Synthesis and photovoltaic properties of low bandgap dimeric perylene diimide based non-fullerene acceptors

Article abstract of DOI:10.1007/s11426-015-5485-8

Non-fullerene organic acceptors have attracted increasing attention in recent years. One of the challenges in the synthesis of non-fullerene organic acceptors is to tune the absorption spectrum and molecular frontier orbitals, affording low bandgap molecules with improved absorption of the near-infrared solar photons. In this paper, we present the synthesis, optoelectronic and photovoltaic properties of a series of dimeric perylene diimide (PDI) based non-fullerene acceptors. These PDI dimers are bridged by oligothiophene (T) from 1T to 6T. With the increase of the oligothienyl size, the highest occupied molecular orbital (HOMO) energy is raised from -5.65 to -5.10 eV, while that of the lowest unoccupied molecular orbit (LUMO) is kept constant at -3.84 eV, affording narrow bandgap from 1.81 to 1.26 eV. The absorption from the oligothiophene occurs between 350 and 500 nm, which is complementary to that from its bridged PDI units, leading to a wide spectral coverage from 350 to 850 nm. The optimal dihedral angle between the bridged two perylene planes is dependent on the oligothienyl size, varying from 5° to 30°. The solubility of the dimers depends on the oligothienyl size and can be tuned by the alkyl chains on the bridged thienyl units. The possible applications as the solution-processable non-fullerene organic acceptor is primarily studied using commercial P3HT as the blend donor. The photovoltaic results indicate that 1T, 4T and 6T all yield a higher efficiency of ～1.2%, whereas 2T, 3T and 5T all give a lower efficiency of <0.5%. The difference in the cell performance is related with the tradeoff between the differences of absorption, HOMO level and film-morphology between these dimers.

Full text of DOI:10.1007/s11426-015-5485-8

SCIENCE CHINA
Chemistry
February 2016 Vol.59 No.2: 209–217
doi: 10.1007/s11426-015-5485-8
• ARTICLES •
Synthesis and photovoltaic properties of low bandgap dimeric
perylene diimide based non-fullerene acceptors
Xin Zhang, Jiannian Yao & Chuanlang Zhan*
Beijing National Laboratory of Molecular Science; Key Laboratory of Photochemistry, Chinese Academy of Sciences; Institute of Chemistry,
Chinese Academy of Sciences, Beijing 100190, China
Received June 9, 2015; accepted July 6, 2015; published online October 22, 2015
Non-fullerene organic acceptors have attracted increasing attention in recent years. One of the challenges in the synthesis of
non-fullerene organic acceptors is to tune the absorption spectrum and molecular frontier orbitals, affording low bandgap mol-
ecules with improved absorption of the near-infrared solar photons. In this paper, we present the synthesis, optoelectronic and
photovoltaic properties of a series of dimeric perylene diimide (PDI) based non-fullerene acceptors. These PDI dimers are
bridged by oligothiophene (T) from 1T to 6T. With the increase of the oligothienyl size, the highest occupied molecular orbital
(HOMO) energy is raised from 5.65 to 5.10 eV, while that of the lowest unoccupied molecular orbit (LUMO) is kept con-
stant at 3.84 eV, affording narrow bandgap from 1.81 to 1.26 eV. The absorption from the oligothiophene occurs between
350 and 500 nm, which is complementary to that from its bridged PDI units, leading to a wide spectral coverage from 350 to
850 nm. The optimal dihedral angle between the bridged two perylene planes is dependent on the oligothienyl size, varying
from 5° to 30°. The solubility of the dimers depends on the oligothienyl size and can be tuned by the alkyl chains on the
bridged thienyl units. The possible applications as the solution-processable non-fullerene organic acceptor is primarily studied
using commercial P3HT as the blend donor. The photovoltaic results indicate that 1T, 4T and 6T all yield a higher efficiency
of 1.2%, whereas 2T, 3T and 5T all give a lower efficiency of <0.5%. The difference in the cell performance is related with
the tradeoff between the differences of absorption, HOMO level and film-morphology between these dimers.
non-fullerene acceptor, organic photovoltaic cell, perylene diimide, solution-processed, bulk-heterojunction
Citation:
Zhang X, Yao JN, Zhan CL. Synthesis and photovoltaic properties of low bandgap dimeric perylene diimide based non-fullerene acceptors. Sci
–
China Chem, 2016, 59: 209 217, doi: 10.1007/s11426-015-5485-8
been synthesized, for example, 9,9′-bifluorenylidene [2],
benzothiadiazole [3,4], naphthalene diimide [5–8], quina-
cridone [9], fluoranthene-fused imide [10,11], decacyclene
triimide [12], diketopyrrolopyrrole [13], electron-deficient
pentacene [14], and indandione [15].
In 2013, the effort leads to significant advances. Because
the PDI chromophore exhibits promising optoelectronic
attributes such as strong absorption in the visible region of
450 to 650 nm [16], good electron affinity with tunable
low-lying lowest unoccupied molecular orbital (LUMO,
about 4.0 eV) [17], high electron mobility of 103–10
cm2/(V s) [18], and high chemical, thermal, and photo-
chemical stabilities [19], it is considered as a potential can-
1 Introduction
Non-fullerene organic photovoltaic (OPV) cells use non-
fullerene organic n-type molecule (small or polymeric) as
the acceptor, which is blended with small molecule or pol-
ymer donor to form the photoactive blend layer. In 1986,
Tang [1] used a phthalocyanine compound as the donor (D)
and a perylene diimide (PDI) derivative as the acceptor (A),
proposing the first OPV cell with a bilayered planar cell
structure. Since then, a number of organic acceptors have
*Corresponding author (email: clzhan@iccas.ac.cn)
© Science China Press and Springer-Verlag Berlin Heidelberg 2015
chem.scichina.com link.springer.com

