3MBT:PC71BM blend ratios
Voc
Jsc
FF
PCE
[%]
Hole mobility
[cm2 V−1 s−1
Rs
Rsh
[Ω cm2]a)
[V]
[mA cm−2
]
]
[Ω cm2]a)
1:0.5
1:1
0.72
0.65
0.63
9.15
7.98
7.06
0.63
0.57
0.54
4.15
2.96
2.40
3.83 × 10−4
1.82 × 10−4
7.95 × 10−5
3.2
6.1
635
582
560
1:2
12.2
a)Rs, series resistance; Rsh, shunt resistance. The series and shunt resistances were deduced by the inverse gradient of the J–V curves.
best performance, in which the open-circuit voltage (Voc)
of the device can reach 0.72 V, with a short-circuit cur-
4. Experimental Section
The P3MBT polymer was synthesized as reported in the Supporting
Information. The molecular weight (Mn) of P3MBT was 16 800
with a PDI of 1.12. Details about the dissipative particle dynamics
method are described in the Supporting Information.
rent density (Jsc) of 9.15 mA cm−2 and a fill factor (FF)
of 0.63, resulting in a power-conversion efficiency (PCE)
of 4.15% (Figure 8b). The high PCE may result from
the short branched side chain of P3MBT, enhancing
the charge separation efficiency. Note that regardless
of the blend ratio, the results in Table 1 show that our
Optimized solar cell devices used in this study were prepared
by dissolving P3MBT and PC71BM (purchased from Nano-C) in
ODCB in the weight ratios of 1:0.5, 1:1, and 1:2 with a P3MBT
concentration of 20 mg mL−1 . For device fabrication, a solution
was stirred for more than 48 h to ensure complete dissolution.
Patterned ITO-coated glass substrates (Kintec Company, 15 Ω per
square) were successively cleaned by ultrasonication in 1% neutral
detergent in water, then deionized water, followed by acetone and
finally ethanol for 20 min each. The substrates were then dried and
O2-plasma-cleaned immediately prior to the deposition of a 35-nm-
thick layer of PEDOT:PSS, as verified by a Dektak 3030 profilom-
eter. Deposition of the PEDOT:PSS layer was followed by baking
at 140 °C for 20 min. Substrates were subsequently transferred to
an inert N2-filled glove box (<0.1 ppm O2 and H2O), and the active
layer was spin-coated. The wet film was slowly dried in a covered
Petri dish for a certain time and subsequently annealed at 150 °C
for 10 min in the glove box. The thicknesses of the P3MBT:PC71BM
composite films were adjusted to ∼70–100 nm by controlling the
spin-coating rate. The coated substrates were then transferred to
a thermal evaporator and evacuated to ≤1 × 10−6 Torr before a
30-nm-thick calcium layer followed by a 100-nm aluminum elec-
trode layer were deposited. The thin interfacial layer of calcium has
been reported to enhance the fill factor of the devices.[10,14,16]
TM-AFM images were taken on a NanoScope IIIa controller
(Veeco Metrology Group/Digital Instruments, Santa Barbara, CA),
using NanoScope software (version V6.13R1) to capture images. TEM
images and SAED patterns were obtained using a Philips FEI Tecnai
F20 G2 fi eld emission transmission electron microscope. GIXRD pat-
terns were collected using a Rigaku-TTRAX III diffractometer, oper-
ating at an accelerating voltage of 50 kV and current of 300 mA. A
Ni-filtered Cu Kα radiation source (0.154 nm) was used. A 0.4° inci-
dent angle, a 0.01° sampling step, a 5° min−1 scanning rate, and a
∼4–30° of scanning range at 2θ were carefully chosen to allow for
complete X-ray penetration into the polymer film. The absorption
spectra were obtained on a Hitachi U-4100 spectrophotometer. Hole
mobility was measured according to a similar method described in
the literature,[40–42] using a diode configuration of ITO/PEDOT:PSS/
polymer/Al. The SCLC current was measured under dark conditions
using a Keithley 2400 Source Meter. The fabricated device was
encapsulated in a nitrogen-filled glove box with UV epoxy and cover
