Pore size dependent hysteresis elimination in perovskite solar cells based on highly porous TiO2 films with widely tunable pores of 15-34 nm

Article abstract of DOI:10.1021/acs.chemmater.6b03445

Pore size and porosity of the porous materials play an important role in catalysis, dye-sensitized solar cells and mesoscopic perovskite solar cells (PSC), etc. Increasing pore size and porosity of mesoporous TiO₂ is crucial for facilitating pore-filling of perovskite, charge extraction on TiO₂/CH₃NH₃PbI₃ interface and thus cell performance enhancement. Highly porous TiO₂ films (TFs) with a large pore size that extends the limit of particle size have been achieved through a novel TiO₂ paste using copolymer P123 as a pore-adjusting agent and 2-butoxyethyl acetate as a solvent. A highly porous structure with the pore size of 34.2 nm and porosity of 73.5% has been obtained, the porosity of which is the largest that has ever been reported in the screen-printed TiO₂ thick films. The pore size and porosity of TFs can be successively adjusted in a certain range by tuning the P123 content in the pastes. As particle size and surface area of TFs are kept almost constant, the specific investigation on the effect of varied pore size on the performance of bilayer-structured PSCs becomes possible. The hysteresis phenomenon, the notorious problem of PSCs, is found to depend greatly on pore size and porosity of TFs, that is, pore-filling of perovskite. The suppressing effect of highly porous TFs on hysteresis by avoiding charges accumulation on the interface due to enhanced interfacial contact is proved by the invariable photocurrent response after prebias treatment. A hysteresis-free solar cell with an efficiency of 15.47% was achieved by depositing a 242 nm-thick perovskite capping layer upon 350 nm-thick TF with a pore size of 34.2 nm. This method developed for the preparation of highly porous TFs provides a new way to fabricate hysteresis-free PSCs and is widely applicable for the fabrication of other mesoporous metal oxide films with large pore sizes.

Full text of DOI:10.1021/acs.chemmater.6b03445

Subscriber access provided by United Arab Emirates University | Libraries Deanship
Article
The Pore Size Dependent Hysteresis Elimination in Perovskite Solar Cells
Based on Highly Porous TiO2 Films with Widely Tunable Pores of 15-34nm
Jun Shao, Songwang Yang, Lei Lei, Qipeng Cao, Yu Yu, and Yan Liu
Chem. Mater., Just Accepted Manuscript • DOI: 10.1021/acs.chemmater.6b03445 • Publication Date (Web): 20 Sep 2016
Downloaded from http://pubs.acs.org on September 23, 2016
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Chemistry of Materials is published by the American Chemical Society. 1155 Sixteenth
Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 15
Chemistry of Materials
1
2
3
4
5
6
7
8
9
The Pore Size Dependent Hysteresis Elimination in Perovskite
Solar Cells Based on Highly Porous TiO2 Films with Widely
Tunable Pores of 15-34nm
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Jun Shao†,‡, Songwang Yang*,†, Lei Lei†, Qipeng Cao†, Yu Yu†,‡ and Yan Liu*,†
†
CAS Key Laboratory of Materials for Energy Conversion, Shanghai Institute of Ceramics, Chinese Academy of
Sciences, 588 Heshuo Road, Shanghai, 201899, P. R. China.
‡
University of Chinese Academy of Sciences, Beijing 100039, P. R. China.
ABSTRACT: Pore size and porosity of the porous materials play an important role in catalysis, dye-sensitized solar cells,
and mesoscopic perovskite solar cells (PSC), etc. Increasing pore size and porosity of mesoporous TiO2 is crucial for
facilitating pore-filling of perovskite, charge extraction on TiO2/CH3NH3PbI3 interface and thus cell performance
enhancement. Highly porous TiO2 films (TFs) with a large pore size that extends the limit of particle size have been
achieved through a novel TiO2 paste using co-polymer P123 as pore-adjusting agent and 2-butoxyethyl acetate as solvent.
A highly porous structure with the pore size of 34.2 nm and porosity of 73.5% has been obtained, the porosity of which is
the largest that has ever been reported in the screen-printed TiO2 thick films. The pore size and porosity of TFs can be
successively adjusted in a certain range by tuning the P123 content in the pastes. As particle size and surface area of TFs
are kept almost constant, the specific investigation on the effect of varied pore size on the performance of
bilayer-structured PSCs becomes possible. Hysteresis phenomenon, the notorious problem of PSCs, is found to depend
greatly on pore size and porosity of TFs, that is, pore-filling of perovskite. The suppressing effect of highly porous TFs on
hysteresis by avoiding charges accumulation on the interface due to enhanced interfacial contact is proved by the
invariable photocurrent response after pre-bias treatment. A hysteresis-free solar cell with an efficiency of 15.47% was
achieved by depositing 242-nm-thick perovskite capping layer upon 350-nm-thick TF with a pore size of 34.2 nm. This
method developed for the preparation of highly porous TFs provides a new way to fabricate hysteresis-free PSCs and is
widely applicable for the fabrication of other mesoporous metal oxide films with large pore sizes.
mp-TiO2 films are stable to full-solar-spectrum
irradiation and it will be possible for the PSC devices
INTRODUCTION
to withstand stressful long-term light irradiation4.
Porous materials are of great interest because of their
Until
bilayer-structured PSCs with
mesoporous layer and capping layer, are still the most
popular choice for highly efficient PSCs10-12
now,
mesoscopic
PSCs,
particularly
wide applications in adsorption, separation, catalysis,
biotechnology, sensing, energy storage and
conversion1. Design and synthesis of mesoporous
TiO2 (mp-TiO2) materials with controllable pore size
have extending implications for the particular
applications in photocatalysts, lithium-ion batteries,
quantum dot-sensitized solar cells, dye-sensitized
solar cells (DSSCs), and hybrid lead halide perovskite
solar cells (PSCs). PSCs, which employing the new
light-harvesting material, e.g., methyl ammonium
lead tri-halide (CH3NH3PbX3), have seen tremendous
power conversion efficiencies (PCEs) exceeding 20%2.
As CH3NH3PbX3 is found to be good at transporting
both electrons and holes3, PSCs with high
performance are also achieved with insulating Al2O3
scaffold4 or with planar structure abandoning the
mesoporous layer at all5, 6. Previous studies have
pointed out that mp-TiO2 materials are extremely
unstable to UV light irradiation and may cause
degradation of perovskite as they are typical
photocatalysts7-9. Fortunately, it appears that the
a
combination of
.
The mp-TiO2 layer functions as both scaffold for
the infiltration of perovskite and the electron
injection layer for the effective separation of
photoelectron and holes. Properties of mp-TiO2, such
as their crystallinity13, morphologies14-16, and particle
size17, 18, are widely investigated in DSSCs and
mesoscopic PSCs. Recent research demonstrate that
the infiltration of PbI2 or CH3NH3PbI3 into the TiO2
scaffold is crucial for effective perovskite conversion12,
successful charge separation19 and thus better cells
performance. Besides the improvement of active layer
forming procedure20, 21, enlarging the pore size and
porosity of TiO2 scaffold is an alternative way to
accomplish
the
target
of
complete
perovskite-infiltration. It has been pointed out that
the device performance of PSCs is affected by the
particle size of TiO2. TiO2 with larger particle size is
ACS Paragon Plus Environment

