Simple mono-halogenated perylene diimides as non-fullerene electron transporting materials in inverted perovskite solar cells with ZnO nanoparticle cathode buffer layers

Article abstract of DOI:10.1039/c7ta02617j

We have synthesized and characterized three perylene diimides, X-PDI, where X = H, F, or Br. These three compounds have been tested for the substitution of [6,6]-phenyl C₆₁ butyric acid methyl ester (PC₆₁BM) in inverted perovskite solar cells (PVSCs). Although the efficiency of PDI derivative-based conventional PVSCs is as high as 17.6% (J. Mater. Chem. A, 2016, 4, 8724), the corresponding performance of inverted PVSCs is still lagging behind and improvement is necessary. For electron accepting (from CH₃NH₃PbI₃ perovskite) and electron transporting properties of the materials, UV-visible absorption spectroscopy, electrochemical cyclic voltammetry, direct current conductivity, space-charge limited current (SCLC) electron mobility, atomic force microscopy (AFM) surface morphology, and photoluminescence (PL) spectroscopy gauging charge-trapping have been studied. The HOMO/LUMO energy levels are 5.94/4.00, 5.83/3.96, 6.00/3.97, and 5.48/3.73 eV for H-PDI, F-PDI, Br-PDI, and PC₆₁BM, respectively. Direct current conductivities of H-PDI, F-PDI, Br-PDI, and PC₆₁BM are 8.71 × 10^-8, 1.18 × 10^-9, 2.2 × 10^-6, and 8.42 × 10^-6 S cm^-1, respectively. The SCLC electron mobility of H-PDI, F-PDI, Br-PDI, and PC₆₁BM are 1.12 × 10^-4, 8.31 × 10^-6, 1.08 × 10^-3, and 5.00 × 10^-3 S cm^-1, respectively. PL spectroscopy of CH₃NH₃PbI₃ perovskite provides emission at wavelength of 781 nm, which is the same as for the perovskite layer covered with a thin film of H-PDI or F-PDI. However, the emission wavelength was blue-shifted to 773-774 nm when the perovskite layer was covered with a thin film of Br-PDI or PC₆₁BM. Using UV-visible absorption spectroscopy, the solubility (in chloroform) was determined as 1.2 × 10^-2, 8.7 × 10^-2 and >10^-1 mol L^-1 for F-PDI, H-PDI, and Br-PDI, respectively. In thin film state, UV-visible absorption spectroscopy indicated that the extent of molecular aggregation was F-PDI ? H-PDI > Br-PDI, which is consistent with the AFM-estimated root-mean-square roughness of F-PDI > H-PDI > Br-PDI ～ PC₆₁BM. Without the solution processed ZnO NP cathode buffer layer (CBL), the power conversion efficiency (PCE) of H-PDI, F-PDI, Br-PDI, and PC₆₁BM PVSCs is ～1%, ～0%, 3.2%, and 4.1%, respectively. With the ZnO NP CBL, PCE is ～7.8%, ～0%, 10.5%, and 11.1% for H-PDI, F-PDI, Br-PDI, and PC₆₁BM PVSCs, respectively. Through this study, we have demonstrated that the simple mono-bromine substituted perylene diimide (Br-PDI), is solution processable and has potential for use as a non-fullerene electron accepting and electron transporting material in inverted PVSCs.

Full text of DOI:10.1039/c7ta02617j

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. Chen, J. Wu,
W. Huang, Y. Chang, B. Tsai, Y. Hsiao, C. Chang and C. Chen, J. Mater. Chem. A, 2017, DOI:
10.1039/C7TA02617J.
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 11
View Article Online
DOI: 10.1039/C7TA02617J
Journal of Materials Chemistry A
PAPER
Simple mono-halogenated perylene diimides as non-fullerene
electron transporting materials in inverted perovskite solar cells
with ZnO nanoparticle cathode buffer layer
Jhao-lin Wu,a Wen-Kuan Huang,b Yu-Chia Chang,b Bo-Chou Tsai,b Yu-Cheng Hsiao,b Chih-Yu Chang,
*b Chin-Ti Chen,*c and Chao-Tsen Chen*a
Received 00th January 20xx,
Accepted 00th January 20xx
DOI: 10.1039/x0xx00000x
www.rsc.org/
We have synthesized and characterized three perylene diimides, i.e., X-PDI, X = H, F, or Br. These three compounds are
tested for the substitution of [6,6]-phenyl C61 butyric acid methyl ester (PC61BM) in inverted perovskite solar cells (PVSCs).
Although efficiency of PDI derivative-based conventional PVSCs is as high as 17.6% (J. Mater. Chem. A, 2016, 4, 8724),
corresponding performance of inverted PVSCs is still lagged behind and the improvement is necessary.
For electron
accepting (from CH3NH3PbI3 perovskite) and electron transporting properties of the materials, The UV-visible absorption
spectroscopy, electrochemical cyclic voltammetry, direct current conductivity, space-charge limited current (SCLC) electron
mobility, atomic force microscopy (AFM) surface morphology, and photoluminescence (PL) spectroscopy gauging charge-
trap have been studied. The HOMO/LUMO energy levels are 5.94/4.00, 5.83/3.96, 6.00/3.97, and 5.48/3.73 eV for H-PDI, F-
PDI, Br-PDI, and PC61BM, respectively. Direct current conductivities of H-PDI, F-PDI, Br-PDI, and PC61BM are 8.71  10-8, 1.18
 10-9, 2.2  10-6, and 8.42  10-6 S cm-1, respectively. The SCLC electron mobility of H-PDI, F-PDI, Br-PDI, and PC61BM are
1.12  10-4, 8.31  10-6, 1.08  10-3, and 5.00  10-3 S cm-1, respectively. PL spectroscopy of CH3NH3PbI3 perovskite provides
emission wavelength peaked at 781 nm, which is the same as perovskite layer covered with a thin film of H-PDI or F-PDI.
However, the emission wavelength is blue-shifted to 773-774 nm when the perovskite layer was covered with a thin film of
Br-PDI or PC61BM. By UV-visible absorption spectroscopy, solution (in chloroform) solubility was determined as 1.2 × 10-2,
8.7 × 10-2, and > 10-1 mol L-1 for F-PDI, H-PDI, and Br-PDI, respectively. As thin film sate, UV-visible absorption spectroscopy
indicated the extent of molecular aggregation is F-PDI >> H-PDI > Br-PDI, which is consistent with the AFM-estimated root-
mean-square roughness of F-PDI > H-PDI > Br-PDI ~ PC61BM. Without solution processed ZnO NP cathode buffer layer
(CBL), power conversion efficiency (PCE) of H-PDI, F-PDI, Br-PDI, and PC61BM PVSCs is ~1%, ~0%, 3.2%, and 4.1%,
respectively. With the ZnO NP CBL, PCE is ~7.8%, ~0%, 10.5%, and 11.1% for H-PDI, F-PDI, Br-PDI, and PC61BM PVSCs,
respectively. Through this study, we have demonstrated that the simple mono-bromine substituted perylene diimide (Br-
PDI), which is solution processable and is potential for a non-fullerene electron accepting and electron transporting material
in inverted PVSCs.
ethylenedioxythiophene) polystyrene sulphonate (PEDOT:PSS)
and [6,6]-phenyl C61 butyric acid methyl ester (PC61BM) as HTL
and ETL, respectively.6 In addition to the compatible solution
Introduction
Since 2013, there has been great development in the field of
perovskite solar cells (PVSCs), which have the advantages of low
cost, simple fabrication and high efficiency.1–4 To date, the
certified maximum power conversion efficiency (PCE) of
perovskite solar cells has reached 21.1%.5 Based on either
mesoscopic or planar device architectures, PVSCs are typically
composed of a light harvesting perovskite layer sandwiched
between an electron transport layer (ETL) and hole transport
layer (HTL). For an inverted construction of planar PVSCs, the
process with perovskite material, polymer PEDOT:PSS has been
widely used as an HTL for inverted PVSCs due to its good
conductivity (0.014 S cm-1)7 and small work function of 5.0 eV.
On the other hand, C60 fullerene derivative PC61BM is usually
chosen as solution processed ETL because (i) the high
conductivity (0.016 mS cm-1)7 or impressive electron mobility
(space-charge limited mobility 5 × 10-3 cm2 V-1 S-1)8 which is
expected to reduce the current density (J)–voltage (V)
hysteresis with respect to the scan direction because of the
balanced charge carrier of devices; (ii) a relatively deep lowest
unoccupied molecular orbital (LUMO) energy level of -4.15 eV9
which is energetically favourable for extracting electron from
perovskite exciton. Due to these good physical features, few
other organic materials are better than PC61BM as ETL for
inverted PVSCs. However, PC61BM is not photochemically stable
and it is prone to aggregate resulting a thin film with poor
conventional
transport
layers
are
poly
(3,4-
a.Department of Chemistry, National Taiwan University, Taipei, Taiwan 10617,
R.O.C. (*E-mail: chenct@ntu.edu.tw)
b.Department of Materials Science and Engineering, Feng Chia University,
Taichung, Taiwan 40724, R.O.C. (*E-mail: changcyu@fcu.edu.tw)
c. Institute of Chemistry, Academia Sinica, Taipei, Taiwan 11529, R.O.C. (*E-mail:
chintchen@gate.sinica.edu.tw)
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 1
Please do not adjust margins