210
Zhang et al. Sci China Chem February (2016) Vol.59 No.2
didate of n-type organic semiconducting moieties to synthe-
size non-fullerene organic acceptors. However, the key fac-
tor that limits its applications in OPVs is the aggregation
tendency, which is caused by the large -system of the
perylene core. When blended with a donor, PDI acceptor
often forms large aggregates, typically with a size of over
100 nm, severely limiting the charge-dissociation [20]. Af-
ter reduction of the aggregation tendency by the formation
of twisted dimeric [21–28] or trimeric [29,30] backbone, or
by imidization using branched alkyl amine such as
1-ethylpropylamine [31–33] or by incorporation of large
conjugated -systems, for example, on the bay region [34],
the power conversion efficiency (PCE) from the PDI based
small molecule acceptors has been raised up to 3%4%
[21–26]. With respect to the conventional cell configuration,
an inverted cell structure takes advantages of the carrier-
extraction favourable donor/acceptor (D/A) distribution in
the top and buried surface of the active layer, which leads to
a higher PCE value (4.34% vs. 3.28%) [35]. On another
aspect, the inverted cell structure can trap more light and a
PCE of 5.9% was achieved by using an inverted cell struc-
ture with a fullerene self-assembled monolayer (C60-SAM)
on the ZnO electron selective layer [36]. Very recently, the
PCE value from the small molecule acceptor based non-
fullerene OPVs has been increased up over 6% either from
perylene derivatives or other n-type small molecule semi-
conductors [37–40].
With respect to the fullerene acceptor, organic counter-
part normally shows absorption in the visible region of
wavelength. Generation and consequent exploitation of the
acceptor’s excitons is thus the same important as the do-
nor’s for efficient non-fullerene solar cells. This requires
improving the light harvesting ability. Moreover, this re-
quires the acceptor molecule having suitable HOMO level
to match with that of the blend donor, affording efficient
hole transfer from the acceptor to the blend donor. Accord-
ingly, it is an issue in the research of non-fullerene solar
cells to tune the light-harvesting ability and the HOMO en-
ergy level of the organic acceptor via the backbone modifi-
cations.
level are both tuned simply by the size of the bridged oli-
gothiophene. The photovoltaic properties of these organic
acceptors are primarily checked by using the commercial
P3HT as the blend donor.
2 Experimental
2.1 Materials and instruments
The polymer of P3HT was purchased from Solarmer com-
1
pany. H NMR and 13C NMR spectra were recorded by a
Bruker DMX-400 (Germany) spectrometer with CDCl3 as a
solvent and tetramethylsilane as an internal reference.
MALDI-TOF mass spectra were recorded by a Bruker
BIFLEXIII (Germany). Transmission electron microscopy
(TEM) tests were performed on a JEM-2011F (JEOL, Japan)
which is operated at 200 kV. The cyclic voltammetry (CV)
was performed using a Zahner IM6e electrochemical work-
station in a 0.1 mol/L tetrabutylammoniumhexafluorophos-
phate (Bu4NPF6) dichloromethane (DCM) solution with a
scan speed at 0.1 V/s. A Pt wire and Ag/AgCl were used as
the counter and reference electrodes, respectively. The con-
centration of the dimer is 1104 mol/L in chromatographic
pure DCM. Absorption spectrum was measured on Hitachi
U-3010 (Japan) UV-Vis spectrophotometer. All spectro-
scopic measurements were carried out at room temperature.
The film spectrum was obtained via the transmission mode.
The thickness of the solid films was measured by Bruker
Dektak XT Profilometer (USA). Thermogravimetric analy-
sis (TGA) was performed on a Perkin-Elmer TGA-7 (USA)
at a heating rate of 10 °C/min under nitrogen flow.
2.2 Measurements of solubility
A working plot was first built for each dimer by plotting the
absorbance at 560 nm versus concentration, which is con-
trolled from 1×106 to 1×105 mol/L. Secondly, a saturated
solution was prepared for each dimer at room temperature
and it was then diluted to this concentration range. The real
concentration of this dilute solution was estimated accord-
ing to its absorbance, from which the solubility was finally
measured.
In this paper, we report a series of PDI dimer based ac-
ceptors (Figure 1), in which the absorption and HOMO
2.3 Quantum chemical calculations
Density functional theory calculations were performed with
the Gaussian 09 program, using the B3LYP functional.
All-electron double- valence basis sets with polarization
functions 6-31G* were used for all atoms. Geometry opti-
mizations were performed with full relaxation of all atoms
in gas phase without solvent effects. Vibrational frequency
calculations were performed to check that the stable struc-
tures had no imaginary frequency.
Figure 1 Molecular structures of the six PDI dimers.