Chemistry of Materials
Page 2 of 15
favourable in PSCs as it creating wider tunnels which probable pore size, calculated from pore size
are beneficial for the perovskite penetration17. In such
situation, the device performance is essentially
affected by the optimization of pore size and porosity
of mp-TiO2. However, further increase of the TiO2
particle size reduces the efficiency as a result of
decreased porosity22. Another way to tune the pore
size of TiO2 film is by surfactants or polymers. For
distribution curve) and a high porosity of 73.5% are
achieved by applying P123 as pore-adjusting agent
and 2-butoxyethyl acetate (BCA) as solvent in the
screen-printable paste. As pore size and porosity of
TFs enlarges, perovskite deposited on scaffold
exhibits better infiltration by the evidence of thinner
capping layer. The effect of varying pore size of TFs
on the performance of bilayer-structured PSCs is
investigated specifically by excluding other factors
such as morphology or particle size of TiO2
nanoparticles. PSC devices fabricated on such highly
porous TFs show higher PCE and diminished
hysteresis with larger pore-size and better pore-filling
of perovskite. PSCs are pre-treated under forward
bias or reverse bias in dark to accelerate ion
migration before photocurrent-voltage and steady
state photocurrent characterization. PSCs based on
highly porous TFs suppress slow photocurrent
response after pre-bias treatment and provide the
evidence that the hysteresis problem is eliminated by
avoiding charges accumulation on the interface of
TiO2/CH3NH3PbI3 due to effective infiltration of
perovskite. Hysteresis-free PSC with a PCE of 15.47%
was finally achieved by depositing a dense and
fully-covered perovskite onto the TF with
34.2-nm-pore-size and 73.5-%-porosity (calculated by
the Barrett-Joyner-Halenda (BJH) method). The novel
method for preparing screen-printable paste to form
highly porous TiO2 films we present here has a bright
application prospect and research significance as it
can be extended to electronically conductive
adhesives, ceramic materials, mesoporous catalyst
and other fields.
1
2
3
4
5
6
7
8
example,
Triblock-copolymer,
9
HO(CH2CH2O)20(CH2CH(CH3)O)70(CH2-CH2O)20H
(Pluronic P123), is commonly used to form highly
porous TiO2 films23-25. However, the average pore size
achieved by applying P123 as template in sol-gel,
hydrolysis or hydrothermal method is limited to ~10
nm, which is not large enough for the complete
infiltration of perovskite particles. On the other hand,
the low crystallinity of TiO2 achieved by using
polymer in sol-gel method hinders fast electron
injection and efficiency enhancement26. To the best
of our knowledge, TiO2 thick films above 10 µm with
high porosity above 70% prepared by printing viscous
pastes haven't been reported. Therefore, further
tuning the pore size and porosity to exceed the limit
of particle size is challenging.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Despite the pursue of high PCE through the
optimization of chemical and physical properties of
mp-TiO2, there are more fierce challenges which
hinder the real commercialization of PSCs such as
degradation of perovskite material itself and
hysteresis phenomenon27. Abnormal hysteresis which
strongly depending on the scan direction and rate
during J-V measurement affects the accurate
measurement of performance and long-term stability
of PSC devices. Different magnitudes of hysteresis
can be observed for a variety of device architectures28,
RESULTS AND DISCUSSION
29
(planar PSCs30, mesoscopic PSCs on mp-TiO2,
superstructured PSCs on Al2O331,
and inverted
32
Porous TiO2 film. Porous TiO2 films were prepared
from a novel paste, which contains the hydrothermally
synthesized TiO2 nanoparticles, ethyl cellulose (EC), P123
and BCA. We firstly characterized the physical properties
of synthesized TiO2 nanoparticles. Particle sizes of the
well-crystalized TiO2 nanoparticles are 20.5 nm calculated
from the transmission electron microscopy (TEM) image
(Figure S1) and 22.4 nm calculated from the powder X-ray
diffraction (XRD) pattern (Figure S2), respectively. A pore
size of 19.3 nm and a porosity of 54.0% were obtained by
the random packing of the raw TiO2 nanoparticles (Figure
S3), which was not large enough for the complete
penetration of perovskite. So we developed a novel
system of TiO2/EC/P123/BCA for screen-printable TiO2
pastes, in which P123 acted as the pore-adjusting agent
and BCA as the solvent. It is worth pointing out that
the novel TiO2 paste we have developed here has
appropriate rheology as seen in Figure S4, permitting
application in forming smooth and homogeneous
films by screen-printing process. Highly porous TiO2
films (denoted as TFs) were prepared by those pastes with
varied P123 content and sintered at 510 ℃. As seen in
table 1, pore size from 16.0 to 34.2 nm and porosity from
57.1 to 73.5% can be tuned successively by adjusting P123
content in the pastes. It's notable that the surface area of
polymeric PSCs33). Devices without significant
hysteresis have been successfully fabricated34, but
understanding the origin of hysteresis is equally
important. Several origins of hysteresis have been
proposed such as ferroelectric characteristics, trap
sites and ion migration27. Ion migration is confirmed
to occur in perovskite material by many direct and
indirect evidences and is broadly believed to be the
cause of notorious hysteresis problem in PSCs35-38. It
has been confirmed by many reports that
bilayer-structured PSCs employing mp-TiO2 exhibit
diminished hysteresis compared to planar PSCs and
superstructured PSCs29, 39, 40, but sometimes hysteretic
behavior is empirically observed even though
mesoporous TiO2 layer is used29. So it is of significant
importance to systematically study the relationship
between pore size and porosity of mp-TiO2 and
hysteresis behaviour of the corresponding PSCs.
In this work, we propose a novel approach to
modulate the pore size and porosity of porous TiO2
films consecutively while the specific surface area and
particle size of TiO2 films being fixed. Highly porous
TiO2 films (TFs) with a large pore size of 50 nm (most
ACS Paragon Plus Environment
2

Page 3 of 15
Chemistry of Materials
TFs is kept constant within the margin of error. The
invariable surface area is caused by the constant particle
size of the same batch of synthesized TiO2 nanoparticles.
In a word, the desired porosimetry properties of TFs are
successfully tuned by adjusting the P123 content in the
pastes in a certain range.
—
Terpineo
l
BCA
BCA
BCA
BCA
66.5
16.0
57.1
TF-16
1
2
3
4
5
6
7
8
TF-20
TF-25
TF-28
TF-31
—
1:3
1:2
1:1
66.8
65.2
64.1
66.9
20.4
25.9
28.6
31.9
63.2
67.9
69.9
73.1
TF-34
2:1
BCA
66.3
34.2
73.5
a Surface area calculated from the linear portion of the Brunauer–
Emmett–Teller (BET) equation (P/P0=0.05–0.30)
Table 1 Structural characteristics of sintered TiO2
films as achieved by the adjustment of P123
content in TiO2 pastes
b Pore size calculated by the BJH method using desorption branches
9
c Porosity calculated from pore volume by the BJH method using
desorption branches
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Porosityc
P123:TiO2
(weight
ratio)
Surfac
e areaa
(m2/g)
Pore
sizeb
(nm)
Sampl
e
Solvent
(％)
Figure 1 Top-view SEM images of porous TiO2 films TF-16 (a), TF-25 (b) and TF-34 (c) and the corresponding
cross-section SEM image of TF-34 (d). Pore size distribution curves (e) and nitrogen adsorption-desorption isotherms (f)
for sintered TiO2 films.
Figures 1 (a)-(c) show the top-view scanning
electron microscopy (SEM) images of TF-16, TF-25
and TF-34 (-16, -25 and -34 indicating the actual
average pore size of TFs as shown in table 1).
Compared to the top morphology of pristine TF-16
without added P123, mesopores in the scale of tens of
nanometers and micropores in the scale of hundreds
of nanometers exhibited in TF-25 and TF-34
corroborate the fact that the added P123 leaves much
larger voids in those films after sintering and thus
causes consequential larger pore size and higher
porosity. The large pores in the scale of hundreds of
nanometers shown in the cross section image of
Figure 1 (d) also confirm the highly porous property
of TFs, which is beneficial for the fully pore-filling of
perovskite. According to the pore size distribution
curves in Figure 1 (e), the most probable pore size
rises up with respect to the P123 content. The wider
pore size distribution (Figure 1 (e)) and sharp
increase in the region of higher relative pressures
(Figure 1 (f)) with the increase of added P123 content
indicate that more large pores exist in those films,
which is consistent with the former observations
from SEM images. The largest average pore size of the
TFs obtained by this method we demonstrate is 34.2
nm and the corresponding porosity is 73.5%, of which
the porosity is the highest parameter that has ever
been achieved for thick TiO2 films to the best of our
knowledge. The thickness of TF with porosity of 73.5%
we obtained here can be up to 10 µm without
cracking by screen-printing process. Further increase
of the added P123 content makes the obtained TFs
tend to crack and thus there is no further
improvement on the pore size and porosity (Figure
S5).
For common screen-printable paste, terpineol is
usually used as the solvent. However, in the
preliminary experiment, films prepared by the system
of TiO2/EC/P123/Terpineol only show slight increase
in pore size from 15 to 18 nm as listed in table S1.
Besides, the films are prone to crack during sintering
procedure and no further enlargement of pore size is
obtained by further increase of the added P123
ACS Paragon Plus Environment
3