Journal of Materials Chemistry A
Page 2 of 11
Please do not adjust margins
Journal Name
ARTICLE
morphology that impairs the physical feathers mentioned PVSCs based on PC61BM/ZnO NP. Whereas PVSCsVibewasAertdicleoOnnlHine-
above. Therefore, non-fullerene type ETLs, particularly those PDI are inferior to Br-PDI ones, F-PDI PVSDCOs Ie: x10h.i1b0i3t9v/Cir7tTuAa0ll2y61n7oJ
suitable for inverted PVSCs, are highly demanded.
photovoltaic effect regardless ZnO NPs. We have conducted a
The perylene diimide (PDI) compounds are n-type series of experiments to find out the reasons why Br-PDI is
semiconductors with relatively low LUMO energy levels in a better than H-PDI and why F-PDI is so poor for PVSCs.
range of 3.5-4.3 eV10,11 due to the electron affinity of its imide
parts. Their high electron mobility (space-charge limited
mobility of 0.2 ~ 1.3 cm2 V-1 S-1)12-14 and conductivity (4.4 × 10-2
~ 3.3 × 10-3 S cm-1)15,16 originate from the rigid and planar
π-
skeleton, which are prone to ordered molecular packing. In
addition to three reports about bis(N,N’-
dimethylaminopropyl)perylene diimide (N-PDI)17, bis(N,N’-2,4-
difluorophenyl)perylene diimide (dFPh-PDI)18, and N,N′-Bis(3-
(dimethylamino)propyl)-5,11-dioctylcoronene-2,3,8,9-
tetracarboxdiimide19 as ETL in conventional PVSCs, there are
only two reports about inverted PVSCs, which are more
pertinent to the study herein. They are N,N-dioxide of
bis(N,N’-dimethylaminopropyl)perylene diimide (PDINO)20,21
and 1,1’-di-perylene diimide (diPDI).22 The inverted
configuration of PVSCs are ITO/PEDOT:PSS/CH3NH3PbI3-
Fig 1 Energy levels of each layer alignment of PVSCs (left) and chemical structures of X-
PDI (right).
Results and discussion
xClx/PC61BM/PDINO/Ag
and
ITO/PEDOT:PSS/CH3NH3PbI3/
diPDI:DMBI/TiO2/Al, respectively. In the case of PDINO PVSC,
PCE as high as 14.0% was achieved, although PDINO was
actually used as a cathode buffer layer (CBL) or cathode
interlayer (CIL) instead of ETL.21 In the case of diPDI PVSC, PCE
as high as 10.0% was achieved but an n-type dopant (1,3-
dimethyl-2-phenyl-2,3-dihydro-1H-benzoimidazole: DMBI) has
to be included in the ETL. Otherwise, PCE of 7.1% was the best
result without DMBI dopant in diPDI ETL.
Synthesis and characterization
Each of X-PDI was prepared consecutively in a three-step
synthetic procedure (Scheme 1). The synthesis started from
the imidization of a commonly available commercial product of
perylene-3,4,9,10-tetracarboxylic dianhydride. From the
reaction with 2-ethyl-1-hexylamine under refluxing condition in
N-methyl-2-pyrrolidione (NMP), H-PDI was obtained with near
quantitative isolation yields (97%). In fact, H-PDI is a literature
known compound.29-31 Following the same literature procedure,
Br-PDI was afforded from a room temperature reaction of H-
PDI and bromine in dichloromethane, although 34% was the
best isolation yields after the purification by column
chromatography. F-PDI is a previously unknown compound
and its synthetic procedure was modified from a literature
There are a number of literature reports having ZnO as CBL
or CIL in polymer-based bulk heterojuction conventional solar
cells.23-25 In addition to the low cost and ambient stability of ZnO,
a
combination effect of hole-blocking (reducing charge
recombination), enhancing election extraction (or selection),
preventing diffusion of cathode metal into active layer through
the interspace of incomplete ETL, and the optical spacer effect
has been attributed to the improvement of the device
performance. To our surprise in surveying literature, only
three of our previous reports have mentioned solution
processed ZnO nanoparticles (NPs) as CBL or CIL in a series of
inverted PVSCs.26-28 Potentially, there could be an
interdissolving problem of the perovskite material when
applying ZnO nanoparticles (NPs) solution (in ethanol) by a
solution process, if the ETL thin film has an low surface coverage
and forms serious aggregate with large interspace between
aggregate domains. The solvent of ZnO NPs may leak through
the ETL thin film reaching perovskite layer underneath and may
generate a straight up tunnel facilitating the direct contact of
deposited cathode metal.
report of
a
1,7-difluoro-substituted perylene diimide
derivative.32 Synthetic details and the characterization data of
F-DPI can be found in the section of experimental.
Scheme 1 consecutive synthesis of three X-DPI compounds
.
In this report, we demonstrate two structure simple mono-
halogen-substituted X-PDI, i.e., F-PDI ) and Br-PDI
in addition to the parent H-PDI (X = H) (Figure 1), to replace the
(
X
=
F
(X = Br),
Absorption spectroscopic properties
Figure 2 shows the UV-visible absorption spectra for both X-PDI
chloroform solution (10-5 mol L-1) and spin-coated thin film.
Their data are summarized in Table 1. The spectra (300-600 nm)
of X-PDI in solution mainly consist of a π-π* electronic
transitions associated with a series of characteristic vibronic
commonly used PC61BM as ETL in inverted MAPbI3 (MA:
methylammonium) PVSCs. Adding ZnO NPs as a CBL or CIL on
the top of the ETL, we have successfully fabricated PVSCs with
improved PCEs. A respectable PCE of 10.5% has been reached
by Br-PDI/ZnO NP PVSC, which is comparable with 11.1% of
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 2
Please do not adjust margins

Page 3 of 11
Journal of Materials Chemistry A
Please do not adjust margins
Journal Name
ARTICLE
structures.33, 34 They are 0-0, 0-1, 0-2, and maybe 0-3 transitions
with absorption λmax at 520-525, 484-489, 454-459, 430-433
nm, respectively. The absorption spectra of the spin-coated thin
film
all show broadening, splitting, and red-shifting absorption
bands, of which relative intensity of the vibronic structures is
quite different from those of solution spectra. All these
Table 1 Summary of the UV-visible absorption spectroscopy ofVXie-PwDAIrticle Online
DOI: 10.1039/C7TA02617J
Solution
max onset
[nm] [nm] [eV]a


