Effect of molar mass and regioregularity on the photovoltaic properties of a reduced bandgap phenyl-substituted polythiophene

Article abstract of DOI:10.1002/pola.25970

Among the numerous reduced bandgap polymers currently being developed, poly[3-(4-octylphenyl)thiophene)]s (POPT) may present attractive properties for organic solar cells due to its facile preparation and improved absorption with respect to poly(3-hexylthiophene). This article appraises methods of preparation, including the use of diphenyl ether as a reaction medium, and discusses the effects of variations in molar masses, from about 3200 to 65,000 g mol^-1 and regioregularity on its optoelectronic properties. The photovoltaic properties of POPT with [6,6]-phenyl C₆₁ butyric acid methyl ester (PCBM) in bulk heterojunction devices are also discussed in the light of morphological variations, as indicated by atomic force microscopy characterizations. With an initial screening of conditions, namely POPT:PCBM ratios and deposition solvent, a power conversion efficiency of 1.58% was obtained using a relatively high molar mass POPT sample.

Full text of DOI:10.1002/pola.25970

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
Effect of Molar Mass and Regioregularity on the Photovoltaic Properties
of a Reduced Bandgap Phenyl-Substituted Polythiophene
1
1
2
2
3
ꢀ
ꢀ
Farid Ouhib, Guillaume Dupuis, Remi de Bettignies, Severine Bailly, Abdel Khoukh,
4
1
3
1
ꢀ
`
Herve Martinez, Jacques Desbrieres, Roger C. Hiorns, Christine Dagron-Lartigau
1
ꢀ
IPREM (EPCP) UMR 5254, Universite de Pau et des Pays de l’Adour, 2 avenue President Angot, 64053 Pau, France
2
ꢀ
INES-RDI DRT/LITEN/DTS/LCS Savoie Technolac BP 332, 50 avenue du lac Leman, 73377 Le Bourget du Lac, France
3CNRS, IPREM (EPCP) UMR 5254, 2 avenue President Angot, 64053 Pau, France
4
ꢀ
IPREM (ECP) UMR 5254, Universite de Pau et des Pays de l’Adour, 2 avenue President Angot, 64053 Pau, France
Correspondence to: C. Dagron-Lartigau (E-mail: christine.dagron-lartigau@univ-pau.fr)
Received 20 September 2011; accepted 24 January 2012; published online 23 February 2012
DOI: 10.1002/pola.25970
ABSTRACT: Among the numerous reduced bandgap polymers
currently being developed, poly[3-(4-octylphenyl)thiophene)]s
(POPT) may present attractive properties for organic solar cells
due to its facile preparation and improved absorption with
respect to poly(3-hexylthiophene). This article appraises meth-
ods of preparation, including the use of diphenyl ether as a
reaction medium, and discusses the effects of variations in
molar masses, from about 3200 to 65,000 g molꢀ1 and regio-
regularity on its optoelectronic properties. The photovoltaic
properties of POPT with [6,6]-phenyl C61 butyric acid methyl
ester (PCBM) in bulk heterojunction devices are also discussed
in the light of morphological variations, as indicated by atomic
force microscopy characterizations. With an initial screening of
conditions, namely POPT:PCBM ratios and deposition solvent,
a power conversion efficiency of 1.58% was obtained using a
C
V
relatively high molar mass POPT sample. 2012 Wiley Periodi-
cals, Inc. J Polym Sci Part A: Polym Chem 50: 1953–1966, 2012
KEYWORDS: chain growth polymerization polythiophenes; con-
jugated polymers; Grignard metathesis chain growth polymer-
izations; [6,6]-phenyl C61 butyric acid methyl ester (PCBM);
photovoltaic devices; poly[3-(4-octylphenyl)thiophene]
INTRODUCTION Their low cost and ease of processing have
pushed polymer-based solar cells organic photovoltaics
(OPVs) to the forefront of the search for energy sources that
do not pollute the environment with CO2.1–3 It is expected
that as soon as 10 year, 10% efficiency OPVs can be mass-
produced, there will be a paradigm shift in the market
place.4,5
of the so-called low bandgap polymers which display
increased absorptions in the about 600–900 nm range,12–15
such that OPV efficiencies of around 8% are now feasi-
ble.16,17 However, these materials remain relatively complex
to prepare, and an objective remains to find new materials
that can deliver the requisite properties at low costs.
Polythiophene-based materials are of interest due to their
ease of preparation and the high degree of control over
polymer molar masses afforded by the Grignard metathesis
(GRIM) polymerization route,18–23 which can be facilely
performed at room temperature.24
In OPVs, such as that in Figure 1, the dominant process is
of light absorption by the polymer to form excitons that
can move around 10 nm, and on meeting an interface with
an acceptor molecule, fast electron transfer results in a pos-
itively charged polymer and a negatively charged acceptor.6–8
The conduction of these charges to the respective electro-
des gives electricity. The absorption qualities of the materi-
als, along with the positions of their electronic bands, and
the nanomorphology of the materials in the active layer,
impinge strongly on the OPV qualities.9,10 A considerable
wealth of knowledge has been gained through the use of
an active-layer based on poly(3-hexylthiophene) (P3HT)
and [6,6]-phenyl C61 butyric acid methyl ester (PCBM),
which have shown efficiencies of around 5%.11 More
recently, great gains have been made through the use
Modification of the position-3 in the thiophene group is rela-
tively facile and leads to considerable changes in optoelec-
tronic properties. For example, Oshimizu et al. introduced
phenyl or pyridinyl groups into the side-chains and found
that the former gave rise to more crystalline polymers.
Indeed, both showed a decrease in HOMO and bandgaps (Eg)
with respect to P3HT, although the phenyl-derivative showed
a greater decrease due to the increased p-conjugation.25
Reduced bandgaps and improved hole-mobilities have also
been obtained through introducing, for example, conjugated
thiophene-2,5-diyl pendent moieties or crosslinking
Additional Supporting Information may be found in the online version of this article.
C
V
2012 Wiley Periodicals, Inc.
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 1953–1966
1953

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
sities. And although the properties of POPT, such as its rela-
tively high value of kmax and good thermal stability, make it
of interest for use in photovoltaic cells. Previous results,
showed 1.9% (simulated solar spectrum at 77 mW cmꢀ2
)
was obtained on a bilayer device with a cyano poly(p-phen-
ylene vinylene) derivative acceptor.36 The same acceptor
with POPT in a bulk heterojunction resulted in average
power conversion efficiencies (PCEs) of about 2%.37 Devi-
ces based on blends of POPT with a vinazene derivative
yielded to an average PCE of 1.4%, whereas 1.1% was
obtained with blends of P3HT with this acceptor.38 More
recently, the effect of side-chain densities on packing struc-
ture and photovoltaic performances with PCBM as an
acceptor was studied on POPT and POPT derivatives with
two or three thiophene spacer units. Best results were
achieved (2.77%) with the latter.39
FIGURE 1 ‘‘Normal’’ OPV device structure.
It is evident that the poor control over POPT molar masses
has hampered its use in devices. More recent, albeit initial,
work showed that it was possible to prepare relatively
high molar masses of POPT using the GRIM route,40 and
furthermore, that POPT can be combined with P3HT as
block copolymers using this method.41 We therefore
decided to explore the control over the GRIM-based poly-
merization leading to POPT and furthermore study the
effect of its molar mass and regioregularity on its photo-
voltaic properties. A series of comparable reactions were
thus performed using the oxidative polymerization of 3-(4-
octylphenyl)thiophene (3) with Fe3þ, and the GRIM poly-
merization of 2,5-dibromo-3-(4-octylphenyl)thiophene (4)
in tetrahydrofuran (THF) or in diphenyl ether (DPE), as
described in Figure 2. The various POPTs were then com-
pared with respect to their molar mass distributions and
their thermal, optical, and photovoltaic properties. To bet-
ter understand the effects of thermal annealing, deposition
solvent, and POPT:PCBM ratios on photovoltaic behaviors,
a study of the morphology of the active layer was per-
formed using atomic force microscopy (AFM).
agents.26–29 Inter-chain branches have similarly been
explored.30 These works have demonstrated that the optoe-
lectronic properties and the way in which the materials self-
organize is exceptionally sensitive to the nature and degree
of side-chain modification. And while the work on the so-
called ‘‘double-cable’’31 polythiophenes was also susceptible
to disorganizing the polythiophene backbone, it did demon-
strate the exceptional degree of modifications that can be
made at the 3-thiophene position, notably through the intro-
duction of fullerene32 and PCBM-like moieties.33
Poly[3-(4-octylphenyl)thiophene] (POPT) was first prepared
in 1992 as a variation to the already well-known poly(3-
alkylthiophene)s (P3AT). It was thought that the introduc-
tion of a phenyl ring into the side chains would enhance
conjugation and thermal stability.34 The initial reactions
leading to POPT were based on the facile oxidation poly-
merization of 3-(4-octylphenyl)thiophene using Fe3þ (Fig.
2).35 Samples suffered though from relatively low-molecular
weights (up to around 23,000 g molꢀ1) and high disper-
FIGURE 2 Synthetic routes to POPTs.
1954
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 1953–1966