Chemistry of Materials
Page 4 of 15
Figure 2 Schematic representation of the formation
mechanism of the highly porous TiO2 films presented in
our proposed synthesis strategy.
content as seen in Figure S6. By using BCA instead of
terpineol, the pore size and porosity of porous TiO2
films are successively regulated with increasing P123
content and no cracking appears until the weight
ratio of P123 to TiO2 is above 2:1.
1
2
3
4
PSCs performance. The prepared highly porous
TF-25 and TF-34 as well as pristine TF-16 were
employed to fabricate mesoscopic PSCs. Figures 3
(a)-(f) exhibit the cross-section architectures of PSCs
deposited on one-layer (1L) and two-layer (2L) TFs
with varied pore size, respectively. A typical bilayer
structure, which combines underneath mesoporous
layer infiltrated with perovskite and upper perovskite
capping layer is observed. Thickness of 1L TiO2 film is
~250nm while thickness of 2L TiO2 film is ~450nm.
For both 1L and 2L TFs, thickness of perovskite
capping layer gradually reduces with increasing pore
size and porosity of TFs, which implies a better
pore-filling of perovskite into TiO2 scaffold as the
precursor content and spin coating speed are kept
constant. It should be pointed out that the white dots
shown in Figure 3 (e) were the fragments of silver
contacts that dropped onto the HTM layer during
sample preparation before SEM characterization. The
corresponding top-view SEM images of the perovskite
capping layers are exhibited in Figures 3 (g)-(l), which
clearly display the surface morphology and coverage
of perovskite films. By enhancement of pore size and
porosity of TFs, voids appear on the surface of
perovskite capping layer and cause poor coverage of
perovskite as seen in Figure 3. As pore size and
porosity of TFs increase, more perovskite particles
infiltrate into the TiO2 scaffold, which makes it
difficult for the less raw material of perovskite left
upon the scaffold forming a dense layer with
complete coverage. This phenomenon we observe
here is also an evidence that the penetration of
perovskite into TiO2 scaffold is improved by
increasing pore size and porosity of TiO2 films. It is
expected that a dense and fully-covered capping layer
would be achieved by further improvement of
deposition method.
5
6
7
8
Why can highly porous TiO2 films be successfully
achieved by this novel system of TiO2/EC/P123/BCA? It’s
known that chemicals with -OH group such as butanol,
pentanol and hexanol, act as co-surfactants or swelling
agents, are located at hydrophilic-hydrophobic interface
to swell the block copolymer micelles by reducing
interfacial curvatures41. So the adding of terpineol with
-OH group, results in larger swelling micelles as seen in
the scheme of Figure 2 based on our experiment results.
As pore-forming agent of EC also exists in this system, the
pore volume becomes so large that the scaffold
construction collapses during removing organic
components by sintering procedure and therefore the
films crack and no obvious increase of pore size is
observed, which could be approved by the fact that the
cracking problem can be solved by washing with ethanol
to remove P123 before the adding of terpineol. On the
other hand, while using BCA without -OH group instead
of terpineol, P123 acts as pore-adjusting agent merely
because of its high molecular weight. Due to the steric
effect of P123, large voids are formed inside the films after
removal of additives by sintering as shown in Figure 2.
Adding more EC component is not an alternative choice
as high viscous EC makes the films hard to print and
inhomogeneous. Finally the successful porous properties
adjustment of TiO2 films is accomplished by this novel
screen-printable paste containing P123 as pore-adjusting
agent and BCA as solvent. Crack-free TiO2 thick films
with high porosity of 73.5% that have never been literally
reported are achieved by this method. It should be
mentioned that this method is also suitable for other
highly porous metal oxide films.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment
4

Page 5 of 15
Chemistry of Materials
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3 Cross-section SEM images of the PSC devices based on 1L TF-16 (a), TF-25 (b) and TF-34 (c) and 2L TF-16 (d),
TF-25 (e) and TF-34 (f). Corresponding top-view SEM images of the perovskite films deposited on 1L TF-16 (g), TF-25 (h)
and TF-34 (i) and 2L TF-16 (j), TF-25 (k) and TF-34 (l).
In order to confirm the better interconnection
between TiO2 and perovskite and therefore the better
charges extraction due to better pore-filling by
increased pore size and porosity, steady-state
photoluminescence (PL) spectra of samples
FTO/bl-TiO2/CH3NH3PbI3,
typical PL peaks centered at ~765 nm, corresponding
to the excitation wavelength of 457 nm, which is in
agreement with the previously reported data42. For
the perovskite films deposited on mp-TiO2, the PL
intensity decreases significantly due to the injection
of electrons from perovskite to TiO2. The PL intensity
further reduces with increasing pore size and film
thickness of TF, which indicates more effective
electron-injection from perovskite to TiO2 caused by
the improved interconnection on the interface of
TiO2/CH3NH3PbI3 because of better pore-filling.
FTO/bl-TiO2/1L-TF-16-CH3NH3PbI3/CH3NH3PbI3,
FTO/bl-TiO2/2L-TF-16-CH3NH3PbI3/CH3NH3PbI3, and
FTO/bl-TiO2/2L-TF-P34-CH3NH3PbI3/CH3NH3PbI3
were measured, where bl-TiO2 is TiO2 blocking layer. As
can be seen in Figure 4 (a), all the samples show
ACS Paragon Plus Environment
5

Chemistry of Materials
Page 6 of 15
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4 Steady-state photoluminescence spectra of perovskite films deposited on blocking layer (dash line) and on 1L TF-16
(black line), 2L TF-16 (blue line) and 2L TF-34 (red line), respectively (a). Schematic device architecture of bilayer-structured PSC
(b). Dependence of JSC VOC, FF and PCE on the pore size of 1L 250-nm-thick TFs (c, e) and 2L 450-nm-thick TFs (d, f),
PSCs fabricated via one-step solution method (c, d) and two-step VASP method (e, f).
,
We
further
investigated
the
current
change trends of PCEs on pore sizes are similar for
these two different device fabrication processes,
which strongly proves that the device performance is
mostly affected by the pore-filling of perovskite. It
could be found out that PCE is mostly affected by the
variation of FF. FF increases firstly with increasing
pore size from 16.0 nm to 25.9 nm, indicating
effective electron injection on the interface because
of sufficient pore-filling. However, The average
coordination number of TiO2 particles reduces with
increasing porosity43, which causes discontinuity of
TiO2 and block charge transport pathway inside TiO2
scaffold. Therefore FF reduces with further enlarged
pore size of 34.2 nm. The JSC increases at first and
gradually reduces with further increasing pore size
and thickness of mesoporous layer. We propose that
the JSC is affected by the combination of
electron-extraction of mp-TiO2 and coverage of
perovskite capping layer. Better penetration of
perovskite into the TiO2 scaffold could be achieved
because of wide tunnels in mesoscopic framework
with large pore size and high porosity, which leads to
the improvement of JSC due to more effective charge
injection on the interface of TiO2/CH3NH3PbI3.
density-voltage characteristics for PSCs based on
TF-16, TF-25 and TF-34, consisting of a device
structure
FTO/bl-TiO2/mp-TiO2-CH3NH3PbI3/CH3NH3PbI3/spir
o-OMeTAD/Ag, where spiro-OMeTAD is
transport material (HTM)
of
a
hole
of
(2,2’,7,7’-tetrakis-N,N-di-4-methoxyphenylamino)-9,9’-spir
obifluorene, and Ag is the silver contact (Figure 4 (b)).
The deposition of perovskite was carried out by using
two methods, one-step solution method and two-step
vapour-assisted solution method (VASP), in order to
confirm that the photovoltaic performance and
hysteresis behaviour were related to the porous
properties of TFs but not the forming quality of
perovskite layers. Photovoltaic parameters of
short-circuit photocurrent-density (JSC), open-circuit
voltage (VOC), fill factor (FF) and power conversion
efficiency (PCE) as a function of pore size of TFs are
plotted in Figures 4 (c)-(f), where each dataset is
carried out from at least 3 devices. Tables S2 and S3
list detailed parameters of the champion devices and
calculated average PCEs of PSCs under standard
simulated solar radiation (AM 1.5, 100mW/cm2). The
ACS Paragon Plus Environment
6