Film
onset
[nm]
637
op
op
Eg
max
[nm]
471
Eg
[eV]a
1.94
H-PDI
F-PDI
Br-PDI
PC61BM
525
520
524
510
541
545
544
630
2.29
2.28
2.28
1.97
467
662b
610
710b
1.87b
2.03
469
530b
1.75b
changes of spectroscopic feature indicate
a significant
molecular interaction of X-PDI taking place in solid state (as thin
film). In addition, F-PDI has a most intense absorption band
around 633 nm in Figure 2b, which can be ascribed to the strong
b
a optical energy gap (Egop) corresponding to absorption onset wavelength.
estimated value.
molecular interactions (-F-, F-F interactions) of F-PDI in
Table 2 Cyclic voltammetry data and HOMO/LUMO energy levels of X-
solid state. Moreover, the high-rise base line observed in the
spectrum of F-PDI thin film can be attributed to the scattering
of near infrared and the long wavelength of visible light, which
is consistent with the pronounced inhomogeneity of F-PDI thin
film. On the other hand, molecular interaction of X-PDI also
reflects in solution solubility. Based on the Beer’s law of
absorption spectroscopic measurement, F-PDI has the worst
solubility (1.2 × 10-2 mol L-1) followed by H-PDI (8.7 × 10-2 mol L-
PDI and PC61BM
.
Solution
ELUMO EHOMO
[ev] [ev]
Film
ELUMO EHOMO
[ev]
red1
op
red1
op
E1/2
E1/2
[V]
[V]
[ev]
H-PDI
F-PDI
Br-PDI
-1.05 -4.05 -6.34
-1.02 -4.08 -6.36
-1.00 -4.10 -6.38
-1.10 -4.00 -5.94
-1.14 -3.96 -5.83
-1.13 -3.97 -6.00
-1.37 -3.73 -5.48
PC61BM -1.08 -4.02 -5.99
1), which is relatively better than F-DPI
PDI (> 10-1 mol L-1) is at least an order of magnitude higher than
F-PDI In this regard, F-PDI has the strongest molecular
. The solubility of Br-
a E1/2red1 is the half-wave potential of the first reduction; ELUMO = -(E1/2red1 + 5.10);
op
EHOMO
=
ELUMO – (Egop in Table 1).
.
interaction in solid state, whereas Br-PDI has the weakest
molecular interaction in solid state. The bulky bromine
substituent inhibits the - stacking of Br-PDI in solid state and
a higher solubility observed for Br-PDI than H-PDI in solution.
Considering the solubility and molecular interaction, the
homogeneity or quality of solution-processed Br-PDI thin film
should be better than that of H-PDI or F-PDI. Such speculation
has been verified in the AFM study (see the results and
discussion of thin film surface morphology later).
Cyclic voltammetry - energy level determination
The electrochemical properties of X-PDI and PC61BM were
studied by cyclic voltammetry (CV). The corresponding
electrochemical data are summarized in Table 2. The cyclic
voltammograms of X-PDI and PC61BM measured in solution
(dichloromethane) and as thin film coated on electrode are
shown in Fig. 3.
-1.00
1.2
-0.95
(a)
(b)
H-PDI
(a)
H-PDI
F-PDI
Br-PDI
H-PDI
E1/2red1 -1.10 V
E1/2red1 -1.05 V
520,524,525 nm
1.0
0.8
0.6
0.4
-1.1
-0.97
-1.25
-0.99
F-PDI
E1/2red1 -1.02 V
484,489,489 nm
F-PDI
E1/2red1 -1.14 V
-1.07
-0.94
-1.28
-0.95
Br-PDI
E1/2red1 -0.99 V
454,458,459 nm
Br-PDI
-1.05
-1.03
0.2
E1/2red1 -1.13 V
430,433,433 nm
PC61BM
E1/2red1 -1.08 V
0.0
300
-1.31
-1.32
400
500
600
700
800
900
PC61BM
E1/2red1 -1.37 V
Wavelength(nm)
-1.12
-1.43
1.2
1.0
0.8
0.6
0.4
0.2
0.0
(b)
H-PDI
F-PDI
Br-PDI
465-495 nm
-1.5 -1.0 -0.5 0.0
-1.5 -1.0 -0.5 0.0
+
Potential (V vs Fc/Fc+)
Potential(V vs Fc/Fc)
537-577 nm
Fig 3 Cyclic voltammetry of X-PDI in chloroform solution (a) and as thin film (b).
633 nm
As can be seen in Fig. 3a, all voltammograms of X-PDI or
PC61BM in dichloromethane show two reversible or semi-
reversible reduction signals in the negative potential around -
???
none
1.0 V, relative to the oxidation potential of ferrocene (Fc

Fc+).
Based on the solution first reduction potential (E1/2red1), all X-
PDI have a slightly weaker electron affinity than that of PC61BM,
indicating three X-PDI reported here are potential as electron
acceptor like PC61BM for perovskite material. Here, we have
300
400
500
600
700
800
900
Wavelength (nm)
Fig 2 UV-Vis absorption spectra of X-PDI in chloroform solution (a) and as thin film (b).
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 3
Please do not adjust margins

Journal of Materials Chemistry A
Page 4 of 11
Please do not adjust margins
Journal Name
ARTICLE
The charge carrier transport of X-PDI and PC61BM ETLs was studied
DOI: 10.1039/C7TA02617J
by the space-charge limited current (SCLC) approach. The electron
found that the mono fluorine or bromine substituent little
affects the reduction potential of H-PDI. However, in the thin
film measurement (Figure 3b), there seems only one reduction
signal of three X-PDI and PC61BM in voltammograms around -
1.10~-1.14 V and they are all less negative than -1.37 V of
View Article Online
mobilities determined from a series of electron–dominated devices
are summarized in Table 3. The electron–dominated devices studied
herein are ITO/ZnO/X-PDI (or PC61BM)/Ca/Al and ITO/ZnO/X-
PDI (or PC61BM)/ZnO NP/Ca/Al, with and without ZnO NP layer,
respectively.
Figure 5a shows the J–V characteristics of electron–dominated
device of X-PDI and PC61BM without ZnO NP. The electron mobility
of Br-PDI is 1.08  10-3 cm2 V-1 s-1, which is significantly better than
that of H-PDI or F-PDI, about 10 and 130 times higher, respectively.
The electron mobility of Br-PDI is in a same order of magnitude as
that of PC61BM. In fact, electron mobility determined from SCLC
red1
PC61BM. Just considering the energy level of LUMO or E1/2
in solid state, all X-PDI are better electron acceptor than PC61BM
for perovskite material. Comparing solution and thin film data,
it is interesting to note that the shifting of energy levels is more
pronounced on HOMO than on LUMO energy levels. Moreover,
the shifting of HOMO energy level is F-PDI > H-PDI > Br-PDI
,
which is consistent with the order of molecular interaction and
solubility found for X-PDI series.
follows the same trend of DC σ0 found for X-PDI and PC61BM. We
also took the X-PDI and PC61BM in the fabrication of electron–
dominated device with ZnO NP top contact layer. As data shown in
Table 3 and figures of Figure 5b and 5d, the electron mobility of X-
PDI and PC61BM was improved 1.55 ~17 and ~5 times, respectively,
comparing with those without ZnO NP top contact layer. Our SCLC
measurements have demonstrated that the thin film electron
mobility of Br-PDI is far more superior to that of F-PDI. Since the
LUMO energy level of X-PDI is not much different from each other,
we ascribe the difference of determined electron mobility to the
morphology of the thin film, of which the surface roughness was
found the lowest in the case of Br-PDI (see the following AFM
section). In addition, due to the improvement of electron mobility
by ZnO NP, we expect that adopting ZnO NP as CBL in Br-PDI or
PC61BM PVSCs can greatly decrease the series resistance (Rs) and
largely enhance the fill factor (FF).
Direct current conductivity of X-PDI and PC61BM
The DC (direct current) conductivity (σ0) can be determined
from the slope of the current-voltage (I–V) plot. Figure 4 shows
the I–V characteristics of ITO/X-PDI or PC61BM/Al devices. The
thicknesses of the thin films are 200, 260, 250, and 280 nm for
PC61BM, H-PDI
corresponding DC σ0s of PC61BM, H-PDI
8.42
10-6, 8.71 10-8, 1.18 10-9, and 2.2
respectively. Among X-PDI Br-PDI has the highest DC σ0, which
,
F-PDI
,
and Br-PD
,
respectively. The
F-PDI, and Br-PDI are
10-6 S cm-1,
,
×
×
×
×
,
is in the same order of magnitude comparing with PC61BM.
Therefore, it can be expected that the charge carriers generated
in the MAPbI3 layer will be more efficiently transported through
PC61BM or Br-PDI than H-PDI or F-PDI
.
On the other hand,
among X-PDI F-PDI has the lowest DC σ0, which can be
,
attributed to its poor thin film morphology and the relatively
low solubility in solution.
0.01
1E-3
1E-4
1E-5
1E-4
1E-5
1E-6
1E-7
-24
-25
-26
-27
-28
-29
-30
-31
-32
-33
-34
H-PDI
F-PDI
Br-PDI
PC61BM
H-PDI
F-PDI
(a)
(b)
1
0.1
0.01
1E-3
1E-4
H-PDI
F-PDI
Br-PDI
PC61BM
1E-6
0.1
1E-8
1
0.01
Br-PDI
0.1
1
0.01
PC61BM
0.1
1
1
2
3
4
5
200
300
400
500
600
(VL-1)0.5
Bias Voltage (V)
0.01
1E-3
1E-4
1E-5
0.1
0.01
1E-3
1E-4
10
(c)
-22
-24
-26
-28
-30
-32
-34
(d)
1
0.1
PC61BM/ZnO NP
PC61BM+ZnO NPs
Br-PDI/ZnO NP
F-PDI/ZnO NP
H-PDI/ZnO NP
Br-PDI+ZnO NPs
F-PDI+ZnO NPs
H-PDI+ZnO NPs
0.01
0.1
1
0.01
0.1
1
0.01
1E-3
Bias Voltage (V)
Bias Voltage (V)
Fig 4 I–V characteristics of ITO/X-PDI or PC61BM/Al devices.
4
5
6
7
8
9
10
200
300
400
500
600
(VL-1)0.5
Bias Voltage (V)
Table 3 SCLC electron mobilities of X-PDI and PC61BM with and without ZnO NP.
Fig 5 The effect of ZnO NPs top contact layer on electron transport property of X-PDI and
PC61BM. The dark J–V characteristics of devices without ZnO NPs (a, b) and with ZnO NPs
(c, d).
w/o ZnO NP
w/ ZnO NP
e (cm2 V-1 s-1
4.57  10-4
1.29 × 10-5
1.84 × 10-2
2.54 × 10-2
μ
e (cm2 V-1 s-1
)
μ
)
1.12  10-4
8.31  10-6
1.08  10-3
5.00  10-3
H-PDI
F-PDI
Br-PDI
PC61BM
Thin film surface morphology of X-PDI, PC61BM, and ZnO NP
To investigate the changes in morphology of X-PDI solution
spin-coated MAPbI3 perovskite films, atomic force microscopy
(AFM) measurements were performed in tapping mode. A two-
step sequential deposition method was used to prepare high-
quality MAPbI3 perovskite film, and its preparation details can
Space charge limited current - electron mobility
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 4
Please do not adjust margins

