The Importance of End Groups for Solution-Processed Small-Molecule Bulk-Heterojunction Photovoltaic Cells

Article abstract of DOI:10.1002/cssc.201501626

End groups in small-molecule photovoltaic materials are important owing to their strong influence on molecular stability, solubility, energy levels, and aggregation behaviors. In this work, a series of donor-acceptor pentads (D₂-A-D₁-A-D₂) were designed and synthesized, aiming to investigate the effect of the end groups on the materials properties and photovoltaic device performance. These molecules share identical central A-D₁-A triads (with benzodithiophene as D₁ and 6-carbonyl-thieno[3,4-b]thiophene as A), but with various D₂ end groups composed of alkyl-substituted thiophene (T), thieno[3,2-b]thiophene (TT), and 2,2′-bithiophene (BT). The results indicate a relationship between conjugated segment/alkyl chain length of the end groups and the photovoltaic performance, which contributes to the evolving molecular design principles for high efficiency organic solar cells.

Full text of DOI:10.1002/cssc.201501626

DOI: 10.1002/cssc.201501626
Full Papers
The Importance of End Groups for Solution-Processed
Small-Molecule Bulk-Heterojunction Photovoltaic Cells
Ruomeng Duan,[a] Yong Cui,[b] Yanfei Zhao,[a] Chen Li,[a] Long Chen,[a] Jianhui Hou,[b]
Manfred Wagner,[a] Martin Baumgarten,[a] Chang He,*[b] and Klaus Müllen*[a]
End groups in small-molecule photovoltaic materials are im-
portant owing to their strong influence on molecular stability,
solubility, energy levels, and aggregation behaviors. In this
work, a series of donor–acceptor pentads (D2–A–D1–A–D2)
were designed and synthesized, aiming to investigate the
effect of the end groups on the materials properties and pho-
tovoltaic device performance. These molecules share identical
central A–D1–A triads (with benzodithiophene as D1 and 6-car-
bonyl-thieno[3,4-b]thiophene as A), but with various D2 end
groups composed of alkyl-substituted thiophene (T),
thieno[3,2-b]thiophene (TT), and 2,2’-bithiophene (BT). The re-
sults indicate a relationship between conjugated segment/alkyl
chain length of the end groups and the photovoltaic per-
formance, which contributes to the evolving molecular design
principles for high efficiency organic solar cells.
Introduction
Conjugated small molecules can be precisely purified and char-
acterized by standard laboratory techniques with high batch-
to-batch reproducibility, which is advantageous in large-scale
application of organic photovoltaics (OPV) compared with con-
jugated polymers.[1] Small molecules as donor materials in bulk
heterojunction (BHJ) devices should meet several require-
ments: they must (1) have sufficient environmental stability,
(2) harvest as many photons as possible, (3) have appropriate
energy levels, and (4) possess sufficiently high charge-carrier
mobility. Intensive investigations on polymeric donors give
some valuable guidelines to achieve these goals with small
molecules. For example: (1) the donor–acceptor (D–A) alternat-
ing structure is an effective way to decrease the energy-gap;[2]
(2) 2D conjugated structures are promising for broad absorp-
tion and high charge-carrier mobility values;[3] (3) extending
backbones is favorable for improving film quality,[4] and
(4) branched alkyl side chains are more effective in improving
solubility than straight carbon chains, although the latter has
the beneficial effect of improving close molecular packing.[5]
Following these guidelines, many excellent conjugated small
molecules were developed,[6] leading to power conversion effi-
ciency (PCE) values of up to 10.1%.[7]
In comparison with high molecular weight polymers, end
groups in oligomeric small molecules naturally play a much
more important role.[8] They not only determine the stability of
molecules, but also affect the energy-level properties and the
self-assembly behaviors, which are all related directly to device
performance. Therefore, we designed and synthesized a series
of conjugated small molecules with a common central unit
and different end groups. They are namely M1, M2, M3-C8,
M3-C6, and M3-C4, as shown in Figure 1. Their identical cen-
tral core is an acceptor–donor–acceptor (A–D1–A) alternating
structure, with a 2D benzodithiophene (2D-BDT) as D1 and
(thieno[3,4-b]thiophene-2-yl)-2-ethylhexan-1-one (TT-C) as A.