Zhang et al. Sci China Chem February (2016) Vol.59 No.2
211
intercept value of ln(900/8) by linearly plotting ln(JL3/V2)
vs. (V/L)0.5.
2.4 Fabrications and characterizations of organic solar
cells
Solar cell devices with a typical configuration of ITO/
PEDOT:PSS/active layer/Ca/Al was fabricated as follows:
The ITO glass was pre-cleaned with deionized water,
CMOS grade acetone and isopropanol in turn for 15 min.
The organic residues were further removed by treating with
UV-ozone for 1 h. Then the ITO glass was modified by
spin-coating PEDOT:PSS layer (30 nm) on it. After dried in
oven at 150 °C for 15 min, the active layer was spin-coated
on the ITO/PEDOT:PSS surface with a blend solution of
PDI dimer and P3HT (40 mg/mL in o-DCB). The electrode
with Ca (20 nm) and Al (80 nm) was then thermally evapo-
rated onto the active layer under the vacuum of 1104 Torr.
The active area of the device was 0.06 cm2, and the thick-
nesses of the active films were 100 nm. The devices were
characterized in nitrogen atmosphere under the illumination
of simulated AM 1.5 G, 100 mW/cm2 using a xenon-lamp-
based solar simulator (AAA grade, XES-70S1). The
current-voltage (J-V) measurement of the devices was con-
ducted on a computer-controlled Keithley 2400 Source
Measure Unit. The external quantum efficiency (EQE)
measurements were performed in air using QE-R3011 (Enli
Technology Co. Ltd., Taiwan, China) with a scan increment
of 10 nm per point.
2.6 General procedure for synthesis of 2d–2f
A mixture of 5,5′-dibromo-4,4′-dihexyl-2,2′-bithiophene (5,
492.4 mg, 1.00 mmol) and 2-tributylstannylthiophene (6,
821.0 mg, 2.20 mmol) was dissolved in dry toluene (20 mL).
A catalytic amount of Pd[P(C6H5)3]4 (69.3 mg, 0.06 mmol)
was added and the reaction mixture was stirred at 110 °C
for 36 h. The mixture was extracted with dichloromethane
(DCM) and then washed with water. Then the DCM layer
was dried over Na2SO4. After removal of DCM, the residue
was subjected to chromatography with petroleum ether/
DCM as eluents to afford 3′,4″-dihexyl-2,2′:5′,2″:5″,2″′-
quaterthiophene (2d) (399.1 mg, 0.80 mmol, 80%) as a yel-
1
low solid; H NMR (400MHz, CD2Cl2)  ppm: 7.32 (d,
J=5.20 Hz, 2H), 7.14 (m, 2H), 7.07 (m, 2H), 7.02 (s, 2H),
2.72 (t, J=8.00 Hz, 4H), 1.65 (m, 4H), 1.38 (m, 4H), 1.31
(m, 8H), 0.88 (t, J=6.76 Hz, 6H); 13C NMR (100 MHz,
CDCl3)  ppm: 140.5, 136.1, 135.1, 129.7, 127.6, 126.5,
126.0, 125.5, 31.8, 30.7, 29.5, 29.4, 22.8, 14.2; TOF MS:
m/z=499.2 [M+H]+.
3″-Hexyl-2,2′:5′,2″:5″,2″′:5″′,2″″-quinquethiophene (2e).
The general procedure was followed, and flash column
chromatography (silica gel, petroleum ether/DCM) afforded
2e (372.1 mg, 0.75 mmol, 75%) as a yellow solid; 1H NMR
(400 MHz, CD2Cl2)  ppm: 7.24 (m, 2H), 7.20 (m, 2H),
7.13 (s, 1H), 7.08 (m, 2H), 7.04 (m, 4H), 2.76 (t, J=8.00 Hz,
2H), 1.66 (m, 2H), 1.39 (m, 2H), 1.31 (m, 4H), 0.88 (t,
J=6.76 Hz, 3H); 13C NMR (100 MHz, CDCl3)  ppm: 140.6,
137.3, 137.2, 137.2, 136.3, 136.0, 134.9, 129.6, 128.0,
126.7, 126.4, 124.6, 124.5, 124.3, 124.1, 123.8, 123.7, 31.8,
30.6, 29.6, 29.4, 22.8, 14.2; TOF MS: m/z=497.1 [M+H]+.
3″,4″′-Dihexyl-2,2′:5′,2″:5″,2″′:5″′,2″″:5″″,2″″′-sexithio-
phene (2f). The general procedure was followed, and flash
column chromatography (silica gel, petroleum ether/DCM)
afforded 2f (496.6 mg, 0.75 mmol, 75%) as an orange solid;
1H NMR (400MHz, CD2Cl2)  ppm: 7.23 (dd, J=0.88, 4.88
Hz, 2H), 7.20 (dd, J=0.90, 3.48 Hz, 2H), 7.13 (d, J=3.76 Hz,
2H), 7.04 (m, 4H), 7.00 (s, 2H), 2.77 (t, J=8.00 Hz, 4H),
1.70 (m, 4H), 1.43 (m, 4H), 1.36 (m, 8H), 0.93 (t, J=6.76
Hz, 6H); 13C NMR (100 MHz, CDCl3)  ppm: 140.6, 137.2,
137.1, 135.1, 135.0, 129.6, 128.0, 126.7, 126.5, 124.6,
124.1, 123.8, 31.8, 30.6, 29.7, 29.4, 22.8, 14.3; TOF MS:
m/z=663.1 [M+H]+.
2.5 Mobility measurements by space-charge limited
current (SCLC) method
The electron-only devices were fabricated with a configura-
tion of ITO/titanium (diisopropoxide) bis(2,4-pentanedio-
nate) [41] (TIPD, 20 nm)/blend (150 nm)/Al (100 nm).
Since the HOMO and LUMO energy levels of TIPD are of
–3.91 and –6.0 eV, respectively, it can be used to fabricate
the electron-only SCLC device. The TIPD buffer layer was
prepared by spin-coating a 3.5 wt% TIPD isopropanol solu-
tion onto the pre-cleaned ITO substrate and then baked at
150 °C for 10 min to convert TIPD into TOPD. Subse-
quently, the blend was spin-coated on it under the same
condition as preparation of the optimal solar cell. The Al
layer was thermally deposited on the top of the blend in
vacuum. The Al layer was deposited at a speed of 1 Å/s.
The electron mobility was extracted by fitting the current
density-voltage curves using the equation J=9/80hV2/L3
×exp[0.89(V/E0L)0.5] [42], where  is the dielectric constant
of the polymer (here, 3 is used), 0 is the permittivity of the
vacuum (8.85419×10–12 CV–1 m–1), h is the zero-field mo-
bility, E0 is the characteristic field, J is the current density, L
is the thickness of the films, and V=VapplVbi, here Vappl is
the applied potential, and Vbi is the built-in potential which
results from the difference in the work function of the anode
and the cathode (in the electron-only device, Vbi=0 V). The
electron mobility of the solar cell blend is deduced from the
2.7 General procedure for synthesis of 3a–3f
A solution of n-BuLi (2.20 mol/L in hexane, 1.00 mL, 2.20
mmol, 2.20 equiv.) was added dropwise to a solution of 2a
(68.1 mg, 1.00 mmol, 1.00 equiv.) in anhydrous tetrahydro-
furan (THF, 15 mL) in a Schlenk flask at 78 °C. After 1 h
at 78 °C, the reaction mixture was stirred at ambient tem-