Page 5 of 11
Journal of Materials Chemistry A
Please do not adjust margins
Journal Name
ARTICLE
be found in our previous work.35,36 X-PDI or PC61BM chloroform Accordingly, it is foreseeable that Br-PDI and PC61BM thin films
View Article Online
DOI: 10.1039/C7TA02617J
solution was spin-coated on the perovskite substrate. As shown will have a more extensive contact or a higher surface coverage
in Figure 6a, 6b, 6c, and 6d, an inhomogeneous surface of perovskite layer, which facilitates a more effective electron
morphology with high root-mean-square (RMS) roughness as transfer across the MAPbI3/ETL interface in PVSCs. Except for F-
high as 18.8 and 25.4 nm was observed for H-PDI and F-PDI
,
PDI, the ZnO NP CBL or CIL can further enhance such electron
respectively. Notably better thin films with lower RMS transfer and consequent electron transport to the cathode
roughness of 9.5 and 10.4 nm were found for Br-PDI and because thin film surface roughness all got more or less reduced.
PC61BM, respectively. A similar trend of RMS roughness was also
Charge-trap estimation of MAPbI3/X-PDI interface by
photoluminescence
found for ZnO NP-covered X-PDI or PC61BM thin film (Figure 6e,
6f, 6g, and 6h), i.e., the F-PDI sample has the highest surface
roughness. Whereas the surface roughness of H-PDI sample is
in between, the Br-PDI sample has the least surface roughness.
The Br-PDI and PC61BM thin films showed the most flat surface
with a height variation less than ~50 nm, which is substantially
smaller than ~100 nm and ~150 nm of H-PDI and F-PDI thin films,
respectively. The F-PDI has the most uneven and rough surface
morphology that may be attributed to the strong molecular
interaction due to the fluorine substituent, causing high degree
of aggregation in thin film. This is consistent with what we found
for the broadening and red-shifting thin film absorption
spectrum of F-PDI (Figure 2b) and its relatively low solubility in
solution.
Photoluminescence (PL) is a simple method to study the charge-
trap effect between the surface of perovskite film and ETL.
Huang and co-workers37,38 reported that the existence of
surface trap states on the perovskite surface or grain boundary
could cause band bending close to the surface of the film,
lowering the surface bandgap than the bulk one. Therefore, we
performed steady-state PL measurement to study the
MAPbI3/ETL interface in devices with a configuration of
ITO/PEDOT:PSS/MAPbI3/ETL/air. Using an excitation light of 532
nm, which has a power of penetration length shorter than 100
nm much less than the thickness of the perovskite films (~300
nm), we can ensure a surface probing experiment. As shown in
Figure 8a, with the incident of excitation light from the air side,
the pristine perovskite film showed a PL peaked at 781 nm,
which is the same wavelength as the perovskite film covered
with H-PDI or F-PDI but not Br-PDI or PC61BM. The perovskite
film covered with Br-PDI or PC61BM exhibited a blue-shifted PL
peaked at 774 and 773 nm, respectively. Such PL wavelength
difference did not happen, if the excitation light was from ITO
side (Figure 8b). Our PL study indicates that both Br-PDI and
PC61BM passivate the trap states close to the top surface of the
perovskite film. Accordingly, both Br-PDI and PC61BM have a
better interfacial contact (and hence more efficient electron
extraction) of perovskite material than H-PDI or F-PDI
.
ITO Side
(b)
PC61BM+MAPbI3
Br-PDI+MAPbI3
F-PDI+MAPbI3
H-PDI+MAPbI3
Neat MAPbI3
Air Side
(a)
781nm
773nm
Fig 6 AFM images and height profile of X-PDI (a, b, c) or PC61BM (d) and ZnO NP-covered
X-PDI (e, f, g) or PC61BM (h) thin films, which were prepared from chloroform solution
(X-PDI layer) and ethanol solution (ZnO NP layer) on ITO/PEDOT:PSS/MAPbI3 substrates.
The scanning size of the images is 5 μm × 5 μm. The scale of colour bar is 0-125 nm (for
a, b, c, d) and 0-300 nm (for e, f, g, h).
781nm
781nm
774nm
781nm
781nm
781nm
781nm
781nm
Corresponding SEM images of MAPBI3/H-PDI, MAPBI3/F-PDI
,
MAPBI3/Br-PDI, and MAPBI3/PC61BM are shown in Figure 7a, 7b,
7c, and 7d, respectively. Apparently, Figure 7 reals a similar
trend of surface roughness in terms of the number of dark
domain in the image of X-PDI or PC61BM thin films on perovskite
active layer.
700
800
900
700
800
900
Wavelength (nm)
Wavelength (nm)
Fig 8 PL spectra of MAPbI3/X-PDI and MAPbI3/PC61BM with excitation light from
different incident direction, air side (a) and ITO side (b).
Photovoltaic characterization
The PVSCs with a configuration of ITO/PEDOT:PSS/MAPbI3/X-
PDI or PC61BM/Ag and ITO/PEDOT:PSS/MAPbI3/X-PDI or
PC61BM/ZnO NP/Ag were fabricated and characterized. The
detailed fabrication procedures of these devices can be found
in the experimental section. The characterization data of all
Fig 7 SEM images (~5 x 5 m) of X-PDI or PC61BM thin films on perovskite active layer.
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 5
Please do not adjust margins