On both ends of the central A–D1–A structure, varied end
groups were installed. A systematic study of end groups was
carried out to explore the relationship between the molecular
properties and end groups. Upon going from M1 to M2 to
M3-C8, the end groups are thiophene (T), thieno[3,2-b]thio-
phene (TT), and 2,2’-bithiophene (BT), all decorated with
n-octyl chains on the terminal site. The comparison of M1, M2,
and M3-C8 should reveal the effect of the conjugated segment
length of the end groups. Next, to gain understanding of the
effect of the length of terminal alkyl chains, we considered the
series M3-C8, M3-C6 up to M3-C4, all with identical conjugat-
ed pentads, but varied alkyl chains of n-octyl, n-hexyl, and
n-butyl, respectively. From the results of photovoltaic studies
of these five compounds, we found that the aromatic unit
length of the D2 end group can strongly influence the thermal
stability, energy levels, and molecular packing, with measured
PCEs ranging from 1.1% for M1 to 3.7% for M2 to 5.2% for
M3-C8. From the M3 series, we see no effect of terminal alkyl
chain length on thermal stability or the energy levels, but
minor difference on phase separation of the active layer contri-
buting to an increase in the PCE to 5.5% for M3-C6 and 6.0%
for M3-C4.
[a] R. Duan, Y. Zhao, C. Li, L. Chen, M. Wagner, M. Baumgarten,
Prof. Dr. K. Müllen
Max Planck Institute for Polymer Research
Ackermannweg 10, 55128 Mainz (Germany)
E-mail: muellen@mpip-mainz.de
[b] Y. Cui, J. Hou, Dr. C. He
State Key Laboratory of Polymer Physic and Chemistry
Institute of Chemistry
Chinese Academy of Science
Zhongguancun North First Street 2, 100190 Beijing (P.R. China)
E-mail: hechang@iccas.ac.cn
Supporting Information for this article can be found under http://
dx.doi.org/10.1002/cssc.201501626.
ChemSusChem 2016, 9, 973 – 980
973
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Figure 1. Molecular structures of M1, M2, M3-C8, M3-C6, and M3-C4.
constitutionally isomeric A–D1–A triad cores 10, 11, and 12
(Scheme 1).[1b,12] Fortunately, a single isomer (10) could be iso-
lated in good yield, as confirmed by 1H NMR and 1H–1H 2D
NOESY-NMR spectroscopy (Figure 2). The structure was con-
firmed as follows: First, the 1H NMR spectrum contains only
four aromatic proton signals (Figure S2 in the Supporting Infor-
mation), which excludes the lower symmetry compound 12
with its six unique aromatic protons. Next, in the 1H–1H 2D
NOESY-NMR spectrum, coupling is observed that indicates the
proximity of H1 to H2 and H3, thus eliminating 11 and con-
firming 10 as the isolated triad. Although the coupling of
2D-BDT (9) and TT-C (8) was widely reported in polymers, here
it is the first applied to prove the construction and symmetry
of the A–D–A oligomer 10.[3,9,10,13]
Results and Discussion
M1, M2, and M3-C8 all contain a 2D-BDT segment as D1, TT-C
as A, and varied T, TT, or BT as D2, respectively. The 2D-BDT
core was selected considering its large conjugation plane and
excellent solubility.[6d,c] TT-C was chosen as it was shown to be
very effective for lowering the energy gap and imparting supe-
rior performance in polymer materials.[10,11] Three end units
based on n-octyl-subsituted thiophene derivatives with weakly
electron donating effects were introduced to extend the conju-
gated backbone. Because the terminal a-proton of the thio-
phene end groups would be a site of instability, n-octyl chains
were implanted to make the molecules stable as well as to im-
prove solubility and modify packing.
The D2–A–D1–A–D2 pentads were synthesized according to
Scheme 1. First, thiophene derivatives 1, 3, or 6 were depro-
tonated with n-butyllithium in ultra-dry THF solution, followed
by quenching with trialkyltin chlorides at 08C, which afforded
products 2, 4, and 7, respectively. The key intermediate 10 was
prepared under Stille coupling conditions from 9 and an
excess of 8 in toluene at 808C with a yield of 60%. Finally the
three target molecules M1, M2, and M3-C8 were synthesized
through hetero-coupling reaction between dibromo 10 and
compounds 2, 4, or 7 using tetrakis(triphenylphosphine)palla-
dium [Pd(PPh3)4] mediated reactions in toluene at 110 8C. Good
yields (60–70%) were obtained after purifying using silica-gel
column chromatography and preparative-gel permeation chro-
matography (GPC). Synthetic details are available in the Sup-
porting Information. The structures and purities of all final tar-
gets were unambiguously characterized with MALDI-TOF mass
spectrometry, 1H and 13C NMR spectroscopy, and elemental
analysis. All molecules show good solubility in common organ-
ic solvents, such as dichloromethane, chloroform, toluene,
chlorobenzene, and dichlorobenzene.