212
Zhang et al. Sci China Chem February (2016) Vol.59 No.2
perature for 30 min. Then, Me3SnCl (1.00 mol/L in THF,
2.50 mL, 2.50 mmol, 2.50 equiv.) was added and the stir-
ring was maintained for 1 h. After dilution with ethyl ace-
tate (EtOAc, 50 mL), the mixture was washed with a satu-
rated aqueous solution of KF (20 mL) and then water, dried
over MgSO4 and solvent was removed by distillation under
vacuum. The residue was recrystallized from methanol to
give 2,5-bis(trimethylstannyl)thiophene (3a) (369.8 mg,
0.90 mmol, 90%) as a white crystal.
5,5′-Bis(trimethylstannyl)-2,2′-bithiophene (3b). 3b
(442.7 mg, 0.90 mmol, 90%) was recrystallized from meth-
anol as a white flake crystal.
5,5″-Bis(trimethylstannyl)-2,2′:5′,2″-terthiophene (3c).
3c (528.1 mg, 0.92 mmol, 92%) was recrystallized from
methanol as a pale yellow needle crystal.
5,5″′-Bis(trimethylstannyl)-3′,4″-dihexyl-2,2′:5′,2″:5″,2″′-
quaterthiophene (3d). 3d (750.2 mg, 0.91 mmol, 91%) was
recrystallized from methanol as a yellow needle crystal.
5,5″″-Bis(trimethylstannyl)-3″-hexyl-2,2′:5′,2″:5″,2″′:5″′,
2″″-quinquethiophene (3e). 3e (674.4 mg, 0.82 mmol, 82%)
was recrystallized from methanol as a yellow needle crystal.
5,5″″′-bis(trimethylstannyl)-3″,4″′-dihexyl-2,2′:5′,2″:5″,2
″′:5″′,2″″:5″″,2″″′-sexithiophene (3f). 3f (840.4 mg, 0.85
mmol, 85%) was recrystallized from methanol as a yellow
needle crystal.
analysis for C94H98N4O12S2: Calcd. C, 72.57; H, 6.21; N,
3.45; S, 5.93; Found C, 72.75; H, 6.17; N, 3.47; S, 5.92.
TOF MS: m/z=1622.1 [M+H]+.
1
4T was got in a yield of 85%. H NMR (400 MHz,
CD2Cl2)  ppm: 9.49–9.32 (m, 2H), 8.57–8.02 (m, 10H),
7.18–6.99 (m, 6H), 4.67–4.52 (m, 4H), 4.15–3.93 (m, 12H),
3.68–3.54 (m, 6H), 2.77–2.62 (m, 4H), 1.96–1.81 (m, 4H),
1.71–1.57 (m, 4H), 1.45–1.16 (m, 44H), 0.98–0.84 (m,
24H), 0.84–0.69 (m, 6H); 13C NMR (100 MHz, CDCl3) 
ppm: 164.0, 163.6, 163.5, 157.0, 143.9, 141.3, 138.8, 135.4,
134.1, 133.2, 131.8, 129.3, 127.7, 127.0, 123.8, 122.1,
121.4, 117.5, 70.9, 69.4, 59.6, 44.6, 38.2, 31.7, 31.0, 30.4,
29.8, 29.3, 29.0, 28.9, 24.3, 23.3, 22.7, 14.3, 14.2, 10.8,
10.8; Elemental analysis for C114H126N4O12S4: Calcd. C,
73.12; H, 6.78; N, 2.99; S, 6.85; Found C, 72.98; H, 6.76; N,
3.00; S, 6.86. TOF MS: m/z=1872.2 [M+H]+.
1
5T was got in a yield of 81%. H NMR (400 MHz,
CD2Cl2)  ppm: 9.47–9.18 (m, 2H), 8.51–7.92 (m, 10H),
7.22–6.99 (m, 9H), 4.67–4.50 (m, 4H), 4.14–3.84 (m, 12H),
3.69–3.56 (m, 6H), 2.81–2.65 (m, 2H), 1.95–1.74 (m, 4H),
1.74–1.58 (m, 2H), 1.45–1.16 (m, 38H), 0.96–0.78 (m,
27H); 13C NMR (100 MHz, CDCl3)  ppm: 163.8, 163.6,
163.4, 156.9, 143.0, 139.7, 135.7, 134.0, 133.2, 132.4,
131.7, 130.2, 129.2, 128.7, 127.5, 124.5, 123.8, 123.5,
122.1, 121.3, 120.5, 117.5, 70.9, 69.4, 59.5, 44.6, 38.1, 31.8,
31.0, 30.5, 29.4, 28.9, 24.2, 23.3, 23.2, 22.8, 14.3, 10.8.;
Elemental analysis for C112H116N4O12S5: Calcd. C, 71.92; H,
6.25; N, 3.00; S, 8.57; Found C, 71.58; H, 6.26; N, 3.01; S,
8.56. TOF MS: m/z=1870.7 [M+H]+.
2.8 General procedure for synthesis of PDI dimers 1
from 2T to 6T
1
The synthetic procedure is similar to that of bis-PDI-T-EG
(1T), that we have previously reported [21,22].
6T was got in a yield of 75%. H NMR (400 MHz,
CD2Cl2)  ppm: 9.41–9.22 (m, 2H), 8.49–7.96 (m, 10H),
7.21–6.84 (m, 10H), 4.65–4.47 (m, 4H), 4.15–3.86 (m,
12H), 3.70–3.54 (m, 6H), 2.81–2.60 (m, 4H), 1.93–1.75 (m,
4H), 1.75–1.58 (m, 4H), 1.45–1.16 (m, 44H), 0.98–0.73 (m,
30H); 13C NMR (100 MHz, CDCl3)  ppm: 163.9, 163.6,
163.4, 156.9, 143.0, 140.8, 135.9, 135.1, 134.0, 133.0,
131.7, 129.5, 129.2, 128.5, 126.7, 124.8, 123.8, 123.5,
122.1, 121.3, 120.5, 117.4, 70.9, 69.3, 63.6, 59.6, 58.4, 38.2,
31.8, 30.9, 30.5, 29.8, 29.4, 28.9, 24.2, 23.2, 22.8, 14.3,
10.8; Elemental analysis for C122H130N4O12S6: Calcd. C,
71.94; H, 6.43; N, 2.75; S, 9.45; Found C, 72.18; H, 6.45; N,
2.74; S, 9.45. TOF MS: m/z=2036.2 [M+H]+.
2T was got in a yield of 83%. 1H NMR (400 MHz, CDCl3)
 ppm: 9.59–9.45 (m, 2H), 8.80–8.11 (m, 10H), 7.22–7.11
(m, 4H), 4.72–4.55 (m, 4H), 4.23–4.04 (m, 8H), 4.04–3.95
(m, 4H), 3.66–3.55 (m, 6H), 2.04–1.84 (m, 4H), 1.48–1.19
(m, 32H), 1.02–0.80 (m, 24H); 13C NMR (100 MHz, CDCl3)
 ppm: 164.1, 164.0, 163.7, 163.6, 157.1, 144.0, 139.1,
135.4, 134.3, 133.5, 133.3, 133.2, 131.7, 131.5, 130.4,
129.4, 128.8, 128.7, 128.2, 127.9, 127.8, 125.9, 124.0,
123.7, 122.3, 121.8, 121.5, 120.8, 117.8, 70.9, 69.5, 59.6,
44.6, 38.2, 30.9, 28.9, 24.2, 23.2, 14.3, 10.8; Elemental
analysis for C94H98N4O12S2: Calcd. C, 73.32; H, 6.41; N,
3.64; S, 4.16; Found C, 72.95; H, 6.37; N, 3.67; S, 4.16.
TOF MS: m/z=1539.6 [M+H]+.
1
3 Results and discussion
3T was got in a yield of 85%. H NMR (400 MHz,
CDCl3)  ppm: 9.62–9.41 (m, 2H), 8.70–8.16 (m, 10H),
7.23–7.17 (s, 2H), 7.16–7.05 (m, 4H), 4.70–4.56 (m, 4H),
4.24–4.05 (m, 8H), 4.05–3.95 (m, 4H), 3.69–3.54 (m, 6H),
2.04–1.82 (m, 4H), 1.49–1.14 (m, 32H), 1.04–0.71 (m,
24H); 13C NMR (100 MHz, CDCl3)  ppm: 163.8, 163.6,
163.5, 157.0, 143.3, 139.5, 136.2, 135.1, 134.1, 133.3,
133.0, 132.2, 131.6, 131.4, 130.2, 129.3, 128.7, 128.5,
128.0, 127.8, 127.6, 125.3, 125.2, 123.9, 123.6, 122.2,
121.8, 121.5, 121.3, 120.6, 117.6, 70.9, 69.4, 59.6, 44.5,
44.4, 38.1, 31.0, 28.9, 24.2, 23.2, 14.3, 10.8; Elemental
3.1 Synthesis
As shown in Scheme 1, the distannyl product of thiophene
(3a), bithiophene (3b) and trithiophene (3c) was synthesized
from commercial oligothiophenes involved no alkyl side
chains. Solubility of oligothiophene decreases with the size.
To possibly increase solubility of synthesized PDI dimers,
we introduced one or two n-hexyl side chains onto the
-position of the central thiophene on quaterthiophene,

