Boosting the photovoltaic thermal stability of fullerene bulk heterojunction solar cells through charge transfer interactions

Article abstract of DOI:10.1039/c7ta06530b

Fullerene-based bulk heterojunction organic solar cells (BHJ-OSCs) represent one of the current state-of-the-art organic solar cells. Nonetheless, most of these devices still suffer from adverse performance degradation due to thermally induced morphology changes of active layers. We herein demonstrate that the photovoltaic performance stability of BHJ-OSCs can be profoundly enhanced with an appositely functionalized 9-fluorenylidene malononitrile. The latter, through charge transfer (CT) interactions with a donor polymer, enables the formation of a frozen 3-dimensional mesh-like donor polymer matrix, which effectively restrains free movement of embedded fullerene molecules and suppresses their otherwise uncontrolled aggregation. 9-Fluorenylidene malononitrile derivatives with multiple CT interaction sites are particularly effective as preservation of a power conversion efficiency of over 90% under severe thermal stress has been accomplished. The generality of this novel strategy has been affirmed with several common donor polymers, manifesting it to be hitherto the most efficient approach to stabilized fullerene-based BHJ-OSCs.

Full text of DOI:10.1039/c7ta06530b

View Article Online
View Journal
Journal of
Materials Chemistry A
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: C. H. Y. Ho, H.
Cao, Y. Lu, T. Lau, S. H. Cheung, H. LI, H. Yin, K. L. Chiu, L. Ma, Y. Cheng, S. W. TSANG, X. Lu, S. So and B.
S. Ong, J. Mater. Chem. A, 2017, DOI: 10.1039/C7TA06530B.
This is an Accepted Manuscript, which has been through the
Volume
4 Number 1 7 January 2016 Pages 1–330
Royal Society of Chemistry peer review process and has been
accepted for publication.
Journal of
Materials Chemistry A
Accepted Manuscripts are published online shortly after
Materials for energy and sustainability
www.rsc.org/MaterialsA
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the
author guidelines.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
standard Terms & Conditions and the ethical guidelines, outlined
in our author and reviewer resource centre, still apply. In no
event shall the Royal Society of Chemistry be held responsible
for any errors or omissions in this Accepted Manuscript or any
consequences arising from the use of any information it contains.
ISSN 2050-7488
PAPER
Kun Chang, Zhaorong Chang et al.
Bubble-template-assisted synthesis of hollow fullerene-like
MoS2 nanocages as a lithium ion battery anode material
rsc.li/materials-a

Please do not adjust margins
Journal of Materials Chemistry A
Page 1 of 9
View Article Online
DOI: 10.1039/C7TA06530B
Journal Name
ARTICLE
Boosting Photovoltaic Thermal Stability of Fullerene Bulk
Heterojunction Solar Cells through Charge Transfer Interactions
Carr Hoi Yi Ho,a,† Huanyang Cao,b,† Yong Lu,b,† Tsz-Ki Lau,c Sin Hang Cheung,a Ho-Wa Li,d Hang Yin,a
Ka Lok Chiu,a Lik-Kuen Ma,a Yuanhang Cheng,d Sai-Wing Tsang,d Xinhui Lu,c Shu Kong So,a,* Beng S.
Ongb,*
Received 00th January 20xx,
Accepted 00th January 20xx
DOI: 10.1039/x0xx00000x
www.rsc.org/
Fullerene-based bulk heterojunction organic solar cells (BHJ-OSCs) represent one of the current state-of-the-art organic
solar cells. Nonetheless, most of these devices still suffer from adverse performance degradation due to thermally induced
morphology changes of active layers. We herein demonstrate that the photovoltaic performance stability of BHJ-OSCs can
be profoundly enhanced with an appositely functionalized 9-fluorenylidene malononitrile. The latter, through charge
transfer (CT) interactions with the donor polymer, enabling formation of a “frozen” 3-dimensional mesh-like donor
polymer matrix, which effectively restrains free movement of embedded fullerene molecules and suppresses their
otherwise uncontrolled aggregation. 9-Fluorenylidene malononitrile derivatives with multiple CT interaction sites are
particularly effective as preservation of power conversion efficiency of over 90% under severe thermal stress has been
accomplished. The generality of this novel strategy has been affirmed with several common donor polymers,
manifestingthis to be hitherto the most efficient approach to stabilized fullerene-based BHJ-OSCs.