glass. The J–V curves were measured with a Newport-Oriel AM 1.5 G
light source operating at 100 mW cm−2 and independently cross-
checked using a 300-W AM1.5 G source operating at 100 mW cm−2
nanowire-based devices have
a higher Voc compared
to that obtained from a P3HT:PC61BM device (Voc
=
0.55–0.60). The high Voc value is mainly due to the energy
level difference between the highest occupied molecular
orbital (HOMO) level of the donor (–5.23 eV) and the
lowest unoccupied molecular orbital (LUMO) level of the
acceptor (–4.02 eV). Moreover, upon further vertical phase
separation, the top contact from the 1:0.5 nanowires to
the electrode increases the shunt resistance by eliminating
shorting paths present in the active layer. A similar effect on
Voc has been observed in other organic BHJ structures.[45]
3. Conclusion
In summary, based on a DPD approach, we present the
previously unrecognized feature that branched alkyl side
chains on a rodlike polymer tend to form nanowire bundles
that are more thermodynamically favorable than those with
linear alkyl side chains. The theoretical prediction is exempli-
fied by P3MBT, which successfully demonstrates highly dense
nanowire formation spontaneously at room temperature,
free from any stringent requirements. Subsequently, we fabri-
cated BHJ solar cells using as-prepared, highly dense P3MBT
nanowires blended with PC71BM to achieve short and contin-
uous pathways in the photoactive layer, a key factor contrib-
uting to the high performance of PSCs by facilitating efficient
hole transfer. The as-prepared BHJ solar cells offer several
advantages such as relatively high Voc (∼0.72 V), high IPCE
(∼68%) and low P3MBT molecular weight (Mn = ∼16 800 with
a polydispersity index (PDI) of 1.12). The optimum conver-
sion efficiency of 4.15% for the P3MBT:PC71BM device was
mainly limited by absorption spectral coverage of P3MBT.
Interestingly, it is recently shown that a PSC using diketo-
pyrrolopyrrole with highly branched side chains has a PCE
of 5.5% and also involves spontaneous polymer fiber forma-
tion,[46] further reaffirming our theoretical prediction in this
study. We hope this seminal concept and its demonstration
will stimulate future work on the development of novel
nanowire-based polymers, such as branched or dendritic side
chains for photovoltaic applications.
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1106 wileyonlinelibrary.com
small 2011, 7, No. 8, 1098–1107

Superiority of Branched Side Chains in Spontaneous Nanowire Formation
[18] A. Haugeneder, M. Neges, C. Kallinger, W. Spirkl, U. Lemmer,
J. Feldmann, U. Scherf, E. Harth, A. Gügel, K. Müllen, Phys. Rev. B
1999, 59, 15346–15351.
[19] J. Nelson, J. J. Kwiatkowski, J. Kirkpatrick, J. M. Frost, Acc. Chem.
Res. 2009, 42, 1768–1778.
[20] T. Dufaux, J. Boettcher, M. Burghard, K. Kern, Small 2010, 6,
1868–1872.
[21] H. Xin, F. S. Kim, S. A. Jenekhe, J. Am. Chem. Soc. 2008, 130,
5424–5425.
for verification. The light intensity was determined by a monosilicon
detector (with KG-5 visible color filter) calibrated by the National
Renewable Energy Laboratory (NREL) to minimize spectral mismatch.
The IPCE spectra were measured using a lock-in amplifier with a cur-
rent preamplifier under short-circuit conditions. The devices were
illuminated by monochromatic light from a xenon lamp passing
through a monochromator with a typical intensity of 30 μW. A cali-
brated monosilicon diode with known spectral response was used
as a reference.
[22] S. Berson, R. de Bettignies, S. Bailly, S. Guillerez, Adv. Funct.
Mater. 2007, 17, 1377–1384.
[23] P. J. Hoogerbrugge, J. M. V. A. Koelman, Europhys. Lett. 1992, 19,
155–160.