Zhang et al. Sci China Chem February (2016) Vol.59 No.2
213
Scheme 2 Synthesis of dimeric PDIs 1T−6T bridged with oligothio-
phenes with 1−6 monomer units. Reagents and conditions: i) 3% Pd(PPh3)4,
toluene, 110 °C.
isomers. However, the perylene-H atoms from the 1,6- and
1,7-isomers overlap around 9.5 ppm. Therefore, a broad
1
band is seen in this position in the H NMR spectrum of
each product of 1T–3T.
3.2 Thermostability
Thermal stability of each dimer was investigated with ther-
mogravimetric analysis (TGA) under a nitrogen atmosphere,
as outlined in Figure S1 (Supporting Information online).
The results reveal that these dimeric PDIs have excellent
thermal stabilities with thermal decomposition temperatures
(Td, 5% weight loss) over 400 °C.
Scheme 1 Synthetic procedure of oligothiophenes 2d−2f, and bis(trime-
thylstannyl)oligothiophenes 3a−3f. Reagents and conditions: i) n-BuLi,
THF, 78 °C; Me3SnCl, r.t.; ii) 1% PdCl2(PhCN)2, AgNO3/KF, DMSO,
60 °C; iii) 3% Pd(PPh3)4, toluene, 110 °C; iv) NBS, CHCl3, 0 °C.
quinquethiophene, and sexithiophene. The palladium-
catalyzed C–H homocoupling reaction of the commercial 4
gave 5 in the presence of silver(I) nitrate and potassium
fluoride at 60 °C for 6 h in 96% yield. Then, 2d was synthe-
sized through a Stille coupling reaction between 5 and the
commercial 6 in a yield of 80%. 8 was prepared by the di-
bromination of the commercial 7 under mild conditions with
N-bromosuccinimide (NBS) in nearly quantitative yield.
Palladium-catalyzed Stille coupling reaction of 8 with 9,
and 5 with 9 yielded 2e and 2f, respectively. Stannylation of
2a−2f, respectively, afforded 3a−3f in 82%−92% yield.
Stille coupling between the bis(trimethylstannyl)oligo-
thiophene and 1-(2-methoxyethoxyl)-7-brominated PDI (10)
gave the target products of the perylene diimide dimers of
1T6T, as shown in Scheme 2. Because PDI 10 was syn-
thesized from 1,7-dibrominated PDI by following the nu-
cleophilic substitution of one of the two bromo atoms [43],
the product of PDI 10 will be certainly a mixture containing
the unsymmetrically substituted 1-(2-methoxyethoxyl)-6-
brominated PDI as the minor component [44]. The 1,6- and
1,7-isomers are difficult to be separated, and, therefore each
final compound of 1T–6T is surely a mixture, coupling from
the 1,6- and 1,7-isomers [21,22]. The peaks appearing
around 9.3 ppm in the 1H NMR spectrum of each product of
4T–6T are indicative of the 1,6-isomers. Indeed, the prod-
uct of 1T–3T also contains the components from the 1,6-
3.3 Solubility
All dimers are soluble in the commonly used solvents such
as 1,2-dichlorobenzene (o-DCB). The solubility in o-DCB
is measured by spectrophotometric method and the results
are collected in Table 1. The solubility is of 30 mg/mL for
1T and about 20 mg/mL for 2T, 4T, and 6T, while reduces
down to about 9 mg/mL for 3T and 5T. The higher solubil-
ity of 4T and 6T than 3T is due to the introduction of solu-
ble alkyl chains on the bridged thiophene units.
3.4 Molecular modeling
The optimal conformations of the dimers were calculated on
Gaussian 09 at the B3LYP/6-31G level of theory in the gas
phase. Figures 2 and S2 collect the optimal conformations
and dihedral angles of the six PDI dimers. The dihedral an-
gle between two PDI planes varies as the change of the
bridge size. As indicated from Table 1, the dihedral angle
between the two covalent PDI units increases slightly from
20.5° to 22.9° and 30.3° with the thienyl going from 1T to
3T and 5T, while from 2T to 4T and 6T, that value goes
from 5.7° to 17.7° and 30.0°. The dihedral angle between
the PDI plane and the directly attached thienyl plane is 50°

214
Zhang et al. Sci China Chem February (2016) Vol.59 No.2
Table 1 Solubility, conformations and energy levels of the six dimers
cv
Solubility
(mg/mL) a)

