All-donor poly(arylene-ethynylene)s containing anthracene and silole-based units: Synthesis, electronic, and photovoltaic properties

Article abstract of DOI:10.1002/pola.26914

The manuscript deals with the synthesis and properties of four new all-donor alternating poly(arylene-ethynylene)s DBSA, DBSTA, DTSA, and DTSTA. The polymers have been obtained by a Sonogashira cross-coupling of 9,10-diethynyl-anthracene with the dibromo-derivatives of 9,9-dioctyl- dibenzosilole (DBSA), 2,7-dithienyl-9,9-dioctyl-dibenzosilole (DBSTA), 4,4-dioctyl-dithienosilole (DTSA), or 2,6-dithienyl-9,9-dioctyl-dithienosilole (DTSTA). The polymers exhibited absorption profiles and frontier orbital energies strongly dependent on their primary structure. Density functional theory calculations confirmed experimental observations and provided an insight into the electronic structure of the macromolecules. In particular, the effects exerted by the thiophene units in DBSTA and DTSTA on the optical properties of the corresponding polymers could be rationalized with respect to DBSA and DTSA. Preliminary photovoltaic measurements have established that the performance of DTSA is among the highest reported for an all-donor polymer. Moreover, UV irradiation of DTSA films under air evidenced a remarkable photostability of this material, providing further evidence that ethynylene-containing electron-rich systems are promising donors for organic solar cells applications.

Full text of DOI:10.1002/pola.26914

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
All-Donor Poly(arylene-ethynylene)s Containing Anthracene and
Silole-Based Units: Synthesis, Electronic, and Photovoltaic Properties
Roberto Grisorio,1 Gian Paolo Suranna,1 Piero Mastrorilli,1,2 Giovanni Allegretta,3
Anna Loiudice,4,5 Aurora Rizzo,5,6 Giuseppe Gigli,4,5,6 Kyriaki Manoli,3 Maria Magliulo,3
Luisa Torsi3
1DICATECh: Dipartimento di Ingegneria Civile, Ambientale, del Territorio, Edile e di Chimica, Polytechnic of Bari,
Campus Universitario, via Orabona 4, 70125 Bari, Italy
2Consiglio Nazionale delle Ricerche, Istituto di Chimica dei Composti Organometallici (ICCOM-CNR), via Orabona 4, 70125 Bari, Italy
3
ꢀ
Universita degli Studi “Aldo Moro” di Bari, Dipartimento di Chimica, via Orabona 4, 70126 Bari, Italy
4
ꢀ
Dipartimento di Matematica e Fisica “E. De Giorgi”, Universita del Salento, via Monteroni, 73100 Lecce, Italy
5CBN - Center for Biomolecular Nanotechnologies, Italian Institute of Technology - Energy Platform, via Barsanti sn-73010,
Arnesano (LE), Italy
6NNL CNR-Istituto Nanoscienze, c/o Distretto Tecnologico, via per Arnesano km 5, 73100 Lecce, Italy
Correspondence to: G. P. Suranna (E-mail: surannag@poliba.it)
Received 4 July 2013; accepted 15 August 2013; published online 18 September 2013
DOI: 10.1002/pola.26914
ABSTRACT: The manuscript deals with the synthesis and prop-
erties of four new all-donor alternating poly(arylene-ethyny-
lene)s DBSA, DBSTA, DTSA, and DTSTA. The polymers have
been obtained by a Sonogashira cross-coupling of 9,10-dieth-
ynyl-anthracene with the dibromo-derivatives of 9,9-dioctyl-
dibenzosilole (DBSA), 2,7-dithienyl-9,9-dioctyl-dibenzosilole
(DBSTA), 4,4-dioctyl-dithienosilole (DTSA), or 2,6-dithienyl-9,9-
dioctyl-dithienosilole (DTSTA). The polymers exhibited absorp-
tion profiles and frontier orbital energies strongly dependent
on their primary structure. Density functional theory calcula-
tions confirmed experimental observations and provided an
insight into the electronic structure of the macromolecules. In
particular, the effects exerted by the thiophene units in DBSTA
and DTSTA on the optical properties of the corresponding
polymers could be rationalized with respect to DBSA and
DTSA. Preliminary photovoltaic measurements have estab-
lished that the performance of DTSA is among the highest
reported for an all-donor polymer. Moreover, UV irradiation of
DTSA films under air evidenced a remarkable photostability of
this material, providing further evidence that ethynylene-
containing electron-rich systems are promising donors for
C
V
organic solar cells applications.
2013 Wiley Periodicals, Inc.
J. Polym. Sci., Part A: Polym. Chem. 2013, 51, 4860–4872
KEYWORDS: anthracene; bulk hetero-junction solar cells; conju-
gated polymers; dibenzosilole; dithienosilole; light harvesting
donors;
photophysics;
polyaromatics;
poly(arylene-
ethynylene)s
the figures of merit of the photovoltaic devices, namely the
short circuit current (Jsc), the fill factor (FF), and the open
circuit voltage (VOC).2
INTRODUCTION The p-conjugated polymers have widely
been employed as light-harvesting electron donor materials
in bulk heterojunction (BHJ) solar cells, since they can be
suitably designed to provide the necessary light harvesting
as well as a favorable mixing with fullerene-based electron
acceptors.1 An ideal donor material for BHJ solar cells should
possess the following requisites: (i) a broad absorption spec-
trum in the range of wavelengths where the solar photon
flux is maximum (i.e., 500–800 nm); (ii) a thermodynami-
cally feasible electron transfer to the acceptor materials
(usually fullerene derivatives); (iii) an efficient hole trans-
port; (iv) a relatively deep highest occupied molecular orbital
(HOMO) energy. All these properties independently influence
In particular, in order to maximize the light harvesting proper-
ties of the active layer, the obtainment of conjugated materials
with suitably low band-gap is crucial. The band-gap engineer-
ing of a conjugated material is extremely important to
enhance the solar cell photocurrent, since the amount of
absorbed light depends both on the absorption wavelength
and on its extinction coefficient. From the chemical point of
view, the donor–acceptor approach3 has proved remarkably
effective for the synthetic design of narrow band-gap
Additional Supporting Information may be found in the online version of this article.
C
V
2013 Wiley Periodicals, Inc.
4860
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2013, 51, 4860–4872