[24] R. D. Groot, P. B. Warren, J. Chem. Phys. 1997, 107, 4423–4435.
[25] R. Groot, T. Madden, J. Chem. Phys. 1998, 108, 8713–8724.
[26] Y. J. Sheng, T. Y. Wang, W. M. Chen, H. K. Tsao, J. Phys. Chem. B
2007, 111, 10938–10945.
Supporting Information
[27] Y. K. Levine, A. E. Gomes, A. F. Martins, A. Polimeno, J. Chem.
Phys. 2005, 122, 144902.
[28] S. Salaniwal, S. K. Kumar, A. Z. Panagiotopoulos, Langmuir 2003,
19, 5164–5168.
Supporting Information is available from the Wiley Online Library
or from the author.
[29] D. W. Cheong, A. Z. Panagiotopoulos, Langmuir 2006, 22,
4076–4083.
Acknowledgements
[30] J. A. Merlo, C. D. Frisbie, J. Phys. Chem.
B 2004, 108,
19169–19179.
This work was supported by the National Science Council and
Ministry of Economic Affairs of Taiwan. We thank Yueh-Lien Lee,
National Taiwan University, for the grazing-incidence X-ray diffrac-
tion measurements.
[31] W. D. Oosterbaan, V. Vrindts, S. Berson, S. Guillerez, O. Douhéret,
B. Ruttens, J. D’Haen, P. Adriaensens, J. Manca, L. Lutsen,
D. Vanderzande, J. Mater. Chem. 2009, 19, 5424–5435.
[32] J. Liu, M. Arif, J. Zou, S. I. Khondaker, L. Zhai, Macromolecules
2009, 42, 9390–9393.
[33] M. T. Rispens, A. Meetsma, R. Rittberger, C. J. Brabec,
N. S. Sariciftci, J. C. Hummelen, Chem. Commun. 2003,
2116–2118.
[34] H. Sirringhaus, P. J. Brown, R. H. Friend, M. M. Nielsen,
K. Bechgaard, B. M. W. Langeveld-Voss, A. J. H. Spiering,
R. A. J. Janssen, E. W. Meijer, P. Herwig, D. M. de Leeuw, Nature
1999, 401, 685–688.
[35] Y. Kim, S. Cook, S. M. Tuladhar, S. A. Choulis, J. Nelson,
J. R. Durrant, D. D. C. Bradley, M. Giles, I. McCulloch, C. S. Ha,
M. Ree, Nat. Mater. 2006, 5, 197–203.
[36] B. D. Cullity, S. R. Stock, Elements of X-ray Diffraction, 3rd ed.,
Prentice Hall, Englewood Cliffs, NJ 2001.
[37] K. Sakurai, H. Tachibana, N. Shiga, C. Terakura, M. Matsumoto,
Y. T o k u r a , Phys. Rev. B 1997, 56, 9552–9556.
[38] N. Hosaka, H. Tachibaba, N. Shiga, M. Matsumoto, Y. Tokura,
Phys. Rev. Lett. 1999, 82, 1672–1675.
[39] M. Campoy-Quiles, T. Ferenczi, T. Agostinelli, P. G. Etchegoin,
Y. Kim, T. D. Anthopoulos, P. N. Stavrinou, D. D. C. Bradley,
J. Nelson, Nat. Mater. 2008, 7, 158–164.
[40] V. D. Mihailetchi, H. Xie, B. de Boer, L. J. A. Koster, P. W. M. Blom,
Adv. Funct. Mater. 2006, 16, 699–708.
[41] K. Maturová, S. S. van Bavel, M. M. Wienk, R. A. J. Janssen,
M. Kemerink, Adv. Funct. Mater. 2011, 21, 261–269.
[42] C. Goh, R. J. Kline, M. D. McGehee, E. N. Kadnikova, J. M. J. Fréchet,
Appl. Phys. Lett. 2005, 86, 122110.