HOMO
(eV)
LUMO
(eV)
Eg
PDI
() b)
20.49
(eV)
1.81
1.70
1.52
1.40
1.39
1.26
1T
2T
3T
4T
5T
6T
28.0
19.3
8.5
–5.65
–5.54
–5.36
–5.24
–5.23
–5.10
–3.84
–3.84
–3.84
–3.84
–3.84
–3.84
5.66
22.87
17.72
30.33
30.03
21.6
9.8
18.3
a) Measured in o-DCB; b) dihedral angle between two PDI plane.
Figure 2 Optimal conformations of PDI dimers of 1T (a), 2T (b), 3T (c),
4T (d), 5T (e), 6T (f).
Figure 3 (a) Cyclic voltammograms of the six PDI dimers in DCM with
0.1 mol/L Bu4NPF6. Scan rate was 100 mV/s, Ag/AgCl was used as refer-
ence electrode; (b) energy diagram of P3HT, PC61BM and the six PDI
dimers.
for all the six dimers. The dihedral angles between the two
naphthalene planes of one PDI unit are of 20° and 15°, re-
spectively, along the side attached to the bridged oligoth-
ienyl and along the other side attached with the 2-methoxyl-
ethoxyl unit. Otherwise, the oligothineyl twists in a different
way with the increase of the size. This means that the in-
crease of the size varies the twisted nature of the covalently
linked two PDI units and also changes the coplanarity of the
oligothienyl bridge.
3.5 Electrochemical properties
The electrochemical properties of the dimers were analyzed
by using cyclic voltammetry (CV) with Ag/AgCl as refer-
ence electrode, as shown in Figure 3(a). The energy levels
of the HOMO and LUMO were determined from the onset
of oxidation and reduction waves, respectively, by using an
offset of 4.40 eV for the Ag/AgCl versus the vacuum level
(Table 1). Figure 3(b) depicts the energy level diagrams of
the synthetic PDIs, P3HT [21], and PC61BM. The HOMO
level is raised from 5.65 to 5.10 eV as the electron-
donating ability increases with the size of the bridged oli-
gothiophene going from 1T to 6T. The LUMO level is held
constant at 3.84 eV. This is because the highly twisted
conformation between the bridged-oligothiophene and ter-
minal perylene chromophoric unit significantly reduced the
electron conjugation between them, and hence the LUMO
energy is mainly determined by the electron-accepting
perylene and the HOMO energy is, however, mainly related
to the oligothiophene. This is indicated by the nearly iden-
tical absorption band from the oligothienyl bridge of the
PDI dimer (Figure 4(a)) to that of the corresponding oli-
gothiophene (Figure S3), both in the solution state. The
Figure 4 UV-Vis absorption spectra of the six PDI dimers in a dilute
DCM with a concentration of 1106 mol/L (a) and in thin films on the
quartz plates (b).
cv
electrochemical band gaps (Eg ) are estimated from the CV
data and listed in Table 1. The band gap decreases from
1.81 to 1.26 eV. The HOMO and LUMO levels of the PDI
dimers are both lower than those of P3HT, allowing for the
use as the acceptor material.
3.6 Absorption spectra
Figure 4(a) displays the absorption spectra of the six PDI
dimers of 1T–6T in dilute DCM solutions, and the optical
properties are shown in Table 2. The oligothiophene seg-
ments 2c–2f are chosen for comparison (Figure S3 and Ta-
ble S1). 1T–6T all show two well-separated absorption

Zhang et al. Sci China Chem February (2016) Vol.59 No.2
215
Table 2 Optical properties of the six PDI dimers
collects the photovoltaic data under various optimizing con-
ditions. The best D:A weight ratio is of 1:1. The host sol-
vent is 1,2-dichlorobenzene (o-DCB).
Solution
Film
max/103 (L/(mol cm))
000, 51.0, 56.4
39.8, 53.9, 54.8
53.7, 58.0, 46.0
56.8, 59.6, 47.4
63.4, 58.1, 000
70.5, 58.0, 000
max (nm)
λonset (nm)
PDI

max (nm)

Treatment of the 1:1 (w/w) blend film of P3HT:1T by
following a solvent vapor annealing or a thermal annealing
process gives a PCE of 0.89% and 0.47%, respectively (Ta-
ble S1). As the annealing process is applied firstly with the
solvent vapor annealing and followed by the thermal an-
nealing, a higher PCE of 1.16% is obtained with a short-
circuit current-density (Jsc) of 3.18 mA/cm2, an open-circle
voltage (Voc) of 0.57 V, and a fill factor (FF) of 64.3%. The
optimal processing conditions for 1T were then applied for
other PDI dimers. Table 3 collects the photovoltaic parame-
ters of the best cells and Figure 5 displays the representative
current density-voltage (J-V) curves. Among the six PDI
dimers, 1T, 4T and 6T gives a PCE higher than 1%, where-
as another three have a PCE lower than 0.5%. The low Jsc is
a major factor for the low PCE.
The accuracy of the J-V measurement is confirmed by the
EQE spectra (Figure 6(a)) of the best cells. The EQE re-
sponses cover a wide wavelength range from 350 to 700 nm,
and match with the corresponding absorption spectrum of
the blend film (Figure 6(b)). The shoulder before 500 nm,
for example, that around 410, 440 and 500 nm for 3T, 4T
and 6T indicates the contributions from the PDI dimer, par-
ticularly, from the oligothienyl bridge, to the photon-to-
1T 000, 530, 562
2T 362, 514, 566
3T 397, 523, 562
4T 404, 535, 562
5T 421, 536, 000
6T 432, 536, 000
000, 522, 645
338, 524, 572
398, 528, 568
415, 541, 580
432, 542, 662
472, 560, 671
732
752
785
821
838
862
bands (indicating by the dashed green arrow): the right one
is mainly benefited from the PDI chromophores and origi-
nated from its electronic S0S1 transition (BPDI), whereas
the left one is mainly from the bridged oligothiophene (BT)
segment. The absorptivity of the BT increases with the in-
crease of the size and its peak red-shifts simultaneously. The
maximum extinction coefficient of the BPDI is 55103
60103 L/(mol cm) for all the dimers. While, its shape varies
with the bridge size. 1T exhibits two absorption peaks around
530 and 562 nm, respectively. Their intensity is close to
each other. 2T shows a broader band with two peaks at 514
and 566 nm. The intensity of these two peaks is identical.
3T6T, however, gives a hypochromically shifting peak at
523 nm for 3T and around 535 nm for 4T6T. The lift of
the long-wavelength edge of the BPDI, as observed from 2T
6T in between 600700 nm, is due to the intramolecular
charge-transfer (ICT) transition from the electron-donating
oligothiophene to the electron-accepting PDI unit.
Table 3 Photovoltaic parameters of the best cells using P3HT as the
blend donor
As the dimer transitions from the dilute solution to the
thin film, the absorption band of BPDI is broadened obvi-
ously towards the long-wavelength and, for example, there
appears an apparent shoulder around 645 nm for 1T and a
visible band around 700 nm for 6T, as shown in Figure 4(b).
The long-wavelength edge of the BPDI (onset) extends to 732,
752, 785, 821, 838 and 862 nm for 1T6T, respectively.
The estimated optical bandgap from the edge is in line with
that from the electrochemical data. Conspicuously, the larg-
er bridge size leads to a more red-shifted absorption band
and a lower bandgap. The band of BT of 5T and 6T be-
comes much stronger than that of BPDI and the BT Peak of
T6 is bathochromatically shifted from 432 nm in solution to
472 nm in film.
PCE (%)
e
Voc
(V)
Jsc
FF
(%)
PDI
(mA/cm2)
av. a)
1.10
0.34
0.20
0.98
0.10
1.18
105 (cm2/(V s))
Best
1T
2T
3T
4T
5T
6T
0.57
0.64
0.38
0.66
0.51
0.62
3.18
1.15
1.50
2.76
0.66
4.57
64.3
52.6
44.4
59.7
32.5
44.2
1.16
0.38
0.25
1.08
0.11
1.26
72
15
5.3
57
1.7
39
a) av.=average value, estimated from 10 devices.
3.7 Photovoltaic property
The photovoltaic property of the PDI dimer was primarily
characterized by selecting the commercial P3HT as the
blend donor. The solution-processed bulk-heterojunction
(BHJ) solar cells were fabricated with a conventional con-
figuration of ITO glass/PEDOT:PSS/P3HT:PDI dimer/
Ca/Al and were tested under an AM 1.5G simulated solar
light at 100 mW/cm2. Here, PEDOT:PSS is poly (3,4-
ethylenedioxy thiophene):poly (styrenesulfonate). Table S1
Figure 5 The J-V curves of the best devices based on the 1:1 blended
film of P3HT and PDI dimer.