The thermal behavior of materials can be one of the most
important considerations for real applications. The thermal
properties of M1, M2, and M3-C8 were investigated using ther-
mogravimetric analysis (TGA, Figure S3a) and differential scan-
ning calorimetry (DSC, Figure S4). As shown in Figure S3, all
three molecules exhibited good thermal stability up to 3708C
under nitrogen at a heating rate of 108CminÀ1. DSC analysis
for the three molecules revealed phase-transition peaks at 148,
233, and 1628C for M1, M2, and M3-C8, respectively.
Harvesting of photons is the precondition to utilize solar
energy. The UV/Vis absorption spectra of M1, M2, and M3-C8
in chloroform and in thin films are shown in Figure 3. In solu-
tion, the molecules exhibit broad absorption in the visible
region from 350–700 nm with a peak maximum at ~570 nm.
With an increase of the conjugated segment length of the end
group, the peak shifts from 560 nm for M1, to 572 nm for M2,
and to 578 nm for M3-C8. Along with conjugated segment
length increase, extinction coefficients (e) increase from
57000mÀ1 cmÀ1 for M1 to ~78000mÀ1 cmÀ1 for M2 and M3-
C8. Films on quartz plates used for UV/Vis absorption spectros-
copy measurements were prepared by spin coating from
chloroform solutions of M1, M2, and M3-C8 (10 mgmLÀ1). In
As the acceptor building block 8 is asymmetric, its regio
selective Stille coupling with the 9 could lead to three possible
ChemSusChem 2016, 9, 973 – 980
www.chemsuschem.org
974
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Scheme 1. Synthetic routes of M1, M2, and M3-C8.
Figure 2. 1H–1H 2D NOESY-NMR spectroscopy of compound 10.
comparison with the absorption spectra in solution, the ab-
sorption bands of all three compounds are broader and red-
shifted in the solid state. The broadening and red-shifting of
the absorption bands in the films are usually ascribed to the
formation of aggregates or ordered p–p stacking, which are
helpful for improving charge-carrier mobility of the resulting
films. In film, M1 exhibits an absorption maximum at about
656 nm, which is red-shifted by approximately 96 nm com-
pared to its solution spectrum. M2 and M3-C8 in film also dis-
played bathochromic absorption peaks at about 667 and
681 nm, respectively. From the onset of the thin film absorp-
tion, the optical energy gaps were estimated to be 1.7, 1.7,
and 1.6 eV for M1, M2, and M3-C8, respectively. The optical
properties of the molecules are summarized in Table 1. It is
ChemSusChem 2016, 9, 973 – 980
www.chemsuschem.org
975
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
redox couple was measured under the same conditions, and is
located at 0.38 V versus the Ag/Ag+ electrode. If the absolute
energy level of Fc/Fc+ is set at À4.8 eV relative to vacuum, the
energy levels of the HOMO and LUMO could be estimated by
the following equations: EHOMO (eV)=À (foxÀfFc=Fcþ +4.8);
ELUMO (eV)=À(fredÀfFc=Fcþ +4.8), where fox and fred are the ox-
idation and reduction potentials versus Ag/Ag+, respectively.
The HOMO and LUMO energies were estimated to be À5.2
and À3.2 eV for M1, À5.0 and À3.2 eV for M2, and À5.0 and
À3.2 eV for M3-C8, respectively. The changes in end groups
show little effect on the LUMO values for these molecules. The
similarities of LUMO values originate from the identical accept-
or units, whereas varying HOMO energies arise from the differ-
ent electron-donating properties of the end groups. From the
most simplistic viewpoint, extended conjugation leads to
HOMO destabilization (shallower HOMO energies) as was ob-
served here for M2 and M3-C8. The estimated energy levels of
the three molecules and [6,6]-phenyl-C71-butyric acid methyl
ester (PC71BM) are summarized in Figure S5. Here, it was appar-
ent that the LUMO levels of all three donating molecules are
approximately 0.8 eV above that of PC71BM. This value is suffi-
cient to produce exciton dissociation in the interlayer between
donor and PC71BM.
Figure 3. UV/Vis absorption spectra of M1, M2, and M3-C8.