Page 7 of 15
Chemistry of Materials
scan (from open circuit to short circuit, RS) and the
However, larger pore size and thicker film reduce the
1
2
3
4
5
6
7
8
capping layer thickness and cause voids on the
surface of perovskite as seen in top-view images of
Figure 3, and this ultimately results in reduced JSC due
to recombination of photoelectrons and holes
because of possible contact of TiO2 and
spiro-OMeTAD. No reduction on VOC is observed
until the pore size is enlarged to 34.2 nm. As better
pore-filling achieved with thick porous film of high
porosity, more electrons injected to TiO2 lower the
Fermi energy level at equilibrium between EF(TiO2)
and EF(CH3NH3PbI3)44, leading to a lower VOC. As a
opposite way forward scan (from short circuit to open
circuit, FS). Figure 5 demonstrates the typical
hysteresis behavior of bilayer-structured PSCs
fabricated on 1L TF-16, 2L TF-25 and 2L TF-34 with
varied pore size, porosity and film thickness. The
more detailed photovoltaic parameters and J-V curves
for each cell of 1L (and 2L) TF-16, TF-25 and TF-34 are
shown in table S4, Figures S7 and S8. Here, we qualify
the hysteresis extent in J-V curve by defining
hysteresis index (HI) as eq 1 according to the
literature45.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
result of a compromise between VOC, JSC and the FF,
ꢄ
ꢊ
ꢄ
ꢊ
ꢈꢉ
ꢅ.ꢆꢇ
ꢋꢁ
ꢌꢃ
ꢅ.ꢆꢇ
ꢊ
HI ꢀ ꢁ
(1)
ꢂꢃ
ꢈꢉ
the champion performances among the fabricated
PSCs are 15.77% with a JSC of 22.55 mA/cm2, a VOC of
1.039 V, and an FF of 67.34% via one-step solution
method, and 12.24% with a JSC of 20.34 mA/cm2, a VOC
of 0.910 V, and an FF of 66.12% via two-step VASP
method.
ꢄ
ꢁ
ꢂꢃ
ꢅ.ꢆꢇ
ꢈꢉ
It is noteworthy that hysteresis extents of mesoscopic
PSCs depend strongly on the porous properties of TFs
as seen in Figure 5 and table S4. According to the
estimated HI, the hysteresis extent successively
diminishes with increasing pore size and film
thickness of TFs. We attribute the reason to the fast
charge extraction by improved infiltration of
perovskite into mp-TiO2 that suppresses charge
accumulation on the interlayer of TiO2/CH3NH3PbI346
which will be discussed later.
Dependence of hysteresis on pore size. To
further explore the importance of pore-filling of
perovskite for the diminished hysteresis in PSCs, we
investigated the effect of varying pore size of TFs on
,
the hysteresis extents of J-V curves under reverse
Figure 5 Scan direction dependent J-V curves for the PSCs based on 1L 250-nm-thick TF-16 (a, d), 2L 450-nm-thick TF-25 (b, e)
and 2L 450-nm-thick TF-34(c, f) via one-step solution method (a, b and c) and two-step VASP method (d, e and f).
PSCs fabricated by VASP method show much
lower HI due to the fact that the VASP method would
further facilitate the pore-filling of perovskite into
mp-TiO2 as the CH3NH3I vapour could easily get into
the bottom of the scaffold. By using one-step solution
procedure, although forming of intermediate
MAI·PbI2·DMSO retards fast-evaporation of DMF,
crystallization occurs during diethyl ether dripping
and subsequent annealing on hotplate is in the time
scale of minutes. Immediate formation of perovskite
due to strong ionic interactions between the metal
cations and halogen anions hinders excellent
infiltration of perovskite into the bottom of TiO2
scaffold. By contrast, VASP procedure avoids the
extremely high reaction rate of perovskite often
observed in solution-processed method42. Smooth
and uniform PbI2 framework is firstly formed inside
the TiO2 scaffold and subsequently exposed to
CH3NH3I vapour at 110
℃ under a reduced pressure
(100 Pa) for several hours to complete perovskite
transformation. As diffusion of the raw material
CH3NH3I is improved in the case of VASP, a fully
pore-filling is yielded. Schematic for comparison on
these two diverse methods is shown in Figure S9.
Ultimately a hysteresis-free PSC is achieved by
depositing perovskite on 450-nm-thick TF-34 via
ACS Paragon Plus Environment
7

Chemistry of Materials
Page 8 of 15
two-step VASP method. The lower PCEs of redistribution of ions reach equilibrium. As a result,
1
2
3
4
5
6
hysteresis-less PSCs are caused by the poor coverage
of perovskite on the scaffold as seen in Figures 3 (i),
(k) and (l). We expect that PSCs with both high PCE
and less hysteresis could be achieved by depositing a
fully-covered perovskite thin layer over the mp-TiO2
with completely-infiltrated perovskite.
the transient photocurrent that being recorded in J-V
characterization is underestimated under FS and
overestimated under RS, and the hysteresis is
observed.
By applying porous TiO2 scaffold as seen in Figure
6 (b), mp-TiO2 is capable of extracting electrons from
I- and resulting in I, which would avoid charge
accumulation on the interface and the consequent
screen field. By improving the pore-filling of
perovskite with highly porous TFs, hysteresis would
be further diminished and eliminated because of
greatly accelerated charge transfer from I- to TiO2 at
the interface of TiO2/CH3NH3PbI3. Planar PSCs
containing compact TiO2 layer suffer from fierce
hysteresis because of inefficient charge transfer from
perovskite to compact TiO2 layer due to much smaller
interfacial contact39. Severe hysteresis observed in
mesosuperstructured PSCs is consistent with the fact
that more charges accumulate on the interface due to
the insulating characteristics of Al2O3 scaffold29. And
the null hysteresis that always being observed in
inverted PSCs with PCBM is caused by the superior
charge extraction properties of PCBM with respect to
TiO239. In a word, enlarging the interfacial contact
between perovskite and TiO2 by effective pore-filling
suppresses the charge accumulation caused by ion
migration and diminishes the possibility of hysteresis.
7
8
9
Suppression of charge accumulation by
pore-filling. The above-mentioned results have
clearly demonstrated that hysteresis is strongly
influenced by varied pore size and porosity of TiO2
scaffold. In other words, hysteresis largely depends
on the pore-filling of perovskite. We attribute such a
phenomenon to the suppression of charge
accumulation on the TiO2/CH3NH3PbI3 interface by
using highly porous TFs. We propose a plausible
mechanism of the suppressive effect on hysteresis by
highly porous TFs as illustrated in Figure 6. Ions of I-
and CH3NH3+ travel to the anode (electron transport
layer, ETL) and the cathode (hole transport material,
HTM) under increasing external bias from 0 to VOC of
FS, and migrate back into the bulk perovskite under
reducing external bias of RS as seen in Figure 6 (a).
The accumulated charges on the interfaces screen the
intrinsic build-in electric field and affect the
collection of photogenerated charge carriers on both
contacts36. The photocurrent under a certain bias will
not get to a steady state until the migration and
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment
8

Page 9 of 15
Chemistry of Materials
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Figure 6 Mechanism of ion migration in bulk perovskite (a) and suppressing effect of mp-TiO2 by electron extraction
(b).
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 7 Steady state photocurrent for PSCs fabricated on 250-nm-thick TF-16 (a), 250-nm-thick TF-25 (b) and
450-nm-thick TF-34 (c). Devices were measured at a bias voltage equal to voltage corresponding to maximum power
ACS Paragon Plus Environment
9