216
Zhang et al. Sci China Chem February (2016) Vol.59 No.2
ized with TEM, as shown in Figure 7. The blend film from
1T, 2T, 4T and 6T each shows nanoscale interpenetrating
networks with smaller phase size, and the white/black do-
main size if of 20/30, 35/65, 35/50, 35/50 nm, respectively.
In contrast, each of 3T and 5T gives much different film-
morphology with relatively larger phase size (>100 nm).
4 Conclusions
Except for the reported dimer of 1T, another five new PDI
dimers were successfully synthesized by using oligoth-
iephene as the covalent aromatic bridge. The bridged oli-
gothiophene segment shows an absorption band before 500
nm, which is complementary to that of the covalent PDI
chromophore. The ICT absorption from the electron-
donating oligothiophene segment to the electron-accepting
PDI unit enhances with the increase of the oligothienyl size,
leading to a raise of absorption in the near IR region from
650 to 900 nm. Therefore, the absorption from the oligoth-
ienyl, PDI chromophore and the ICT between oligothienyl
and PDI together contributes to a wide spectral coverage
from 350 to 900 nm. The HOMO level increases from 5.65
to 5.10 eV as the size of the oligothiophene increases,
while, the LUMO level is held constant at 3.84 eV. This
leads to a decrease of the band gap going from 1.81 to 1.26
eV as the covalent bridge goes from 1T to 6T. Taken to-
gether, a series of low bandgap and wide spectral coverage
small molecule acceptors are reported. The photovoltaic
property of each dimer is checked using P3HT as the donor,
while the dimer as the non-fullerene acceptor.
Figure 6 (a) EQE curves of the best cells; (b) absorption spectra of 1:1
(w/w) blend films with 1T6T as the acceptor.
electron conversion. The lift of the EQE tail between 650
and 800 nm, for example, as observed from 6T, further
supports the contribution from the PDI acceptor. These cur-
rent density integrated from the EQE spectrum is 2.91, 1.07,
1.38, 2.55, 0.61, 4.20 mA/cm2 for 1T–6T, respectively,
which is in good agreement with the Jsc values (Table 3).
The Voc is generally related to the energy difference be-
tween the LUMO of the acceptor and the HOMO of the
donor. Otherwise, it is also affected by the film-morphology.
As we have observed that the Voc is raised with the increase
of the phase size [22]. The Voc value is about 0.60 V for the
dimer of 1T, 2T, 4T and 6T, respectively, which is close to
the value reported from other systems of P3HT:PDI dimer
[21,28]. These four dimers has a solubility value of about or
larger than 20 mg/mL (Table 1). Good quality blend films
can be obtained from these dimers. However, 3T and 5T
has a much smaller solubility, only about 9 mg/mL. Indeed,
we observed that the blend film from these two dimers is
worse than that from other four dimers. In fact, the film-
quality dependent Voc is a normal phenomenon in the fabri-
cation of organic solar cells.
Although the PCE from non-fullerene organic acceptor
based solution-processed solar cells has been raised over
6%, this value is still yet largely lagged from the fullerene-
based counterparts, from which a PCE of 10%–11% has
been reported from several donor-acceptor systems with a
single-junction solution-processed device strcuture [45–48].
3.8 Electron mobilities
The electron mobilities were carried out by the SCLC
method with an electron-only device of ITO/TIPD/active
layer/Al. Figure S4 shows the plots of ln(JL3/V2) vs. (V/L)0.5,
which are extracted from the experimental electron-only J-V
data, and Table 3 gives the estimated electron mobilities.
The solar cell blend of P3HT:1T, P3HT:4T, and P3HT:6T
each shows a higher electron mobility, while each of
P3HT:4T and P3HT:6T has a much lower electron mobility.
The change trend of the electron mobility is in accordance
with the varying tendency of the Jsc.
3.9 Film-morphology
Figure 7 TEM images of the solar cell films based on 1:1 blended P3HT
and 1T (a), or 2T (b), or 3T (c), or 4T (d), or 5T (e), or 6T (f), respectively,
fabricated under the optimal conditions of the best cells.
The film-morphology of the solar cell blend was character-