Table 1. Optical and electrochemical data for M1, M2, and M3-C8.
sol
max
film
max
[nm][b]
CV
HOMO
CV
LUMO
Material
l
e
l
Eogpt
E
E
ECgV
[nm][a] [MÀ1 cmÀ1
]
[eV][c] [eV][d] [eV][d] [eV][e]
M1
M2
M3-C8
560
572
578
57000
77000
78000
603; 656 1.7
610; 667 1.7
626; 681 1.6
À5.2 À3.2 2.0
À5.0 À3.2 1.8
À5.0 À3.2 1.8
[a] Measured on 10À5 m solutions in chloroform. [b] Samples were spin
XRD was used to investigate the packing of the molecules in
solid films. Films of M1, M2, and M3-C8 were spin-coated from
chloroform solutions onto silica substrates. As shown in
Figure 5, the pure M1-, M2- and M3-C8-based films displayed
a single sharp reflection peak (100) at low diffraction angles at
2q of 4.68, 4.28, and 4.78, respectively, indicating their strong
molecular packing.[6c]
coated on quartz plates. [c] Estimated from the long wavelength onset of
the thin-film absorption. [d] CV values are referenced to Ag/Ag+. Fc/Fc+
Fc=Fcþ
1=2
was used as an external reference (E
=0.38 V vs. Ag/Ag+, EHOMO
=
À4.80 eV). [e] ECgV =ELUMOÀEHOMO
.
worth noting that a polymer (PBDTTT-C-T) reported by Hou
et al.[9] with donor and acceptor units identical to those in this
work, had a thin-film absorption profile (maxima at 631 and
692 nm and an optical gap of approximately 1.58 eV)[9] nearly
equal to M3-C8. This result indicates that, by adopting larger
conjugated end groups, the absorption of small molecules can
be improved to behave as well as polymeric materials.
Appropriate energy levels are a key factor to realize the con-
version from light-to-electrical energy. The electrochemical
properties of M1, M2, and M3-C8 were investigated by cyclic
voltammetry in solution (CV, Figure 4). For calibration, the
redox potential of the ferrocene/ferrocenium (Fc/Fc+, fFc=Fcþ
)
Figure 5. The XRD profiles of M1, M2, and M3-C8.
To investigate the effect of the end groups on photovoltaic
performance, single layer BHJ photovoltaic cells based on the
three molecules as the donor component and PC71BM as the
acceptor material were fabricated with an ITO/PEDOT:PSS/
MX:PC71BM/Al architecture (see the Supporting Information for
details). The photovoltaic data were collected under AM1.5G
at 100 mWcmÀ2. To obtain improved device performance, the
device fabrication conditions were optimized to D/A ratios of
3:1 and an active layer thickness of 100 nm without annealing
(see the Supporting Information for details). The current–
Figure 4. CV traces of M1, M2, and M3-C8 in CH2Cl2 (1”10À3 m) with 0.1m
Bu4NPF6 as the supporting electrolyte measured at a scan rate of 50 mVsÀ1
.
ChemSusChem 2016, 9, 973 – 980
www.chemsuschem.org
976
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
C8, an increase in the PCEs was demonstrated from M1, to M2
to M3-C8. It is worth noting that the variation in end groups
caused only a slight difference in the molecular properties
(e.g., UV/Vis absorption and electrochemistry as observed by
CV), but remarkable changes in efficiency were observed in the
photovoltaic performance.
Figure 7 shows the IPCE response of the three molecules
based on optimal device conditions. Because of the main con-
tribution of the IPCE originating from the donor absorption,
Figure 6. J–V characteristics of BHJ devices fabricated using M1, M2, and
M3-C8 as the donor and PC71BM as the acceptor measured under AM1.5G
illumination (100 mWcmÀ2) and in the dark. These devices were made with
D/A ratios of 3:1 without processing additives or thermal annealing.
voltage (J–V) curves for the resulting devices are shown in
Figure 6. The optimized PCEs are 1.1, 3.7, and 5.2% for M1,
M2, and M3-C8 (average over 20 devices), respectively. To
make a clear comparison, the corresponding results are listed
in Table 2.
Figure 7. IPCE of the corresponding devices.
Table 2. Photovoltaic properties of BHJ solar cells based on M1, M2, M3-
C8, M3-C6, or M3-C4 with PC71BM.
M2 and M3-C8 show broader and stronger absorption than
M1 in the 300–750 nm range, leading to an increased IPCE re-
sponse in this region. The integral current density is an evalua-
tion of the devices ability to harvest photons and transfer
charges. The integral currents of M1, M2, and M3-C8 are 2.38,
8.95, 10.06 mAcmÀ2, respectively, which is in agreement with
the photovoltaic test results.