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
EXPERIMENTAL
was for 5 min. Following drying under reduced pressure at
2 ꢂ 10ꢀ7 mbar for 1 h, devices were completed by deposi-
tion of LiF/Al (0.8 and 100 nm, respectively) cathode
through a shadow mask with 6 mm diameter openings at
about 2 ꢂ 10ꢀ7 mbar. Cell illumination was through the ITO
side. g of the devices was calculated using eqs 1 and 2:
Materials
Chemicals were obtained from Aldrich (France) and used as
received. Solvents (Baker, France) were distilled from over
standard drying agents under dry nitrogen. DPE was dried
over CaH2 under nitrogen and filtered via a syringe equipped
R
with a WhatmanV type ‘‘B’’ glass filter before use. All reac-
VocJsc
g ¼ FF
(1)
(2)
tions were performed under dry nitrogen.
Pin
Vmax max
J
Instrumentation
FF ¼
;
13C (100 MHz) and 1H (400 MHz) NMR spectra were
VocJsc
R
recorded on a BrukerV Avance 400 spectrometer. Polymer
where Pin is the power of incident light and Vmax and Jmax
represent the voltage and current densities at the maximum
power point, respectively.
molecular weights were estimated against polystyrene (PS)
standards using gel permeation chromatography (GPC) as
detailed in ref. 41. Ultraviolet (UV)-visible absorption spectra
were recorded on a Shimadzu UV-21021PC spectrometer
using solutions or spin-coated films. Differential Scanning
Calorimetry (DSC) was performed using a TA Instruments
DSC Q100 with samples on their second thermal run and at
Syntheses
Poly[3-(4-octylphenyl)thiophene] via FeCl3 (POPT1)
To 3 (5.51 mmol) dissolved in chloroform (30 mL) was added
a suspension of FeCl3 (27.57 mmol) in chloroform (70 mL)
over a period of 4 h. The final concentration of monomers
and FeCl3 was 0.05 and 0.27 M, respectively. The reaction
mixture was stirred at room temperature for 20 h and poured
into methanol (700 mL). The precipitate was filtered off,
washed with methanol, dissolved in chloroform, and dedoped
with ammonia solution (1 M, aq.). The mixture was stirred for
24 h, and the water phase replaced with fresh ammonia solu-
tion. This procedure was repeated another four times, and
then the chloroform solution was washed twice with ethylene-
diamine tetraacetic acid (0.05 M, aq.). The chloroform solution
was washed with water and poured into methanol (400 mL).
The precipitated polymer was Soxhlet washed with diethyl
ether and recovered with chloroform. After evaporation of the
solvent, the polymer labeled POPT-1 was dried under reduced
10 C minꢀ1. Thermal Gravimetric Analyses (TGA) were per-
ꢁ
ꢀ1
ꢁ
formed using a TA Instruments TGA Q50 at 10 C min and
R
under air. AFM imaging was performed with a VeecoV (now
R
BrukerV) MultiMode Scanning Probe Microscope with a
Nanoscope IIIA and Quadrex controller. A preliminary study
optimizing the working conditions found that ‘‘contact’’ mode
resulted in poorer topographic definition than ‘‘tapping’’
mode. Moreover, various tips were tested to find a good
compromise between a low radius of curvature, permitting a
high resolution and sufficient inflexibility for the weakest
possible abrasion during scanning. Taking into account these
preliminary results, all AFM images were acquired in inter-
mittent contact mode (tapping) to obtain topographic (with
a scanning frequency from 0.4 to 0.8 Hz), phase and deflec-
tion (error signal) images using a MPP-11100 Veeco probe
(quoted probe radius of curvature less than 10 nm). Phos-
phorus (n) doped Si with a force of 20–80 N mꢀ1 at a reso-
nance frequency of about 290 kHz was used. This instru-
ꢁ
ꢁ
pressure at 70 C for 24 h and 1 h at 130 C (yield 46.6%).
1
ꢁ
H NMR (C2D2Cl4, 100 C): 7.33 (2 H, dd, J3 ¼ 8 Hz), 7.25 (2
H, dd, J3 ¼ 8 Hz), 6.9 (1 H, s), 2.73 (2 H, t, J3 ¼ 7.29 Hz),
1.75 (2H, m), 1.40 (10H, m), and 0.99 (3 H, t, J3 ¼ 7.2 Hz).
13
ꢁ
C NMR (C2D2Cl4, 100 C): 143.1, 140.1, 134.8, 133.6, 131.4,
130.3, 129.5, 128.9, 36.1, 32.3, 31.5, 29.8, 29.7, 29.6, 23, and
14.4 ppm. Molecular weights for this polymer and other
details are given in Table 1.
ment was used in
a glove box under dried argon
atmosphere. To complete the quantitative analysis, the size
of plate sides was evaluated by taking into account 30 meas-
urements with a typical deviation of 1 nm.
Preparation and characterization of the solar cell devices
were performed at the French National Institute for Solar
Poly[3-(4-octylphenyl)thiophene] using Ni(dppp)Cl2 in THF
(POPT2)
ꢀ
Energy (INES, Chambery). The current–voltage characteris-
In a 100 mL three-necked round bottomed flask, 4 (2.32
mmol) was dissolved in THF (12 mL). Isopropylmagnesium
chloride (iPMgCl) in THF (2 M, 2.32 mmol) was added, and
tics and PCE values of the solar cells were measured in an
inert gas atmosphere via a computer-controlled Keithley
SMU 2400 unit using 100 mW cmꢀ2 AM1.5 simulated white
light from a Steuernagel Solar Constant 575 simulator. A
mono-crystalline silicon solar cell, calibrated at the
Fraunhofer-Institut fu¨r Solare Energiesysteme (Freiburg, Ger-
many) was used as the reference cell to confirm stabilization
of the 100 mW cmꢀ2 illumination. Devices (28 mm2 active
surface) were prepared as detailed in ref. 41 with the excep-
tion that the active POPT-blend-PCBM (1:1 by weight ratio)
layer was deposited by spin-coating from an anhydrous
chlorobenzene solution (10 mg mLꢀ1) under dry nitrogen to
obtain films about 40 nm thick. Annealing, where performed,
ꢁ
the mixture was stirred at 40 C for 2 h. Ni(dppp)Cl2 (0.046
mmol; ratio Ni:monomer was 1:50, respectively) was added
ꢁ
in one portion, and the reaction mixture was stirred at 40 C
for an additional 60 h. The polymerization was terminated
by the addition of methylmagnesium chloride in THF (3 M,
0.01 mmol) and stirring for 1 h. The polymer was precipi-
tated in methanol, then Soxhlet washed with diethyl ether
(to remove unreacted monomer, small oligomers, and Ni
salts), and recovered with chloroform. Removal of the sol-
vent and drying under reduced pressure at 70 ꢁC for 24 h
gave a near negligible yield of POPT2. 1H NMR (CDCl3,
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 1953–1966
1955