Journal of Materials Chemistry A
Page 6 of 11
Please do not adjust margins
Journal Name
ARTICLE
PVSCs are summarized in Table 4. The corresponding current PDI has strong molecular interactions (
density (J) and voltage (V) plots of PVSCs without ZnO NP CBL and hence a serious aggregation in solid state, whereas both H-
-F-, F-FViinewteArrtaiccletiOonnlinse)
DOI: 10.1039/C7TA02617J
are illustrated in Figure 9.
PDI and Br-PDI are less serious. The molecular aggregation
causes defected thin film, an inhomogeneous and incomplete
thin film, of ETL on top of MAPbI3. In this regard, Br-PDI
outperforms H-PDI or F-PDI and its device performance is
comparable with that of PC61BM. Both Br-PDI and PC61BM
PVSCs show more reasonable Rs, FF, short circuit current (Jsc),
and hence PCE than those of H-PDI and F-PDI PVSCs.
Attempting to improve the low PCE of X-PDI or PC61BM
PVSCs, we introduce solution processed ZnO NP as CBL to
overcome or alleviate the problems associated with the ETL.
In addition to the data summarized in Table 4, the J-V plots of
Table 4
Summary of the photovoltaic properties of X-PDI- and PC61BM-based PVSCs
a
PCEavg
[%]
Voc
[V]
Jsc
FF
[%]
Rsh
Rs
J0
ETL
nid,dark
[mA/cm2]
(kΩ cm2)
(Ω cm2)
(mA cm-2)
H-PDI
F-PDI
Br-PDI
PC61BM
0.66
~0
3.23
4.13
0.64
0.17
0.75
0.75
6.65
0.12
14.64
13.76
15.4
21.7
29.5
40.4
0.511
51.78
0.118
0.091
274.1
460.6
37.60
24.52
-
-
-
-
W/O
ZnO NPs
1.77 × 10-3
2.49 × 10-1
3.19
2.47
H-PDI
F-PDI
7.78
~0
10.50
10.20
11.07
10.24
0.73
0.06
0.83
0.84
0.97
0.95
17.79
0.60
18.90
18.80
17.21
16.33
59.9
20.9
66.9
64.6
66.3
66.0
0.887
0.049
0.269
0.336
0.761
0.814
10.29
11.84
4.726
6.250
6.960
6.212
2.09 × 10-2
4.04 × 10-1
1.96 × 10-4
-
3.64
5.41
1.56
-
1.13
-
Br-PDI (forward)
Br-PDI (reverse)
PC61BM (forward)
PC61BM (reverse)
W/
ZnO NPs
6.15 × 10-4
-
a The average PCE is obtained from over 15 devices.
PVSCs with ZnO NP CBL are illustrated in Figure 10a (solar AM
1.5 G irradiation with 100 mW cm−2) and 10b (dark).
0
-5
(a) 0
-10
-4
-8
-15
H-PDI
F-PDI
Br-PDI
-20
-12
PC61BM
H-PDI
F-PDI
Br-PDI
PC61BM
-25
-16
-20
-24
0.0
0.2
0.4
0.6
0.8
1.0
Voltage (V)
Fig 9 J−V characteristics of X-DPI and PC61BM PVSCs without CBL of ZnO NP under
simulated AM 1.5G solar irradiation of 100 mW cm−2
.
0.0
0.2
0.4
0.6
0.8
1.0
Voltage (V)
First, as the frontier energy levels of each layer shown in
Figure 1a, the stepwise alignment of energy levels from
CH3NH3PbI3 perovskite layer to Ag cathode facilitates transport
and collection of electrons. Energy wise, this is true for each
X-PDI, PC61BM, and ZnO NP. Since the LUMO energy level of
each X-PDI is lower than that of PC61BM; the HOMO energy level
of each X-PDI is higher than that of PC61BM, electron extraction
and hole blocking are presumably better for MAPbI3/X-PDI than
for MAPbI3/PC61BM. However, the molecular interaction
(aggregation in solid state) and the thin film morphology have a
more critical role for the ETL studied herein.
For the devices with PC61BM ETL (as the control PVSL), an
average PCE of 4.13% was achieved. For the devices with X-
PDI ETL, very low average PCE of 0.66% was recorded for H-PDI
PVSCs and virtually no photovoltaic effect was found for F-PDI
PVSCs. As evident in spectroscopic and microscopic studies, F-
(b)1000
100
10
1
0.1
0.01
1E-3
1E-4
1E-5
1E-6
H-PDI
F-PBI
Br-PDI
PC61BM
-1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0
Voltage (V)
Fig 10 J−V characteristics of X-DPI and PC61BM PVSCs with CBL of ZnO NP under simulated
AM 1.5G solar irradiation (a) and dark (b).
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 6
Please do not adjust margins

Page 7 of 11
Journal of Materials Chemistry A
Please do not adjust margins
Journal Name
ARTICLE
As shown from the results, whereas the performance of the experimental section. Two key modelling parameters, the
View Article Online
DOI: 10.1039/C7TA02617J
F-PDI devices still remains very poor, PCE of Br-PDI, H-PDI, and reverse saturation current density (J0) and the ideality factor
PC61BM PVSCs with CBL of ZnO NP have notably improved to (nid,dark), of all four devices (with CBL of ZnO NP) are summarized
10.50%, 7.78%, and 11.07%, respectively. The improvement is in Table 4. Much smaller values of J0 and nid,dark were obtained
comprehensively in all aspects, including Voc, JSC, FF, shunt for the Br-PDI and PC61BM PVSCs compared to that of F-PDI
resistance (Rsh), and Rs (Table 4). We can logically infer that the devices, whereas J0 and nid,dark of H-PDI devices are in between
CBL of ZnO NP has an effective function of interspace-filling those of Br-PDI and F-PDI ones. Such modelling results signify
upon the aggregate domains of Br-PDI, H-PDI, and PC61BM thin that Br-PDI and PC61BM are better materials than H-PDI and
films. This reduces the direct contact of Ag cathode metal to much better than F-PDI for electron selection at the interface of
perovskite material and increases the effective contact area for MAPbI3/ETL because of less charge recombination loss.
electron transfer from MAPbI3 perovskite, and the electron Higher values of Voc observed for Br-PDI and PC61BM than H-PDI
mobility (also direct current conductivity) of ETL is also elevated. and F-PDI PVSCs also can be explained by the less interfacial
However, why F-PDI PVSCs were not beneficial from the CBL of trap-assisted charge recombination,41 consistent to the
ZnO NP?
Inspired by the AFM image of the thin film, we have a
modelling results obtained herein.
It is interesting to know that Br-PDI PVSCs always have a
proposal to explain why CBL of ZnO NP works effectively for Br- higher Jsc output than PC61BM PVSCs whether or not devices
PDI, H-PDI, and PC61BM but does not work for F-PDI devices. have ZnO NP (Table 4). Considering the DC conductivity,
As schematically shown in Figure 11a and 11c, an interspace- electron mobility, photoluminescence of perovskite/ETL
filling of ETL by ZnO NPs prevents the diffusion of cathode metal interface, AFM thin film surface roughness of two materials, we
(Ag) through the defect of ETL, which will form a direct contact can infer that the better energy alignment, i.e., lower LUMO and
of MAPbI3, facilitating the undesired charge recombination HOMO energy levels of Br-PDI than PC61BM (see data listed in
process. Such interspace-filling by ZnO NPs is not effective for F- Table 2) are a main factor. The better energy alignment enables
PDI. Due to the large and many interspace between the Br-PDI to be a more effective material in electron extraction and
aggregate domains of F-PDI, the solvent (ethanol) used for ZnO more efficient in hole-blocking from perovskite material.
NPs goes through the ETL readily and washes away part of However, larger energy difference between HOMO (of MAPbI3)
perovskite material underneath the interspace of aggregate and LUMO (of ETL) energy levels renders a higher Voc of PC61BM
domains, enabling the direct contact of deposited cathode PVSCs, which offsets the Jsc deficiency of the devices.
metal (Ag) (Figure 11b), which is seriously detrimental to the Depending on FF, the PCE of Br-PDI and PC61BM PVSCs having
performance of PVSCs.
ZnO NP CBL is comparable with each other.
Comparing the hysteresis of PVSCs with ZnO NP (Figure
12), ~3% and ~0.5% are the degree of hysteresis effect for Br-
PDI and PC61BM devices, respectively, based on the calculation
of 2  (FFforward - FFreverse)/(FFforward + FFreverse) with corresponding
data listed in Table 4. We are pleased to know that the
hysteresis effect of Br-PDI PVSCs is just a little bit more than
that of PC61BM PVSCs.
0
Br-PDI forward scan
Br-PDI reverse scan
PC61BM forward scan
-4
PC61BM reverse scan
-8
-12
-16
-20
0.0
0.2
0.4
0.6
0.8
1.0
Fig 11 Schematic depict of the cross-sectional view of H-PDI (a), F-PDI (b), and Br-PDI or
PC61BM (c) ETLs (yellow) as well as their interspace-filling by ZnO NPs (dark cyan), which
prevent the contact of cathode metal Ag (light gray) with MAPbI3 perovskite (red).
Voltage (V)
Fig 12 The reverse (1.0 V → - 0.3 V, step 0.02 V, delay time 500 ms) and forward (- 0.3V
→ 1.0V, step 0.02 V, delay time 500 ms) J−V scans of Br-PDI and PC61BM PVSCs with CBL
or CIL of ZnO NP.
The inhibited charge recombination process of Br-PDI and
PC61BM PVSCs with CBL of ZnO NP gets evidences from a
significantly weaker dark current density compared to that of H- Conclusions
PDI and F-PDI PVSCs with CBL of ZnO NP (Figure 10b).
We have developed and demonstrated three solution
processable perylene diimides, i.e., X-PDI , or Br, as non-
fullerene electron accepting and electron transporting
Moreover, the dark J-V curves shown in Figure 10b can be
modelled using the Shockley diode equation for a single
junction device.39,40 Detailed modelling method can be found in
, X = H, F
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 7
Please do not adjust margins

Journal of Materials Chemistry A
Page 8 of 11
Please do not adjust margins
Journal Name
ARTICLE
materials in inverted PVSCs. Whereas H-PDI or F-PDI performs previously unknown F-PDI was synthesized according to the
View Article Online
DOI: 10.1039/C7TA02617J
unsatisfactorily, our best PVSCs is based on Br-PDI exhibiting procedures shown in Scheme 1. Patterned ITO-coated glass
PCE of 3.23%, which is just a bit shy of 4.13% of fullerene substrates with a sheet resistance of 15 ohm sq−1 were
(PC61BM) PVSCs. Through a series of physical, spectroscopic, purchased from Ruilong Tech. PEDOT:PSS aqueous solution
and microscopic studies, we have understood that the low (Heraeus CleviosTM P VP Al 4083) was purchased from Heraeus.
solubility of F-PDI is a major factor causing poor quality of the Methylammonium iodide (MAI, CH3NH3I, >99.5%) and lead
thin film, rendering virtually no photovoltaic effect of F-PDI
.
iodide (PbI2) were purchased from Lumtec and Acros,
Although the solubility is better than F-PDI, the inferior electron respectively. The fullerene PC61BM (>99.5%) was purchased
mobility and conductivity compared with those of Br-PDI make from Solenne. Unless otherwise stated, all chemicals were
H-PDI PVSCs relatively worse performance.
Having the purchased from Sigma-Aldrich and used as received.
highest solubility, electron mobility, and conductivity among X-
PDI, Br-PDI-based PVSCs are almost as efficient as PC61BM- Synthesis
of
5-fluoro-2,9-bis(2-ethylhexyl)-anthra[2,1,9-
based PVSCs. Except for F-PDI, PCE of PVSCs based on X-PDI def:6,5,10-d',e',f']diisoquinoline-1,3,8,10(2H,9H)-tetraone (F-
or PC61BM all get significantly improved to 7.78%, 10.50%, and PDI)
11.07%, respectively, once solution processed ZnO NP is A mixture of Br-PDI (2 g, 2.82 mmol), KF (2 g, 34.4 mmol), and
included in the PVSCs as a CBL. We infer that the CBL of ZnO DMF (125 mL) was stirred at 130°C for 12 h under argon
NP has a function of interspace-filling on the defect of H-PDI
,
atmosphere. After the removal of solvent, the residue was
Br-PDI, or PC61BM thin film, reducing the direct contact of Ag subjected to purification by silica gel column chromatography,
cathode to the perovskite material. Due to the strong molecular using CH2Cl2/hexane (3/1, v/v) as an eluent, affording F-PDI
interaction, F-PDI aggregates seriously in thin film creating too (0.82 g) as a red solid in a yield of 50%. 1H NMR (400 M Hz, CDCl3)
many and too large defects to be remedied or improved with or δ ppm: 9.06 (d, 2H, J = 8.0 Hz), 8.68 (m, 5H), 8.49 (d, 2H, J = 13.6
without CBL of ZnO NP. Very poor electron mobility and Hz), 4.17–4.09 (m, 4H), 1.97 (m, 2H), 1.24–1.53 (m, 16H), 0.94
conductivity of F-PDI are two other factors devastating its (t, 6H, J = 7.6 Hz), 0.89 (t, 6H, J = 7.6 Hz). 13C NMR (100 M Hz,
PVSCs.
CDCl3) δ ppm: 163.62, 163.45, 163.32, 162.57, 162.52, 162.49,
134.35, 133.83, 131.75, 130.97, 130.33, 128.85, 128.61, 128.30,
128.05, 126.29, 126.06, 123.61, 123.31, 123.08, 122.67, 122.00,
121.68, 44.49, 44.44, 44.37, 37.97, 37.94, 30.79, 30.75, 28.71,
28.67, 24.10, 24.07, 23.08, 23.06, 14.08, 10.63, 10.60. MALDI-
MS for C41H43FN2O4, m/z = 646.79, found, m/z = 646.32.
Experimental
General Instruments
1H and 13C NMR spectra were recorded on a Brucker AV-400 MHz or
AV-500 MHz Fourier transform spectrometer at room temperature.
Fast atom bombardment (FB) high resolution mass spectroscopy
(HRMS) was performed by Mass Spectroscopic Laboratory as in-
house service of the Institute of Chemistry, Academia Sinica. UV-
visible absorption spectra were recorded on a Perkin-Elmer lambda
900 spectrometer and Hewlett-Packard 8453 diode array
spectrometer for solution and thin film samples, respectively. The PL
spectrum was measured by LP920 flash photolysis spectrometer
(Edinburgh Instrument, Kirkton Campus, Livingston, England) at
room temperature. A 532-nm green laser (Brilliant B laser, Quantel
Company, Newbury, Berkshire, UK) was used as the excitation source
in PL measurement. The thin film AFM images were obtained by
Digital Instrument D3100CL. The thin film SEM images were
characterized by FEI Nova NanoLab™ 600 DualBeam (FIB/SEM).
Synthesis of ZnO NPs. ZnO NPs were synthesized by a solution-
precipitation process according to literature procedures.43
Briefly, zinc acetate dihydrate (2.95 g) was dissolved in
methanol (125 mL) at room temperature. A potassium
hydroxide solution (1.48 g in 65 mL methanol) was then added
dropwise within 30 min and stirred for 3 hr at 65 °C. After
cooling, the solution was then decanted and the precipitate was
washed twice with ethyl acetate and ethanol, respectively.
Afterward, ethanol was added to disperse the precipitates and
produce ZnO NPs solution with a concentration of 6 mg mL-1.
Before use, the ZnO NPs solution was filtered through a 0.45
mm PVDF syringe filter.
Cyclic voltammetry of X-DPI
Using an Electrochemical Analyser BAS C3 operated at a
scanning rate of 50 mV s-1, the cyclic voltammetry (CV) was
Materials
All chemicals were purchased from Aldrich, Alfa Aesar, Acros,
and TCI Chemical Co., and they were used without further
purification. Solvents such as dichloromethane (CH2Cl2), N-
methylpyrrolidinone (NMP), N,N-dimethylformamide (DMF)
were distilled after drying with appropriate drying agents. The
dried solvents were stored over 4 Å molecular sieves before use.
2,9-bis(2-ethylhexyl)-anthra[2,1,9-def:6,5,10-
d',e',f']diisoquinoline-1,3,8,10(2H,9H)-tetraone (H-PDI) and 5-
bromo-2,9-bis(2-ethylhexyl)-anthra[2,1,9-def:6,5,10-
d',e',f']diisoquinoline-1,3,8,10(2H,9H)-tetraone (Br-PDI) were
synthesized according to the procedures of literature.29,42 The
obtained in
(Bu4NPF6, 0.1 M) supported CH2Cl2 solution containing X-PDI or
PC61BM (1.0
10-3 M) at room temperature. The measurement
a tetrabutylammonium hexafluorophosphate
×
was conducted with a glassy carbon rod as the working
electrode and a platinum wire as the counter electrode. The
Ag–Ag+ (0.1 M of AgNO3 in acetonitrile) was used as the
reference electrode. The ferrocene/ferrocenium (Fc/Fc+) was
used as the reference to calibrate the redox potentials. Fc/Fc+
potential in our CV measurement was +0.10 V relative to Ag-
Ag+. For a thin film CV measurement, X-PDI or PC61BM thin film-
coated platinum electrode was prepared from a drop-cast
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 8
Please do not adjust margins

Page 9 of 11
Journal of Materials Chemistry A
Please do not adjust margins
Journal Name
ARTICLE
chloroform solution. The CV of X-PDI or PC61BM thin film was mL-1 for PbI2 and 40 mg mL-1 for MAI, respectively. Both solutions and
View Article Online
DOI: 10.1039/C7TA02617J
studied in an anhydrous and nitrogen-saturated solution of PEDOT:PSS-coated ITO substrates were heated at 100 °C for 10 min
acetonitrile containing 0.10 M Bu4NPF6. The HOMO and LUMO before being used. The PbI2 solution was spun on preheated
energy levels of X-PDI or PC61BM were calculated based on substrate (5000 rpm for 40 sec) and then annealed at 70 °C for 10
op
equations, ELUMO = -(E1/2red1 + 5.10) [eV] and EHOMO = ELUMO - Eg
min. The MAI solution was then spun on top of the dried PbI2 film
[eV], where Egop is the optical energy gap corresponding to the (6000 rpm for 30 sec), followed by annealing at 100 °C for 2 hr. The
absorption onset wavelength of absorption spectra; the energy X-PDI solution (15 mg mL-1 in chloroform) was then casted by spin-
level of Fc/Fc+ related to vacuum level is 5.10 eV based on coating (1000 rpm for 60 sec) on top of the preformed perovskite
literature reports.9,31
layers. The thickness of X-PDI was near 200 nm. Afterward, ZnO NPs
layer (about 10-20 nm) was deposited by spin-coating ZnO NP
Determination of direct current conductivity of X-PDI and precursor solution (3000 rpm for 60 sec), then dry in ambient
PC61BM condition without thermal annealing. The opaque Ag cathode (150
The X-PDI or PC61BM thin films were prepared by spin-coating nm) was then deposited from thermal evaporator under high
method from chloroform solution on an ITO bottom electrode. vacuum (<10-6 Torr). Contributions to the Jsc from regions outside the
A layer of Al (100 nm) as top electrode was then deposited active area were eliminated using illumination masks with an
through an area (0.2 x 0.5 cm2) patterned mask in a thermal aperture size of 0.12 cm2. The bulk resistances, Rsh and Rs, of samples,
evaporation chamber under high vacuum (<10-6 Torr). The were calculated from Ohm's law, R = V I-1, with at least ten separate
electrical conductivity was analysed on the current-voltage measurements made for each sample.
characteristics of the thin film. The DC (direct current)
conductivity (
0), which is in unit of siemens per metre (S m-1), illuminated under light intensity of 100 mW cm−2 by using a 150 W
of the thin film material in such device can be obtained by Class AAA solar simulator (XES-100S1, SAN-EI). The light intensity was
Claude Pouillet law, R = L (
0A)-1, where the resistance (R) of a determined by using a standard silicon photodiode (Hamamatsu
For the evaluation of the device performance, the devices were


given thin film increases with length (L) but decreases with S1133) calibrated by the National Renewable Energy Laboratory. The
increasing cross-sectional area (A). The 0 is defined as the J−V curves were measured using a Keithley 2400 digital source meter.
inverse of resistivity (ρ, in unit of Ω m), i.e.,

0 = ρ-1.
Unless otherwise stated, the scan rate was set at 0.15 V s−1.
From the J-V characteristics of solar cells in dark circumstance,
the dark current densities (JD) of a solar cell can be estimated
Electron mobility determination by SCLC
In the case of a field dependent mobility as described by the according to the thermionic emission model,39,40 JD = J0[exp(qV (nkT)-
Poole-Frenkel effect, the space-charge limited (SCL) current 1)-1], where n is the dark ideality factor, J0 is the reverse saturation
density (JSCL) is given by the modified Mott-Gurney equation,44 current, q is the elementary charge, and kT is the thermal energy
JSCL = 9ε0εrμVin2exp(γV1/2 L-1/2)(8L3)-1, where ε0 is the permittivity (Boltzmann’s constant and temperature). It is well known that the
of free space (8.854 × 10−12 F m−1); εr is the relative dielectric ideality factor of PVSC is normally between 1 and 2.41 The ideality
constant of the thin film, which was assigned to be 3; μ is the factor is close to 1 for an ideal trap-free device, while it is closer to or
charge carrier mobility; L is the thickness of the device; Vin is the larger than 2 for the case of the deep and dominant charge
voltage dropped across the sample given by Vin
= V - Vbi - Vrs, recombination induced by the trap.
where V is the applied voltage, Vbi the built-in voltage, and Vrs is
the voltage drop due to the series resistance of the contacts; γ
is the electric field-activation factor of mobility and accounts for
the degree of disorder, particularly the energetic level
distribution of the carrier hopping sites in the material.
Accordingly, in the region of a log J - log V curve having a slope
Acknowledgements
This research was support in part by Sustainability Science
Research Program of Academia Sinica, Institute of Chemistry of
Academia Sinica, National Taiwan University, and the Ministry
of Science and Technology of Taiwan.
~2, the
L-1)1/2 plot (see Figure 5b and 5d), i.e.,

can be calculated from the intercept of a ln(JL2 V-2)-(V
=
Notes and references
exp(intercept)8L(9ε0εr)-1. The expression of a field dependent
SCL current (SCLC) yields a reasonably good fit to the measured
J–V curves of single-carrier devices (see the fitting curves in
Figure 5a and 5c).45
1
J. Burschka, N. Pellet, S. Moon, R. Humphry-Baker, P. Gao, M. K.
Nazeeruddin and M. Grätzel, Nature, 2013, 499, 316.
H. Zhou, Q. Chen, G. Li, S. Luo, T. Song, H.-S. Duan, Z. Hong, J.
You, Y. Liu and Y. Yang, Science, 2014, 345, 542.
2
3
4
M. Liu, M. B. Johnston and H. J. Snaith, Nature, 2013, 501, 395.
A. Mei, X. Li, L. Liu, Z. Ku, T. Liu, Y. Rong, M. Xu, M. Hu, J. Chen,
Y. Yang, M. Grätzel, H. Han, Science, 2014, 345, 295.
M. Saliba, T. Matsui, J. Seo, K. Domanski, J. P. Correa-Baena, M.
K. Nazeeruddin, S. M. Zakeeruddin, W. Tress, A. Abate, A.
Hagfeldtd, M. Grätzel, Energy Environ. Sci., 2016, 9, 1989.
J.-Y. Jeng, Y.-F. Chiang, M.-H. Lee, S.-R. Peng, T.-F. Guo, P. Chen,
T.-C. Wen, Adv. Mater., 2013, 25, 3727.
Fabrication and characterization of perovskite solar cells
ITO-coated glass substrates were cleaned stepwise in detergent,
water, acetone, and isopropyl alcohol under ultra-sonication for 20
min each and subsequently pre-treated by UV-ozone for 60 min. The
PEDOT:PSS layer (25 nm) was spin-coated on the ITO surface and
then annealed at 120 °C for 15 min. The CH3NH3PbI3 perovskite layer
(~380 nm) was prepared following two-step solution deposition, as
described in our previous work.35,36 Briefly, PbI2 and MAI were
dissolved into DMF and 2-propanol with concentrations of 450 mg
5
6
7
J. H. Heo, H. J. Han, D. Kim, T. K. Ahnb, S. H. Im, Energy Environ.
Sci., 2015, 8, 1602.
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 9
Please do not adjust margins

Journal of Materials Chemistry A
Page 10 of 11
Please do not adjust margins
Journal Name
ARTICLE
8
9
H. Azimi, A. Senes, M. C. Scharber, K. Hingerl, C. J. Brabec, Adv.
Energy Mater., 2011, 1, 1162.
J. Wu, Y. Ma, N. Wu, Y. Lin, J. Lin, L. Wang, C.-Q. Ma, Org.
Electron., 2015, 23, 28.
28 C.-Y. Chang, W.-K. Huang, Y.-C. Chang, K.-T. Lee, C.V-iTe.wCAhrteicnle,OJ.nline
Mater. Chem. A, 2016, 4, 640.
29 L. Zhang, L. Wang, G. Zhang, J. Yu, X. Cai, M. Teng, Y. Wu, Chin. J.
Chem. 2012, 30, 2823.
DOI: 10.1039/C7TA02617J
10 C. B. Nielsen, S. Holliday, H.-Y. Chen, S. J. Cryer, I. McCulloch,
Acc. Chem. Res., 2015, 48, 2803.
11 Th. B. Singh, S. Erten, S. Günes, C. Zafer, G. Turkmen, B. Kuban,
Y. Teoman, N.S. Sariciftci, S. Icli, Org. Electron., 2006, 7, 480.
12 X. Zhan, A. Facchetti, S. Barlow, T. J. Marks, M. A. Ratner, M. R.
Wasielewski, S. R. Marder, Adv. Mater., 2011, 23, 268.
30 C.-W. Ge, C.-Y. Mei, J. Ling, J.-T. Wang, F.-G. Zhao, L. Liang,H.-J.
Li, Y.-S. Xie, W.-S. Li, Polym. Sci. Part A Polym. Chem., 2014, 52,
1200.
31 J. Yi, Y. Ma, J. Dou, Y. Lin, Y. Wang, C.-Q. Ma, H. Wang, Dyes
Pigments, 2016, 126, 86.
32 L. Perrin, P. Hudhomme, Eur. J. Org. Chem. 2011, 2011, 5427.
13 Z. An, J. Yu, S. C. Jones, S. Barlow, S. Yoo, B. Domercq, P. Prins, L. 33 R. Mercadantea, M. Trsicas, J. Duffb, R. Aroca, J. Mol. Struct.-
D. A. Siebbeles, B. Kippelen, S. R. Marder, Adv. Mater., 2005, 17,
2580.
14 M. Funahashi, A, Sonoda, Org. Electron., 2012, 13, 1633.
15 Y. Chen, Y. Feng, J. Gao, M. Bouvet, J. Colloid Interf. Sci., 2012,
368, 387.
16 B. A. Gregg, R. A. Cormier, J. Phys. Chem. B, 1998, 102, 9952.
17 H. Zhang, L. Xue, J. Han, Y. Q. Fu, Y. Shen, Z. Zhang, Y. Li, M.
Wang, J. Mater. Chem. A, 2016, 4, 8724.
18 J. Huang, Z. Gu, L. Zuo, T. Ye, H. Chen, Solar Energy, 2016, 133,
331.
Theochem., 1997, 394, 215.
34 J. D. Fernandesa, P. H. B. Aoki, R. F. Arocac, W. D. M. Juniora, A.
E.-d. Souzaa, S. R. Teixeiraa, M. L. Braungera, C.-d. A. Olivatia, C.
J. L. Constantino, Mater. Res., 2015, 18(Suppl 2), 127.
35 C.-Y. Chang, W.-K. Huang, J.-L. Wu, Y.-C. Chang, K.-T. Lee, C.-T.
Chen, Chem. Mater., 2016, 28, 242.
36 C.-Y. Chang, Y.-C. Chang, W.-K. Huang, W.-C. Liao, H. Wang, C.
Yeh, B.-C. Tsai, Y.-C. Huang, C.-S. Tsao, J. Mater. Chem. A, 2016,
4, 7903.
37 C. Sun, Z. Wu, H.-L. Yip, H. Zhang, X.-F. Jiang, Q. Xue, Z. Hu, Z.
Hu, Y. Shen, M. Wang, F. Huang, Y. Cao Adv. Energy Mater.,
2016, 6, 1501534.
19 Z. Zhu, J.-Q. Xu, C.-C. Chueh, H. Liu, Z. Li, X. Li, H. Chen, A. K.-Y.
Jen, Adv. Mater. 2016, 28, 10786.
20 Z.-G. Zhang, B. Qi, Z. Jin, D. Chi, Z. Qi, Y. Li, J. Wang, Energy
Environ. Sci., 2014, 7, 1966.
38 Y. Shao, Z. Xiao, C. Bi, Y. Yuan, J. Huang, Nat. Comm., 2014, 5,
5784.
39 X. Li, H. Zhu, K. Wang, A. Cao, J. Wei, C. Li, Y. Jia, Z. Li, X. Li, D.
Wu, Adv. Mater., 2010, 22, 2743.
40 X. Yu, X. Shen, X. Mu, J. Zhang, B. Sun, L. Zeng, L. Yang, Y. Wu, H.
He, D. Yang, Sci. Rep., 2015, 5, 17371.
41 A. Zekry, G. Eldallal, Solid-State Electron., 1998, 31, 91.
42 Á. J. Jiménez, M. Sekita, E. Caballero, M. L. Marcos, M. S.
Rodríguez-Morgade, D. M. Guldi, T. Torres, Chem. Eur. J., 2013,
19, 14506.
21 J. Min, Z.-G. Zhang, Y. Hou, C. O. R. Quiroz, T. Przybilla, C.
Bronnbauer, F. Guo, K. Forberich, H. Azimi, T. Ameri, E. Spiecker,
Y. Li, C. J. Brabec, Chem. Mater., 2015, 27, 227.
22 S. S. Kim, S. Bae, W. H. Jo, RSC Adv., 2016, 6, 19923.
23 P.S. Mbule, T.H. Kim, B.S. Kim, H.C. Swart, O.M. Ntwaeaborwa,
Solar Energy Mater. Solar Cells, 2013, 112, 6.
24 Y. Jouane, S. Colis, G. Schmerber, P. Kern, A. Dinia, T. Heisera,
Y.-A. Chapuis, J. Mater. Chem., 2011, 21, 1953
25 O. M. Ntwaeaborwa, R. Zhou, L. Qian, S. S. Pitale, J. Xue, H.C.
Swart, P.H. Holloway, Physica B, 2012, 407, 1631.
26 C.-Y. Chang, K.-T. Lee, W.-K. Huang, H.-Y. Siao, Y.-C. Chang,
Chem. Mater., 2015, 27, 5122.
43 D. Liu, T. L. Kelly, Nat. Photon., 2014, 8, 133.
44 P. N. Murgatroyd, J. Phys. D: Appl. Phys., 1970, 3, 151.
45 L. C. Palilis, M. Uchida, Z. H. Kafafi, IEEE J. Sel. Top. Quantum
Electron., 2004, 10, 79.
27 C.-Y. Chang, W.-K. Huang, Y.-C. Chang, Chem. Mater., 2016, 28,
6305.
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 10
Please do not adjust margins

Page 11 of 11
Journal of Materials Chemistry A
View Article Online
DOI: 10.1039/C7TA02617J
Graphical abstracts
Mono-halogenated perylene diimides as solution-processable electron
transporting layer in inverted perovskite solar cells with ZnO nanoparticle
cathode buffer layer
C2H5
C4H9
O
O
N
O
H-PDI
(X = H)
X
F-PDI
(X = F)
Br-PDI
(X = Br)
N
O
C4H9
C2H5