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
p-conjugated macromolecules. On the other hand, their wider
band-gap all-donor counterparts, with the exception of regio-
regular poly(3-hexyl)thiophene (P3HT),4 did not thus far
exhibit impressive performances in BHJ applications. In
donor–acceptor materials, the band-gap contraction is the
result of an intramolecular charge transfer from the electron-
rich to the electron-poor unit. Furthermore, the presence of
electron poor units in a p-conjugated backbone could bring
advantages in terms of VOC, lowering the HOMO energy of the
corresponding polymer. However, some features of the donor–
acceptor concept can result detrimental for the BHJ perform-
ances, since the electron-withdrawing units might impair the
hole transport within the active layer of the BHJ solar cell
and, concentrating the electron density at the excited state,
inhibit the electron transfer to the fullerene.5
and improved hole mobility. Moreover, the possible function-
alization of the silicon atom with two alkyl chains adds a
degree of freedom to the solubility and processability of the
corresponding polymers and allows a further control of the
intermolecular polymer/fullerene interactions, aimed at
favoring the formation of a suitable interpenetrating network
in the BHJ active layer.
Having previously studied anthracene-based oligomers and
polymers for both field-effect transistor and organic photo-
voltaics applications,12 we have decided to synthesize alter-
nating poly(arylene-ethynylene)s containing 9,10-diethynyl-
anthracene and suitably functionalized dithienosilole (DTSA)
or dibenzosilole (DBSA) repeating units. The properties of
these materials were compared with those of the corre-
sponding polymers containing dithienyl-dithienosilole and
dithienyl-dibenzosilole (DTSTA and DBSTA, respectively)
units that were synthesized aiming at improving optical
response and hole mobility. While better hole mobilities
were obtained with DBSTA in field effect transistors, the
best power conversion efficiencies (1.5%, among the highest
for all-donor macromolecules in BHJ solar cells) were
achieved using DTSA in blend with PC61BM.
Aiming at the obtainment of low band-gap semiconductors,
the choice of the appropriate class of polymers is the first
step. In poly(arylene-ethynylene)s,6 the presence of the ethy-
nylene bridges can attenuate some steric and conformational
constraints, favoring the achievement of relatively planar
structures with a consequent band gap lowering. Anthracene
is a smaller acene that, due to its higher stability with
respect to tetracene and pentacene is receiving a growing
attention as building block for its manifold applications in
organic electronics.7 The combination between the rigid
electron-rich anthracene core and the ethynylene bridges
give rise to the 9,10-diethynyl-anthracene unit, which not
only promotes the planarization of the polymeric backbone,
but also favors strong intermolecular p–p stacking with pro-
spected beneficial influence on the hole transport.8 Moreover,
flanking another electron-rich moiety to the anthracene can
lead to the fine tuning of the absorption properties, frontier
orbital energies, and hole transport of the corresponding
material. It has been shown how the incorporation of 9,10-
diethynyl-anthracene units into a poly(p-phenylene-vinylene)
backbone could lead to good hole carrier mobilities and
power conversion efficiencies of up to 3.8%.9 However, a
strategy based on the mere construction of a p-conjugated
material made of all-donor units could bring to an increase
of the HOMO energy, negatively affecting the VOC of the cor-
responding solar cell. It is therefore advisable to properly
choose the further donor unit flanking the anthracene unit,
in order to avoid the aforementioned drawback, and in this
respect, silacyclopentadiene-based materials are gaining con-
siderable attention. Silacyclopentadiene (silole) is endowed
with a low lowest unoccupied molecular orbital (LUMO)
energy as a consequence of the interaction between the r*
orbital of the silylene moiety, effectively interacting with the
p* orbital of the butadiene fragment.10 As a result, the low
band gap of silole-containing materials does not occur at the
expenses of the increase in HOMO level, thereby preserving
a condition for a sufficiently high VOC. It is therefore not sur-
prising that dithienosilole and dibenzosilole building blocks
are being especially employed for the preparation of host
material for BHJ solar cells.11 Part of this interest is due to
the fact that the higher CASi bond distance (ꢀ0.3 Å) within
the polymer backbone allows for better interchain packing
EXPERIMENTAL
General Remarks
9,10-Diethynylanthracene,13 3,7-dibromo-5,5-di-n-octyl-dibenzo
[b,d]silole14 (1), 3,30-dibromo-2,20-bithiophene (4),14 and
15
Pd(PPh3)4 were synthesized according to literature proce-
dures, while the other reactants were purchased from standard
commercial sources and used as received. All solvents used
were carefully dried and freshly distilled according to standard
laboratory practice. All manipulations were carried out under
inert nitrogen atmosphere. Flash chromatography was per-
formed using a silica gel of 230–400 mesh. 1H and 13C{1H}
NMR spectra were recorded on a Bruker Avance 700 MHz
instrument. The purity of all title compounds was judged to
1
be > 95 % as determined by a combination of H and 13C NMR
analyses. FTIR measurements were carried out on a JASCO
FTIR 4200 spectrophotometer. UV–vis spectra were recorded
on a Jasco V-670 instrument. Photoluminescence (PL) spectra
were obtained on a Varian Cary Eclipse spectrophotometer. Gel
permeation chromatography (GPC) analyses were carried out
on an Agilent Series 1100 instrument equipped with a Pl-gel
5 lm mixed-C column. THF solutions for GPC analysis were
eluted at 25 ꢁC at a flow rate of 1.0 mL min21 and analyzed
using a multiple wave UV–vis detector. Number average molec-
ular weights (Mn), weight average molecular weights (Mw) and
polydispersity index (PDI) are relative to polystyrene standards.
Thermogravimetric analyses (TGA) were performed under a
nitrogen atmosphere (flow of 40 mL min21) with a Perkin-
ꢁ
Elmer Pyris 6 TGA in the range from 30 to 800 C with a heat-
ing rate of 10 ꢁC min21. Melting points were measured on a
€
Buchi B-545 instrument. Elemental analyses were obtained on
a EuroVector CHNS EA3000 instrument. Cyclic voltammetry
(CV) was carried out on a Metrohm Autolab PGSTAT 302-N
potentiostat. The materials were drop cast on a glassy carbon-
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2013, 51, 4860–4872
4861

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
working electrode from a 1 mg mL21 chloroform solution.
Measurements were carried at 25 ꢁC in acetonitrile solution
containing tetrabutylammonium tetrafluoroborate (0.025 M) as
supporting electrolyte with a scan rate of 25 mV s21. The
potentials were measured versus Ag/Ag1 as the quasi-
reference electrode. After each experiment, the potential of the
Ag/Ag1 electrode was calibrated against the ferrocene/ferroce-
nium (Fc/Fc1) redox couple. The electrochemical energy gap
mination of one-sun (AM 1.5G). The atomic force microscopy
(AFM) topography images were acquired with a XE-100 PSIA
Park Scanning Probe Microscope in noncontact mode.
Organic thin-film transistor devices were fabricated in top-
contact configurations using a highly n-doped silicon wafer
(resistivity: 20 X cm) as a gate electrode on which 300 nm
of dielectric (SiO2) was thermally grown. After suitable sur-
face cleaning, films of the relevant polymers were deposited
by spin-coating from 2.0 mg mL21 chloroform solutions.
Drain and source electrodes were fabricated on the organic
semiconducting layer by thermal evaporation through a
shadow mask. The devices were measured with an Agilent
4155 C semiconductor parameter analyzer in ambient condi-
was determined as the difference between the onsets of the
onset
oxidation and reduction potentials (Egelc 5 Eox
– Eredonset).
The HOMO and LUMO energy values were estimated from the
onset potentials of the first oxidation and reduction event,
respectively. After calibration of the measurements against Fc/
Fc1, the HOMO and LUMO energy levels were calculated
according to the following equations:
tions. The field-effect mobility (l) and threshold voltage
1/2
(Vth) were extracted from the relevant (IDS
)
versus gate
Â
À
Á
Ã
onset
ox
EHOMO ðeV Þ5– E
–E1=2 Fc =Fc1 14:8
voltage (Vg) plot in the saturation regime obtained from the
device transfer characteristics.
Â
À
Á
Ã
–E1=2 Fc =Fc1 14:8
onset
ELUMO ðeV Þ5– E
red
3,7-Di(thiophen-2-yl)-5,5-di-n-octyl-dibenzo[b,d]silole (2)
A mixture of 1 (0.60 g, 1.06 mmol), n-tributyl(thiophen-2-
yl)stannane (0.89 g, 2.35 mmol), and Pd(PPh3)4 (70 mg, 0.06
mmol) in toluene (10 mL) was refluxed for 24 h under vig-
orous stirring. The mixture was cooled to room temperature,
diluted with chloroform (100 mL), washed with water (3 3
50 mL) and dried over Na2SO4. After solvent removal, the
crude product was purified by column chromatography
(SiO2, petroleum ether 40–60 ꢁC:dichloromethane 5 9:1) to
yield 2 (0.55 g, 91%) as an off-white solid.
where E1/2(Fc/Fc1) is the half-wave potential of the Fc/Fc1
couple (the oxidation potential of which is assumed at 14.8
eV below the vacuum level) against the Ag/Ag1 electrode.
Analyses of the ground-state structures for the molecules
were carried out using the density functional theory (DFT).
The B3LYP functional was used in conjunction with the 6–
31G(d,p) or 3–21G(d) basis set. Time-dependent DFT
(TDDFT) calculations were performed to assess the excited-
state transition energies. All calculations were carried out
with the Gaussian09 program16 package and performed on
isolated molecules in vacuo. UV-photodecomposition experi-
ments (>300 nm) were carried out by irradiating the sample
with a 150-W high-pressure Hg lamp for 2 h. The samples
were placed at 1 cm from the UV source.
M.p. 5 76.9–77.9 ꢁC. 1H NMR (CDCl3): d 7.89 (d, J 5 1.5 Hz,
2H), 7.83 (d, J 5 7.9 Hz, 2H), 7.71 (dd, J 5 7.9, 1.5 Hz, 2H),
7.39 (d, J 5 3.7 Hz, 2H), 7.32 (d, J 5 5.0 Hz, 2H), 7.13 (dd,
J 5 5.0, 3.7 Hz, 2H), 1.45–1.40 (m, 4H), 1.33–1.19 (m, 20H),
1.05–1.01 (m, 4H), 0.86 (t, J 5 7.0 Hz, 6H) ppm; 13C{1H}
NMR (CDCl3): d 147.3, 144.7, 138.9, 133.2, 130.6, 128.0,
127.9, 124.6, 122.9, 121.3, 33.3, 31.8, 29.2, 29.1, 23.9, 22.6,
14.0, 12.3 ppm. ELEM. ANAL. Calcd for C36H46S2Si: C, 75.73; H,
8.12. Found: C, 75.71; H, 8.10.
Device Construction
The synthesized polymers were tested in BHJ solar cells
with structure ITO/poly(3,4-ethylenedioxythiophene):poly(s-
tyrenesulfonate) (PEDOT:PSS)/polymer:PC61BM/LiF/Al. The
photovoltaic devices were fabricated on patterned ITO-
coated glass substrates as the anode. ITO glass substrates
were sequentially cleaned by ultrasonication in deionized
water, acetone, isopropyl alcohol, and cleaned for 10 min in
3,7-Bis(2-bromothiophen-5-yl)-5,5-di-n-octyl-
dibenzo[b,d]silole (3)
N-Bromosuccinimide (NBS; 0.25 g, 1.43 mmol) was slowly
added to a solution of 2 (0.40 g, 0.70 mmol) in DMF (15
mL) at room temperature and the obtained mixture was
stirred overnight. The reaction mixture was poured into
water (100 mL) and the formed precipitate was collected by
filtration. The crude product was then purified by column
chromatography (SiO2, petroleum ether 40–60 ꢁC) to afford
3 (0.43 g, 85%) as a pale yellow solid.
ꢁ
a TL-1 solution at 85 C. The ITO surface was then modified
by spin-coating of a conductive PEDOT:PSS thin film (40 nm)
purchased by HC Starck, followed by a baking at 140 ꢁC in
nitrogen atmosphere for 15 min. Subsequently, the active
layer blend (ꢀ100 nm) was deposited by spin coating a
chloroform solution of the relevant compound and [6,6]-phe-
nyl-C61-butyric acid methyl ester (PC61BM, Nano-C) at 700
rpm for 240 s. All the operations concerning the deposition
of the active layer were performed in a glove box under a
nitrogen atmosphere. The LiF (0.6 nm) and Al (150 nm)
electrodes were thermally evaporated at low pressure (<7
3 1026 Torr) through a shadow mask. The active area of
each device (ꢀ0.04 cm2) was accurately determined by an
optical microscope equipped with a ruler. The device meas-
urements were carried out in ambient conditions under illu-
M.p. 5 84.0–85.0 ꢁC. 1H NMR (CDCl3): d 7. 91 (d, J 5 1.5 Hz,
2H), 7.82 (d, J 5 7.9 Hz, 2H), 7.69 (dd, J 5 7.9, 1.5 Hz, 2H),
7.49 (d, J 5 4.8 Hz, 2H), 7.23 (d, J 5 4.8 Hz, 2H), 1.45–1.40
(m, 4H), 1.33–1.19 (m, 20H), 1.05–1.01 (m, 4H), 0.86 (t,
J 5 7.0 Hz, 6H) ppm. 13C{1H} NMR (CDCl3): d 143.9, 140.8,
134.8, 132.4, 132.2, 131.1, 129.3, 129.1, 128.0, 127.6, 33.2,
31.9, 29.2, 29.1, 23.8, 22.5, 14.0, 12.2 ppm. . ELEM. ANAL. Calcd
for C36H44Br2S2Si: C, 59.33; H, 6.09. Found: C, 59.35; H, 6.10.
4862
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2013, 51, 4860–4872