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
TABLE 1 POPT Samples and their Primary Characteristics
kmax (nm)
Polymer
MW
[Ni]/[M] (g molꢀ1 a
Regioregularity kmax (nm)
kmax (nm)
POPT-blend- Optical
PCBM Filmc Eg (eV)d
(%)b in Solution POPT Filmc
-
D
(synthesis)
)
POPT1 (FeCl3)
–
53,000
3,200
2.4 94
1.2 Low
1.4 90
1.5 95
1.5 91
1.6 ꢃ100
1.7 95
477
–
560, 598, 652 568, 614, 670 1.76
POPT2 (GRIM)
50
50
50
–
–
–
POPT3 (GRIM, fraction THF)
POPT4 (GRIM, fraction CHCl3)
11,900
49,000
65,000
50,000
39,000
472
470
473
470
478
547, 580, 640 486
1.80
560, 600, 656 568, 614, 675 1.74
557, 598, 656 570, 618, 678 1.74
563, 606, 664 569, 614, 672 1.76
550, 588, 642 569, 614, 672 1.76
POPT5 (GRIM, fraction C6H5Cl) 50
POPT6 (not fractionated)
500
500
POPT7 (GRIM, fraction CHCl3)
a
d
Determined by GPC against polystyrene standards.
Calculated from 1H NMR.
The optical bandgap was obtained from the formula Eg ¼ 1240/ konset
,
b
c
where konset is the onset value of absorption spectrum of POPT films at
high wavelengths.
UV-visible absorption maximum wavelength obtained on annealed
(120 ꢁC for 5 min) spin-coated films.
RESULTS AND DISCUSSION
ambient): 7.03 (5 H, m), 2.60 (2 H, m), 1.65 (2 H, m), 1.31
(10 H, m), and 0.91 (3 H, t) ppm.
Monomer and Polymer Syntheses
Figure 2 presents the routes to monomers and polymers.
The bromination of octylbenzene to yield 1-bromo-4-octyl-
benzene (2) was performed in the presence of montmoril-
lonite clay42 rather than the more habitually used Fe.43 This
resulted in a slight improvement in both yield and selectivity,
even if 2 was contaminated with about 14% of the isomer 1-
bromo-2-octylbenzene. However, on addition of 2 to 3-bro-
mothiophene, using a well-established route in the presence
of Ni(dppp)Cl,44,45 the resulting 3-(4-octylphenyl)thiophene
(3) was found to be quasi-pure. The isomer was in effect
excluded from the reaction with 3-bromothiophene, perhaps
by steric effects. This is important to note as variations in
the position of the group adjunct to the phenyl ring strongly
impact on the polymer properties.43 The preparation of the
brominated monomer, 2,5-dibromo-3-(4-octylphenyl)thio-
phene (4) was found to be facile using N-bromosuccinimide,
and the reaction gave yields comparable with those found
for alkylated thiophenes.22
Poly[3-(4-octylphenyl)thiophene] using Ni(dppp)Cl2 in DPE
(POPT3 to POPT5)
The following is a representative method. In a 250 mL three-
necked round bottomed flask, 4 (5.02 mmol) was dissolved
in DPE (50 mL). iPrMgCl in THF (2 M, 5.02 mmol) was
added, and the mixture was stirred at 40 ꢁC for 2 h.
Ni(dppp)Cl2 (0.1 mmol, ratio Ni:monomer was 1:50, respec-
tively) was addedꢁin one portion, and the reaction mixture
was stirred at 60 C for an additional 60 h. The polymeriza-
tion was terminated by the addition of CH3MgBr in THF (3
M, 0.03 mmol) and stirring for 1 h. Precipitation in methanol
yielded the crude polymer. This was washed in a Soxhlet
with diethyl ether and then sequentially with THF, CHCl3,
and chlorobenzene, to give POPT3 (220 mg, 16%), POPT4
(295 mg, 21%), and POPT5 (275 mg, 20%), respectively.
Yields given are ꢁfollowing sample drying at 70 ꢁC for 24 h
and then at 130 C for 1 h.
Poly[3-(4-octylphenyl)thiophene] using Ni(dppp)Cl2 in DPE
(POPT6)
Molar Masses and Regioregularities
Polymer molar mass affects OPV properties at several levels:
first, the length of the polymer must be great enough to
avoid electron confinement effects that limit absorption;46
second, high molar mass polymers tend to form fibrillar
structures across the composite that increase p-p interac-
tions between adjacent chains and enhance charge conduc-
tion through the device, while medium and lower molar
mass polymers can enhance interactions with the acceptor
molecules; and third, higher molar mass materials generally
require greater annealing temperatures to attain ordered
structures.47
An additional reaction using a ratio 4 (2.155 mmol) to
Ni(dppp)Cl2 (4.31 ꢂ 10ꢀ3 mmol) of 500 in DPE (40 mL)
was performed. All other reagents were used at the same
ratios with respect to the monomer as in the representative
method, but the temperature for polymerization was kept at
ꢁ
40 C. The obtained polymer was washed with diethyl ether
in a Soxhlet and recovered with dichlorobenzene yielded
POPT6 (150 mg, 25%). POPT3, POPT4, POPT5, and POPT6
gave NMR 1H and 13C spectra as that of POPT1 with varia-
tions in sizes of minor signals.
Poly[3-(4-octylphenyl)thiophene] using Ni(dppp)Cl2 in DPE
(POPT7)
The same procedure as for POPT3 to POPT5, in particular
for reaction temperatures was used. The only difference is
the monomer:Ni ratio of 500. The sample of POPT7 was
recovered after fractionation from chloroform.
POPTs prepared using oxidative polymerizations (POPT1)
and GRIM routes (POPT2–7) have molar masses detailed in
Table 1. The values were obtained against PS standards
using a GPC with THF as eluant and therefore should be
treated with caution; a closely related polymer, P3HT, is
known to have lower molar masses than those indicated by
1956
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 1953–1966

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
were prepared. As POPT4 and POPT5 were only slightly solu-
ble in THF, they were slowly transferred from chlorobenzene
into warm THF (GPC eluent) just before GPC analysis. Another
GRIM polymerization was realized in DPE at 40 ꢁC with a
monomer:Ni ratio ¼ 500:1 to yield a polymer denoted
POPT6. This polymer was expected to have a molecular
weight around 135,000 g molꢀ1. However, GPC experiments
for the nonfractionated POPT6 indicated Mw ¼ 50,000 g
molꢀ1. From the observation of the peak asymmetry and the
presence of a shoulder toward low molar masses (between 5
ꢂ 103 and 104 g molꢀ1), it is reasonable to assume that high
molar mass material is retained during filtration before injec-
tion. Matric assisted-laser desorption/lonisation-time-of-flight
spectrometry (MALDI-TOF) experiments were undertaken on
POPTs, but trials remained unsuccessful even though a num-
ber of matrices were used (namely those based on sinapinic
acid, 3,5-dimethoxy-4-hydroxybenzoic acid, 1,8,9-anthracene
triol or 2,20,50,200-terthiophene with and without a silver ion
donor). We used both ‘‘sandwich’’ techniques and simple drop
cast techniques to prepare the surfaces for ablation, as previ-
ously indicated.47 This was probably due to the difficulty of
ionizing carbocyclic-based structꢁures. Finally, a GRIM based
polymerization performed at 60 C rather than at room tem-
perature, yielded POPT7 with a GPC-indicated Mw ¼ 39,000 g
FIGURE 3 GPC chromatograms of POPT samples 1 to 7 (THF,
UV-254 nm).
GPC under similar circumstances.48 Nevertheless, POPT1 was
found to have Mw ¼ 53,000 g molꢀ1, a bimodal distribution
-
(see Fig. 3), and a high dispersity (D ¼ Mw/Mn) of 2.4; val-
molꢀ1 and D ¼ 1.7. The similar dispersity made it possible to
-
ues that are in accordance with those described in the litera-
ture for the oxidative polymerization.35
compare it directly against samples POPT5 and POPT6.
Andersson et al.35 showed that regioregular (94%) POPT
could be obtained via the use of FeCl3. An estimation of the
degree of regioregularity of the POPT1 was made from the ra-
tio of 1H NMR integrals of signals at 2.73 and 2.62 ppm due to
a-CH2- groups on regular head-to-tail triads and on irregular
head-to-head triads, respectively, as identified in Figure 4.
POPT1 shows a similar degree of regioregularity (ca. 94%).
Given the known nature of the mechanism,18 the GRIM chain-
growth polymerization was expected to deliver POPTs higher
regioregularity and molar masses. POPT2 was prepared with
a ratio monomer to Ni 50:1 and was therefore expected to
have a degree of polymerization of around 50 units. POPT2
had a surprisingly low molar mass of around 3200 g molꢀ1
.
Given the probable difference in hydrodynamic volumes for
PS and POPT, the average number degree of polymerization of
POPT2 was at most around 10, quite below that expected.
POPT2 was most likely poorly solvated by THF and precipi-
tated out from the reaction mixture before the full chain
length could be reached. The use of DPE as a solvent in the
place of THF and for a same ratio monomer to Ni (50:1)
made it possible to obtain a crude polymer sample with a
much higher molar mass, indeed so high that it was only
partly soluble in the GPC solvent THF. To obtain a rough esti-
mate to the molar mass of this crude polymer, an attempt to
modify it using bromination (with the most likely effect of
causing steric-disruption of aggregates) with a small quantity
of Br2 in CCl4 was attempted. GPC analysis of this modified
In contrast, and from comparable P3HTs, one would expect a
much higher degree of regioregularity from POPT prepared
using GRIM. Surprisingly, the 1H NMR spectrum of POPT2 (Fig.
4) indicated a regioregularity (ca. 94%), when compared with
the other samples shown in Table 1. It is understood that the
GRIM polymerization ‘‘inserts’’ irregularities at one chain-end of
the polymer,18,19 and this material suffered from the presence
of a high number of low molar mass chains. This confirmed that
THF was not a good solvent in which to prepare POPT.
The remaining samples were prepared in DPE. Apart from
variations in peak sizes, the 1H and 13C spectra of POPT3,
POPT4, POPT5, and POPT7 were all very similar. Therefore,
only the 1H (Fig. 4) is shown. Again, one would expect rea-
sonably high regioregularities but that of POPT4 (95%) is
comparable with that of POPT1. Indeed, POPT3 and POPT5
were even lower at ꢃ90%. Although the value for POPT3
could be explained by a high content of short polymer chains
in this sample (as indicated in Fig. 3) and Table 1, this is an
unexpected value of regioregularity for the high-molecular
weight sample POPT5. We attributed this to the use of a
higher polymerization temperature (increased from 40 to 60
ꢁC), used to obtain higher molar masses. This caused an
polymer indicated Mw ¼ 21,000 g molꢀ1 and D ¼ 2.1. The
-
high dispersity was probably due to the uncontrolled nature
of the polymerization, probably either due to its rapidity with
respect to the period of initiation, or precipitation. Either way,
this result was a definite improvement on those obtained
using THF. To obtain a selection of resolvable samples, the
crude polymer was sequentially fractionated in a Soxhlet with
solvents of increasing solvating power. Thus, THF (POPT3,
27% w/w of the crude polymer), CHCl3 (POPT4, ca. 38% w/
w), and chlorobenzene (POPT5, 35% w/w) soluble fractions
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 1953–1966
1957

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
1
FIGURE 4 H NMRs of POPT1, POPT4, and POPT6 (C2D2Cl4, 100 ꢁC) and POPT2 (CDCl3, ambient temperature).
increased introduction of defects, probably due to secondary
trans-metallation reactions.
HSQC spectrum of POPT4 was also performed (Supporting
Information Fig. S2). In the region of the aliphatic protons,
the methylene protons in a-position of the phenyl ring (tri-
plet at 2.7 ppm) are correlated with the carbon peak at 36.1
Nevertheless, POPT6, prepared using a ratio monomer to Ni
ꢁ
500:1 and in DPE at 40 C was expected to display, due to a
1
ppm. The peak H at 1.74 ppm, assigned to -CH2-b, is corre-
low concentration of chain-ends, a higher degree of regiore-
gularity. This was confirmed by 1H NMR spectrum, where it
is shown (Fig. 4) to be quasi-100% regioregular. This con-
firmed the role of the temperature during polymerization.
1
lated with carbon at 31.5 ppm and the peak H at 1.45 ppm
of the protons -CH2-c is correlated with carbon with 32.3
ppm. The three signals 13C at 29.8, 29.7, and 29.6 ppm are
correlated, respectively, with the protons Hd, He, and Hn. The
Further Characterizations—13C NMR
signals H due to -CH2-y and the methyl terminal groups are
1
A
13C spectrum of a representative sample of (POPT4) is
correlated with carbons resonating at 23.0 and 14.4 ppm,
respectively. In the aromatic region, the 1H singlet at 6.89
ppm assigned to H4 is correlated with the peak 13C at 130.3
shown in the Supporting Information Figure 1, and the sig-
nals confirmed the expected structure. A 2D 1H–13C NMR
1958
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 1953–1966

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
ppm. The phenyl protons (H7 and H8), giving a double dou-
blet in the 1H spectrum, are correlated with the signals at
128.9 and 129.5 ppm in the 13C spectrum. In Figure S2(a),
the total decoupled 13C NMR spectrum, the undecoupled 13C
spectrum and the 13C NMR spectrum with selective decou-
pling of the a-CH2- protons of carbons C4, C7, and C8 are pre-
sented. These spectra allow to confirm the assignments of
the peak due to the carbons C7 and C8. Indeed, the single
signals at 128.9 ppm due to C8 and the peak at 129.5 ppm
due to C7 observed in the decoupled spectrum become,
respectively, a doublet of a quadruplet and a double-doublet
in the undecoupled 13C spectrum. The multiplicity of C8 can
be explained by two different couplings, with H8 (J1 ꢃ150
Hz), and with 2Ha and H8 (J3 ꢃ 6 Hz). In addition, the selec-
0
tive decoupling of the a-CH2- protons only simplify the dou-
blet-quadruplet in simple double-doublet, confirming that C8
was coupled with the a-CH2- protons. The double-doublet
found for C7 in the undecoupled spectrum is due to the cou-
plings with H7 (J1ꢃ 150 Hz) and with H7 (J3 ꢃ 8 Hz). It was
0
noted that in the case of phenyl derivatives, a J2(C-H) cou-
pling constant could be estimated at 1 Hz and a J3(C-H) at 8
Hz.49 Thus, in the case of POPT, the J2 (C-H) coupling con-
stants of the carbons of the phenyl ring with their adjacent
protons are too weak to be observed on the undecoupled
13C spectrum. The Supporting Information Figure S2(b,c)
present the decoupled and undecoupled 13C spectra of the
quaternary carbon region (C2, C3, C4, C6, and C9). The single
signals at 131.4 and 134.8 ppm observed on the total
decoupled spectrum, respectively, became a broad doublet
with a coupling constant around 10 Hz and a doublet with a
weak coupling constant (5 Hz) on the undecoupled spec-
trum. The first signal was allotted to C2, as this carbon is
coupled with H4 with a large coupling constant J3 (C2-H4)
and the second peak corresponded to C5 coupled with the
H4 with a weaker coupling constant J2 (C5-H4). The peak at
133.6 ppm obtained as a triplet with a coupling constant of
about 7 Hz, could be assigned to a carbon coupled with two
protons with a J3 coupling. This corresponds to C9 supposed
to be coupled with both protons H7 with a J3 C9-H3 coupling.
From Figure 5(c), the assignments of carbons C6 and C3 can
be made. The carbon C6 is coupled, respectively, with both
protons H8 and with the proton H4 with a similar J3 (8–10
Hz) coupling, leading to an expected triplet in the unde-
coupled spectrum. The carbon C3 is coupled with both pro-
tons H7 (J3 ꢃ 8–10 Hz) and proton H4 (J2 ꢃ 2-5 Hz), respec-
tively. Thus, C3 presents a narrower peak than that of C6 in
the undecoupled 13C spectrum. Thus, the signal at 140.1
ppm can be attributed to carbon C3 and the one at 143.1
ppm to C6.
FIGURE 5 (a) Absorption UV-visible spectra of POPTs in solu-
tion in chloroform; (b) Absorption UV-visible spectra of
annealed (120 ꢁC for 5 min) POPTs on spin-coated films; and
(c) Absorption UV-visible spectra of POPT:PCBM (1:1) blends
on spin-coated films annealed at 120 ꢁC for 5 min. Note the
low kmax for POPT3 due to its low molar mass.
Thermal Properties
initial step was due to the loss of octyl chains, given that it
yielded to a total loss of around 40% by weight. As to why
the polymers exhibited different rates of loss is not clear. It
is possible that the high concentration of high molar mass
chains in POPT1 and POPT5 (Mw > 50,000 g molꢀ1) and the
bimodal distribution of molar mass led to a faster first deg-
radation. In addition, the distribution of irregularities
throughout its structure, rather than essentially at chain
The Supporting Information Figure S3 shows the TGA plots
of the POPTs. Two clear steps can be seen in the thermal
decomposition of the POPTs. The initial step started at ꢁabout
ꢁ
280 C for POPT1, POPT5, and POPT7 and about 300 C for
POPT3, POPT4, and POPT6. This first degradation was more
rapid for POPT1, POPT5, and POPT7, closely followed by
POPT3 and then POPT4 and POPT6. It is probable that this
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 1953–1966
1959

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
TABLE 2 Electrochemical Properties of the POPTs
ence of the gap determined by CV (2.2 eV) and by optical
absorption (1.76 eV), but this discrepancy has been observed
elsewhere and is explained by considering that as the proc-
esses involved are not the same. In the case of optical transi-
tions, there is no formation of free charge carriers, only a
bound exciton. The formation of charge carriers will then
require a higher energy. This argues that the optical bandgap
is probably underestimated. In the case of CV, electrochemical
doping processes are occurring.54
p-Doping
HOMO
n-Doping
LUMO
Eg
Polymer
Eox (V)a
(eV)
Ered (V)
(eV)
(eV)b
POPT1
POPT4
a
0.52
0.66
ꢀ5.23
ꢀ5.22
ꢀ1.73
ꢀ1.65
ꢀ2.98
ꢀ2.91
2.25
2.31
Potential referred versus Fc/Fcþ (see text).
b
Eg ¼ (Eox ꢀ Ered).
The UV-visible spectra of POPTs in chloroform solution are
shown in Figure 5(a). All spectra present a maximum absorp-
tion in the visible region, around 470 nm, characteristic of the
p–p* transition of the dissolved polythiophene backbones.
However, it is interesting to note that there is a slight shift
due to the difference in molar masses. For photovoltaic appli-
cations, the absorbance of polymers in thin films, especially
the absorbance in the visible region, is an important parame-
ter. Solid-state absorption spectra of the POPT samples
obtained after optimal thermal annealing are shown in Figure
5(b), and maximum values are recorded in Table 1. For all the
samples, the absorption spectra present a red-shift compared
with those obtained for P3HTs (kmax around 550 nm).43 More-
over, compared with the absorption spectra obtained from
chloroform solutions, the maximum absorption wavelength of
the thin films (spin-coated) are red-shifted by 100 to 150 nm,
and an additional vibrational fine structure, tailing up to 800
nm, was observed. This indicates that the slightly higher
degree of main chain conjugation and the presence of consid-
erable inter-chain interactions between the polymer chains in
the solid film are due to the presence of the pendent phenyl
rings. It was shown that the phenyl group adds disorder to
the polymer chain when considered in an isolated state, but
on going to the solid state the chains become more planar.55
An exception is POPT3, which did not contain any high molar
mass material and exhibited a lower kmax (540 nm). The opti-
cal band gaps deduced from the absorption edges of thin film
spectra for all samples, around 1.7 eV, were lower than those
found for P3HT (1.9 eV).
ends as was the case for POPT3, may have enhanced insta-
bilities. The second degradation process, which took place in
a temperature range of 400–650 ꢁC for POPT1, and POPT5
and 470–650 ꢁC for POPT3, POPT4, POPT6, and POPT7 can
be attributed to the polymer backbone degradation. How-
ever, from these TGA results, it is important to note that the
thermal stability of the POPTs is adequate for the fabrication
processes of photovoltaic devices.
POPT thermal properties were also studied by DSC to deter-
mine the nature of possible phase transitions. Heating and
cooling cycles were performed using drop-cast thin films
with a preliminary cycle to remove thermal histories. As the
Supporting Information Figure S4 shows, no melting transi-
tions were observed, a property in contrast to that of
P3HT50,51 and indicating a negligible concentration of crys-
tallites. Nevertheless, the thermogram indicated that on cool-
ing, there was a wide disorder-order transition with a peak
at around 230 ꢁC. Although this was a relatively large exo-
thermic transition, it was difficult to assign it to a crystalliza-
tion as the heating cycle showed no clear melting peak nor
order–disorder transition for that matter. A weak Tg was
ꢁ
detected at 160 C, most likely due to in-chain rotations.
Optical and Electrochemical Properties
Once deposited on a platinum electrode, the electrochemical
behavior of POPT1 and POPT4 were here considered repre-
sentative of the two types of polymers in this series investi-
gated by cyclic voltammetry (CV) in propylene carbonate with
0.1 M Bu4NBF4. Both polymers gave a reversible response in
the positive potential range ascribed to a p-doping/dedoping
process and an irreversible response in the negative potential
range corresponding to irreversible n-doping process. The
energy values of the lowest unoccupied molecular orbital
(LUMO) and the highest occupied molecular orbital (HOMO)
could thus be estimated for each polymer according to the fol-
lowing equations52: HOMO ¼ ꢀ(Eox þ 4.71) (eV), LUMO ¼
ꢀ(Ered þ 4.71) (eV). The potential values referred to are ver-
sus the ferrocene/ferrocenium redox couple, Fc/Fcþ (E1/2 Fc/
Fcþ ¼ 0.15 mV vs. Ag/10 mM AgNO3). The redox data and
energy levels are listed in Table 2. The HOMO and LUMO lev-
els obtained for POPT1 and POPT4 are comparable. An aver-
age value of ꢀ5.2 eV for the HOMO level is obtained and is
quite similar to that of P3HT.53 However, the LUMO level is
lower with an average value of ꢀ3 eV compared with ꢀ3.3 eV
for P3HT. The values obtained with POPTs are in good agree-
ment with the ones already reported.38 We noticed a differ-
Figure 5(c) shows the absorption spectra of POPT-blend-
PCBM films (1:1 by weight) prepared by spin coating from
hot solutions inꢁ chlorobenzene (130 ꢁC). The films were
annealed at 120 C for 5 min. The absorption bands are due
to the polymer, except for the band at 340 nm which is char-
acteristic of PCBM. POPT1, 4, 5, and 6 showed a slight
increase in the wavelengths of the fine vibronic bands (from
10 to 20 nm) when going from films of the polymer alone to
those with PCBM (as detailed in Table 1). It can be assumed
that POPT, through the presence of pendent phenyl rings, is
considerably more rigid than P3HT, and therefore any move-
ment enhancing p-stacking is restricted in the solid state.
However, the results seem to indicate that PCBM allows
POPT a greater freedom of movement so that it can better
self-organize to increase conjugation lengths and inter-chain
interactions. Conversely, though in the case of POPT3 which
contains low molar mass samples, the kmax decreased in the
presence of PCBM, and no vibrational fine structure was
observed. This is an effect similar to that observed with
1960
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 1953–1966

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
FIGURE 6 Height AFM images of spin coated thin films from chlorobenzene (a) POPT as deposited; (b) POPT:PCBM 1:1 (wt ratio)
as deposited; (c) POPT:PCBM 1:1 5 min at 90 ꢁC; (d) POPT:PCBM 1:1 5 min at 120 ꢁC; and (e) POPT:PCBM 1:1 5 min at 150 ꢁC.
P3HT, where close mixing of the two components is pro-
moted by facile movement and the dislocation of the short
polymer chains.47
Morphological and Photovoltaic Properties
The surface morphology of thin films was studied by AFM.
First, thin films only based on POPT were studied to
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 1953–1966
1961

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
ꢁC, a slight phase segregation between polymer and PCBM
TABLE 3 Photovoltaic Characteristics of ITO/PEDOT:PSS/
POPT:PCBM (1:1)/LiF/Al Devices under AM 1.5 Illumination at
an Irradiation of 100 mW cm22 POPT:PCBM Blends Deposited
from Chlorobenzene Solutions
ꢁ
was observed. By heating at 120 C, the morphology remains
ꢁ
virtually unchanged, but with a heating at 150 C, the PCBM
seems to be redispersed again in the polymer matrix. How-
ever, given that these are surface characterizations only, it is
not possible to state anything about the underlying struc-
ture. There may have been a preference for the bulk-N2
interface (these characterizations were performed under dry
gas) by one component, or more likely, the lower molar
mass parts of the relatively disperse polymer (POPT6) mak-
ing their way to the surface at higher temperatures. Similar
effects have been observed elsewhere in different systems.56
POPT:
Temperature
(ꢁC)
Voc
Jsc
(mA cmꢀ2
FF
PCE
(%)
PCBM (1:1)
(V)
)
(%)
POPT1
POPT3
POPT4
POPT5
POPT6
POPT7
25
90
25
90
25
80
25
90
25
80
25
100
0.63
0.61
0.54
0.43
0.67
0.63
0.62
0.59
0.57
0.58
0.61
0.57
3.05
4.65
1.28
1.32
3.26
4.40
4.02
5.49
4.69
4.37
4.75
5.26
36
46
28
29
40
46
40
49
42
47
44
48
0.69
1.30
0.19
0.16
0.87
1.26
1.01
1.58
1.14
1.19
1.29
1.43
To study the photovoltaic performance of the POPTs, OPVs
were fabricated with
a
structure ITO/PEDOT:PSS/
POPT:PCBM (1:1)/LiF/Al. Preliminary results obtained on
some POPTs without optimizations were previously
reported.40 The open-circuit voltage (Voc), short circuit cur-
rent density (Jsc), fill factor (FF), and power conversion effi-
ciency (PCE) (PCE of the OPVs based on POPT1 and POPT3–
7 (deposited from chlorobenzene solutions) under AM1.5
conditions (100 mW cmꢀ2), before and after heat treatments,
are summarized in Table 3.
examine if this polymer could exhibit a fibrillar structure
such as P3HT. The height image of a thin film as deposited
from chlorobenzene is presented in Figure 6(a). It shows
that there is no specific structuration, the film seemed to be
amorphous.
The performance parameters of devices from POPT1 to
POPT7 increase with appropriate thermal treatments. A
slight improvement in performances is observed after
annealing POPT6 and POPT7. For the low molar mass sam-
ple (POPT3), no significant variation of the photovoltaic pa-
The surface of thin films obtained from blends of
POPT:PCBM (1:1 by weight) was followed with various ther-
mal annealing processes [Fig. 6(c–f)]. These images showed
that the roughness of the films is around 10 nm and that the
morphology is dependent on the annealing temperature. In
the case of P3HT, there is usually an increase in the fibrillar
structure with thermal annealing. In the case of POPT, it is
mainly the dispersion of PCBM in the polymer matrix that is
modified. Indeed, as deposited, the blend is homogenous and
seems amorphous. With a thermal annealing of 5 min at 90
rameters are observed with annealing (Voc ¼ 0.55 V, Jsc
¼
1.28 mA cmꢀ2, PCE ¼ 0.19%). But, the PCE and Jsc of POPT3
are reduced by approximately 80% compared POPTs with
greater molar masses. Therefore, it can be assumed that the
reduction of current density of POPT3 was probably due to
the blue-shift in absorption spectrum [Fig. 5(c)] and its poor
self assembly behavior with PCBM. The photovoltaic compo-
sites with low molar mass polymers have, in general, a
reduced value of Jsc. Before heat treatment, the performance
FIGURE 7 Height AFM images of POPT:PCBM cast from chlorobenzene: (a) POPT:PCBM 1:1.5 (wt ratio) and (b) POPT:PCBM 1:2
(wt ratio).
1962
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 1953–1966

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
TABLE 4 Photovoltaic Characteristics of ITO/PEDOT:PSS/
parameters of the devices increased on going from POPT5
(Voc ¼ 0.62 V, Jsc ¼ 4.02 mA cmꢀ2, FF ¼ 40%, PCE ¼
1.01%), to POPT6 (Voc ¼ 0.57 V, Jsc ¼ 4.69 mA cmꢀ2, FF ¼
42%, PCE ¼ 1.14%) and to POPT7 (Voc ¼ 0.61 V, Jsc ¼ 4.75
mA cmꢀ2, FF ¼ 44%, PCE ¼ 1.29%). The difference between
these three samples is mainly the molar mass that decreases
from POPT5 to POPT7. It seems that greater annealing is
necessary when using high molar mass samples, in agree-
ment for results found for the comparable P3HT.41 Although
POPT6 presents a high degree of regioregularity (ꢃ100%)
comparatively to POPT5 (91%). This indicates that the incor-
poration of a small amount of irregularities into the polymer
backbone does not greatly affect the performances of the
POPT solar cells.
POPT:PCB/LiF/Al Devices under AM 1.5 Illumination at an
Irradiation of 100 mW cm22 POPT:PCBM Blends Deposited
from p-xylene (Xyl) or Chlorobenzene (Clbz) Solutions, and for
Different Ratios
POPT:
T
Voc
Jsc
(mA cmꢀ2
FF
PCE
Solvent PCBM (ꢁC) (V)
)
(%) (%)
POPT5 Xyl
1:1
25 0.38 5.40
80 0.28 5.10
37
33
40
49
38
33
30
34
44
42
44
48
34
35
34
35
0.80
0.49
1.01
1.58
0.46
0.79
0.37
0.61
0.72
1.25
1.29
1.43
0.96
0.73
0.84
0.81
Xyl
Clbz
1:1
1:1
25 0.62 4.02
90 0.59 5.49
25 0.50 2.59
80 0.48 4.90
25 0.47 2.60
80 0.45 4.01
25 0.64 2.60
80 0.63 4.60
25 0.61 4.75
100 0.57 5.26
25 0.73 3.90
100 0.47 4.50
25 0.69 3.60
100 0.51 4.50
Clbz
1:1
Clbz
1:1.5
1:1.5
1:2
Clbz
After heat treatment, the devices based on samples that con-
tain high molar mass polymers and a small amount of irreg-
ularities (POPT1, 4, and 5) exhibited a more pronounced
improvement in the photovoltaic parameters (Jsc, FF, PCE),
while Voc remains nearly constant slightly above 0.60 V.
Clbz
Clbz
1:2
POPT7 Xyl
Xyl
1:1
1:1
Clbz
1:1
Generally, Voc is related to the difference between the LUMO
energy level of acceptor and the HOMO energy level of the
conjugated polymer. Therefore, in the case of POPTs, Voc attain-
able is in the same range of value as that of P3HT.57 In con-
trast, the PCE rꢁeached a maximum value at 80 ꢁꢁC for POPT4
(1.26%), at 90 C for POPT1 (1.3%), and at 90 C for POPT5
(1.58%). The value of Jsc after annealing followed the order
POPT5 (5.49 mA cmꢀ1), POPT1 (4.65 mA cmꢀ1), and POPT4
(4.40 mA cmꢀ1). The same extending annealing condition does
Clbz
1:1
Clbz
1:1.5
1:1.5
1:2
Clbz
Clbz
Clbz
1:2
the polymer matrix and the phase separation was more
clearly visible. This observation was confirmed in the case of
POPT:PCBM 1:2, where the size of both domains was
increased and estimated around 50 nm. This morphology
would seem to be more appropriate for photovoltaic applica-
tions as there are phases which can transfer changes away
from their point of formation.60
not strongly affect the device performance of POPT6 (Jsc
¼
4.37 mA cmꢀ2, PCE ¼ 1.19%). In the case of POPT7, the ther-
ꢁ
mal annealing at 100 C improved Jsc and FF the most, leading
to an increase of PCE from 1.29% to 1.43%. It can be seen
that while maximizing regioregularity in POPT may provide
the best photovoltaic properties in the pristine material (with-
out annealing), the best (annealed POPT5) photovoltaic per-
formance is obtained with a POPT with a reduced degree of
regioregularity. This is similar as a result to that demonstrated
elsewhere for P3HT, where a reduced tendency to crystallize
can result in a more stable interpenetration of polythiophene
and PCBM-based phases.58,59 As presented in Figure 3, POPT5
contained considerably higher molar mass polymer chains
than POPT1, POPT4, and POPT6. However, the sample that
contains high molar mass material (POPT5) exhibits a higher
value of Jsc after heat treatment (Jsc ¼ 5.5 mA cmꢀ2). This can
indicate that the presence of high molar mass macromolecules
ensures high charge mobilities and forms more thermally sta-
ble POPT:PCBM bulk-heterojunction devices.
Ratios of 1:1.5 and 1:2 were chosen to test in solar devices
for POPT5 and POPT7. The photovoltaic characteristics are
summarized in Table 4.
The behavior of both polymers appeared to be very different
with the increase of PCBM content. For POPT5, this had the
effect to decrease the Voc which is around 0.47 V whatever
the ratio and the annealing temperature. Furthermore, Jsc
increased with annealing leading to a higher PCE, as the FF
was not greatly modified. In this series, the highest PCE
(0.79%) was obtained with the ratio 1:1.5 after annealing at
ꢁ
80 C. In the case of POPT7, the PCEs were higher on nonan-
nealed devices with both tested ratios. And as for POPT5,
the best result was obtained with the 1:1.5 ratio (0.96%).
This PCE is higher than that obtained with POPT5 for the
same ratio, due to a high value of Voc (0.73 V). However,
whatever the case, the usual 1:1 ratio remained the most
suitable ratio for these two polymers as the FF obtained in
this case was above 47%, compared with values around
34% for higher contents of PCBM.
Effect of POPT:PCBM Ratio
As AFM images showed a very homogeneous dispersion of
PCBM in the polymer matrix for the blend 1:1 without
annealing, which would tend to enhance charge-recombina-
tion rather than charge-transfer to the electrodes, higher
PCBM contents were attempted. As shown in Figure 7, by
increasing the ratio of PCBM compared with polymer, the
phase separation was clearly modified. Indeed, by raising the
ratio to 1:1.5, the PCBM was no-longer perfectly mixed with
Effect of Solvent for Deposition of the Active Layer
It has been shown that the morphology of the active layer is
highly dependent on the solvent used for deposition.61 para-
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 1953–1966
1963

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
FIGURE 8 Height AFM images of: (a) POPT cast from p-xylene; (b) POPT:PCBM (1:1 wt) p-xylene; and (c) POPT/PCBM (1:2 wt) p-
xylene.
Xylene is a poorer solvent than chlorobenzene for polythio-
phenes and can favor the formation of aggregates or fibrils
in the case of P3HT.62,63 With this in mind, p-xylene was
used to prepare thin films of POPT and than POPT:PCBM
blends. The AFM height image obtained in this case [Fig.
8(a)] showed a structuration of ‘‘platelets,’’ probably induced
by a crystallization obtained from solution. Two different
POPT:PCBM ratios were tested. In the case of 1:1 ratio, a
strong difference of morphology was observed between thin
film prepared from chlorobenzene and the one prepared
from p-xylene. In the latter, large domains of POPT were
formed up to 100 nm, probably due to a phase separation
even in solution. When doubling the PCBM content (1:2 ra-
tio), the phase separation was reduced approximately by
two, with an average domain size of 50 nm. One explanation
could be that by increasing the content of PCBM, it disturbed
the aggregation of POPT in p-xylene and thus diminished the
domain size.
Tests in solar devices were performed for the 1:1 ratio for
POPT5 and POPT7. In the case of POPT5, the Voc was sur-
prisingly low, 0.38 V at room temperature. All the parame-
ters were decreased with annealing, leading to a decrease of
PCE from 0.8% to 0.49%. In the case of POPT7, the Voc was
higher 0.64 V, but with a low Jsc (2.60 mAꢁ cmꢀ2), thus pro-
viding a PCE of 0.72%. An annealing at 80 C did not change
the Voc but significantly increased the Jsc leading to a PCE of
1.25%. These results showed
a
clear difference of
1964
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 1953–1966

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
parameters between POPT5 and POPT7 that could be
explained by the difference of molar masses of the two poly-
mers. Indeed, POPT5 (Mn ¼ 65,000 g molꢀ1, was probably
poorly soluble in p-xylene, thus inducing aggregation in solu-
tion. This could affect the Jsc with a high content of charge
recombination due to a large phase separation. Results with
POPT7 were quite promising, even if the PCE remained
lower with this solvent (1.25% after annealing) than that
with chlorobenzene (1.58% after annealing). In all cases,
POPT:PCBM blends prepared from chlorobenzene gave the
best performances.
Agency of Research (ANR) through the project NANORGYSOL
ꢀ
and the Conseil Regional d’Aquitaine.
REFERENCES AND NOTES
1 IPCC 4th Assessment Report: Climate Change. Available at:
http://www.ipcc.ch/publications_and_data/publications_and_data_
reports.shtml. Accessed July 5, 2011.
2 Lewis, N. S.; Nocera, D. G. Proc. Natl. Acad. Sci. U.S.A. 2006,
103, 15729–15735.
3 Dennler, G.; Scharber, M. C.; Brabec, C. J. Adv. Mater. 2009,
21, 1323–1338.
4 Brabec, C. J.; Gowrisanker, S.; Halls, J. J. M.; Laird, D.; Jia,
S. J.; Williams, S. P. Adv. Mater. 2010, 22, 3839–3856.
CONCLUSIONS
It has been shown that the GRIM chain-growth polymeriza-
tion can be used to prepare POPTs with high molar masses
by changing the usual solvent, THF, for DPE. Regioregular-
ities obtained with this method are comparable with P3HTs
and those obtained by oxidative polymerization using FeCl3.
As the GRIM reaction introduces mainly irregularities at
chain ends, high molar mass polymers were targeted to di-
minish the irregularity content and to obtain improved solar
cell performances. However, it was found that slightly
reduced regioregularities had little impact on photovoltaic
device characteristics. We have shown that the introduction
of the phenyl ring on thiophene units improves the spectral
absorption of the polythiophene toward the higher wave-
lengths with respect to P3HT and also improves the thermal
stability. The bathochromic effect brought by the phenyl
group is even increased in POPT:PCBM (1:1) blends with an
extended absorption up to 750 nm. Electrochemical meas-
urements have afforded the determination of energy levels,
indicating that the HOMO level is quite similar to that of
P3HTs. This can explain the relatively high values of Voc
obtained up to 0.67 V. Finally, the photovoltaic results
obtained in solar cells with POPTs were encouraging with a
PCE up to 1.58% with the higher molar mass POPT and with
usual conditions (1:1 ratio and chlorobenzene as solvent).
The limiting parameter is the Jsc which achieves at most 5.5
mA cmꢀ2. Changing conditions, such as component ratios
and solvent, did not allow device improvements, even while
AFM images indicated a positive change of morphology with
an increase of the domain size. The solar cell results indi-
cated that the molar masses of POPTs strongly affect the PCE
and that each sample required independent optimization
with respect to thermal annealing. The morphological study
based on AFM images showed an unusual evolution of the
morphology of the POPT:PCBM blend (1:1) on thermal
annealing. After deposition, thin films were completely amor-
phous without phase separation, optimized thermal anneal-
ing promoted an interesting phase separation allowing an
increase of photovoltaic parameters. But, in the case of
POPTs, contrary to P3HTs, the PCBM seemed to be redis-
persed in the polymer matrix by continuing the increase in
temperature.
5 OPV’s with efficiencies of ꢃ 10% ‘‘OE-A Roadmap for Organic
and Printed Electronics’’ (3rd ed., 2009). Available at: http://
www.vdma.org/wps/portal/Home/en. Accessed July 5, 2011.
6 Kirova, N. Polym. Int. 2008, 57, 678–688.
7 Deibel, C.; Dyakonov, V. Rep. Progr. Phys. 2010, 73, 096401.
8 Clarke, T. M.; Durrant, J. R. Chem. Rev. 2010, 110, 6736–6767.
9 Gunes, S.; Neugebauer, H.; Sariciftci, N. S. Chem. Rev. 2007,
107, 1324–1338.
10 Thompson, B. C.; Frechet, J. M. J. Angew. Chem. Int. Ed.
Engl. 2008, 47, 58–77.
11 Ma, W.; Yang, C.; Gong, X.; Lee, K.; Heeger, A. J. Adv.
Funct. Mater. 2005, 15, 1617–1622.
12 Boudreault, P. L. T.; Najari, A.; Leclerc, M. Chem. Mater.
2011, 23, 456–469.
13 Cheng, Y. J.; Yang, S. H.; Hsu, C. S. Chem. Rev. 2009, 109,
5868–5923.
14 Zhan, X. W.; Zhu, D. B. Polym. Chem. 2010, 1, 409–419.
15 Liang, Y. Y.; Xu, Z.; Xia, J. B.; Tsai, S. T.; Wu, Y.; Li, G.; Ray,
C.; Yu, L. P. Adv. Mater. 2010, 22, E135-E138.
16 Konarka. Available at: http://www.konarka.com/index.php/
site/pressreleasedetail/konarkas_power_plastic_achieves_worl-
d_record_83_ efficiency_certification_fr. Accessed July 5, 2011.
17 Solarmer. Available at: http://www.solarmer.com/index.php.
Accessed July 5, 2011.
18 Loewe, R. S.; Khersonsky, S. M.; McCullough, R. D. Adv.
Mater. 1999, 11, 250–253.
19 Iovu, M. C.; Sheina, E. E.; Gil, R. R.; McCullough, R. D. Mac-
romolecules 2005, 38, 8649–8656.
20 Yokozawa, T.; Yokoyama, A. Chem. Rev. 2009, 109,
5595–5619.
21 Osaka, I.; McCullough, R. D. Acc. Chem. Res. 2008, 41,
1202–1214.
22 Yokoyama, A.; Miyakoshi, R.; Yokozawa, T. Macromolecules
2004, 37, 1169–1171.
23 Sheina, E. E.; Liu, J.; Iova, M. C.; Laird, D. W.; McCullough,
R. D. Macromolecules 2004, 37, 3526–3528.
24 Hiorns, R. C.; Khoukh, A.; Gourdet, B.; Dagron-Lartigau, C.
Polym. Int. 2006, 55, 608–620.
25 Oshimizu, K.; Takahashi, A.; Rho, T.; Higashihara, T.; Ree,
M.; Ueda, M. Macromolecules 2011, 44, 719–727.
26 Li, Y.; Zou, Y. Adv. Mater. 2008, 20, 2952–2958.
27 Zhou, E.; Tan, Z.; Yang, Y.; Huo, L.; Zou, Y.; Yang, C.; Li, Y.
Macromolecules 2007, 40, 1831–1837.
The authors acknowledge Dr. Martial Billon and Dr Renaud
Demadrille for the electrochemical characterizations of POPTs.
This work was financially supported by the French National
28 Hou, J.; Tan, Z.; Yan, Y.; He, Y.; Yang, C.; Li, Y. J. Am.
Chem. Soc. 2006, 128, 4911–4916.
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 1953–1966
1965

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
29 Zhou, E.; Tan, Z.; Yang, C.; Li, Y. Macromol. Rapid. Com-
mun. 2006, 27, 793–798.
47 Hiorns, R. C.; de Bettignies, R.; Leroy, J.; Baily, S.; Firon, M.;
Sentein, C.; Khoukh, A.; Preud’homme, H.; Dagron-Lartigau, C.
Adv. Funct. Mater. 2006, 16, 2263–2273.
30 Tu, G.; Bilge, A.; Adamczyk, S.; Forster, M.; Heiderhoff, R.;
Balk, L. J.; Muhlbacher, D.; Morana, M.; Koppe, M.; Scharber,
¨
M. C.; Choulis, S. A.; Brabec, C. J.; Scherf, U. Macromol. Rapid.
Commun. 2007, 28, 1781–1785.
48 Liu, J.; Loewe, R. S.; McCullough, R. D. Macromolecules
1999, 32, 5777–5785.
49 Levy, G. C.; Nelson, G. L. Carbon 13 Nuclear Magnetic Reso-
nance for Organic Chemists; Wiley: New York, 1972.
31 Cravino, A. Polym. Int. 2007, 56, 943–956.
32 Ouhib, F.; Khoukh, A; Ledeuil, J.-B.; Martinez, H. Desbrie`res,
J.; Dagron-Lartigau, C. Macromolecules 2008, 41, 9736–9743.
50 Cugola, R.; Giovanella, U.; Di Gianvincenzo, P.; Bertini, F.;
Catellani, M.; Luzzati, S. Thin Solid Films 2006, 511–512,
489–493.
33 Tan, Z.; Hou, J.; He, Y.; Zhou, Y.; Yang, C. Li, Y. Macromole-
cules 2007, 40, 1868–1873.
51 Ze, A.; Saphioannikova, M.; Neher, D.; Grenzer, J.; Grigor-
ian, S.; Pietsch, U.; Asawapirom, U.; Janietz, S.; Scherf, U.; Lie-
berwirth, I.; Wegner, G. Macromolecules 2006, 39, 2162–2171.
¨
€
€
34 Pei, Q. Jarvinen, H.; Osterholm, J. E.; Inganas, O.; Laakso, J.
Macromolecules 1992, 25, 4297–4301.
€
35 Andersson, M. R.; Selse, D.; Berggren, M.; Jarvinen, H.;
52 Sun, Q. J.; Wang, H. Q.; Yang, C. H.; Li, Y. F. J. Mater.
Chem. 2003, 13, 800–806.
¨
€
€
Hjertberg, T.; Inganas, O.; Wennerstrom, O.; Osterholm, J.-E.
Macromolecules 1994, 27, 6503–6506.
53 Ong, B.; Wu, Y.; Jiang, L.; Liu, P.; Murti, K. Synth. Met.
2004, 142, 49–52.
36 Granstrom, M.; Petritsch, K.; Arias, A. C.; Lux, A.; Ander-
sson, M. R.; Friend, R. H. Nature 1998, 395, 257–260.
54 Johansson, T.; Mammo, W.; Svensson, M.; Andersson, M.
37 Holcombe, T. W.; Woo, C. H.; Kavulak, D. F. J.; Thompson, B.
C.; Frechet, J. M. J. J. Am. Chem. Soc. 2009, 131, 14160–14161.
€
R.; Inganas, O. J. Mater. Chem. 2003, 13, 1316–1323.
ꢀ
55 Dkhissi, A.; Ouhib, F.; Iratc¸abal, P.; Pouchan, C.; Hiorns, R.
C.; Dagron-Lartigau, C. To be submitted for publication.
38 Woo, C. H.; Holcombe, T. W.; Unruh, D. A.; Sellinger, A.;
ꢀ
Frechet, J. M. J. Chem. Mater. 2010, 22, 1673–1679.
56 Hiorns, R. C.; Cloutet, E.; Ibarboure, E.; Vignau, L.; Lemaitre,
N.; Guillerez, S.; Absalon, C.; Cramail, H. Macromolecules 2009,
42, 3549.
39 Cho, C.-H.; Kang, H.; Kang, T. E.; Cho, H.-H.; Yoon, S. C.;
Jeon, M.-K.; Kim, B. J. Chem. Commun. 2011, 47, 3577–3579.
40 Ouhib, F.; Hiorns, R. C.; Bailly, S.; de Bettignies, R.; Khoukh,
A.; Preud’homme, H.; Desbrie`res, J.; Dagron-Lartigau, C. Eur.
Phys. J. Appl. Phys. 2007, 37, 343–346.
57 Scharber, M. C.; Muhlbacher, D.; Koppe, M.; Denk, P.; Wal-
dauf, C.; Heeger, A. J.; Brabec, C. J. Adv. Mater. 2006, 18,
789–794.
¨
41 Ouhib, F.; Hiorns, R. C.; de Bettignies, R.; Bailly, S.; Des-
brie`res, J.; Dagron-Lartigau, C. Thin Solid Films 2008, 516,
7199–7204.
ꢀ
58 Sivula, K.; Luscombe, C. K.; Thompson, B. C.; Frechet, J. M.
J. J. Am. Chem. Soc. 2006, 128, 13988–13989.
59 Woo, C. H.; Thompson, B. C.; Kim, B. J.; Toney, M. F.;
42 Venkatachalapathy, C.; Pitchumani, K. Tetrahedron 1993, 53,
2581–2586.
ꢀ
Frechet, J. M. J. Am. Chem. Soc. 2008, 130, 16324–16329.
60 Yang, X.; Loos, J.; Veenstra, S. C.; Verhees, W. J. H.; Wienk,
M. M.; Kroon, J. M.; Michels, M. A. J.; Janssen, R. A. J. Nano
Lett. 2005, 5, 579–583.
¨
€
43 Jarvinen, H.; Laakso, J.; Taka, T.; Osterholm, J. E.; Pei, Q.;
€
Inganas, O. Synth. Met. 1993, 55–57, 1260–1265.
44 Tamao, K.; Kodama, S.; Nakjima, I.; Kumada, M. Tetrahe-
dron 1982, 38, 3347–3354.
61 Hoppe, H.; Sariciftci, N. S. J. Mater. Chem. 2006, 16, 45–61.
62 Berson, S.; De Bettignies, R.; Bailly, S.; Guillerez, S. Adv.
Funct. Mater. 2007, 17, 1377–1384.
45 Aasmundtveit, K. E.; Samuelsen, E. J.; Mammo, W.; Svens-
€
son, M.; Andersson, M. R.; Pettersson, L. A. A.; Inganas, O.
Macromolecules 2000, 33, 5481–5489.
63 Bertho, S.; Oosterbaan, W. D.; Vrindts, D.; D’Haen, J.; Cleij,
T. J.; Lutsen, L.; Manca, J.; Vanderzande, D. Org. Electron.
2009, 10, 1248–1251.
46 Kline, R. J.; McGehee, M. D.; Kadnikova, E. N.; Liu, J.;
ꢀ
Frechet, J. M. J.; Toney, M. F. Macromolecules 2005, 38, 3312.
1966
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 1953–1966