Zhang et al. Sci China Chem February (2016) Vol.59 No.2
217
20 Rajaram S, Armstrong PB, Kim BJ, Fréchet JMJ. Chem Mater, 2009,
21: 1775–1777
21 Lu ZH, Zhang X, Zhan CL, Jiang B, Zhang XL, Chen LL, Yao JN.
Phys Chem Chem Phys, 2013, 15: 11375–11385
22 Zhang X, Lu ZH, Ye L, Zhan CL, Hou JH, Zhang SQ, Jiang B, Zhao
Y, Huang JH, Zhang SL, Liu Y, Shi Q, Liu YQ, Yao JN. Adv Mater,
2013, 25: 5791–5797
23 Jiang B, Zhang X, Zhan CL, Lu ZH, Huang JH, Ding XL, He SG,
Yao JN. Polym Chem, 2013, 4: 4631–4638
Therefore, efforts on the design and synthesis of new organic
acceptors and understanding of the reasons between the
performance differences should be made to further improve
the electric property of the non-fullerene organic solar cells.
Acknowledgments This work was supported by the National Natural
Science Foundation of China (21327805, 91227112, 91433202, 21221002),
Chinese Academy of Sciences (XDB12010200), Ministry of Science and
Technology of China (2011CB808400, 2013CB933503, 2012YQ120060).
24 Yan QF, Zhou Y, Zheng YQ, Pei J, Zhao DH. Chem Sci, 2013, 4:
4389–4394
Conflict of interest The authors declare that they have no conflict of
interest.
25 Jiang W, Ye L, Li XG, Xiao CY, Tan F, Zhao WC, Hou JH, Wang
ZH. Chem Commun, 2014, 50: 1024–1026
26 Shivanna R, Shoaee S, Dimitrov S, Kandappa SK, Rajaram S, Dur-
rant JR, Narayan KS. Energy Environ Sci, 2014, 7: 435–441
27 Lin YZ, Wang JY, Dai SX, Li YF, Zhu DB, Zhan XW. Adv Energy
Mater, 2014, 4: 1400420
28 Zhang X, Yao JN, Zhan CL. Chem Commun, 2015, 51: 1058–1061
29 Lin YZ, Wang YF, Wang JY, Hou JH, Li YF, Zhu DB, Zhan XW.
Adv Mater, 2014, 26: 5137–5142
30 Zhang X, Jiang B, Zhang SQ, Hou JH, Yao JN, Zhan CL. Proc SPIE,
2014, 9184: 91840C
31 Sharenko A, Proctor CM, van der Poll TS, Henson ZB, Nguyen TQ,
Bazan GC. Adv Mater, 2013, 25: 4403–4406
32 Ye TL, Singh R, Butt HJ, Floudas G, Keivanidis PE. ACS Appl Mater
Inter, 2013, 5: 11844–11857
33 Singh R, Aluicio-Sarduy E, Kan Z, Ye TL, MacKenzie RCI, Keiva-
nidis PE. J Mater Chem A, 2014, 2: 14348–14353
34 Zhang XL, Jiang B, Zhang X, Tang AL, Huang JH, Zhan CL, Yao JN.
J Phys Chem C, 2014, 118: 24212–24220
35 Lu ZH, Jiang B, Zhang X, Tang AL, Chen LL, Zhan CL, Yao JN.
Chem Mater, 2014, 26: 2907–2914
36 Zang Y, Li CZ, Chueh CC, Williams ST, Jiang W, Wang ZH, Yu JS,
Jen AKY. Adv Mater, 2014, 26: 5708–5714
37 Zhong Y, Trinh MT, Chen RS, Wang W, Khlyabich PP, Kumar B,
Xu QZ, Nam CY, Sfeir MY, Black C, Steigerwald ML, Loo YL,
Xiao SX, Ng F, Zhu XY, Nuckolls C. J Am Chem Soc, 2014, 136:
15215–15221
38 Zhang X, Zhan CL, Yao JN. Chem Mater, 2015, 27: 166–173
39 Zhao JB, Li YK, Lin HR, Liu YH, Jiang K, Mu C, Ma TX, Lai JYL,
Yan H. Energy Environ Sci, 2015, 8: 520–525
40 Lin YZ, Wang JY, Zhang ZG, Bai HT, Li YF, Zhu DB, Zhan XW.
Adv Mater, 2015, 27: 1170–1174
41 Tan ZA, Zhang WQ, Zhang ZG, Qian DP, Huang Y, Hou JH, Li YF.
Adv Mater, 2012, 24: 1476–1481
42 An ZS, Yu JS, Jones SC, Barlow S, Yoo S, Domercq B, Prins P,
Siebbeles LDA, Kippelen B, Marder SR. Adv Mater, 2005, 17: 2580–
2583
Supporting information The supporting information is available online
at chem.scichina.com and link.springer.com/journal/11426. The supporting
materials are published as submitted, without typesetting or editing. The
responsibility for scientific accuracy and content remains entirely with the
authors.
1
2
Tang CW. Appl Phys Lett, 1986, 48: 183–185
Brunetti FG, Gong X, Tong M, Heeger AJ, Wudl F. Angew Chem Int
Ed, 2010, 49: 532–536
3
4
5
Holcombe TW, Norton JE, Rivnay J, Woo CH, Goris L, Piliego C,
Griffini G, Sellinger A, Bredas JL, Salleo A, Fréchet JM. J Am Chem
Soc, 2011, 133: 12106–12114
Bloking JT, Han X, Higgs AT, Krastrop JP, Pandey L, Norton JE,
Risko C, Chen CE, Brédas JL, McGehee MD, Sellinger A. Chem
Mater, 2011, 23: 5484–5490
Ahmed E, Ren G, Kim FS, Hollenbeck EC, Jenekhe SA. Chem Mater,
2011, 23: 4563–4577
6
7
Ren G, Ahmed E, Jenekhe SA. Adv Energy Mater, 2011, 1: 946–953
Wang X, Huang JH, Niu ZX, Sun YX, Zhan CL. Tetrahedron, 2014,
70: 4726–4731
8
9
Fernando R, Mao Z, Muller E, Ruan F, Sauvé G. J Phys Chem C,
2014, 118: 3433–3442
Zhou TL, Jia T, Kang BN, Li FH, Fahlman M, Wang Y. Adv Energy
Mater, 2011, 1: 431–439
10 Zhou Y, Ding L, Shi K, Dai YZ, Ai N, Wang J, Pei J. Adv Mater,
2012, 24: 957–961
11 Zhou Y, Dai YZ, Zheng YQ, Wang XY, Wang JY, Pei J. Chem
Commun, 2013, 49: 5802–5804
12 Pho TV, Toma FM, Chabinyc ML, Wudl F. Angew Chem Int Ed,
2013, 52: 1446– 1451
13 Lin YZ, Li YF, Zhan XW. Adv Energy Mater, 2013, 3: 724–728
14 Shu Y, Lim YF, Li Z, Purushothaman B, Hallani R, Kim JE, Parkin
SR, Malliaras GG, Anthony JE. Chem Sci, 2011, 2: 363–368
15 Winzenberg KN, Kemppinen P, Scholes FH, Collis GE, Shu Y,
Singh TB, Bilic A, Forsythb CM, Watkins SE. Chem Commun, 2013,
49: 6307–6309
43 Zhang X, Pang SF, Zhang ZG, Ding XL, Zhang SL, He SG, Zhan CL.
Tetrahedron Lett, 2012, 53: 1094–1097
16 Zhan CL, Li ADQ. Curr Org Chem, 2011, 15: 1314–1339
17 Schmidt R, Oh JH, Sun YS, Deppisch M, Krause AM, Radacki K,
Braunschweig H, Könemann M, Erk P, Bao ZN, Würthner F. J Am
Chem Soc, 2009, 131: 6215–6228
44 Zhang X, Zhan CL, Zhang XL, Yao JN. Tetrahedron, 2013, 69:
8155– 8160
45 Li YF. Sci China Chem, 2015, 58: 188
46 Wu HB. Sci China Chem, 2015, 58: 189
18 Zhao Y, Guo YL, Liu YQ. Adv Mater, 2013, 25: 5372–5391
19 Würthner F. Chem Commun, 2004: 1564–1579
47 Huang F. Sci China Chem, 2015, 58: 190
48 Li YF. Sci China Chem, 2015, 58: 191