[43] J. Huang, G. Li, Y. Yang, Appl. Phys. Lett. 2005, 87, 112105.
[44] V. D. Mihailetchi, H. Xie, B. de Boer, L. M. Popescu, J. C. Hummelen,
P. W. M. Blom, Appl. Phys. Lett. 2006, 89, 012107.
[45] S. Heutz, P. Sullivan, B. M. Sanderson, S. M. Schultes, T. S. Jones,
Sol. Energy Mater. Sol. Cells 2004, 83, 229–245.
[46] J. C. Bijleveld, V. S. Gevaerts, D. D. Nuzzo, M. Turbiez,
S. G. J. Mathijssen, D. M. de Leeuw, M. M. Wienk, R. A. J. Janssen,
Adv. Mater. 2010, 22, E242–E246.
[1] S. R. Forrest, Nature 2004, 428, 911–918.
[2] J. Peet, A. J. Heeger, G. C. Bazan, Acc. Chem. Res. 2009, 42,
1700–1708.
[3] F. Huang, K. S. Chen, H. L. Yip, S. K. Hau, O. Acton, Y. Zhang,
J. Luo, A. K. Y. Jen, J. Am. Chem. Soc. 2009,131, 13886–13887.
[4] B. C. Thompson, J. M. J. Fréchet, Angew. Chem., Int. Ed. 2008, 47,
58–77.
[5] I. C. Wu, C. H. Lai, D. Y. Chen, C. W. Shih, C. Y. Wei, B. T. Ko, C. Ting,
P. T. Chou, J. Mater. Chem. 2008,18, 4297–4303.
[6] L. M. Chen, Z. Hong, G. Li, Y. Yang, Adv. Mater. 2009, 21,
1434–1449.
[7] B. Friedel, C. R. McNeill, N. C. Greenham, Chem. Mater. 2010, 22,
3389–3398.
[8] C. V. Hoven, X. D. Dang, R. C. Coffin, J. Peet, T. Q. Nguyen,
G. C. Bazan, Adv. Mater. 2010, 22, E63–E66.
[9] J. Mei, K. Ogawa, Y. G. Kim, N. C. Heston, D. J. Arenas,
Z. Nasrollahi, T. D. McCarley, D. B. Tanner, J. R. Reynolds,
K. S. Schanze, ACS Appl. Mater. Interfaces 2009, 1, 150–161.
[10] H. Y. Chen, J. Hou, S. Zhang, Y. Liang, G. Yang, Y. Yang, L. Yu, Y. Wu,
G. Li, Nat. Photon. 2009, 3, 649–653.
[11] J. H. Lee, D. W. Kim, H. Jang, J. K. Choi, J. Geng, J. W. Jung,
S. C. Yoon, H. T. Jung, Small 2009, 5, 2139–2143.
[12] Y. J. Cheng, S. H. Yang, C. S. Hsu, Chem. Rev. 2009, 109,
5868–5923.
[13] G. Lu, L. Li, X. Yang, Small 2008, 4, 601–606.
[14] G. Li, V. Shrotriya, J. Huang, Y. Yao, T. Moriarty, K. Emery, Y. Yang,
Nat. Mater. 2005, 4, 864–868.
[15] A. Gadisa, W. D. Oosterbaan, K. Vandewal, J. C. Bolsée, S. Bertho,
J. D’Haen, L. Lutsen, D. Vanderzande, J. V. Manca, Adv. Funct.
Mater. 2009, 19, 3300–3306.
[16] Y. Liang, D. Feng, Y. Wu, S. T. Tsai, G. Li, C. Ray, L. Yu, J. Am. Chem.
Soc. 2009, 131, 7792–7799.
[17] J. J. M. Halls, K. Pichler, R. H. Friend, S. C. Moratti, A. B. Holmes,
Appl. Phys. Lett. 1996, 68, 3120–3122.
Received: December 6, 2010
Revised: January 18, 2011
Published online: March 21, 2011
small 2011, 7, No. 8, 1098–1107
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
wileyonlinelibrary.com 1107