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
4,4-Di-n-octyl-silolo[3,2-b:4,5-b0]dithiophene (5)
A mixture of 4 (9.72 g, 30.00 mmol) in anhydrous THF (50
mL) was added dropwise to a n-BuLi solution (1.6 M in hex-
2,6-Bis(5-bromothiophen-2-yl)-4,4-di-n-octyl-silolo
[3,2-b:4,5-b0]dithiophene (8)
NBS (0.30 g, 1.66 mmol) was added portionwise to a solution
of 7 (0.40 g, 0.83 mmol) in DMF (20 mL) at room tempera-
ture and the obtained mixture stirred overnight. The reaction
was quenched with water (50 mL) and extracted (3 3 50
mL) with diethyl ether. The combined organic phases were
dried over Na2SO4. After solvent removal, the crude product
was purified by column chromatography (SiO2, petroleum
ether 40–60 ꢁC) to afford 8 (0.50 g, 81%) as an orange oil.
ꢁ
anes, 24.0 mL, 60.00 mmol) in THF (50 mL) kept at 278 C
under a vigorous stirring. After 1 h reaction at 278 ꢁC, the
formation of a white suspension was observed. Subsequently,
a solution of di-n-octyl-dichlorosilane (8.07 g, 30.00 mmol)
in THF (100 mL) was added dropwise. The mixture was
stirred for further 5 h at 278 ꢁC, then allowed to reach
room temperature. After overnight stirring, the reaction was
quenched by adding a saturated aqueous NH4Cl solution
(300 mL). The aqueous layer was extracted with diethyl
ether (3 3 100 mL). The organic phases were then com-
bined, washed with water and dried over MgSO4. After sol-
vent removal, the crude product was purified by column
chromatography (SiO2, petroleum ether 40–60 ꢁC) to give 5
(6.80 g, 68% yield) as a greenish oil.
1H NMR (CDCl3): d 7.46 (d, J 5 4.5 Hz, 2H), 7.24 (s, 2H), 7.20
(d, J 5 4.5 Hz, 2H), 1.32–1.22 (m, 24H), 0.88–0.86 (m, 10H)
ppm. 13C{1H} NMR (CDCl3): d 146.8, 145.9, 139.3, 137.5,
127.5, 126.2, 124.9, 123.4, 33.3, 31.9, 29.1, 28.9, 24.2, 22.6,
14.1, 11.8 ppm.
Poly(9,9-di-n-octyl-dibenzosilole-2,7-diyl-ethynylene-
anthracen-9,10-diyl-ethynylene) (DBSA)
1H NMR (CDCl3): d 7.21 (d, J 5 4.6 Hz, 2H), 7.07 (d, J 5 4.6
Hz, 2H), 1.43–1.37 (m, 4H), 1.33–1.21 (m, 20H), 0.94–0.88
(m, 10H) ppm. 13C{1H} NMR (CDCl3): d 149.2, 141.6, 129.6,
125.0, 33.2, 31.9, 29.2, 29.1, 24.2, 22.7, 14.1, 11.9 ppm.
A mixture of 1 (0.29 g, 0.51 mmol), 9,10-diethynylanthracene
(0.12 g, 0.51 mmol), Pd(PPh3)4 (59 mg, 5.4 3 1022 mmol), CuI
(10 mg, 5.0 3 1022 mmol), triethylamine (10 mL), and toluene
(10 mL) was stirred at 90 ꢁC for 15 min. The reaction mixture
was cooled to room temperature, diluted with chloroform, and
washed with water. After solvent removal, the crude product
was dissolved into the minimum amount of chloroform and
precipitated twice with methanol. The collected powder was
washed in a Soxhlet apparatus using methanol and acetone,
then dissolved in chloroform. The volume of the chloroform
fraction was reduced before precipitation with methanol. The
obtained solid was filtered and eventually dried under vacuum
overnight to afford DBSA (51% yield) as a red powder.
2,6-Dibromo-4,4-di-n-octyl-silolo[3,2-b:4,5-b0]
dithiophene (6)
NBS (1.98 g, 11.00 mmol) was added portionwise to a solu-
tion of 5 (2.25 g, 5.38 mmol) in DMF (40 mL) at room tem-
perature. The mixture was stirred for 10 min before the
addition of water (50 mL). Subsequently, the compound was
extracted with diethyl ether (3 3 50 mL) and the combined
organic phases were dried over Na2SO4. After solvent
removal, the crude product was purified by column chroma-
ꢁ
tography (SiO2, petroleum ether 40–60 C) to afford 6 (2.82
1H NMR (CDCl3): d 8.79–7.58 (m, 14H), 1.42–1.24 (m, 24H),
1.10–1.01 (m, 4H), 0.92–0.85 (m, 6H) ppm. ELEM. ANAL. Calcd
for (C46H48Si)n: C, 87.84; H, 7.69. Found: C, 86.76; H, 7.37. IR
(KBr): m 3058, 2954, 2919, 2879, 2184, 1456, 1403, 762
cm21. GPC: Mn 5 16,200 Da; PDI 5 2.0.
g, 91%) as a greenish oil.
1H NMR (CDCl3): d 7.01 (s, 2H), 1.38–1.32 (m, 4H), 1.31–
1.20 (m, 20H), 0.91–0.87 (m, 10H) ppm. 13C{1H} NMR
(CDCl3): d 148.9, 141.1, 132.2, 111.4, 33.1, 31.8, 29.2, 29.1,
24.0, 22.6, 14.1, 11.6 ppm.
Poly[2,7-di(thiophen-2-yl)-9,9-di-n-octyl-dibenzosilole-
50,500-diyl-ethynylene-anthracen-9,10-diyl-ethynylene]
(DBSTA)
Following the procedure reported for DBSA, the polymer
DBSTA was obtained by reacting 3 (0.91 g, 1.24 mmol),
9,10-diethynylanthracene (0.28 g, 1.24 mmol), Pd(PPh3)4 (63
mg, 5.4 3 1022 mmol), CuI (10 mg, 5.4 3 1022 mmol), trie-
thylamine (15 mL), and toluene (15 mL). The material was
obtained as a deep red solid (48% yield).
2,6-Di(thiophen-2-yl)-4,4-di-n-octyl-silolo
[3,2-b:4,5-b0]dithiophene (7)
A mixture of 6 (0.56 g, 0.97 mmol), tri-n-butyl(thiophen-2-
yl)stannane (0.71 g, 1.94 mmol), and Pd(PPh3)4 (112 mg,
0.97 3 1021 mmol) in toluene (20 mL) was refluxed for 24
h under vigorous stirring. Upon cooling down the reaction to
room temperature, the mixture was diluted with dichlorome-
thane and washed with water, and the obtained organic
phase was dried over Na2SO4. After solvent removal, the
crude product was purified ꢁ by column chromatography
(SiO2, petroleum ether 40–60 C) to give 7 (0.46 g, 82%) as
an orange oil.
1H NMR (CDCl3): d 8.87–7.61 (m, 18H), 1.43–1.21 (m, 24H),
1.12–1.00 (m, 4H), 0.94–0.85 (m, 6H) ppm. ELEM. ANAL. Calcd
for (C54H52S2Si)n: C, 81.77; H, 6.61. Found: C, 81.06; H, 5.99.
IR (KBr): m 3058, 2954, 2919, 2879, 2185, 1455, 1403, 763
cm21. GPC: Mn 5 7400 Da; PDI 5 1.3.
1H NMR (CDCl3): d 7.21 (dd, J 5 4.9, 1.0 Hz, 2H), 7.18 (dd,
J 5 3.6, 1.0 Hz, 2H), 7.14 (s, 2H), 7.04 (dd, J 5 4.9, 3.6 Hz,
2H), 1.39–1.33 (m, 4H), 1.31–1.20 (m, 20H), 0.92–0.87 (m,
10H) ppm. 13C{1H} NMR (CDCl3): d 147.7, 142.9, 138.1,
137.7, 127.8, 126.4, 123.9, 123.3, 33.2, 31.8, 29.1, 29.0, 24.1,
22.6, 14.1, 11.8 ppm.
Poly[4,4-di-n-octyl-dithienosilole-2,6-diyl-ethynylene-
anthracen-9,10-diyl-ethynylene] (DTSA)
Following the procedure reported for DBSA, the polymer
DTSA was obtained by reacting 6 (0.29 g, 0.50 mmol), 9,10-
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2013, 51, 4860–4872
4863