fullerene acceptors to aggregate into undesirably large and
discontinuous domains in the active layer.12ꢀ13 Such polymer
donor/fullerene acceptor phase segregation not only leads to
1. Introduction
Bulk heterojunction (BHJ) organic solar cells (OSCs) with
active layers comprising a polymer electron donor and a
fullerene electron acceptor have been a prevalent topic of
intense research.1ꢀ2 One of the primary driving forces stems
from their perceived facile solution processability, which would
enable highꢀthroughput rollꢀtoꢀroll fabrication leading to lowꢀ
cost manufacturing.3ꢀ4 Other potential attributes relate to their
lightweight and structural flexibility characteristics which may
be useful for portable devices and other niche applications.5ꢀ7
Fullerene acceptors have been widely studied in bulk
heterojunction organic solar cells (BHJꢀOSCs) owing to their
reasonably high electron mobility, solution processability and
compatibility with many donor polymer materials.8ꢀ10 They
have also provided excellent photovoltaic properties with stateꢀ
ofꢀtheꢀart donor polymers. However, most of the fullerene
based BHJꢀOSCs suffer from rapid performance degradation
over time, particularly upon exposure to elevated
temperatures.11ꢀ14 This is largely due to the propensity of
reduced donorꢀacceptor (DꢀA) interfaces with detrimental
effects to exciton separation and diffusion, but also disruption
of charge conduction pathways. Accordingly, the morphology
of active layers plays a decisive role in impacting the
photovoltaic performance of BHJꢀOSCs.14ꢀ21 While many
strategies including solvent annealing, incorporation of ternary
components, crossꢀlinkable fullerene derivatives, etc. have been
engaged to optimize and stabilize active layer morphology with
varying degrees of success, it remains extremely challenging to
sustain thermal stability without sacrificing the photovoltaic
performance.22ꢀ25 A favorable active layer morphology, besides
being compositionally graded in vertical direction for charge
collection and having favorable molecular orientation for
charge transport, should possess a continuous network of
nanodomains of semiconducting materials for efficient exciton
dissociation and charge transport.26ꢀ30
We herein report our structural engineering strategy to
enhance the thermal stability of active layer morphology,
thereby helping preserve and sustain photovoltaic performance
of fullereneꢀbased BHJꢀOSCs under greatly exaggerated
thermal exposure. Our strategy involves “freezing” of donor
polymer matrix through charge transfer (CT) interactions with
an appropriate electronꢀdeficient compound such as 9ꢀ
fluorenylidene malononitrile derivative, leading to formation of
a
presumably threeꢀdimensional (3D) “meshꢀlike” donor
polymer matrix network in the active layer. When uniformly
dispersed in this meshꢀlike polymer network structure, fullerene
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 1
Please do not adjust margins

Please do not adjust margins
Journal of Materials Chemistry A
Page 2 of 9
View Article Online
DOI: 10.1039/C7TA06530B
ARTICLE
Journal Name
acceptor molecules can be visualized as being captured within
the network “cavities”, capable of forming nanodomains but are
substantially restrained in movement from aggregating to
undesirably large domains even under prolonged thermal
exposure. This would enable stable and continuous fullerene
nanodomain structures for efficient exciton dissociation and
transport, thus sustained photovoltaic performance over
prolonged thermal exposure. The beneficial effects of such a
3D meshꢀlike polymer matrix in sustaining thermal
photovoltaic performance stability have been unequivovally
supported by the experimental results of charge transport, trap
density, and morphological studies of active layer in the
presence of an appropriate 9ꢀfluorenylidene malononitrile
derivative.
2. Results and discussion
Through systematic studies on the electronic interactions of
electronꢀdeficient compounds with electron donor polymers, we
observed that certain 9ꢀfluorenylidene malononitrile derivatives
Figure 1 Chemical structures of 9-fluorenylidene malononitrile derivatives.
manipulated through substitution on the fluorenylidene
malononitrile nucleus with electron withdrawing functionalities
such as carbonyl, carboxylic, nitro groups, etc. to achieve the
desirable electron affinity. Its physical properties such as
polymer compatibility and solubility characteristics to facilitate
solution processing can also be modulated with appropriate
alkyl or carboxylic ester substitution. Accordingly, 9ꢀ
fluorenylidene malononitrile derivatives are attractive
electronic agents which can be employed to interact with the
electron donorꢀacceptor (DꢀA) polymers commonly utilized as
donor polymers in BHJꢀOSCs. Specifically, with a reasonably
(
Figure 1) were capable of interacting with donor polymers
through CT interactions. Strong CT interactions of this nature
may be expected to lead to physical “crosslinking” of donor
polymer, thus
a suitably functionalized 9ꢀfluorenylidene
malononitrile derivative with a lowꢀlying lowest unoccupied
molecular orbital (LUMO) or high electron affinity may satisfy
as a suitable CT crosslinking or polymer matrix freezing agent.
From this perspective, a dimeric or trimeric acceptor containing
9ꢀfluorenylidene malononitrile moieties would be particularly
efficient. These materials, which carry multiple active sites for
CT interactions within a single molecule, would enable
formation of a 3D meshꢀlike polymer network structure. This
“cavities” containing network would impede the movements of
embedded fullerene molecules, suppressing their diffusion and
otherwise uncontrolled aggregation into relatively large
domains. Our results showed that incorporation of a minute
quantity of an appropriate 9ꢀfluorenylidene malononitrilie
derivative in the active layers of fullereneꢀbased BHJꢀOSCs
effectively hampered fullerene aggregation, resulting in
low lying LUMO,
a 9ꢀfluorenylidene malononitrile can
effectively form CT bonding with the donor moieties of a DꢀA
polymer, leading to a physically crosslinked donor polymer
network. Depending on the mode and efficacy of CT
interactions, the crosslinked donor polymer matrix with varying
degrees of crosslink density can exert modulating effects on the
movements of fullerene molecules. This will have decisive
consequences on the distribution and stability of fullerene
acceptor dispersions, thus the microꢀ and nanostructural
morphology of active layer.
The fluorenylidene malononitrile derivatives that were
utilized in our studies as electronic agents or additives are
depicted in Scheme S1. Details of their synthesis and
characterizations are described in the experimental section. We
selected carboxylic ester and nitro substitutions to modulate the
solubility and LUMO of 9ꢀfluorenylidebne malononitrile
chromophore as these offered ease of synthesis and attainment
of desirable LUMOs. Three small molecular 9ꢀfluorenylidene
sustained
photovoltaic
performance
characteristics.