½cŠ
[a]
JV [a,b]
SC
[mAcmÀ2
EQE
SC
Donor VOC
J
J
FF[a]
[%]
PCE[a]
[%]
[V]
]
[mAcmÀ2
]
M1
M2
0.87Æ0.01
0.87Æ0.01
2.66Æ0.03
8.60Æ0.13
2.4
8.9
47.4Æ0.50 1.11Æ0.03
50.0Æ0.30 3.70Æ0.10
57.2Æ1.08 5.20Æ0.09
53.3Æ0.40 5.50Æ0.04
55.0Æ0.76 6.00Æ0.08
M3-C8 0.84Æ0.01 10.83Æ0.21 10.1
M3-C6 0.82Æ0.01 12.73Æ0.13 11.7
M3-C4 0.78Æ0.01 13.92Æ0.05 12.9
To gain some insight in the differences in JSC, hole mobility
values were measured using the space-charge limited current
(SCLC) method and found to be 2.79”10À5, 1.13”10À4, and
1.25”10À4 cm2 VÀ1 SÀ1 for M1, M2, and M3-C8, respectively. It
would appear that hole mobility values indeed increase as
a function of the conjugated-segment length of the end
groups, as longer conjugated end groups facilitate to extend
delocalization and increase intermolecular electron hopping.
To further understand the device performance, the morphol-
ogy of three blend layers were investigated by AFM. As shown
in Figure 8, the height images show obvious differences in
morphologies with root mean square (RMS) roughnesses of
1.69, 0.82, and 0.95 nm for the M1, M2, and M3-C8 active
layers, respectively. The M1/PC71BM layer shows the roughest
surface among the three blend layers. From their phase
images, it can be seen that the blend films of M2/PC71BM and
M3-C8/PC71BM show nano-scale phase separation with desira-
ble domain sizes around 20 nm, whereas the blend film of M1/
PC71BM shows less desirable, larger domain sizes around
100 nm. In contrast, the proper phase separation in blended
films of M3-C8/PC71BM is favorable for the production and col-
lection of charges from excitons, leading to a good IPCE re-
sponse, high hole mobility, and large PCE values.
[a] Average value and standard deviation based on the measurement of
20 duplicate devices under simulated AM1.5G illumination
(100 mWcmÀ2). [b] Calculated from the J–V curve. [c] Calculated from the
EQE spectrum.
The open circuit voltages (VOC) were estimated from the J–V
curves to be 0.87, 0.87, and 0.84 V for M1, M2, and M3-C8, re-
spectively. As the VOC can be estimated by the gap between
the LUMO of PC71BM and the HOMO of the donor, the results
also indicated this the trend (i.e., M2 and M3-C8 had lower
HOMOs than that of M1, and thus the photovoltaic devices of
M2 and M3-C8 exhibited higher VOC).[14] The short-circuit cur-
rent densities (JSC) determined from the J–V curves were 2.66,
8.60 and 10.83 mAcmÀ2 for M1, M2, and M3-C8, respectively.
The variation in JSC is consistent with the trend of incident
photon-to-current efficiency (IPCE) and hole mobility values, as
discussed below. In this case, M1 yielded the weakest IPCE and
lowest hole mobility; therefore, M1 also exhibited the lowest
JSC. On the contrary, M3-C8 gave a highest IPCE and hole mo-
bility and it achieved highest JSC in this series. Combined with
a fill factor (FF) of 47% for M1, 50% for M2, and 57% for M3-
ChemSusChem 2016, 9, 973 – 980
www.chemsuschem.org
977
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Figure 8. AFM height images of active layer blends: (a) M1/PC71BM (3:1), (c) M2/PC71BM (3:1), and (e) M3-C8/PC71BM (3:1). AFM phase images of active layer
blends: (b) M1/PC71BM (3:1), (d) M2/PC71BM (3:1), and (f) M3-C8/PC71BM (3:1).
The effect of length of terminal alkyl chains
The photovoltaic performances of M3-C8, M3-C6, and M3-
C4 were investigated in the same fashion as described above.
Their J–V curves are shown in Figure 10 and their PCE data are
collected in Table 2. From the J–V curves, it was observed that
as the terminal alkyl chains became shorter, the VOC decreased
The alkyl chains attached to these molecules also are an im-
portant factor for material performance, as they can influence
the active layer phase separation behavior. To investigate the
effect of length of substituents on the device performance,
two analogues of M3-C8 were prepared with shorter terminal
alkyl chains and denoted M3-C6 and M3-C4 (Figure 1). Their
chemical nomenclature and synthetic details are reported in
the Supporting Information. The conjugated pentad structure
of M3-C8 was chosen as it demonstrated the best performance
in preliminary studies.