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
SCHEME 1 Synthesis of monomers 3, 6, and 8 (R 5 n-octyl).
diethynylanthracene (0.11 g, 0.50 mmol), Pd(PPh3)4 (58 mg,
5.0 3 1022 mmol), CuI (10 mg, 5.0 3 1022 mmol), triethyl-
amine (15 mL), and toluene (15 mL). The material was iso-
lated as a deep violet solid (65% yield).
CuCl2 was added to promote the formation of the corre-
sponding homo-coupling product 3,30-dibromo-2,20-bithio-
phene (4). The subsequent step consisted in the formation
of the silole ring, that was obtained by treating 4 with n-
butyl lithium (BuLi) at 280 ꢁC, followed by the addition of
the di(n-octyl)dichloro-silane, that afforded the dithienosilole
1H NMR (CDCl3): d 8.82–7.39 (m, 10H), 1.34–1.21 (m, 24H),
0.89–0.85 (m, 10H) ppm. ELEM. ANAL. Calcd for (C42H44S2Si)n:
C, 78.70; H, 6.92. Found: C, 78.16; H 6.87. IR (KBr): m 3058,
2954, 2920, 2878, 2166, 1349, 1167, 759 cm21. GPC:
Mn 5 10,800 Da; PDI 5 1.8.
derivative
5 in good yield (68%). The corresponding
dibromo-derivative 6, to be used in the polymerization reac-
tions, was obtained in 91% yield by bromination of 5 with
NBS. The introduction of the thiophene units was carried out
by a using Pd(PPh3)4-catalyzed Stille reaction between 6 and
tri-n-butyl(thiophen-2-yl)stannane to obtain the dithienyl-
derivative 7 in 82% yield. The reaction of 7 with NBS gave
the corresponding dibromo-derivative 8 in 81% yield.
Poly[4,4-di-n-octyl-2,6-di(thiophen-2-yl)-dithienosilole-
50,500-diyl-ethynylene-anthracen-9,10-diyl-ethynylene]
(DTSTA)
Following the procedure reported for DBSA, the polymer
DTSTA was obtained by reacting 8 (0.11 g, 0.15 mmol),
9,10-diethynylanthracene (34 mg, 0.15 mmol), Pd(PPh3)4 (17
mg, 1.5 3 1022 mmol), CuI (3 mg, 1.5 3 1022 mmol), trie-
thylamine (3 mL), and toluene (3 mL). The material was iso-
lated as a deep red solid (57% yield).
The synthetic sequence for the obtainment of the polymers
is reported in Scheme 2. All materials were prepared by a
Pd-catalyzed Sonogashira coupling between the silole-based
monomers 1, 3, 6, and 8 and an equivalent amount of 9,10-
diethynylanthracene. The reactions were carried out at reflux
in the presence of CuI as co-catalyst in toluene/triethylamine
as solvent. Due to the fast polymerization rate, the reactions
had to be stopped after a few minutes, in order to avoid the
formation of insoluble materials notwithstanding the pres-
ence of the n-octyl chains. Confirmation of the polymer
structure was obtained by 1H NMR, FTIR, as well as by ele-
mental analyses. The number-average molecular weights
(Mn) and PDIs of the synthesized polymers were evaluated
by GPC. The thiophene-containing polymers DBSTA and
DTSTA showed lower molecular weights (Mn 5 7400 and
5600 Da, respectively) compared to the corresponding poly-
mers DBSA and DTSA (Mn 5 16,200 and 10,800 Da, respec-
tively). This can be due to the lower solubility of DBSTA and
DTSTA in the reaction medium, induced by the presence of
the thiophene units, hampering the chain growth.
1H NMR (CDCl3): d 8.82–7.49 (m, 14H) ppm, 1.34–1.21
(m, 24H), 0.89–0.85 (m, 10H) ppm. ELEM. ANAL. Calcd for
(C50H48S4Si)n: C, 74.58; H, 6.01. Found: C 74.16; H 5.87. IR
(KBr):
m 3058, 2954, 2919, 2879, 2167, 1348, 1166,
760 cm21. GPC: Mn 5 5600 Da; PDI 5 1.4.
RESULTS AND DISCUSSION
Polymers Preparation and Characterization
The synthetic approach for the obtainment of the silole-
based monomers is described in Scheme 1. The preparation
of 2,7-dithienyl-9,9-di-n-octyl-dibenzosilole (2) was achieved
by
a Pd(PPh3)4-catalyzed Stille cross-coupling of 2,7-
dibromo-9,9-di-n-octyl-dibenzosilole (1) with tri-n-butyl(thio-
phen-2-yl)stannane. The obtained product was submitted to
a
bromination with NBS to afford the corresponding
TGA revealed a satisfactory thermal stability of the synthe-
sized polymers, as shown in Figure 1. The 5% weight loss
decomposition temperatures for the dibenzosilole-containing
polymers DBSA and DBSTA are lower (389 and 357 ꢁC,
respectively) compared to those obtained for the
dibromo-derivative (3), to be used in the polymerization
reaction. For the preparation of dithienosilole monomers, the
commercially available 3-bromothiophenꢁe was reacted with
lithium diisopropylamide in THF at 280 C. After metalation,
4864
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2013, 51, 4860–4872

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
SCHEME 2 Synthesis of the polymers. Reaction conditions: Pd(PPh3)4/CuI in toluene triethylamine (R 5 n-octyl). Reaction condi-
tions: Pd(PPh3)4, CuI, NEt3, toluene, reflux.
ꢁ
dithienosilole-containing DTSA and DTSTA (422 and 427 C,
respectively).
absorption peak that, in chloroform solution, is located at
562 nm while in the solid state it is red-shifted up to 601
nm. The absorption spectra of DTSTA shows a maximum at
545 nm in chloroform which is shifted to 586 nm in the
solid state. The bathochromic shifts of the absoption profiles
of all the four polymers as thin films can evidently be attrib-
uted to the enhanced intermolecular interactions occurring
in solid state.
Optical Properties
The UV–vis spectra of DTSA, DTSTA, DBSA, and DBSTA
were measured both in diluted chloroform solution (to rule
out aggregation effects on the optical properties) and in the
solid state (Figs. 2 and 3, respectively) by depositing them
as thin films on quartz substrates. The UV–vis spectra of
DBSA and DBSTA in chloroform solution show absorption
maxima (kmax) at 508 and 499 nm, respectively. The absorp-
tion profiles of DBSA and DBSTA in the solid state are
slightly broadened and sensibly red-shifted with respect to
those recorded in solution, exhibiting kmax at 518 and 558
nm, respectively. The UV–vis spectra of DTSA show an
The introduction of thiophene bridges between the
diethynyl-anthracene and the dibenzosilole units leads to the
increase of the p-conjugation extension of DBSTA with
respect to DBSA as evidenced by the observed decrease in
optical band gap (Egopt, Table 1) measured both in solution
and in the solid state. Counterintuitively, the optical
FIGURE 1 TGA plots obtained for the polymers. [Color figure
can be viewed in the online issue, which is available at
wileyonlinelibrary.com.]
FIGURE 2 UV–vis spectra of the polymers in chloroform. [Color
figure can be viewed in the online issue, which is available at
wileyonlinelibrary.com.]
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2013, 51, 4860–4872
4865