Preservation of photovoltaic performance of over 90% of PCEs
after extended thermal exposure at high temperatures (~ 100 °C
for over 20 hours) had been experimentally accomplished. To
the best of our knowledge, this represents hitherto best
approach to enhancing the thermal stability of fullereneꢀbased
BHJꢀOSCs.
2.1 Molecular Design of 9-Fluorenylidene Malononitrile
Additives
As CT interaction involves electron transfer from the malononitrile derivatives, PCFM (I), 1-NPCFM (II), and 2-
highest occupied molecular orbital (HOMO) of an electron NPCFM (III), differing structurally in the number of nitro
donor to the LUMO of an electron acceptor, a higher HOMO substitutions, and thus their LUMO levels in the descending
level and a lower LUMO level would generally favor CT order of ꢀ3.87 eV, ꢀ4.13 eV, and ꢀ4.32 eV, respectively from
interactions. The electron affinity of an electron acceptor, vacuum were investigated. Since crosslinking of a polymer
which measures its capacity to accept an electron, correlates would be far more effective with a divalent or polyvalent agent,
directly with its LUMO level. The LUMO level of
a
we therefore also designed appropriate fluorenylidene
fluorenylidene malononitrile acceptor can be readily malononitrile dimer and trimer for our studies.
2 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Please do not adjust margins
Journal of Materials Chemistry A
Page 3 of 9
View Article Online
DOI: 10.1039/C7TA06530B
Journal Name
ARTICLE
Figure 2 Photovoltaic thermal stability of PTB7:PC71BM BHJ-OSCs with different 9-fluorenylidene malononitrile derivatives. (a-d) J-V characteristics, (e)
summary plots of photovoltaic characteristics
.
2.2 Photovoltaic Performance
7.7% to 5.3%). The latter was due largely to degradation in FF
For our photovoltaic evaluation, we utilized a device (from 62% to 47%) and concomitant decrease in Jsc (by 2
structure of ITO/PEDOT:PSS/active layer/LiF/Al. The active mA/cm2). These experimental observations were in agreement
layer of control device was composed of donor polymer, with our hypothesis that the CT interaction of PCFM with
poly{4,8ꢀbis[(2ꢀethylhexyl)oxy]benzo[1,2ꢀb:4,5ꢀb']dithiopheneꢀ donor polymer, PTB7, led to donor polymer crosslinking. The
2,6ꢀdiylꢀaltꢀ3ꢀfluoroꢀ2ꢀ[(2ꢀethylhexyl)carbonyl]thieno[3,4b]thio crosslinked donor polymer matrix effectively suppressed
pheneꢀ4,6ꢀdiyl} or PTB7 and fullerene acceptor, [6,6]ꢀphenylꢀ thermally induced migration and ensuing aggregation of
C71ꢀbutyric acid methyl ester or PC71BM, while for those of fullerene molecules into undesirably large domains, thus
experimental devices under investigation, less than 1 weight effectively preventing disruption in exciton dissociation and
percent (wt%) of 9ꢀfluorenylidene malononitrile derivative was charge transport. On the other hand, utilization of stronger
incorporated. The devices were thermally annealed at different electron acceptors such as 1ꢀNPCFM (II) or 2ꢀNPCFM (III) as
temperatures for 20 hours each before electrical active layer additives invoked over 20% drop in PCE even
characterization was carried out using AM 1.5G illumination without thermal treatment (Figure S1 and Table S1). The
with a 100 mW/cm2 solar simulator. Figure 2 shows the negative effects of 1ꢀNPCFM and 2ꢀNPCFM could be
currentꢀvoltage (JꢀV) characteristics of control device and 9ꢀ attributed to their much higher electron affinities (or
fluorenyildene malononitrileꢀincorporated devices at different significantly lower LUMOs), which rendered them excellent
annealing temperatures. The device parameters and detailed electron traps, adversely interfering in charge transport as well
photovoltaic performance [short circuit current density (Jsc), as mitigating their CT interactions with donor polymer in
open circuit voltage (Voc), fill factor (FF) and power conversion crosslinking upon electron acceptance.
efficiency (PCE)] are summarized in Table S1. When added in
minor quantity (<1 wt%) into the active layer, PCFM (I) was derivatives, BDFB (IV) and TDFA (V)
The dimeric and trimeric 9ꢀfluorenylidene malononitrile
Table 1), which
(
found to be beneficial in preserving the thermal stability of would greatly enhance CT interactions with donor polymers,
photovoltaic performance of fullerene BHJꢀOSC devices, and were also synthesized for investigation. These materials, which
the optimum quantity was found to be about 0.8% by weight. carried similar 9ꢀfluorenylidene malononitrile moieties within
Specifically, both the asꢀprepared control and PCFMꢀ their molecules, possessed approximately the same LUMO
incorporated devices without thermal annealing exhibited levels as that of PCFM (ca. ꢀ3.9 to ꢀ3.8 eV), and were not
essentially similar photovoltaic performance characteristics. expected to negatively interfere in charge transport in the active
Upon thermal treatment at 100 °C for 20 hours, the PCFMꢀ layer. As prepared without thermal annealing, both the devices
incorporated device showed preservation of over 85% of its with BDFB and TDFA additives exhibited similar performance
original PCE while the control device exhibited severe characteristics as the control device (Jsc ≈ 16 mA/cm2, Voc
performance deterioration with over 30% drop in PCE (from 0.76 V, FF ≈ 64%, PCE ≈ 7.8%). Upon heating at 100 °C for 20
≈
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 3
Please do not adjust margins

Please do not adjust margins
Journal of Materials Chemistry A
Page 4 of 9
View Article Online
DOI: 10.1039/C7TA06530B
ARTICLE
Journal Name
Figure 3 Plots of current-density versus electric field at room temperature of (a) hole-only and (b) electron-only devices with and
without BDFB additive.
Table 1
Hole and electron mobility as measured at room temperature.
hours, both BDFB and TDFAꢀincorporated devices were thermal annealing at 100 °C for 20 hours, the control BHJ film
observed to perform admirably well, preserving over 90% of showed one order of magnitude drop in hole mobility (10ꢀ4 to
their original PCEs. The dimeric fluorenylidene malononitrile 10ꢀ5 cm2Vꢀ1sꢀ1) and nearly two orders of magnitude decrease in
additive, BDFB, was particularly effective as essentially no electron mobility (10ꢀ4 to 10ꢀ6 cm2Vꢀ1sꢀ1). This imbalance of
change in Jsc was noted with its overall PCE degraded only electron and hole mobility would account for the observed
about 7% from its original value. It should be noted that all Jsc’s lower FF of the control device.33 On the other hand, the BDFBꢀ
were in agreement with their external quantum efficiency incorporated BHJ film suffered comparatively smaller
(EQE) measurements (Figure S2).
degradations in both hole (1.2 × 10ꢀ4 to 7.9 × 10ꢀ5 cm2Vꢀ1sꢀ1)
and electron mobility (1.1 × 10ꢀ4 to 6.4 × 10ꢀ5 cm2Vꢀ1sꢀ1)
2.3 Charge Carrier Transport and Subgap Absorption
Properties
(Figure 3 and Table S1), thus manifesting the stabilization
effect of BDFB additive on the BHJ film morphology.
To understand the role and mechanism of 9ꢀfluorenylidene
The defect and trap densities of thermally annealed BHJ
malononitrile in thermal stabilization of photovoltaic films were analyzed by photothermal deflection spectroscopy
performance of fullereneꢀbased BHJꢀOSCs, we accordingly (PDS) to probe the subgap absorptions.34ꢀ37 For simplicity and
conducted more detailed characterization studies on the devices clarity, we showed only the PDS spectra of each film before
with the most efficient dimeric additive, BDFB. The charge and after annealing at 100 °C for 20 hours (Figure 4; see Figure
transport properties of active layers under thermally annealed S3 for detailed spectral properties). For the control films, we
conditions were investigated using spaceꢀchargeꢀlimitedꢀcurrent observed a noticeable increase in the subgap absorption after
(SCLC) model to measure the current densityꢀvoltage (JꢀV) thermal annealing as compared to BDFBꢀincorporated films.
characteristic, and extract zeroꢀfield hole and electron mobility. The subgap absorption was quantified by estimating the
Corresponding singleꢀcarrier diodes were employed for nominal trap state density (
measurements. Electricꢀfield dependent current of the samples rule:38
was measured and carrier mobility was extracted using
N
) of the films using optical sum
ꢒꢓꢔ
following equation:
ꢑ ꢅ
ꢖ ꢗꢘꢙꢚꢛꢜꢝ ꢄꢜ
2ꢕꢐħꢉꢐ
9
ꢀꢁꢂꢃ ꢄ ꢅ ꢆꢇꢆꢈμꢇꢉꢊꢋꢌ0.89ꢍ√ꢎꢏꢎꢐ
8
E
is the photon energy,
c
is the speed of light,
n
is the refractive
index,
charge,
subgap absorption by Urbach absorption. [ꢗꢘꢙꢚ ꢜ ꢅ ꢗ ꢜ ꢞ
m
is the effective mass of electron,
e is the elementary
where JSCL is the spaceꢀchargedꢀlimited current density,
thickness of tested layer, ε0 is the permittivity of vacuum, εr is
the relative permittivity of the polymer, ꢀ0 is the zeroꢀfield
d is the
ħ
is the Planck constant. αsub represents the subtracted
ꢛ ꢝ ꢛ ꢝ
Eg, and Eu is
⁄
ꢗꢙꢛꢜꢝ, where ꢗꢙ ꢅ ꢗꢇ ꢉꢊꢋꢌꢟꢠ ꢞ ꢜꢡ ꢜꢙꢏ for
E <
the Urbach energy.39] The increased trap densities after thermal
stress are presented in Table S3. As can be noted, the BDFBꢀ
incorporated film exhibited one order lower in trap density
increment as compared to control film. These results suggest
that the additive BDFB impeded trap generation during thermal
mobility,
β is the fieldꢀdependent coefficient, and F is the
average electrical field applied on the sample.31ꢀ32
At room temperature, both the control and BDFBꢀ
incorporated active layer (BHJ) films displayed approximately
similar hole and electron mobility. (~1 × 10ꢀ4 cm2Vꢀ1sꢀ1). Upon
4 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Please do not adjust margins
Journal of Materials Chemistry A
Page 5 of 9
View Article Online
DOI: 10.1039/C7TA06530B
Journal Name
ARTICLE
Figure 4 Photothermal deflection spectra of control and BDFB-incorporated BHJ films. The region shaded in light blue indicates the
raise in subgap trap states upon thermal stress.
stress, and the traps were likely due to unfavorable morphology
Transmission electron microscopy (TEM) was also
changes. These observations were also in agreement with the performed to visualize the effects of BDFB incorporation in the
observed stable FF of BDFBꢀincorporated BHJꢀOSC devices as BHJ film. The TEM images for the BHJ films with and without
similar relationship between FF and trap density had earlier BDFB incorporation are shown in Figure 6. The TEMs of both
been reported.37,40ꢀ41
BHJ films with and without BDFB additive show relatively
uniform morphology before heating. After thermal annealing,
substantial fullerene and donor polymer aggregations were
2.4 Active Layer Morphology Analysis
To shed light on the effect of 9ꢀfluorenylidene malononitrile visibly noted in the film without BDFB (bright and dark
additive on molecular ordering and nanoscale phase separation spots/regions), while the film with BDFB additive retained its
of PC71BM in the BHJ film, grazing incidence wideꢀangle and relatively uniform morphology even after thermal annealing. In
smallꢀangle Xꢀray scattering (GIWAXS and GISAXS) addition, a continuous meshꢀlike network structure of donor
measurements were conducted.42ꢀ44 Twoꢀdimensional (2D) polymer matrix was visually evident in the BHJ film with
GIWAXS patterns of BHJ films, with and without BDFB BDFB additive, which was significantly more pronounced after
additive, thermally annealed at different temperatures are heating. These observations correlated well with the Xꢀray
presented in Figure S4. For all these films, ringꢀlike lamellar scattering results, and provided unequivocal affirmation on the
and pꢀp stacking peaks were observed at approximately the formation of “frozen” 3D meshꢀlike network structure of donor
same position (
q
≈ 0.34 Åꢀ1 and 1.75 Åꢀ1), indicating that polymer matrix in the BDFBꢀincorporated BHJ films.
addition of BDFB did not exert any observable effects on the
molecular packing in the BHJ films. In sharp contrast, the 2.5 Charge Transfer Interactions
influence of BDFB on nanoscale phase separation was
The combined results of charge transport studies and
distinctly noted in GISAXS measurements. Figure 5(a) shows structural morphology analysis of active layer compositions
the 2D GISAXS patterns with the corresponding intensity with and without 9ꢀfluorenylidene malononitrile additive such
integrals along qr direction. The intensity integral [Figure 5(b)
of the control film (without BDFB) shifted to lower q in the malononitrile additive as a capable CT crosslinking agent for
medium q region ( ~ 0.01–0.05 Åꢀ1) and turned up in the low
donor polymer such as PTB7 through CT interactions. The
]
as BDFM unequivocally point to the role of 9ꢀfluorenylidene
q
region (< 0.006 Åꢀ1) after thermal annealing (100 °C for 20 hr), resulted crosslinked donor polymer matrix effectively
demonstrating severe morphology instability of the control film. restrained thermallyꢀinduced fullerene migration and impeded
For the film with BDFB, the intensity integral displayed its microscale aggregation, leading to enhanced stability of BHJ
significantly less changes under similar conditions, reflecting nanophase morphology in the active layer. To substantiate the
the efficacy of BDFB in freezing the film morphology. The existence of CT interaction between BDFM and PTB7, we
estimated domain sizes of aggregated PC71BM clusters conducted spectroscopic and steadyꢀstate photoluminescence
(Rg_PC71BM) from the fitting results, and the amorphous studies of PTB7 films with and without BDFM. Figure 7(a)
intermixing phase are summarized in Table S2. The percentage shows the UVꢀvis absorption spectra of the PTB7 films with
changes in domain size were presented in Figure 5(c). As can and without 0.8 wt% of BDFM. In the presence of BDFM
be noted, the domain size of control film was more than double additive, slight redshifts of vibrational absorption peaks at 608
for both phases at annealing temperature of 100 °C, while for nm (0ꢀ1) and 660 nm (0ꢀ0) to 625 nm and 675 nm respectively,
the BDFBꢀincorporated film, the domain size increments were together with a concomitant increase in (0ꢀ0) vibrational peak
less than about 40% at the same annealing temperature.
intensity at the expense of (0ꢀ1) peak were observed. These
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 5
Please do not adjust margins

Please do not adjust margins
Journal of Materials Chemistry A
Page 6 of 9
View Article Online
DOI: 10.1039/C7TA06530B
ARTICLE
Journal Name
Figure 6
TEM images of control and BDFB-incorporated BHJ films
before and after heating at 100°C for ~20 hours.
van der Waals force are too weak to pull together or physically
crosslink the donor polymers. The CT interactions are much
stronger, much like the hydrogen bonding which has been
shown to crosslink fluorineꢀcontaining donor polymers.25
A schematic diagram portraying CT interactions leading to
a crosslinked donor matrix is depicted in Figure 8. Such a 3D
meshꢀlike donor polymer matrix would effectively impede free
movements of fullerene molecules, preventing them from
forming undesirably large domains even subjected to external
thermal stimulus. This would result in a more stable charge
transport environment with reduced trap density, and better
photovoltaic performance.
To demonstrate the generality of the present approach of
utilizing 9ꢀfluorenylidene malononitrile as an effective BHJ
morphology stabilizer, two other different donor polymers,
namely
poly[4,8ꢀbis(5ꢀ(2ꢀethylhexyl)thiophenꢀ2ꢀyl)benzo[1,
2ꢀb;4,5ꢀb’]ꢀdithiopheneꢀ2,6ꢀdiylꢀaltꢀ(4ꢀ(2ꢀethylhexyl)ꢀ3ꢀfluoro
thieno[3,4ꢀb]thiopheneꢀ)ꢀ2ꢀcarboxylateꢀ2ꢀ6ꢀdiyl)] or PTB7ꢀTh,
and regioregular poly(nꢀhexythiophene) or P3HT, were utilized.
Figure S8 summarizes the thermal stability of photovoltaic
properties of BHJꢀOSC devices using PTB7ꢀTh:PC71BM and
P3HT:PC61BM active layers with and without BDFM additive.
The corresponding JꢀV curves of devices are presented in
Figure S9. With PTB7ꢀTh as the donor polymer, the device
with BDFM additive maintained very good photovoltaic
performance after heating at 100 °C for 20 hours, showing
excellent thermal stability as compared to the device without
BDFM additive. Likewise, the P3HT device with BDFM
additive also demonstrated excellent thermal stability in its
photovoltaic performance after annealing at 70°C. The
corresponding P3HT device without BDFM additive exhibited
a significant drop in PCE under similar thermal annealing
conditions, due largely to a significant loss in FF. It should be
Figure 5 (a) 2D GISAXS patterns of PTB7:PC71BM BHJ films with
and without BDFB additive before and after heating at 100 °C for
~20 hours; (b) corresponding in-plane linecut and best fittings; and
(c) percentage changes in domain size as estimated from GISAXS.
results clearly manifested higher degrees of molecular ordering
in the BDFMꢀincorporated film.45ꢀ46 Moreover, steadyꢀstate
photoluminescence (PL) studies showed that BDFM effectively
quenched the photoluminescence of PTB7 [Figure 7(b)].
Specifically, incorporation of 0.8% by weight of BDFM into
the PTB7 film led to
a
twoꢀthird decrease in its
photoluminescence intensity. It was thus reasonable to assume
that the free movement of fullerene was restricted by the
resulting 3D meshꢀlike donor polymer matrix structure formed
from the CT interactions of BDFB and the donor polymer upon
BDFB addition. Other weak intermolecular interactions such as
6 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Please do not adjust margins
Journal of Materials Chemistry A
Page 7 of 9
View Article Online
DOI: 10.1039/C7TA06530B
Journal Name
ARTICLE
Figure 8 Schematic diagram of PTB7:PC71BM active layer with
BDFB additive showing CT-induced crosslinked donor polymer
matrix with nanophase PC71BM morphology.
impediment to thermallyꢀinduced diffusion and aggregation of
embedded fullerene molecules in the crosslinked donor
polymer matrix, thus imparting significantly enhanced thermal
stability to the BHJ morphology of the active layer. A suitably
low LUMO level of close to ꢀ4 eV would be required for CT
interactions without adversely impacting charge transport
properties; much lower LUMO levels leads to electron
trapping, lower charge transport rates, and thus lower PCEs. 4ꢀ
Alkoxycarbonylꢀ9ꢀfluorenylidene malononitrile derivative,
which has a LUMO level of about ꢀ3.9 eV, represents an
effective agent for the stabilization of fullereneꢀbased BHJꢀ
OSCs with common donor polymers such as PTB7, PTB7ꢀTh
or P3HT. The dimeric 9ꢀfluorenylidene malononitrile
derivative, BDFM, which carries two electronꢀdeficient sites
for CT interaction within a single molecule, is particularly
effective. As much as over 90% of PCE could be preserved
even after severe thermal stress of exposure to 100 °C for 20
hours. To the best of our knowledge, this is hitherto most
effective strategy to enhance thermal stability of fullereneꢀ
based OSCs.
Figure 7 (a) Normalized UV-vis spectra and (b) steady-state
photoluminescence spectra of thin films of PTB7 and PTB7 with 0.8
wt% of BDFB additive (excitation wavelength = 470 nm).
mentioned that in the case of P3HT, 70°C was used as the
annealing temperature as these devices could not withstand
100°C for 20 hours. These results have once again affirmatively
demonstrated the general utility and efficacy of this strategy for
the stabilization of photovoltaic performance of fullereneꢀbased
BHJꢀOSCs.
Acknowledgements
This work was financially supported by
Development Fund (SDF13ꢀ0531ꢀA02) and
a
a
Strategic
Research
Committee grant (RCꢀICRS/15ꢀ16/4AꢀSSK) of Hong Kong
Baptist University, a CRF (C5015ꢀ15GF) and GRF grants
(HKBU211913) from Hong Kong Research Grants Council, a
Shenzhen Science and Technology Innovation Grant
(KQTD20140630110339343), and Hung Hin Shiu Charitable
Foundation.
3. Conclusions
Through molecular design, charge transport, and
spectroscopic studies, together with grazing incidence wideꢀ
angle and smallꢀangle Xꢀray scattering analysis, we have
unequivocally demonstrated that an ultraꢀlow concentration of
appositely functionalized 9ꢀfluorenylidene malononitrile
derivative is extremely efficient in enhancing the thermal
stability of fullereneꢀbased BHJꢀOSCs. The stabilization stems
from the CT interactions of 9ꢀfluorenylidene malononitrile with
the donor polymer in locking up the donor polymer chains,
leading to formation of a frozen 3D meshꢀlike donor polymer
matrix in the active layer. This has provided effective
Author contributions
Device fabrication/characterization, photothermal deflection
spectroscopic and charge transport studies were carried out by
C. H. Y. Ho, H. Yin, K. L. Chiu, L. K. Ma, S. H. Cheung.
Molecular design and synthesis were conducted by H. Cao and
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 7
Please do not adjust margins

Please do not adjust margins
Journal of Materials Chemistry A
Page 8 of 9
View Article Online
DOI: 10.1039/C7TA06530B
ARTICLE
Journal Name
24 B. C. Schroeder, Z. Li, M. A. Brady, G. C. Faria, R. S. Ashraf, C.
J. Takacs, J. S. Cowart, D. T. Duong, K. H. Chiu, C.-H. Tan, J. T.
Cabral, A. Salleo, M. L. Chabinyc, J. R. Durrant, I. McCulloch,
Angew. Chem. Int. Ed., 2014, 53, 12870-12875.
25 P. Cheng, C. Yan, T.-K. Lau, J. Mai, X. Lu, X. Zhan, Adv. Mater.,
2016, 28, 5822-5829.
26 M. Jørgensen, K. Norrman, F. C. Krebs, Sol. Energ. Mat. Sol.,
2008, 92, 686-714.
27 R. Pacios, A.ꢀJ. Chatten, K. Kawano, J.ꢀR. Durrant, D.ꢀD.ꢀC.
Bradley, J. Nelson, Adv. Funct. Mater., 2006, 16, 2117-2126.
28 I. A. Howard, R. Mauer, M. Meister, F. Laquai, J. Am. Chem.
Soc., 2010, 132, 14866-14876.
Y. Lu. T.ꢀK. Lau and X. Lu conducted grazing incidence Xꢀray
scattering experiments and morphology analysis. External
quantum efficiency and steadyꢀstate photoluminescence studies
were conducted by H.ꢀW. Li, Y. Cheng and S.ꢀW. Tsang. B. S.
Ong and S. K. So were responsible for the overall project
direction. The manuscript was prepared with inputs from all
members of the team.
References
29 Z. Xiao, Y. Yuan, B. Yang, J. VanDerslice, J. Chen, O. Dyck, G.
Duscher, J. Huang, Adv. Mater., 2014, 26, 3068-3075.
30 B. Meng, Z. Wang, W. Ma, Z. Xie, J. Liu, L. Wang, Adv. Funct.
Mater., 2016, 26, 226-232.
31 M. A. Lampert, P. Mark, Current Injection in Solids, Academic,
New York, 1970.
32 P. N. Murgatroyd , J. Phys. D., 1970 , 3, 151-156.
33 C. H. Y. Ho, Q. Dong, H. Yin, W. W. K. Leung, Q. Yang, H. K. H.
Lee, S. W. Tsang, S. K. So, Adv. Mater. Interfaces, 2015, 2,
1500166.
34 W. B. Jackson, N. M. Amer, A. C. Boccara, D. Fournier, Appl.
Opt., 1981, 20, 1333-1344.
1
L. Dou, J. You, Z. Hong, Z. Xu, G. Li, R. A. Street, Y. Yang, Adv.
Mater., 2013, 25, 6642–6671.
A. J. Heeger, Adv. Mater., 2014, 26, 10-28.
F. C. Krebs, S. A. Gevorgyana, J. Alstrupa, J. Mater. Chem.,
2009, 19, 5442-5451.
R. Søndergaard, M. Hösel, D. Angmo, T. T. Larsen-Olsen, F. C.
Krebs, Materials Today, 2012, 15, 36-49.
H. Kang, G. Kim, J. Kim, S. Kwon, H. Kim, K. Lee, Adv. Mater.,
2016, 28, 7821-7861.
Q. Lin, H. Huang, Y. Jing, H. Fu, P. Chang, D. Li, Y. Yao, Z. Fan,
2
3
4
5
6
7
J. Mater. Chem. C, 2014,
M. Kaltenbrunner, M. S. White, E. D. Głowacki, T. Sekitani, T.
2, 1233-1247.
35 M. H. Chan, S. K. So, K. W. Cheah, J. App. Phy., 1996, 79
3273-3275.
,
Someya, N. S. Sariciftci, S. Bauer, Nat. Comm., 2012, 3, 770-
776.
36 M. Kuik, J. Vandenbergh, L. Goris, E. J. Begemann, L. Lutsen,
D. J. M. Vanderzande, J. V. Manca, P. W. M. Blom, Appl. Phys.
Lett., 2011, 99, 183305.
37 C. H. Y. Ho, S. H. Cheung, H.-W. Li, K. L. Chiu, Y. Cheng, H. Yin,
M. H. Chan, F. So, S.-W.n Tsang, S. K. So, Adv. Energy. Mater.,
2017, 201602360.
38 M. H. Brodsky, Manuel Cardona, J. J. Cuomo, Phys. Rev. B,
1977, 16, 3556-3571.
39 F. Urbach, Phys. Rev., 1953, 92, 1324.
40 M. M. Mandoc, F. B. Kooistra, J. C. Hummelen, B. de Boer, P.
W. M. Blom, Appl. Phys. Lett., 2007, 91, 263505.
41 H. Yan, J. G. Manion, M. Yuan, F. P. García de Arquer, G. R.
McKeown, S. Beaupré, M. Leclerc, E. H. Sargent, D. S.
Seferos, Adv. Mater., 2016, 28, 6491-6496.
8
9
J. Zhao, A. Swinnen, G. V. Assche, J. Manca, D. Vanderzande,
B. V. Mele, J. Phys. Chem. B, 2009, 113, 1587–1591.
S. Berthoa, I. Haeldermansa, A. Swinnena, W. Moonsa, T.
Martensa, L. Lutsenb, D. Vanderzandea, J. Mancaa, A.
Senesc, A. Bonfiglioc, Sol. Energ. Mat. Sol., 2007, 91, 385-
389.