The basic physical properties, including UV/Vis, TGA, and CV
spectra are shown in Figure S6. Owing to their identical back-
bone structure, the M3 series show similar thermal stability, as
well as (opto)electronic properties. The optical and electro-
chemical properties of M3 are summarized in Table 1. DSC re-
vealed a regular transition point rise as the terminal alkyl
chains shorten, with peak temperatures at 203, 223, and 2588C
for M3-C8, M3-C6, and M3-C4 respectively (Figure 9). This is
a very clear trend of increasing thermal stability of the room-
temperature phase with decreasing the length of terminal
alkyl chain.
Figure 10. J–V characteristics of BHJ devices fabricated using M3-C8, M3-C6,
and M3-C4 as the donor and PC71BM as the acceptor measured under
AM1.5G illumination (100 mWcmÀ2) and in the dark.
slightly, but the JSC increased markedly. With no apparent
change trend in FF, it seems that the increase in JSC is the main
driving force for the growing PCE values of 5.2, 5.5, and 6.0%
for M3-C8, M3-C6, and M3-C4, respectively. Furthermore, their
IPCE (Figure S7) profiles are nearly identical, but the integral is
increasing as the terminal alkyl chain length decreases, with in-
tegral currents of 10.1, 11.7, and 12.9 mAcmÀ2 for M3-C8, M3-
C6, and M3-C4, respectively (Table 2). In AFM height images of
the M3-based active layers, all show quite smooth surfaces.
From the height image, a RMS of 0.82, 0.71, and 0.63 nm for
M3-C8, M3-C6, and M3-C4 can be estimated, respectively
(Figure 11), which indicates that the M3-C4-based active layer
presents the most favorable phase separation. Therefore, ter-
minal butyl chains are more appropriate than hexyl and octyl
Figure 9. DSC curves of M3-C8, M3-C6, and M3-C4.
ChemSusChem 2016, 9, 973 – 980
www.chemsuschem.org
978
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Figure 11. AFM height images of active layer blends: (a) M3-C8/PC71BM (3:1), (c) M3-C6/PC71BM (3:1), and (e) M3-C4/PC71BM (3:1). AFM phase images of
active layer blends: (b) M3-C8/PC71BM (3:1), (d) M3-C6/PC71BM (3:1), and (f) M3-C4/PC71BM (3:1).
a Bruker Reflex II-TOF Spectrometer using a 337 nm nitrogen laser
with TCNQ as matrix. UV/Vis absorption spectra were collected on
a PerkinElmer Lambda 9 spectrophotometer at room temperature.
CV measurements were carried out on a computer-controlled
GSTAT12 in a three-electrode cell in a dichloromethane solution of
tetrabutylammoniumhexafluorophosphate (Bu4NPF6, 0.1m) with
a scan rate of 50 mVsÀ1 at room temperature. A Pt wire, a silver
wire, and a glassy carbon electrode were used as counter elec-
trode, reference electrode, and working electrode, respectively.
DSC was carried out with a Mettler DSC 30 with a heating rate of
10 KminÀ1 from 25 to 2508C. TGA was carried out on a Mettler 500
thermo gravimetry analyzer. AFM experiments were performed on
a Bruker VeecoMultiMode 8 atomic force microscope. Film XRD
was performed on a theta-theta Philips PW 1820 Kristalloflex dif-
fractometer with a graphite-monochromatized CuKa X-ray beam
(l=1.5418 Š) hitting the thin film. The diffraction patterns were re-
corded in the 2q range from 0 to 308.
chains for these conjugated cores, owing to more efficient mo-
lecular packing and thus more efficient intermolecular charge
transport.
Conclusions
Three novel D2–A–D1–A–D2-type conjugated small molecules
with varying donor (D2) end groups, M1, M2, and M3-C8, were
synthesized and applied in solar cells as the donor materials.
As the D2 end groups were varied in conjugated length from
thiophene to thieno[3,2-b]thiophene to 2,2’-bithiophene, the
average device performance dramatically increased from 1.1 to
3.7 to 5.2%. Focusing on the highest performing D2–A–D1–A–
D2 pentad, we found the terminal alkyl chain length could fur-
ther improve the device performance. When the alkyl chains
were shortened from n-octyl to n-hexyl to n-butyl the PCE in-
creased from 5.2 to 5.5 to 6.0%. This result shows that the ex-
tended conjugated length of the end groups in a small mole-
cule donor can effectively adjust the molecular energy levels
and aggregation behaviors of the active layer, thereby improv-
ing the device performance. Moreover, the optimized alkyl
chain length of the end groups can further improve the PCE
while maintaining molecular thermal stability and energy level
positions. This study should contribute to the growing under-
standing of the subtle demands of designing small molecules
for OPV, where both the nature of the p system and the solu-
bilizing side chains are critical.