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
FIGURE 4 Normalized emission spectra of the polymers in
chloroform. [Color figure can be viewed in the online issue,
which is available at wileyonlinelibrary.com.]
FIGURE 3 UV–vis spectra of the polymers as thin films. [Color
figure can be viewed in the online issue, which is available at
wileyonlinelibrary.com.]
properties of DTSA compared to those of DTSTA hint that
the introduction of the thiophene units lead to a lowering of
the p-conjugation extension. This behavior suggests that, in
these poly(arylene-ethynylene)s, the degree of p-conjugation
is influenced by the aromatic units chosen as well as by the
p-conjugated backbone distorsion. This aspect has been
investigated (vide infra) by suitable theoretical calculations.
the introduction of the thiophene units on the optical prop-
erties of the materials. While DBSA and BDSTA exhibited a
vibronically structured emission profile, DTSA and DTSTA
emitted with a broad PL band at longer wavelengths. Fur-
thermore, the relatively low Stokes shifts observed (ranging
from 745 to 2278 cm21) suggest the absence of an intramo-
lecular charge transfer upon excitation, as a consequence of
the electron-rich character of the aromatic units constituting
the primary structures of the polymers. No PL spectra could
be recorded for the copolymers as thin films, suggesting the
occurrence of strong intermolecular interactions in the solid
state. It is plausible that the intermolecular p–p stacking,
probably favored by the anthracene units, leads to aradiative
deactivation pathways of the polymer excited states, with the
consequent quenching of their emission.
The presence of the thiophene units in DTSTA and DBSTA
also strongly influences the molar absorption coefficients (e)
of the polymers, which were calculated in chloroform solu-
tions by considering the Mn as its molecular weight.17 A very
high emax could be observed for DBSA and DTSA (2.4 3 106
M21cm21 and 8.6 3 105 M21cm21, respectively) while the
molar absorbivity is remarkably lower in the case of DBSTA
and DTSTA (emax 5 1.4 3 105 M21cm21 and 1.1 3 105
M21cm21, respectively). This aspect determines the amount
of photons absorbed by the active layer in a BHJ solar cell,
which is higher in the case of DBSA and DTSA with respect
to DBSTA and DTSTA thereby potentially influencing the
photocurrent generation.
Electrochemical Properties
CV measurements allowed the assessment of the HOMO and
LUMO energy levels of the synthesized polymers. Represen-
tative CV curves are shown in Figure 5, while the obtained
HOMO and LUMO values are reported in Table 1. The cyclic
voltammograms of both DTSA and DTSTA showed quasi-
reversible p-doping (oxidation/re-reduction) and quasi-
The PL spectra of the four copolymers in chloroform solu-
tions were recorded and are shown in Figure 4. As reported
in Table 1, the emission maxima (kem) followed the same
trend of the optical band gaps, thus confirming the effects of
reversible n-doping (reduction/reoxidation) processes.
A
similar CV behavior was observed for DBSTA, while DBSA
TABLE 1 Optical and Electrochemical Properties of DTSA, DTSTA, DBSA, and DBSTA: UV–Vis and PL Data Obtained Both
in Chloroform and in the Solid State; HOMO and LUMO Energies Obtained From Cyclic Voltammetry
opt sol
opt film
elc
Eg (eV)
kmax sol (nm)
kmax film (nm)
Eg
(eV)
Eg
(eV)
kem sol (nm)
HOMO (eV)
LUMO (eV)
DBSA
508
499
562
544
518
558
602
586
2.31
2.19
1.87
1.93
2.16
2.06
1.74
1.80
528
552
627
621
25.6
25.4
25.2
25.3
23.3
23.3
23.4
23.4
2.3
2.1
1.8
1.9
DBSTA
DTSA
DTSTA
4866
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2013, 51, 4860–4872

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
given primary structure, the calculations should be carried
out on appropriate finite models, of increasing size, and then
by extrapolating the chosen parameters for an hypothetically
infinite polymer segment. To this purpose, DFT and TDDFT
calculations were carried out on the repeating unit (mono-
mer) of each polymer as well as on oligomers, from dimers
to tetramers (n 5 1–4 in Scheme 2).
It can be assumed that the length of the alkyl chains at the
bridging silicon atom of the dibenzosilole and dithienosilole
units do not exert a significant influence on the geometry of
the selected polymer segments. On this ground, the n-octyl
chains were replaced by ACH3 groups, in order to reduce
the computational load. Geometry optimization was per-
formed in vacuo and carried out using the B3LYP functional
with the 3–21G(d) basis set, also allowing the description of
extended systems containing a large number of atoms. Sub-
sequently, starting from the optimized geometries obtained,
the HOMO and LUMO energy values were determined for
each model structure using the 6–31G(d,p) basis set. The
TDDFT calculations were carried out using the CAM-B3LYP
functional and the 3–21G(d) basis set.
FIGURE 5 CV traces of the synthesized polymers. [Color figure
can be viewed in the online issue, which is available at
wileyonlinelibrary.com.]
exhibited both anodic and cathodic irreversibile events.18
Concerning the HOMO and LUMO energy levels and the elec-
trochemical energy gap (Egelc), the incorporation of the thio-
phene spacers led to a substantial HOMO-level increase
passing from 25.6 eV for DBSA to 25.4 eV for DBSTA, while
the LUMO energy levels of both polymers are the same
Plotting the HOMO, LUMO, and the lowest excitation energies
of the oligomers against 1/n (where n indicates the number
of repeating units of the selected models), it is possible to
extrapolate the values ideally obtainable for a polymer seg-
ment of virtually infinite length, as the intercept with the
ordinate axes, using a linear fit.19 This approach becomes
mandatory for an accurate comparison between polymers
constituted by repeating units containing a different amount
of aromatic units, such as DTSTA (DBSTA) compared to
DTSA (DBSA).
elc
(23.3 eV). As a result, the observed Eg of DBSTA (2.1 eV)
is lower than that of DTSA (2.3 eV).
On the contrary, the insertion of the thiophene units in
DTSTA induced a lowering of the HOMO energy value (25.3
eV) compared to that of DTSA (25.2 eV). Again, no effects
could be observed on the LUMO energy levels (23.4 eV for
elc
both polymers) and consequently the observed Eg
of
DTSTA (1.9 eV) results higher than that of DTSA (1.8 eV).
The trend observed for the Egelc, therefore, follows the one
previously discussed for the Egopt. From the analysis of the
aforementioned results, it can be concluded that the synthe-
sized polymers hold the suitable requisites to efficiently
operate, at least in principle, within a BHJ solar cell: their
HOMO energy levels are in fact favorably aligned with the
ITO workfunction (24.7 eV) and their LUMO energy levels
are compatible with an efficient photoreduction of the com-
mon fullerene-based electron-acceptors (endowed with a
LUMO of ꢀ 24.3 eV) used in the donor–acceptor blend con-
stituting the photoactive layer. It appears, however, that
DTSA and DTSTA hold better features in terms of light har-
vesting properties: their band gap is in fact remarkably
lower with respect to that of DBSA and DBSTA.
The relevant results are shown in Figures 6 and 7 and are
summarized in Table 2. Concerning the HOMO and LUMO
energy values extrapolated for 1/n 5 0, these are in excellent
Theoretical Calculations
Theoretical methods are helpful for gaining insight into the
electronic structure as well as into the optical properties of
conjugated materials. It must be noted however that, for
polymers, several parameters such as the molecular orbital
energies and the excitation vertical transitions, are strongly
influenced by the number of the repeating units of the model
chosen for the calculation. Therefore, in order to draw reli-
able conclusions from the theoretical results obtained for a
FIGURE 6 Calculated linear trends (R > 0.99) of HOMO and
LUMO energy values for the model structures (n 5 1–4) of the
synthesized polymers. [Color figure can be viewed in the
online issue, which is available at wileyonlinelibrary.com.]
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2013, 51, 4860–4872
4867

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
structures characterized by an intramolecular charge trans-
fer, also the LUMOs are evenly distributed along the entire
chain. Furthermore, according to the TDDFT results, the
main transition (S0!S1) of the selected segments of the
polymers is predominantly described by a HOMO!LUMO
transition.
Photovoltaic Characterization
The photovoltaic properties of the synthesized polymers
were investigated in BHJ solar cells of configuration: ITO/
PEDOT:PSS/polymer:PC61BM/LiF/Al. We tested the polymers
in different donor:acceptor ratios. As reported in Table 3, the
best figures of merit in terms of power conversion efficiency
(PCE) for each material were obtained using
a poly-
mer:PC61BM blend ratio of 40:60. Figure 9 shows the charac-
teristic current density–voltage (J–V) plots obtained for the
best devices. In particular, the best performing devices were
obtained with DTSA and afforded a PCE of 1.50%; the short
circuit current density (JSC), open circuit voltage (VOC), and
fill factor (FF) were 5.97 mA cm22, 0.68 V, and 37%, respec-
tively. The devices based on the analogous thiophene-
containing polymer DTSTA were characterized by a decrease
in all figures of merit, leading to a very poor (PCE 5 0.20%)
overall efficiency, substantially ascribable to the poor quality
of the blend films caused by the limited solubility of the
material. In the case of the dibenzosilole-based polymers
DBSA and DBSTA, their poorer performances can be reason-
ably explained by their higher energy gap with respect to
that of DTSA. Analyzing in detail their figures of merit, how-
ever, DBSA showed a less efficient photocurrent generation
FIGURE 7 Calculated linear trends (R > 0.99) of the lowest exci-
tation energies for the model structures (n 5 1–4) of the synthe-
sized polymers. [Color figure can be viewed in the online
issue, which is available at wileyonlinelibrary.com.]
agreement with the experimental estimates of optical and
electrochemical band gaps. Also in the case of the excitation
energies, the extrapolated values are in good agreement with
the found optical band gaps measured for diluted solutions,
as shown in Figure 7. Remarkably, DFT and TDDFT calcula-
tions confirm that the incorporation of the thienylene units
leads to different effects in dibenzosilole- and dithienosilole-
containing polymers, in particular, the higher extrapolated
excitation energy calculated for the DTSTA model justifies
the counterintuitive blue-shifted absorption of DTSTA with
respect to DTSA (vide supra) observed in the UV–vis spectra.
It is very likely that the explanation of this peculiar behavior
may reside in the distortion of the copolymer backbone,
that, limiting the conjugation extension, counterbalances the
band gap contraction promoted by the incorporation of the
low-energy thiophene units.
(JSC 5 2.25 mA cm22
) with respect to that of DBSTA
(JSC 5 3.76 mA cm22) probably ascribable to its slightly
TABLE 2 Calculated HOMO, LUMO, and Lowest Excitation
Energies of DBSA, DBSTA, DTSA, and DTSTA
HOMOa LUMOa Exc. En.b HOMOa LUMOa Exc. En.b
n
(eV)
(eV)
(eV)
(eV)
(eV)
(eV)
In fact, as illustrated in Figure 8, the DBSA and DTSA mod-
els show an extremely planar structure, since the torsion
angle between the silole and anthracene planes in relatively
long segments (trimers) lies between 0.46ꢁ and 4.77ꢁ, due to
the presence of the ethynylene spacers between the aryl
moieties. In the case of DBSTA and DTSTA, conversely, the
dihedral angles between the silole and thiophene units are
higher and comprised between 23ꢁ and 26ꢁ, while the
remaining part of the conjugated segment lies close to
planarity.
DBSA
DBSTA
1
2
3
4
24.99
24.82
24.77
24.74
22.25
22.44
22.51
22.54
22.64
2.98
2.67
2.59
2.54
2.38
24.92
24.75
24.69
24.67
24.58
22.34
22.49
22.52
22.54
22.62
DTSTA
22.43
22.63
22.69
22.70
22.81
2.86
2.60
2.51
2.47
2.34
fit 24.65
DTSA
1
2
3
4
24.83
24.59
24.53
24.48
22.29
22.57
22.64
22.74
22.86
2.79
2.34
2.20
2.11
1.89
24.73
24.60
24.54
24.52
24.45
2.62
2.30
2.20
2.16
2.00
The calculated electron density distributions of the HOMO
and LUMO levels along with the optimized geometry
obtained for selected segments (trimers) of the polymers are
shown in Figure 8. It can be stated that the HOMOs of the
polymers are evenly distributed along the entire chain, hint-
ing a good hole transport ability (at least along the conju-
gated backbone) after the electron transfer to the fullerene
acceptor in a BHJ solar cell. Differently from the polymeric
fit 24.38
aCalculated at the B3LYP/6–31G(d,p) leved based on B3LYP/3–21G(d)
optimized geometries.
b
TDDFT excitation energies calculated on the B3LYP/3–21G(d) optimized
geometries with the CAM-B3LYP functional using the 3–21G(d) basis
set.
4868
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2013, 51, 4860–4872

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
FIGURE 8 Isodensity (0.02) plots of HOMOs and LUMOs for the trimers (n 5 3) of DBSA, DBSTA, DTSA, and DTSTA. [Color figure
can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
narrower and blue shifted (see Fig. 3) absorption profile.
Conversely, due to its lower HOMO energy value, DBSA devi-
ces exhibited a remarkably higher VOC with respect to
DBSTA-based ones (0.75 V respect to 0.49 V for the best
devices).
mer) of the PC61BM amount in the blend ratio (Table 3) led
to lower performances in terms of PCE recorded for each
polymer. This behavior might be ascribable to a less efficient
charge transport deriving from an unfavorable organization
of the interpenetrating network between the polymer and
PC61BM phases. However, the same PCE response was
observed by increasing the polymer amount in the blend
films (up to 60% wt/wt), notwithstanding the plausible
higher efficiency in the light harvesting process. This result
confirms that the optimal tridimensional organization of the
polymer/PC61BM phases exerts a key role in the performan-
ces exhibited by these polymers.
It is known that the variation of the polymer/PC61BM blend
ratios significantly affects the photovoltaic response of the
BHJ solar cell. A slight increase (70% respect to the poly-
TABLE 3 Photovoltaic Parameters Obtained for the BHJ Devi-
ces of Configuration ITO/PEDOT-PSS/Polymer:PC61BM/LiF/Al
Polymer:
PC61BM
A fundamental property to be evaluated for a polymeric
donor material is its hole mobility, since the charge transport
can also influence the performance of a BHJ device. In order
to tackle this issue, suitable top-contact organic field-effect
transistors (OFET) were constructed and their transfer char-
acteristics were measured. As expected, a p-type behavior
for each transistor was observed. Except for DBSA, that did
not exhibit modulation, average extracted hole mobilities
were 1.50 3 1024 cm2 V21 s21 (DBSTA), 3.50 3 1025 cm2
V21 s21 (DTSA) and 6.80 3 1025 cm2 V21 s21 (DTSTA).
The obtained hole mobilities are in line with those exhibited
by other polymers showing similar photovoltaic performan-
ces.20 The transfer curves of the materials (see Supporting
Information) showed a remarkably higher average threshold
voltage (Vth) for the dibenzosilole–containing polymer (229
V for DBSTA) indicating a relatively high density of charge
carrier traps for this material with respect to DTSTA and
Ratio (wt:wt) PCE (%) FF (%) VOC (V) JSC (mA cm22
)
DBSA 30:70
40:60
0.42
0.52
0.20
0.42
0.59
0.35
0.52
1.50
0.55
0.12
0.20
0.19
28
31
28
31
32
30
36
37
29
28
22
29
0.71
0.75
0.75
0.43
0.49
0.40
0.49
0.68
0.39
0.23
0.55
0.32
2.15
2.25
0.95
3.13
3.76
2.88
2.98
5.97
4.85
1.83
1.62
2.03
60:40
DBSTA 30:70
40:60
60:40
DTSA 30:70
40:60
60:40
DTSTA 30:70
40:60
60:40
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2013, 51, 4860–4872
4869

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
Photochemical Stability
In order to qualify a suitably designed p-conjugated polymer
(which is the main responsible for the light-harvesting) as
promising donor for BHJ solar cells, the photoconversion effi-
ciency value reached by the material in the devices has to be
accompanied by a marked optical stability under the operat-
ing conditions. Changes in the absorption profile and inten-
sity of the polymer donor can, in fact, lead to a degradation
of the PCE of the device, deriving from a decreased amount
of absorbed photons by the active layer. The chemical nature
of polyconjugated materials specifically designed as light har-
vesters in organic photovoltaics makes them vulnerable
toward photo-oxidation.22 Although photovoltaic devices can
be efficiently protected from the atmosphere by encapsula-
tion, oxygen traces can, in the course of the device lifetime,
come in contact with the photoactive organic film, triggering
well-established radicalic pathways for organic semiconduc-
tors23 rapidly leading to deterioration of the performances.
On the basis of the aforementioned considerations, acceler-
ated degradation tests carried out under ambient conditions
can provide a wealth of data on the photostability of a donor
material for BHJ solar cells considerably reducing investiga-
tion times on these issues.
FIGURE 9 J–V plots obtained for the devices of configuration
ITO/PEDOT-PSS/polymer:PC61BM/LiF/Al at
a polymer:PC61BM
blend ratio of 40:60 (wt/wt). [Color figure can be viewed in the
online issue, which is available at wileyonlinelibrary.com.]
DTSA, showing Vth of 24 and 218 V, respectively. The on/
off ratios (ꢀ102) were very similar for all the polymers.
From these results, it is apparent that the introduction of the
thiophene units in DBSTA and DTSTA substantially increases
the hole mobility with respect to their counterparts DBSA
and DTSA. This observation can be explicable by admitting a
higher degree of p–p interactions between adjacent conju-
gated backbones, as is usually observed for organic semicon-
ductors for OFET.21 However, especially comparing the field-
effect transistor and the BHJ solar cell performances
obtained for DTSTA with respect to DTSA, the increased
hole mobility of DTSTA did not lead to a consequent
improvement in photovoltaic response. It is reasonable to
suppose that, despite the favorable absorption profile and
hole mobility recorded in OFET, the reasons for the poor BHJ
solar cell figures of merit of DTSTA are solely to be searched
in the formation of a nonsuitable polymer/PC61BM interface
In setting up a photo-degradation accelerated test, it should
be taken into account that the visible portion (400–700 nm)
of the sun-light is partially absorbed by the polymer donor,
giving rise to an excited state the lifetime of which is in
the subnanoseconds range, due to the very fast electron
transfer to the fullerene acceptor.24 As a consequence, poly-
mer degradation is not likely to occur while the polymer
is in its excited state. Conversely, specific degradation path-
ways leading to undesired variations in the absorption spec-
trum of the polymer donor can be triggered by the high-
energy photons in the UV portion (ꢀ5%) of the solar
spectrum.25
For these reasons, it was chosen to study the photochemical
stability of DTSA (the material showing the best figures of
with
a consequent negative influence on the charge
photogeneration.
To this purpose, knowing that nanoscale phase separation of
the photoactive layer on the performance of OPVs is a critical
issue, and in order to corroborate the observations on the
electronic and photovoltaic properties of the new materials
with information on the mesoscopic characterization of the
blends, tapping mode AFM measurements (see Supporting
Information Figs. S8–S11) were carried out on the 40:60 wt/
wt polymer:PC61BM blend film that gave the best results in
terms of PCE. While the DBSA-, DBSTA-, and DTSA-based
blend films revealed similar morphologies, it was observed
that the DTSTA:PC61BM film revealed bigger domain sizes. It
seems rational to assume that the higher tendency toward
interchain interaction of DTSTA, hinted by its higher field-
effect mobility, can also be held responsible for the segrega-
tion of the polymer/PC61BM phases which results detrimen-
tal in terms of its photovoltaic efficiency.
FIGURE 10 Evolution of the UV–visible spectrum of DTSA as
thin films during UV-promoted photo-oxidation.
4870
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2013, 51, 4860–4872

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
radicals in DTSA is ruled out by the presence of the quater-
nary silicon atom,27 which is much less prone to oxidation
with respect to the a-methylene group of the n-hexyl chains
in P3HT. Furthermore, the photochemical stability of DTSA is
reinforced by the 9,10-diethynyl-anthracene units, the pres-
ence of which, as also observed for other polycyclic aromatic
units,28 seems to inhibit the photo-induced degradation of
the corresponding polymers.
CONCLUSIONS
Instead of the mere search for the highest performances
obtainable by light-harvesting materials in BHJ solar cells, the
evaluation of the structure-property relationships between
polymers belonging to a specific class of organic semiconduc-
tors is a valuable tool in order to conceive new donor archi-
tectures for organic photovoltaics. To this purpose, four new
all-donor copolymers containing silole-based units (i.e., diben-
zosilole, dithienyl-dibenzosilole, dithienosilole or dithienyl-
dithienosilole) and 9,10-diethynyl-anthracene units were syn-
thesized by Sonogashira polycondensation reactions. These
alternating poly(arylene-ethynylenes) were characterized by
NMR, IR, UV–vis, PL, and CV. The dithienosilole- and
dithienyldithienosilole-based polymers (DTSA and DTSTA)
were identified as plausible candidates for BHJ applications
due to their light-harvesting properties, and this choice was
corroborated by DFT calculations carried out on suitable finite
model systems of the four polymeric architectures. As a mat-
ter of fact, in BHJ devices DBSA and DBSTA did not exceed a
PCE of 0.59%. Conversely, despite a higher field-effect hole
mobility exhibited by DTSTA with respect to its thiophene-
free conterpart DTSA, organic BHJ solar cells embodying
PC61BM as electron acceptors revealed that DTSA is extremely
more performing with respect to DTSTA (PCE 0.20%) prob-
ably due to the formation of an unsuitable blend between
DTSTA and the acceptor, caused by strong intermolecular
interactions between adjacent polymer chains, as suggested
by AFM investigations. Remarkably, the PCE value (1.50%)
obtained with DTSA represents one of the highest efficiencies
obtained for BHJ solar cells based on all-donor polymers con-
structed without the use of additives or of annealing proc-
esses. Last but not least, UV-promoted degradation of DTSA
films under air evidenced a remarkable photostability of this
material with respect to the well-known P3HT donor. Further
improvements can be expected by an optimization of the
processing conditions as well as by increasing of the molecu-
lar weights and solubility for this class of new organic
semiconductors.
FIGURE 11 Evolution of the UV–visible spectrum of P3HT as
thin films during UV-promoted photo-oxidation.
merit in BHJ devices) by illuminating a spin-coated film of
the polymer with a UV lamp in air. As shown in Figure 10,
the photo-oxidation treatment did not induce substantial
changes in the absorbance of DTSA, while the observed
slight variations in the absorption profile could be attributed
to the polymer chains rearrangements, probably due to the
increase of the temperature during the irradiation.
In order to confirm the photostability of DTSA, the same
experiment was carried out on a film of poly(3-hexyl)thio-
phene (P3HT), which is the prototypical all-donor material
for BHJ solar cells due to its efficiencies in devices. For the
P3HT film,
a continuous decrease in the intensity of
the absorption band accompanied by a slight blue-shift of
the absorption maximum was observed (Fig. 11) during the
irradiation. It is widely accepted that the degradation of
P3HT involves the radical oxidation of the alkyl side chains
with the subsequent stepwise oxidation of the sulfur atom of
the thiophene rings.26 The consequent cleavage of the poly-
mer backbone results in a loss of p-conjugation, leading to a
lowering of the absorbance. The optical instability of P3HT
is due to the loss of the hydrogen atom belonging to the
a-methylene group of the alkyl chain with the formation of a
benzylic-type radical with high mesomerical stability, which
can be considered the “trigger” of the material degradation.
The propagation of the radicalic degradation is not only sug-
gested by the aforementioned modifications of the absorp-
tion spectrum, but is also corroborated by the fact that, after
exposure to the UV irradiation, the resulting material is
insoluble in chloroform, which can be a consequence of a
radical promoted crosslinking between adjacent polymer
chains.
ACKNOWLEDGMENTS
The authors gratefully acknowledge for funding: Regione
Puglia (APQ-Reti di Laboratorio, Project PHOEBUS, cod. 31-
FE1. 20001); MIUR (PRIN 2009 project 2009AZKNJ7); MAAT-
Molecular NAnotechnology for HeAlth and EnvironmenT
(PON02_ 00563_3316357); EFOR-Energia da FOnti Rinnovabili
(Iniziativa CNR per il Mezzogiorno L. 191/2009 art. 2
comma 44).
Conversely, in the case of DTSA, the film recovered after the
UV exposure was almost completely soluble in chloroform,
suggesting that the radical promoted degradation was inhib-
ited for this material. This aspect can be rationalized in view
of the fact that the formation of mesomerically stabilized
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2013, 51, 4860–4872
4871

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
22, 19752–19760; (b) G. Romanazzi, A. Dell’Aquila, G.
P. Suranna, F. Marinelli, S. Cotrone, D. Altamura, C.
Giannini, L. Torsi, P. Mastrorilli, J. Mater. Chem. 2011, 21,
15186–15189.
REFERENCES AND NOTES
1 (a) P.-L. Boudreault, T. A. Najari, M. Leclerc, Chem. Mater.
2011, 23, 456–469; (b) M. Helgesen, R. Sïndergaard, F. C.
Krebs, J. Mater. Chem. 2010, 20, 36–60; (c) S. Gunes, H.
€
13 S. Colella, M. Mazzeo, R. Grisorio, E. Fabiano, G.
Melcarne, S. Carallo, M. D. Angione, L. Torsi, G. P. Suranna, F.
della Sala, P. Mastrorilli, G. Gigli, Chem. Commun. 2010, 6273–
6275.
Neugebauer, N. S. Sariciftci, Chem. Rev. 2007, 107, 1324–1338.
2 (a) H. Zhou, L. Yang, W. You, Macromolecules 2012, 45, 607–
632; (b) C. J. Brabec, S. Gowrisanker, J. J. M. Halls, D. Laird, S.
Jia, S. P. Williams, Adv. Mater. 2010, 22, 3839–3856; (c) R.
Lecover, N. Williams, N. Markovic, D. H. Reich, D. Q. Naiman,
H. E. Katz, ACS Nano 2012, 6, 2865–2870.
14 K. L. Chan, M. J. McKiernan, C. R. Towns, A. B. Holmes, J.
Am. Chem. Soc. 2005, 127, 7662–7663.
15 D. R. Coulson, L. C. Satek, S. O. Grim, Inorg. Synth. 1972,
13, 121–124.
3 (a) N. Blouin, M. Leclerc, Acc. Chem. Res. 2008, 41, 1110–1119;
(b) H. A. M. van Mullekom, J. A. J. M. Vekemans, E. E. Havinga,
E. W. Meijer, Mater. Sci. Eng. R 2001, 32, 1–40; (c) E. E. Havinga,
W. Hoeve, H. Wynberg, Synth. Met. 1993, 229, 55–57.
16 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Men-
nucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P.
Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnen-
berg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa,
M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven,
J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro, M. Bearpark, J.
J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobaya-
shi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S.
S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M.
Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jara-
millo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R.
Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma,
V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S.
4 See for example: G. Li, V. Shrotriya, J. Huang, Y. Yao, T.
Moriarty, K. Emery, Y. Yang, Nat. Mater. 2005, 4, 864–868.
€
5 P. M. Beaujuge, W. Pisula, H. N. Tsao, S. Ellinger, K. Mullen,
J. R. Reynolds, J. Am. Chem. Soc. 2009, 131, 7514–7515.
6 (a) R. S. Ashraf, J. Gilot, R. A. J. Janssen, Sol. Energy Mater.
Sol. Cells 2010, 94, 1759–1766; (b) R. S. Ashraf, M. Shahid, E.
Klemm, M. Al-Ibrahim, S. Sensfuss, Macromol. Rapid Com-
mun. 2006, 27, 1454–1459.
7 J.E. Anthony, Chem. Rev. 2006, 106, 5028–5048.
8 (a) M. Seri, A. Marrocchi, D. Bagnis, R. Ponce, A. Taticchi, T.
J. Marks, A. Facchetti, Adv. Mater. 2011, 23, 3827–3831; (b) D.
K. Susarova, E. A. Khakina, P. A. Troshin, A. E. Goryachev, N.
S. Sariciftci, V. F. Razumov, D. A. M. Egbe, J. Mater. Chem.
2011, 21, 2356–2361; (c) F. Silvestri, A. Marrocchi, M. Seri, C.
Kim, T. J. Marks, A. Facchetti, A. Taticchi, J. Am. Chem. Soc.
2010, 132, 6108–6123; (d) A. Marrocchi, F. Silvestri, M. Seri, A.
Facchetti, A. Taticchi, T. J. Marks, Chem. Commun. 2009, 1380–
1382.
€
Dapprich, A. D. Daniels, O. Farkas, J. B. Foresman, J. V. Ortiz,
J. Cioslowski, and D. J. Fox, Gaussian, Inc., Wallingford CT,
2009.
17 M. Coakley, M. D. McGehee, Chem. Mater. 2004, 16, 4533–
4542.
18 E. Wang, C. Li, W. Zhuang, J. Peng, Y Cao, J. Mater. Chem.
2008, 18, 797–801.
€
€
9 (a) S. Rathbeger, J. Perlich, F. Kuhnlenz, S. Turk, D. A. M.
19 B. P. Karsten, J. C. Bijleveld, L. Viani, J. Cornil, J.
Gierschner, R. A. J. Janssen, J. Mater. Chem. 2009, 19, 5343–
5350.
Egbe, H. Hoppe, R. Gehrke, Polymer 2011, 52, 3819–3826; (b) D.
€
€
A. M. Egbe, S. Turk, S. Rathbeger, F. Kuhnlenz, R. Jadhav, A.
€
Wild, E. Birckner, G. Adam, A. Pivrikas, V. Cimrova, G. Knor, N.
20 See, for example: J. Kim, S. H. Kim, I. H. Jung, E. Jeong, Y.
Xia, S. Cho, I.-W. Hwang, K. Lee, H. Suh, H.-K. Shim, H. Y.
Woo, J. Mater. Chem. 2010, 20, 1577–1586.
S. Sariciftci, H. Hoppe, Macromolecules 2010, 43, 1261–1269;
(c) S. Rathbeger, D. B. de Toledo, E. Birckner, H. Hoppe, D. A.
€
M. Egbe, Macromolecules 2010, 43, 306–315; (d) S. Gunes, A.
21 C. Wang, H. Dong, W. Hu, Y. Liu, D. Zhu, Chem. Rev. 2012,
112, 2208–2267.
Wild, E. Cevik, A. Pivrikas, U. S. Schubert, D. M. Egbe, Sol.
Energy Mater. Sol. Cells 2010, 94, 484–491.
22 M. Jïrgensen, K. Norrman, S. A. Gevorgyan, T. Tromholt, B.
Andreasen, F. C. Krebs, Adv. Mater. 2012, 24, 580–612.
10 S. Yamaguchi, K. Tamao, J. Chem. Soc. Dalton Trans. 1998,
3693–3702.
11 (a) M. C. Scharber, M. Koppe, J. Gao, F. Cordella, M. A. Loi,
P. Denk, M. Morana, H.-J. Egelhaaf, K. Forberich, G. Dennler, R.
Gaudiana, D. Waller, Z. Zhu, X. Shi, C. J. Brabec, Adv. Mater.
2010, 22, 367–370; (b) J. Ding, N. Song, Z. Li, Chem. Commun.
2010, 46, 8668–8670; (c) J. Ohshita, Macromol. Chem. Phys.
2009, 210, 1360–1370; (d) L. Huo, H.-Y. Chen, J. Hou, T. L.
Chen, Y. Yang, Chem. Commun. 2009, 5570–5572; (e) J. Hou,
H.-Y. Chen, S. Zhang, G. Li, Y. Yang, J. Am. Chem. Soc. 2008,
130, 16144–16145; (f) J. Chen, Y. Cao, Macromol. Rapid Com-
mun. 2007, 28, 1714–1742; (g) L. Liao, L. Dai, A. Smith, M.
Durstock, J. Lu, J. Ding, Y. Tao, Macromolecules 2007, 40,
9406–9412; (h) H. Usta, G. Lu, A. Facchetti, T. J. Marks, J. Am.
Chem. Soc. 2006, 128, 9034–9035.
23 R. Grisorio, G. Allegretta, P. Mastrorilli, G. P. Suranna, Mac-
romolecules 2011, 44, 7977–7986.
24 C. J. Brabec, G. Zerza, G. Cerullo, S. De Silvestri, S. Luzzati,
J. C. Hummelen, S. Sariciftci, Chem. Phys. Lett. 2001, 340,
232–236.
€
25 H. Hintz, H.-J. Egelhaaf, L. Luer, J. Hauch, H. Peisert, T.
Chasse, Chem. Mater. 2011, 23, 145–154.
26 M. Manceau, A. Rivaton, J.-L. Gardette, S. Guillerez, N.
^
Lemaıtre, Polym. Degrad. Stab. 2009, 94, 898–907.
ꢁ
27 M. Manceau, E. Bundgaard, J. E. Carle, O. Hagemann, M.
Helgesen, R. Sïndergaard, M. Jïrgensen, F. C. Krebs, J. Mater.
Chem. 2011, 21, 4132–4141.
12 (a) R. Grisorio, G. Allegretta, G. P. Suranna, P. Mastrorilli, A.
Loiudice, A. Rizzo, M. Mazzeo, G. Gigli, J. Mater. Chem. 2012,
28 E. Bundgaard, O. Hagemann, M. Manceau, M. Jïrgensen, F.
C. Krebs, Macromolecules 2010, 43, 8115–8120.
4872
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2013, 51, 4860–4872