Chemistry of Materials
Page 10 of 15
point. J-V curves for PSCs fabricated on 250-nm-thick TF-16 (d), 250-nm-thick TF-25 (e) and 450-nm-thick TF-34 (f).
1
None poling, -1V poling and 1V poling were applied on devices for 30 s under dark beforehand.
2
3
4
5
6
7
8
9
In order to prove the rationality of our proposed
mechanism, we electrically poled PSCs beforehand,
which was able to create a large accumulation of
mobile ions on the interface to imitate ion
migration46. We electrically poled PSCs based on
250-nm-thick TF-16, 250-nm-thick TF-25 and
450-nm-thick TF-34 for 30 s under dark condition
with -1V bias (imitating the ion migration under FS)
or 1V bias (imitating the ion redistribution under RS).
The steady state current and J-V curves were
recorded afterwards under 1 sun illumination.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
As seen in Figure 7 (a), compared to the initial cell
without pre-bias treating (no poling samples), the
slowly increasing current after pre-bias of -1V could
be ascribed to an enhanced build-in field due to ion
migration. And the gradually reducing current after
pre-bias of 1V is caused by the redistribution of
accumulated ions. The transient lower current
observed under pre-biasing of -1V is consistent with
reducing PCE obtained as seen in Figure 7 (d). On the
opposite, a higher transient current under pre-biasing
of 1V and corresponding increased PCE are detected.
As the response of steady state photocurrent after
pre-bias treatment is in the timescale of tens of
seconds shown in Figure 7 (a), the current recorded
during J-V characterization is actual a transient value
that underestimated at FS and overestimated at RS.
The shaky line observed in Figure 7 (b) was caused by
the damaging on perovskite due to bias treatment.
But It's worth mentioning that the variation on PCE
under pre-bias treatment can be recovered by storing
the same cell in dark overnight as seen in Figure S10,
and thus the varied PCEs after bias treatment are not
caused by the degradation of perovskite or the
electron extraction from I-. By enhancement of the
pore size of TFs and pore-filling of perovskite, such
deviations of steady state current and J-V curves are
diminished as seen in Figure 7, which implies
suppression of charge accumulation and hysteresis by
improving infiltration of perovskite into TiO2
scaffold.
Figure
8 Pore size distribution for the TF-34 films
spin-coated by 3wt% TiO2 dilution (black line), 4wt% TiO2
dilution (blue line) and 6wt% TiO2 dilution (red line) (a).
Cross-section images of perovskite deposited on TiO2 scaffold
(b), top-view images of perovskite capping layers (c), and
scan direction dependent J-V curves (d) based on the
corresponding spin-coated films of 290-nm-thick TF-34 (left),
350-nm-thick TF-34 (middle) and 559-nm-thick TF-34
(right).
Hysteresis-free PSCs. To obtain PSCs with both
high PCE and eliminated hysteresis, we deposited a
fully-covered perovskite layer over the mp-TiO2 with
completely-infiltrated perovskite by one-step solution
method due to its low-cost and ease of use. We subtly
adjusted the film thickness of spin-coated TF-34 by
careful control of the TiO2 content in the dilution of
TiO2 paste. The most probable pore size of TF-34
spin-coated by 3wt% dilution is 46.3 nm. And it is
further expanded up to 54.4 nm by increasing the
dilute concentration from 3wt% to 6wt% as shown in
Figure 8 (a). The further increase of pore size along
with increased film thickness is beneficial for the fully
pore-filling of perovskite into the thicker TFs by
one-step solution method. Figure 8 (b) shows the
cross-sectional SEM images of devices incorporating
TF-34 with different thicknesses of 290, 350 and 559
nm achieved by spin-coating 3wt%, 4wt% and 6wt%
TiO2 dilutions, respectively. The corresponding
thicknesses of perovskite capping layers are 308, 242
and 214 nm, respectively. Decent efficiencies above 15%
have been achieved as dense and fully-covered
perovskite thin layers are formed over the
completely-infiltrated TiO2/CH3NH3PbI3 as observed
in Figure 8 (c). As seen in Figure 8 (d), a certain
ACS Paragon Plus Environment
10

Page 11 of 15
Chemistry of Materials
particles in ethanol was stirred overnight to
extent of hysteresis still exists in the cell based on
1
2
3
4
5
6
7
8
290-nm TF-34, which is caused by the severe ion
migration in the thicker perovskite capping layer.
However, the lower PCE for the cell with null
hysteresis that based on 559-nm TF-34 is caused by
the inefficient charge transfer due to the excessively
thick scaffold. Finally the champion cell based on
350-nm TF-34 gives a PCE of 14.85%, a JSC of 21.90
mA/cm2, a VOC of 1.044 V, and a FF of 64.95 at FS, and
a PCE of 15.47%, a JSC of 21.97 mA/cm2, a VOC of 1.064
V, and a FF of 66.18 at RS. The steady-state current
measurement of the same cell shown in Figure S11
gives a stabilized output power of 15.3% under 0.84 V
constant forward bias, which agrees well with the
PCE by J-V measurement.
accomplish a complete dispersion. P123 (Aldrich,
M=5800) and 0.8 g ethyl cellulose (EC, Aldrich, 30-60
mpas) were pre-dissolved in ethanol before adding to
the TiO2/EtOH solution, respectively. 7.2 g Terpineol
or 2-butoxyethyl acetate (BCA, Aladdin, 98%) was
finally added before evaporation. The TiO2/P123 ratio
in grams varied from 3:1 to 1:2 to adjust the pore size
and porosity of TiO2 film. After evaporation to
remove ethanol, the viscous pastes were obtained.
The porous TiO2 films were prepared from those TiO2
pastes with varied P123 content by screen-printing
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
process and sintered at 510
additives.
℃ to remove the organic
Synthesis of CH3NH3I. Methylamine iodide
(CH3NH3I) was synthesized and purified on the basis of a
reported method48. 30 mL methylamine (33%wt in
ethanol, Sigma-Aldrich) and 28 mL hydroiodic acid (57%
CONCLUSIONS
In conclusion, we have demonstrated a novel method
to obtain highly porous TiO2 films with tunable pore
size of 16.0-34.2 nm and porosity of 57.1-73.5%,
porosity of which is the highest that has ever been
literally reported for the TiO2 thick films to the best
of our knowledge. Pore size and porosity of mp-TiO2
is successively regulated by facilely adjusting the
adding amount of P123 in the novel system of
TiO2/EC/P123/BCA for screen-printable paste. The
SEM images of cross-section architectures and PL
measurements confirm better pore-filling of
perovskite into the scaffold and thus faster electron
extraction on the interface with increasing pore size
and improved pore-filling. Bilayer-structured PSCs
performance is affected by the compromise of
improved charge extraction by better pore-filling
with enlarged pore size and increasing charge
recombination by diminishing coverage of capping
layer. Hysteresis extent is sequentially reduced with
increasing pore size and porosity of TFs. Slow
recovery of steady state photocurrent and deviation
of J-V curve measurements after pre-biasing confirm
the link between pore-filling and hysteresis. A
reliable mechanism has been proposed that
improving perovskite infiltration by larger pore size
leads to the fast electron injection on the interface,
which suppresses charge accumulation and thus
diminishes hysteresis. By further achieving both
complete infiltration of perovskite into scaffold and
fully coverage of perovskite capping layer,
hysteresis-free PSC with a PCE of 15.47% is ultimately
obtained. This novel method we demonstrate here for
the preparation of highly porous TiO2 films provides a
new way to fabricate hysteresis-free PSCs. It is also widely
applicable in other fields such as electronically
conductive adhesives, ceramic materials, and
mesoporous catalyst, etc.
in water, Sigma-Aldrich) reacted in
a
250 mL
round-bottomed flask at 0 °С in an ice bath for 2 h with
stirring. The solvent was then removed with a rotary
evaporator by heating the solution at 60 °С under reduced
pressure, and the precipitate was then crystallized. The
product was washed with diethyl ether for three times,
and then dissolved in ethanol, recrystallized using diethyl
ether, and finally dried in a vacuum at 60 °С for 24 h to
yield CH3NH3I.
Fabrication of PSC devices. FTO substrates were
A
cleaned and UV-O3 treated beforehand.
40-nm-thick blocking layer was prepared with sol-gel
method. The sol used here was prepared by mixing
titanium tetraisopropoxide (TTIP) contained solution
A (TTIP, ethanol, acetylacetone) and acid solution B
(ethanol, HCl, H2O). The solution was spin coated
onto FTO at a spin coating speed of 3000 rpm for 20 s.
The substrate was calcined at 510 °С for 30 min in air.
After cooling to room temperature, the compact TiO2
films were treated in 40 mM aqueous solution of
TiCl4 for 40 min at 70 °С, and then rinsed with
deionized water and ethanol. The TiCl4 treated
substrates were again calcined at 510 °С for 30 min.
TiO2 pastes were diluted in ethanol (1:5, weight ratio)
and spin coated onto the substrates at a speed of
3000 rpm for 20 s. A 250-nm-thick mesoporous TiO2
layer was achieved after sintered at 510 °С for 30 min.
The procedure was repeated one more time to
achieve another 450-nm-thick mesoporous TiO2
layer.
Perovskite thin films were prepared with one-step
solution and two-step VASP methods, respectively.
For the one-step solution method, the perovskite
precursor solution was prepared as reported by Ahn
et al49. 461 mg of PbI2 (Sigma-Aldrich, 99%), 159 mg
of CH3NH3I (TCI, 98%), and 78 mg of DMSO
(Sigma-Aldrich, 99.9%) (molar ratio 1:1:1) was mixed
in 600 mg of anhydrous DMF solution at room
temperature with stirring for 1 h. The perovskite
precursor solution was spin-coated on the mp-TiO2
layer at 5000 rpm for 20 s and 0.5 mL of diethyl ether
EXPERIMENTAL SECTION
Preparation of porous TiO2 film. Anatase TiO2
nanoparticles were prepared according to the
literature47. After washed with deionized water and
ethanol for several times, the suspension of 2 g TiO2
ACS Paragon Plus Environment
11

Chemistry of Materials
Page 12 of 15
was dripped on the rotating substrate at the sixth light intensity. The J−V curves were obtained by
1
2
3
4
5
6
7
8
second. The perovskite film was then dried at 100 °С
on a hotplate for 2 min. For the two-step VASP
method50, PbI2 (Aladdin, 99.8%) was dissolved in
DMF at 85 °С (578 mg mL), and spin coated onto the
mp-TiO2 surface at 6500 rpm for 5 s. The substrates
were then dried on a 100 °С hotplate for 2 min to
remove the remaining solvent. The as-prepared PbI2
infiltrated TiO2 film was faced down at a constant
distance of around 3 mm against the CH3NH3I
powder. The reaction was conducted in a vacuum
applying an external voltage bias with a scan rate of
40 mV/s without any pre-condition before
characterization. The active area of cells was fixed at
0.07 cm2. To confirm the ion migration, PSCs were
kept at a forward bias of 1 V or a reverse bias of -1 V
for 30 s in dark before J-V curves and steady state
current characterizations. Steady state current of the
PSCs was measured for more than 2 min at a bias
voltage equal to voltage corresponding to maximum
power point (determined from J-V curve).
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
oven under the pressure of 100 Pa at 110 °С for 6.5 h
.
ASSOCIATED CONTENT
Supporting Information
After the formation of perovskite, the film was rinsed
with isopropanol for 45 s followed by drying with air.
Finally, the perovskite films deposited by different
methods were annealed at 100 °С for 45 min to
enhance the crystallinity. A hole-transporting layer
was deposited onto the perovskite film by spin
coating at 4000 rpm for 30 s. The solution contained
72.3 mg spiro-OMeTAD (Merk), 1 mL chlorobenzene
The supporting information is available free of charge via the
Internet at http://pubs.acs.org.
TEM image, XRD pattern and pore size distribution of
synthesized TiO2 nanoparticles, Rheology diagram of
TiO2 pastes, Comparison of the pore size with further
enhanced amount of P123, Structure characteristics of
terpineol-based TiO2 films, Detailed photovoltaic
parameters, J-V cures under FS and RS, Schematic
figure for comparison of two deposition methods, J-V
cures of the cell pre-bias treated under varied voltage
and time, Steady-state output measurement of the
hysteresis-free cell (PDF)
(Sigma
Aldrich),
17.5
µL
lithium-bis(trifluoromethanesulfonyl)imide (Li-TFSI,
J&K scientific Ltd.) solution (520 mg cm-3 Li-TFSI in
1mL acetonitrile) and 28.5 µL 4-tert-butylpyridine
(TBP, Sigma Aldrich). Finally, 80-nm-thicksilver
acting as the counter electrode was thermally
evaporated on top of the hole-transporting layer.
AUTHOR INFORMATION
Characterization. The crystalline structure of the
hydrothermal synthesized TiO2 nanoparticles was
investigated by powder X-ray diffraction (XRD) using
a Bruker D8 Advance diffractometer. Data were
collected over 2θ values from 20º to 80º, at a scan
speed of 4º /min. Transmission electron microscopy
(TEM) image of TiO2 nanoparticles was carried out
using a FEI Tecnai G2 F20 microscope. Nitrogen
adsorption-desorption isotherms were measured with
a Micromeritics ASAP 2020 nitrogen adsorption
apparatus. The specific surface area was determined
from the linear portion of the Brunauer–Emmett–
Teller (BET) equation (P/P0 = 0.05–0.30). The
pore-size distribution was calculated using the
Barret–Joyner–Halenda (BJH) model. The scanning
electron microscopy (SEM) images of the porous TiO2
films, device architectures were obtained using FEI
Magellan 400. The rheological characteristics of the
pastes were determined using a DVIII+ Brookfield
Programmable Rheometer, equipped with an HB6
RV/H Brookfield spindle, with a disc diameter of
14.62 mm. Viscosity was measured as a function of
Corresponding Author
* S.W.Y.: e-mail, swyang@mail.sic.ac.cn.
* Y.L.: e-mail, liuyan@mail.sic.ac.cn.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENT
This work was financially supported by the National
High Technology Research and Development Program
of China (Grant No.2014AA052002), Science and
Technology Service Network Initiative of the Chinese
Academy of Sciences (KFJ-SW-STS-152), Shanghai
Municipal Natural Science Foundation (Grant No.
16ZR1441000), Shanghai Municipal Sciences and
Technology Commission (Grant No. 12DZ1203900) and
the Shanghai High & New Technology's Industrialization
Major Program (Grant No. 2013-2).
REFERENCES
(1) Li, W.; Wu, Z.; Wang, J.; Elzatahry, A. A.; Zhao, D. A
Perspective on Mesoporous TiO2 Materials. Chem. Mater.
2014, 26, 287-298.
(2) Jesper Jacobsson, T.; Correa-Baena, J.-P.; Pazoki, M.;
Saliba, M.; Schenk, K.; Grätzel, M.; Hagfeldt, A. Exploration
of the compositional space for mixed lead halogen
perovskites for high efficiency solar cells. Energy Environ. Sci.
2016, 9, 1706-1724.
(3) Stranks, S. D.; Eperon, G. E.; Grancini, G.; Menelaou, C.;
Alcocer, M. J. P.; Leijtens, T.; Herz, L. M.; Petrozza, A.;
Snaith, H. J. Electron-Hole Diffusion Lengths Exceeding 1
Micrometer in an Organometal Trihalide Perovskite
Absorber. Science 2013, 342, 341-344.
the
shear
rate.
Steady
photoluminescence
spectroscopy (PL) was recorded on a Horiba-Ltd.
FluoroMax-4 device with an excitation wavelength of
457 nm. All the perovskite samples were deposited
onto the films of mp-TiO2 and bl-TiO2 on FTO glass
as described above. The current density−voltage (J−V
)
curves were measured in air at room temperature
with a solar simulator equipped with a 450 W xenon
lamp and a Keithley-2420 source meter (AM 1.5G, 100
mW/cm2). Light intensity of the measurements was
determined using a calibrated silicon solar cell
(Oriel-91150) as reference for approximating 1 sun
ACS Paragon Plus Environment
12

Page 13 of 15
Chemistry of Materials
(4) Lee, M. M.; Teuscher, J.; Miyasaka, T.; Murakami, T. N.;
(MAPI) perovskite solar cells: charge injection, charge
1
2
3
4
5
6
7
8
Snaith, H. J., Efficient Hybrid Solar Cells Based on
Meso-Superstructured Organometal Halide Perovskites.
Science 2012, 338, 643-647.
(5) Correa Baena, J. P.; Steier, L.; Tress, W.; Saliba, M.;
Neutzner, S.; Matsui, T.; Giordano, F.; Jacobsson, T. J.;
Srimath Kandada, A. R.; Zakeeruddin, S. M.; Petrozza, A.;
Abate, A.; Nazeeruddin, M. K.; Grätzel, M.; Hagfeldt, A.
Highly efficient planar perovskite solar cells through band
alignment engineering. Energy Environ. Sci. 2015, 8,
2928-2934.
(6) Heo, J. H.; Song, D. H.; Han, H. J.; Kim, S. Y.; Kim, J. H.;
Kim, D.; Shin, H. W.; Ahn, T. K.; Wolf, C.; Lee, T.-W.; Im, S.
H. Planar CH3NH3PbI3Perovskite Solar Cells with Constant
17.2% Average Power Conversion Efficiency Irrespective of
the Scan Rate. Adv. Mater. 2015, 27, 3424-3430.
(7) Leijtens, T.; Eperon, G. E.; Pathak, S.; Abate, A.; Lee, M.
M.; Snaith, H. J. Overcoming ultraviolet light instability of
sensitized TiO2 with meso-superstructured organometal
tri-halide perovskite solar cells. Nat. Commun. 2013, 4: 2885
DOI: 10.1038/ncomms3885.
(8) Hwang, I.; Baek, M.; Yong, K. Core/Shell Structured
TiO2/CdS Electrode to Enhance the Light Stability of
Perovskite Solar Cells. ACS Appl. Mater. Interfaces 2015, 7,
27863-27870.
(9) Ito, S.; Tanaka, S.; Manabe, K.; Nishino, H. Effects of
Surface Blocking Layer of Sb2S3 on Nanocrystalline TiO2 for
CH3NH3PbI3 Perovskite Solar Cells. J. Phys. Chem. C 2014, 118,
16995-17000.
(10) Yang, W. S.; Noh, J. H.; Jeon, N. J.; Kim, Y. C.; Ryu, S.;
Seo, J.; Seok, S. I. High-performance Photovoltaic perovskite
layers fabricated through intramolecular exchange. Science
2015, 348, 1234-1237.
(11) Jeon, N. J.; Noh, J. H.; Kim, Y. C.; Yang, W. S.; Ryu, S.;
Seok, S. I. Solvent engineering for high-performance
inorganic–organic hybrid perovskite solar cells. Nat. Mater.
2014, 13, 897-903.
(12) Burschka, J.; Pellet, N.; Moon, S.-J.; Humphry-Baker, R.;
Gao, P.; Nazeeruddin, M. K.; Grätzel, M. Sequential
deposition as a route to high-performance
perovskite-sensitized solar cells. Nature 2013, 499, 316-319.
(13) Lee, J.-W.; Lee, T.-Y.; Yoo, P. J.; Grätzel, M.; Mhaisalkar,
S.; Park, N.-G. Rutile TiO2-based perovskite solar cells. J.
Mater. Chem. A 2014, 2, 9251-9259.
(14) Kim, H.-S.; Lee, J.-W.; Yantara, N.; Boix, P. P.; Kulkarni,
S. A.; Mhaisalkar, S.; Grätzel, M.; Park, N.-G. High Efficiency
Solid-State Sensitized Solar Cell-Based on Submicrometer
Rutile TiO2 Nanorod and CH3NH3PbI3 Perovskite Sensitizer.
Nano Lett. 2013, 13, 2412-2417.
(15) Sung, S. D.; Ojha, D. P.; You, J. S.; Lee, J.; Kim, J.; Lee,
W. I. 50 nm sized spherical TiO2 nanocrystals for highly
efficient mesoscopic perovskite solar cells. Nanoscale 2015, 7,
8898-8906.
(16) Wu, W.-Q.; Huang, F.; Chen, D.; Cheng, Y.-B.; Caruso,
R. A. Thin Films of Dendritic Anatase Titania Nanowires
Enable Effective Hole-Blocking and Efficient
Light-Harvesting for High-Performance Mesoscopic
Perovskite Solar Cells. Adv. Funct. Mater. 2015, 25, 3264-3272.
(17) Yang, Y.; Ri, K.; Mei, A.; Liu, L.; Hu, M.; Liu, T.; Li, X.;
Han, H. The size effect of TiO2 nanoparticles on a printable
mesoscopic perovskite solar cell. J. Mater. Chem. A 2015, 3,
9103-9107.
recombination and solar cell efficiency relationship. J. Mater.
Chem. A 2015, 3, 22154-22161.
(19) Leijtens, T.; Lauber, B.; Eperon, G. E.; Stranks, S. D.;
Snaith, H. J. The Importance of Perovskite Pore Filling in
Organometal Mixed Halide Sensitized TiO2-Based Solar
Cells. J. Phys. Chem. Lett. 2014, 5, 1096-1102.
(20) Wu, Y.; Chen, W.; Yue, Y.; Liu, J.; Bi, E.; Yang, X.; Islam,
A.; Han, L. Consecutive Morphology Controlling Operations
for Highly Reproducible Mesostructured Perovskite Solar
Cells. ACS Appl. Mater. Interfaces 2015, 7, 20707-20713.
(21) Li, N.; Li, H.; Li, Y.; Wang, S.; Wang, L. Improved
charge transport and injection in a meso-superstructured
solar cell by a tractable pre-spin-coating process. Phys.
Chem. Chem. Phys. 2015, 17, 24092-24097.
(22) Abdi-Jalebi, M.; Dar, M. I.; Sadhanala, A.; Senanayak, S.
P.; Giordano, F.; Zakeeruddin, S. M.; Grätzel, M.; Friend, R.
H. Impact of a Mesoporous Titania–Perovskite Interface on
the Performance of Hybrid Organic–Inorganic Perovskite
Solar Cells. J. Phys. Chem.Lett. 2016, 7, 3264-3269.
(23) Zhou, W.; Sun, F.; Pan, K.; Tian, G.; Jiang, B.; Ren, Z.;
Tian, C.; Fu, H. Well-Ordered Large-Pore Mesoporous
Anatase TiO2 with Remarkably High Thermal Stability and
Improved Crystallinity: Preparation, Characterization, and
Photocatalytic Performance. Adv. Funct. Mater. 2011, 21,
1922-1930.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(24) Pan, J. H.; Zhao, X. S.; Lee, W. I. Block
copolymer-templated synthesis of highly organized
mesoporous TiO2-based films and their
photoelectrochemical applications. Chem. Eng. J. 2011, 170,
363-380.
(25) Rapsomanikis, A.; Karageorgopoulos, D.; Lianos, P.;
Stathatos, E. High performance perovskite solar cells with
functional highly porous TiO2 thin films constructed in
ambient air. Sol. Energ. Mater. Sol. C. 2016, 151, 36-43.
(26) Chung, C.-C.; Lee, C. S.; Jokar, E.; Kim, J. H.; Diau, E.
W.-G. Well-Organized Mesoporous TiO2 Photoanode by
Using Amphiphilic Graft Copolymer for Efficient Perovskite
Solar Cells. J. Phys. Chem. C 2016, 120, 9619-9627.
(27) Snaith, H. J.; Abate, A.; Ball, J. M.; Eperon, G. E.;
Leijtens, T.; Noel, N. K.; Stranks, S. D.; Wang, J. T.-W.;
Wojciechowski, K.; Zhang, W. Anomalous Hysteresis in
Perovskite Solar Cells. J. Phys. Chem. Lett. 2014, 5, 1511-1515.
(28) Tress, W.; Marinova, N.; Moehl, T.; Zakeeruddin, S. M.;
Nazeeruddin, M. K.; Grätzel, M. Understanding the
rate-dependent J–V hysteresis, slow time component, and
aging in CH3NH3PbI3 perovskite solar cells: the role of a
compensated electric field. Energy Environ. Sci. 2015, 8,
995-1004.
(29) Wu, B.; Fu, K.; Yantara, N.; Xing, G.; Sun, S.; Sum, T. C.;
Mathews, N. Charge Accumulation and Hysteresis in
Perovskite-Based Solar Cells: An Electro-Optical Analysis.
Adv. Energy Mater. 2015, 5, 1500829.
(30) Unger, E. L.; Hoke, E. T.; Bailie, C. D.; Nguyen, W. H.;
Bowring, A. R.; Heumüller, T.; Christoforo, M. G.; McGehee,
M. D. Hysteresis and transient behavior in current–voltage
measurements of hybrid-perovskite absorber solar cells.
Energy Environ. Sci. 2014, 7, 3690-3698.
(31) Zhang, Y.; Liu, M.; Eperon, G. E.; Leijtens, T. C.;
McMeekin, D.; Saliba, M.; Zhang, W.; de Bastiani, M.;
Petrozza, A.; Herz, L. M.; Johnston, M. B.; Lin, H.; Snaith, H.
J. Charge selective contacts, mobile ions and anomalous
hysteresis in organic–inorganic perovskite solar cells. Mater.
Horiz. 2015, 2, 315-322.
(18) Matas Adams, A.; Marin-Beloqui, J. M.; Stoica, G.;
Palomares, E. The influence of the mesoporous TiO2 scaffold
on the performance of methyl ammonium lead iodide
ACS Paragon Plus Environment
13