Acknowledgements
This work is supported by the DFG priority program SPP1355 and
BASF, National Nature Science Foundation of China (No.
51373181). R.D. gratefully acknowledges the China Scholarship
Council (CSC) for financial support through a scholarship.
Keywords: benzodithiophene · bulk heterojunction · small
molecule · solar cells · thienothiophene
[1] a) Y. S. Chen, X. J. Wan, G. K. Long, Acc. Chem. Res. 2013, 46, 2645–
2655; b) X. Liu, Y. Sun, L. A. Perez, W. Wen, M. F. Toney, A. J. Heeger, G. C.
Bazan, J. Am. Chem. Soc. 2012, 134, 20609–20612; c) R. Fitzner, E.
Mena-Osteritz, A. Mishra, G. Schulz, E. Reinold, M. Weil, C. Korner, H.
Ziehlke, C. Elschner, K. Leo, M. Riede, M. Pfeiffer, C. Uhrich, P. Bauerle, J.
Am. Chem. Soc. 2012, 134, 11064–11067; d) Y. M. Sun, G. C. Welch, W. L.
Leong, C. J. Takacs, G. C. Bazan, A. J. Heeger, Nat. Mater. 2012, 11, 44–
48; e) T. S. van der Poll, J. A. Love, T. Q. Nguyen, G. C. Bazan, Adv. Mater.
2012, 24, 3646–3649; f) J. Y. Zhou, X. J. Wan, Y. S. Liu, Y. Zuo, Z. Li, G. R.
He, G. K. Long, W. Ni, C. X. Li, X. C. Su, Y. S. Chen, J. Am. Chem. Soc.
2012, 134, 16345–16351.
Experimental Section
Materials and methods. All reactions were carried out under
argon atmosphere with the use of standard Schlenk techniques.
2,6-Bis(trimethyltin)-4,8-bis(5-(2-ethylhexyl)thiophen-2-yl)benzo[1,2-
b:4,5-b’]dithiophene (9) was prepared according the reported
method.3c 1-(4,6-Dibromothieno[3,2-c]thiophen-2-yl)-2-ethylhexan-
1-one (8) was purchased from Solarmer Energy, Inc. and used as re-
ceived. Other starting materials were purchased from Alfa-Aesar
and Sigma–Aldrich and used without further purification.
[2] a) E. E. Havinga, W. Tenhoeve, H. Wynberg, Polym. Bull. 1992, 29, 119–
126; b) J. H. Hou, M. H. Park, S. Q. Zhang, Y. Yao, L. M. Chen, J. H. Li, Y.
Yang, Macromolecules 2008, 41, 6012–6018; c) T. S. Qin, W. Zajaczkow-
1H and 13C NMR spectra were recorded on a Bruker DRX 250 or
Bruker DRX 850. MALDI-TOF mass spectrometry was performed on
ChemSusChem 2016, 9, 973 – 980
www.chemsuschem.org
979
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
ski, W. Pisula, M. Baumgarten, M. Chen, M. Gao, G. Wilson, C. D. Easton,
K. Mullen, S. E. Watkins, J. Am. Chem. Soc. 2014, 136, 6049–6055.
[3] a) R. M. Duan, L. Ye, X. Guo, Y. Huang, P. Wang, S. Q. Zhang, J. P. Zhang,
L. J. Huo, J. H. Hou, Macromolecules 2012, 45, 3032–3038; b) L. J. Huo,
L. Ye, Y. Wu, Z. J. Li, X. Guo, M. J. Zhang, S. Q. Zhang, J. H. Hou, Macro-
molecules 2012, 45, 6923–6929; c) L. Ye, S. Q. Zhang, L. J. Huo, M. J.
Zhang, J. H. Hou, Acc. Chem. Res. 2014, 47, 1595–1603.
[4] a) Y. Wu, Z. J. Li, W. Ma, Y. Huang, L. J. Huo, X. Guo, M. J. Zhang, H. Ade,
J. H. Hou, Adv. Mater. 2013, 25, 3449–3455; b) Y. Wu, Z. J. Li, X. Guo,
H. L. Fan, L. J. Huo, J. H. Hou, J. Mater. Chem. 2012, 22, 21362–21365.
[5] a) C. Cabanetos, A. El Labban, J. A. Bartelt, J. D. Douglas, W. R. Mateker,
J. M. J. Frechet, M. D. McGehee, P. M. Beaujuge, J. Am. Chem. Soc. 2013,
135, 4656–4659; b) W. H. Chang, J. Gao, L. T. Dou, C. C. Chen, Y. S. Liu, Y.