10 J. U. Lee, J. W. Jung, J. W. Jo, W. H. Jo, J. Mater. Chem., 2012,
22, 24265-24283.
11 S. B. Dkhil, M. Pfannmöller, M. I. Saba, M. Gaceur, H. Heidari,
C. Videlot-Ackermann, O. Margeat, A. Guerrero, J. Bisquert,
G. Garcia-Belmonte, A. Mattoni, S. Bals, J. Ackermann, Adv.
Energy Mater., 2017, 4, 1601486.
12 L. Derue, O. Dautel, A. Tournebize , M. Drees, H. Pan, S.
Berthumeyrie, B. Pavageau, E. Cloutet, S. Chambon, L.
Hirsch, A. Rivaton, P. Hudhomme, A. Facchetti, G. Wantz,
Adv. Mater., 2014, 26, 5831-5838.
13 O. Oklobia, T.S. Shafai, Sol. Energ. Mat. Sol., 2013, 117, 1-8.
14 M. Campoy-Quiles, T. Ferenczi, T. Agostinelli, P. G. Etchegoin,
Y. Kim, T. D. Anthopoulos, P. N. Stavrinou, D. D. C. Bradley, J.
42 J. Mai, T.-K. Lau, J. Li, S.-H. Peng§, C.-S. Hsu, U-S. Jeng, J.
Zeng, N. Zhao, X. Xiao, X. Lu, Chem. Mater., 2016, 28, 6186–
6195.
43 J. Mai, T.-K. Lau, T. Xiao, C.-J. Su, U-S. Jeng, N. Zhao, X. Xiao,
X. Lu, RSC Adv., 2015, 5, 88500-88507.
44 H.-C. Liao, C.-S. Tsao, Y.-T. Shao, S.-Y. Chang, Y.-C. Huang, C.-
M. Chuang, T.-H. Lin, C.-Y. Chen, C.-J. Su, U-S. Jeng, Y.-F.
Nelson, Nat. Mater., 2008, 7, 158-164.
15 J.ꢀK.ꢀJ. vanꢀDuren, X. Yang, J. Loos, C.ꢀW.ꢀT. Bulle-Lieuwma,
A.ꢀB. Sieval, J.ꢀC. Hummelen, R.ꢀA.ꢀJ. Janssen, Adv. Funct.
Mater., 2004, 14, 425-434.
16 H. Hoppe, N. S. Sariciftci, J. Mater. Chem., 2006, 16, 45-61.
17 Y. Liu, J. Zhao, Z. Li, C. Mu, W. Ma, H. Hu, K. Jiang, H. Lin, H.
Ade, H. Yan, Nat. Comm., 2014, 5, 5293–5300.
Chen, W.-F. Su, Energy Environ. Sci., 2013,
45 Y. Lei, P. Deng, J. Li, M. Lin, F. Zhu, T.-W. Ng, C.-S. Lee, B. S.
Ong, Sci. Rep., 2016, , 24476.
6, 1938-1948.
6
46 Y. Lei, P. Deng, M. Lin, X. Zheng, F. Zhu, B. S. Ong, Adv.
Mater., 2016, 28, 6687-6694.
18 I. A. Howard, R. Mauer, M. Meister, F. Laquai, J. Am. Chem.
Soc., 2010, 132, 14866-14876.
19 Z. Xiao, Y. Yuan, B. Yang, J. VanDerslice, J. Chen, O. Dyck, G.
Duscher, J. Huang, Adv. Mater., 2014, 26, 3068-3075.
20 G. Li, Y. Yao, H. Yang, V. Shrotriya, G. Yang, Y. Yang, Adv.
Funct. Mater., 2007, 17, 1636-1644.
21 J. K. Lee, W. L. Ma, C. J. Brabec, J. Yuen, J. S. Moon, J. Y. Kim,
K. Lee, G. C. Bazan, A. J. Heeger, J. Am. Chem. Soc., 2008,
130, 3619–3623.
22 L. Derue, O. Dautel, A. Tournebize , M. Drees, H. Pan, S.
Berthumeyrie, B. Pavageau, E. Cloutet, S. Chambon, L.
Hirsch, A. Rivaton, P. Hudhomme, A. Facchetti, G. Wantz,
Adv. Mater., 2014, 26, 5831-5838.
23 Y.-J. Cheng, C.-H. Hsieh, P.-J. Li, C.-S. Hsu, Adv. Funct. Mater.,
2011, 21, 1723-1732.
8 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 9 of 9
Journal of Materials Chemistry A
View Article Online
DOI: 10.1039/C7TA06530B
Boosting Photovoltaic Thermal Stability of Fullerene Bulk Heterojunction Solar Cells
through Charge Transfer Interactions
Carr Hoi Yi Ho, Huanyang Cao, Yong Lu, Tsz-Ki Lau, Sin Hang Cheung, Ho-Wa Li,
Hang Yin, Ka Lok Chiu, Lik-Kuen Ma, Yuanhang Cheng, Sai-Wing Tsang, Xinhui
Lu, Shu Kong So and Beng S. Ong
Table of contents:
Charge transfer interaction of donor polymer with an ultra-low concentration of an
appositely functionalized 9-fluorenylidene malononitrile in the active layer leads to
profoundly enhanced thermal stability of fullerene-based bulk heterojunction organic
solar cells. This is attributable to the formation of a “frozen” 3-dimensional mesh-like
donor polymer matrix which impedes fullerene aggregation.