Chemistry of Materials
Page 14 of 15
(32) Chen, H.-W.; Sakai, N.; Ikegami, M.; Miyasaka, T.
(47) Barbe, C. J.; Arendse, F.; Comte, P.; Jirousek, M.;
Lenzmann, D.; Shklover, V.; Gratzel, M. Nanocrystalline
Titanium Oxide Electrodes for Photovoltaic Applications. J.
Am. Ceram. Soc. 1997, 80, 3157-3171.
1
2
Emergence of Hysteresis and Transient Ferroelectric
Response in Organo-Lead Halide Perovskite Solar Cells. J.
Phys. Chem. Lett. 2015, 6, 164-169.
3
4
5
6
7
8
9
(33) Shao, Y.; Xiao, Z.; Bi, C.; Yuan, Y.; Huang, J. Origin and
elimination of photocurrent hysteresis by fullerene
passivation in CH3NH3PbI3 planar heterojunction solar cells.
Nat. Commun. 2014, 5: 5784 DOI: 10.1038/ncomms6784.
(34) Li, G.; Ching, K. L.; Ho, J. Y. L.; Wong, M.; Kwok, H.-S.
Identifying the Optimum Morphology in High-Performance
Perovskite Solar Cells. Adv. Energy Mater. 2015, 5, 1401775.
(35) Xiao, Z.; Yuan, Y.; Shao, Y.; Wang, Q.; Dong, Q.; Bi, C.;
Sharma, P.; Gruverman, A.; Huang, J. Giant switchable
photovoltaic effect in organometal trihalide perovskite
devices. Nat. Mater. 2014, 14, 193-198.
(36) Eames, C.; Frost, J. M.; Barnes, P. R. F.; O’Regan, B. C.;
Walsh, A.; Islam, M. S., Ionic transport in hybrid lead iodide
perovskite solar cells. Nat. Commun. 2015, 6: 7497 DOI:
10.1038/ncomms8497.
(37) Yuan, Y.; Chae, J.; Shao, Y.; Wang, Q.; Xiao, Z.;
Centrone, A.; Huang, J. Photovoltaic Switching Mechanism
in Lateral Structure Hybrid Perovskite Solar Cells. Adv.
Energy Mater. 2015, 5, 1500615.
(38) Leijtens, T.; Hoke, E. T.; Grancini, G.; Slotcavage, D. J.;
Eperon, G. E.; Ball, J. M.; De Bastiani, M.; Bowring, A. R.;
Martino, N.; Wojciechowski, K.; McGehee, M. D.; Snaith, H.
J.; Petrozza, A. Mapping Electric Field-Induced Switchable
Poling and Structural Degradation in Hybrid Lead Halide
Perovskite Thin Films. Adv. Energy Mater. 2015, 5, 1500962.
(39) De Bastiani, M.; Dell'Erba, G.; Gandini, M.;
D'Innocenzo, V.; Neutzner, S.; Kandada, A. R. S.; Grancini,
G.; Binda, M.; Prato, M.; Ball, J. M.; Caironi, M.; Petrozza, A.
Ion Migration and the Role of Preconditioning Cycles in the
Stabilization of the J-V Characteristics of Inverted Hybrid
Perovskite Solar Cells. Adv. Energy Mater. 2016, 6, 1501453.
(40) Kim, H.-S.; Park, N.-G., Parameters Affecting I–V
Hysteresis of CH3NH3PbI3 Perovskite Solar Cells: Effects of
Perovskite Crystal Size and Mesoporous TiO2 Layer. J. Phys.
Chem. Lett. 2014, 5, 2927-2934.
(41) Kim, T.-W.; Kleitz, F.; Paul, B.; Ryoo, R. MCM-48-like
Large Mesoporous Silicas with Tailored Pore Structure: Facile
Synthesis Domain in a Ternary Triblock
Copolymer-Butanol-Water System. J. Am. Chem. Soc. 2005,
127, 7601-7610.
(42) Chen, S.; Lei, L.; Yang, S.; Liu, Y.; Wang, Z.-S.
Characterization of Perovskite Obtained from Two-Step
Deposition on Mesoporous Titania. ACS Appl. Mater.
Interfaces 2015, 7, 25770-25776.
(43) Lagemaat, J. v. d.; Benkstein, K. D.; Frank, A. J. Relation
between Particle Coordination Number and Porosity in
Nanoparticle Filmsꢀ Implications to Dye-Sensitized Solar
Cells. J. Phys. Chem. C 2001, 105, 12433-12436.
(44) Kang, H.-W.; Lee, J.-W.; Son, D.-Y.; Park, N.-G.
Modulation of photovoltage in mesoscopic perovskite solar
cell by controlled interfacial electron injection. RSC Adv.
2015, 5, 47334-47340.
(45) Liu, C.; Fan, J.; Zhang, X.; Shen, Y.; Yang, L.; Mai, Y.
Hysteretic Behavior upon Light Soaking in Perovskite Solar
Cells Prepared via Modified Vapor-Assisted Solution Process.
ACS Appl. Mater. Interfaces 2015, 7, 9066-9071.
(46) Chen, B.; Yang, M.; Zheng, X.; Wu, C.; Li, W.; Yan, Y.;
Bisquert, J.; Garcia-Belmonte, G.; Zhu, K.; Priya, S. Impact of
Capacitive Effect and Ion Migration on the Hysteretic
Behavior of Perovskite Solar Cells. J. Phys. Chemi. Lett. 2015,
6, 4693-4700.
(48) Raga, S. R.; Jung, M.-C.; Lee, M. V.; Leyden, M. R.; Kato,
Y.; Qi, Y. Influence of Air Annealing on High Efficiency
Planar Structure Perovskite Solar Cells. Chem. Mater. 2015,
27, 1597-1603.
(49) Ahn, N.; Son, D.-Y.; Jang, I.-H.; Kang, S. M.; Choi, M.;
Park, N.-G. Highly Reproducible Perovskite Solar Cells with
Average Efficiency of 18.3% and Best Efficiency of 19.7%
Fabricated via Lewis Base Adduct of Lead(II) Iodide. J. Am.
Chem. Soc. 2015, 137, 8696-8699.
(50) Chen, Q.; Zhou, H.; Hong, Z.; Luo, S.; Duan, H. S.;
Wang, H. H.; Liu, Y.; Li, G.; Yang, Y. Planar heterojunction
perovskite solar cells via vapor-assisted solution process. J.
Am. Chem. Soc. 2014, 136, 622-625.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment
14

Page 15 of 15
Chemistry of Materials
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment
15