Yang, Adv. Energy Mater. 2014, 4, 1300864; c) Y. Y. Liang, D. Q. Feng, Y.
Wu, S. T. Tsai, G. Li, C. Ray, L. P. Yu, J. Am. Chem. Soc. 2009, 131, 7792–
7799.
[6] a) J. Y. Zhou, Y. Zuo, X. J. Wan, G. K. Long, Q. Zhang, W. Ni, Y. S. Liu, Z. Li,
G. R. He, C. X. Li, B. Kan, M. M. Li, Y. S. Chen, J. Am. Chem. Soc. 2013, 135,
8484–8487; b) C. H. Cui, J. Min, C. L. Ho, T. Ameri, P. Yang, J. Z. Zhao,
C. J. Brabec, W. Y. Wong, Chem. Commun. 2013, 49, 4409–4411; c) J. H.
Huang, C. L. Zhan, X. Zhang, Y. Zhao, Z. H. Lu, H. Jia, B. Jiang, J. Ye, S. L.
Zhang, A. L. Tang, Y. Q. Liu, Q. B. Pei, J. N. Yao, ACS Appl. Mater. Interfaces
2013, 5, 2033–2039; d) Y. Z. Lin, L. C. Ma, Y. F. Li, Y. Q. Liu, D. B. Zhu,
X. W. Zhan, Adv. Energy Mater. 2014, 4, 1300626; e) J. J. Peng, Y. N. Chen,
X. H. Wu, Q. Zhang, B. Kan, X. Q. Chen, Y. S. Chen, J. Huang, Z. Q. Liang,
ACS Appl. Mater. Interfaces 2015, 7, 13137–13141; f) G. K. Long, B. Wu,
X. Yang, B. Kan, Y. C. Zhou, L. C. Chen, X. J. Wan, H. L. Zhang, T. C. Sum,
Y. S. Chen, ACS Appl. Mater. Interfaces 2015, 7, 21245–21253; g) B. Kan,
Q. Zhang, M. M. Li, X. J. Wan, W. Ni, G. K. Long, Y. C. Wang, X. Yang, H. R.
Feng, Y. S. Chen, J. Am. Chem. Soc. 2014, 136, 15529–15532.
[7] B. Kan, M. M. Li, Q. Zhang, F. Liu, X. J. Wan, Y. C. Wang, W. Ni, G. K. Long,
X. Yang, H. R. Feng, Y. Zuo, M. T. Zhang, F. Huang, Y. Cao, T. P. Russell,
Y. S. Chen, J. Am. Chem. Soc. 2015, 137, 3886–3893.
[8] a) G. K. Long, X. J. Wan, B. Kan, Y. S. Liu, G. R. He, Z. Li, Y. W. Zhang, Y.
Zhang, Q. Zhang, M. T. Zhang, Y. S. Chen, Adv. Energy Mater. 2013, 3,
639–646; b) O. P. Lee, A. T. Yiu, P. M. Beaujuge, C. H. Woo, T. W. Hol-
combe, J. E. Millstone, J. D. Douglas, M. S. Chen, J. M. J. Frechet, Adv.
Mater. 2011, 23, 5359–5363.
[9] L. J. Huo, S. Q. Zhang, X. Guo, F. Xu, Y. F. Li, J. H. Hou, Angew. Chem. Int.
Ed. 2011, 50, 9697–9702; Angew. Chem. 2011, 123, 9871–9876.
[10] H. Y. Chen, J. H. Hou, S. Q. Zhang, Y. Y. Liang, G. W. Yang, Y. Yang, L. P. Yu,
Y. Wu, G. Li, Nat. Photonics 2009, 3, 649–653.
[11] Y. Huang, L. J. Huo, S. Q. Zhang, X. Guo, C. C. Han, Y. F. Li, J. H. Hou,
Chem. Commun. 2011, 47, 8904–8906.
[12] H. L. Zhong, C. Z. Li, J. Carpenter, H. Ade, A. K. Y. Jen, J. Am. Chem. Soc.
2015, 137, 7616–7619.
[13] S. Zhang, L. Ye, W. Zhao, D. Liu, H. Yao, J. Hou, Macromolecules 2014,
47, 4653–4659.
[14] N. K. Elumalai, A. Uddin, Energy Environ. Sci. 2016, 9, 391–410.
Received: December 8, 2015
Revised: January 19, 2016
Published online on March 24, 2016
ChemSusChem 2016, 9, 973 – 980
www.chemsuschem.org
980
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim