Temperature-dependent hysteresis effects in perovskite-based solar cells

Article abstract of DOI:10.1039/c4ta04969a

Staircase voltage sweep measurements were performed on a perovskite solar cell at 250 K, 300 K, and 360 K. Time-dependent photocurrent data reveal the complexity of the signal that cannot be described by a simple mono-exponential function, suggesting that multiple charging-discharging processes are responsible for the complex hysteresis behavior.

Full text of DOI:10.1039/c4ta04969a

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: L. K. Ono, S. R.
Raga, S. Wang, Y. Kato and Y. Qi, J. Mater. Chem. A, 2014, DOI: 10.1039/C4TA04969A.
This is an Accepted Manuscript, which has been through the
Royal Society of Chemistry peer review process and has been
accepted for publication.
Accepted Manuscripts are published online shortly after
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
Information for Authors.
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 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.
www.rsc.org/materialsA

Page 1 of 8
Journal of Materials Chemistry A
View Article Online
DOI: 10.1039/C4TA04969A
Table of contents entry:
Temperature-dependent hysteresis effects on perovskite-based solar cells
Luis K. Ono, Sonia R. Raga, Shenghao Wang, Yuichi Kato, Yabing Qi*
Energy Materials and Surface Sciences Unit (EMSS)
Okinawa Institute of Science and Technology Graduate University (OIST)
1919-1 Tancha Onna-son, Okinawa 904-0495 Japan
*E-mail: yabing.qi@oist.jp
20
360 K
15
300 K
10
250 K
5
0
0.0 0.2 0.4 0.6 0.8 1.0
Voltage (V)
Time- and temperature-dependent photocurrent transient analysis suggests that hysteresis effects are
associated with multiple charging-discharging processes in a perovskite solar cell.

Journal of Materials Chemistry A
Page 2 of 8
View Article Online
DOI: 10.1039/C4TA04969A
Journal of Materials Chemistry A
RSCPublishing
COMMUNICATION
Temperature-dependent hysteresis effects on perovskite-
based solar cells
Cite this: DOI: 10.1039/x0xx00000x
Luis K. Ono, Sonia R. Raga, Shenghao Wang, Yuichi Kato, Yabing Qi*
Received 00th January 2012,
Accepted 00th January 2012
DOI: 10.1039/x0xx00000x
www.rsc.org/
1
2
3
4
5
6
7
Staircase voltage sweep measurements were performed on
perovskite solar cell at 250 K, 300 K, and 360 K. Time-
dependent photocurrent data reveal the complexity of the
signal that cannot be described by a simple mono-
exponential function suggesting multiple charging-
discharging processes are responsible for the complex
hysteresis behavior.
8
9
Organo-lead-halide perovskite (OHP) based solar cells were
reported to achieve energy-to-electricity power conversion
10 efficiency (PCE) as high as ~19.3%,1-8 which combined with
11 reported methods for low-cost, flexible, and large-area solar cell
12 fabrication such as ultra-sonic spray-coating9 and printing
13 technology10 makes OHP cell technology amenable to scaling up
14 to production levels.2, 5 The potential toxic effects of Pb has been
15 pointed out. Alternatives or solutions have been proposed such as
16 Pb-free perovskite based solar cells, encapsulation, and Pb
17 recycling.11-14
18
Methylammonium (MA) lead iodide (CH3NH3PbI3), the most
19 commonly employed material in halide perovskite solar cells, was
20 reported to have both high absorption coefficient (direct bandgap
21 of ~1.55 eV) and high mobilities for electrons (7.5 cm2∂V-1∂s-1)
22 and holes (12.5-66 cm2∂V-1∂s-1) resulting in long carrier diffusion
23 lengths (100 nm - 1 mm).15 Although the amount of incorporated-
24 Cl is still under debate,16 mixed methylammonium-lead halide
25 MAPbI3-xClx is another type of halide perovskite reported with
26 even higher charge-carrier mobilities (~33 cm2∂V-1∂s-1) resulting in
27 carrier diffusion lengths up to 3 mm.17 Despite all the superb
28 properties, perovskite solar cells suffer from strong hysteresis in
29 the current-voltage (I-V) measurements typically conducted under
30 AM1.5G illumination.18-20 Hysteresis in these cells is strongly
31 influenced by the perovskite grain size and structure of the
32 underneath TiO2 layer used as electron transport layer (ETL).
Figure 1. (a) SEM and (b) XRD characterization of the perovskite
film formed from MAI and PbCl2 precursors.
38 planar heterojunction architecture employing a compact-TiO2 (c-
39 TiO2) layer is of particular interest due to its simple cell
40 configuration. However, unlike cells with mesoporous-TiO2 (m-
41 TiO2), in which perovskite is scaffolded onto the mesoporous
42 matrix, the hysteresis effects in planar heterojunction structure are
43 usually strong.18-20 Hysteresis effects in perovskite solar cells are
44 generally associated with high capacitances (on the order of
45 mF/cm2) compared to Si cells (on the order of mF/cm2)19, 23-26 or
33 Efforts now are concentrated on generating perovskite films with
19, 21, 22
34 larger grain sizes to minimize hysteresis.4,
Also, the
35 replacement of MA+ by cations with larger radius of
36 formamidinium (NH2CH=NH2 , FA+)23 and 5-ammoniumvaleric
+
37 acid10 were reported to help reduce the hysteresis effects. The
46 other types of solar cells (as described further in the ESI†). This
This journal is © The Royal Society of Chemistry 2012
J. Name., 2012, 00, 1r3 | 1

Page 3 of 8
COMMUNICATION
Journal of Materials Chemistry A
VJieowuArrtnicalelONnlainme e
DOI: 10.1039/C4TA04969A
2.2
10
5
2.2
20
10
0
2.0
20
(c)
(a)
(e)
300 K
250 K
360 K
0
2.0
1.8
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
2.0
1.8
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
1.8
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
10
0
0.0
0
-5
-5
-10
-20
-30
-40
-50
-60
-10
-15
-20
0
-10
-15
-20
-0.5
-10
-20
-30
-5
-10
Forward
-1.0
0.0
80
100
120
140
80
100
120
140
80
100
Time (s)
120
Time (s)
Time (sec)
2.2
30
20
10
0
2.2
30
20
10
0
2.0
(d)
(b)
(f)
20
300 K
0
250 K
360 K
2.0
1.8
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
2.0
1.8
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
1.8
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
0
10
-5
0
-5
-10
-20
-30
-40
-50
-60
-10
-15
-20
-25
-10
-15
-20
-0.5
-1.0
-10
-20
-30
-10
-20
-30
Reverse
20
40
60
80
20
40
60
80
40
60
80
Time (s)
Time (sec)
Time (s)
Figure 2. Photocurrent and dark current as a function of applied voltage using a staircase function generator. The voltage sweep from zero
to positive voltage and from positive to zero are denoted as forward (a,c,e) and reverse (b,d,f), respectively. The step delayed photocurrent
responses of 8 s long are shown for three different temperatures: 250 K (a-b); 300 K (c-d); 360 K (e-f). The steep jumps observed in the
current signals after the each step-voltage are signatures of the capacitive effects in the device.
47 capacitance effect was hypothesized to be originated from (i) the
48 ferroelectricity or polarization of the perovskite layer; (ii) contact
49 conductivity; (iii) diffusion of excess ion as interstitial defects;
50 and/or (iv) trapping/de-trapping of charge carriers.18-20, 27
81 SEM revealed a complete surface coverage and formation of
82 interconnected crystal domains. The determination of grain
83 sizes was difficult based on those images. Film thickness of
84 ~380 nm was determined by a profilometer. XRD data
51
In this work, we performed staircase voltage sweep
85 showed the characteristic peaks at 14.1è, 28.5è, and 43.3è
52 measurements18-20 at three different temperatures of 250 K,
53 300 K, and 360 K to quantify the photocurrent transient
54 behavior on the complete perovskite cell composed of FTO/c-
55 TiO2/perovskite/spiro-MeOTAD/Au where FTO stands for
56 fluorine doped tin oxide. The hole-collection was through the
57 spiro-MeOTAD hole transport layer (HTL). Our photocurrent
58 data suggest that multiple processes are responsible for the
59 complex transient behavior. Furthermore, solar cell properties
60 can be altered significantly by the changes in environmental
61 conditions such as humidity and temperature.28-31 Thus,
62 device performance characterization of perovskite cells under
63 various conditions is of paramount importance. Also it can
64 provide a fundamental understanding of the charging-
65 discharging processes in perovskite cells if studied under
66 controlled environment (temperature, humidity, gas
67 environment).31 The temperature dependent studies of the
68 steady-state I-V curves provided the trends of open-circuit
69 voltage (Voc), short-circuit current (Isc), and fill factor (FF)
70 electrical parameters as a function of temperature shedding
71 light on the physical processes taking place within the layers
72 of perovskite solar cell. These results highlight the
73 importance to establish a protocol for precise and artifact-free
74 evaluation of perovskite solar cells.
86 corresponding to the (110), (220), and (330) in perovskite
87 structure.3 The as-prepared sample was loaded into a variable
88 temperature probe station chamber (Lake shore, CRX-6.5K)
89 coupled with a pumping station (HiCube 80, Pfeiffer). The
90 pressure inside the probe station was observed to vary
6
5
91 between ~2ì ì
10- and 4 10- Torr depending on the working
92 temperature range (between 250 K and 360 K). Electrical
93 signals from the probes contacting the FTO and Au electrodes
94 were recorded using a semiconductor characterization system
95 (4200-SCS, Keithley). The low-noise level (<1 pA) of our
96 photocurrent data allowed us to unambiguously detect the
97 different charge transport processes. It was difficult to couple
98 a solar simulator with the probe station. Thus, light
99 illumination on the perovskite samples was conducted using a
100 tungsten halogen lamp. The relative light intensity of ~9%
101 compared to the standard AM1.5G was roughly estimated
102 based on the measurements on a reference Si solar cell. This
103 light intensity was strong enough to induce the hysteresis
104 effects (photocurrent in
mA range) in perovskite solar cells
105 during the current-voltage step (I-Vstep) measurements. The
106 perovskite cells were measured under vacuum conditions for
107 three different temperatures (250 K, 300 K, and 360 K). The
108 measurements in vacuum provided high stability of the I-Vstep
75
The perovskite films in this work were prepared from
109 curves. No influences were observed from the pre-
30
110 illumination step in our studies.20,
Most likely the
76 MAI and PbCl2 precursors according to standard literature
77 procedure reported elsewhere.32-34 Detailed sample
78 preparation was described in the supporting information.
79 Sample characterization by X-ray diffraction (XRD) and
80 scanning electron microscopy (SEM) are shown in Fig. 1.
111 photodoping effect is inactive in our measurements because
112 our samples were measured in vacuum, i.e., the influence of
113 reactive gases such as O2 and/or H2O is minimal.20, 30, 31
2 | J. Name., 2012, 00, 1r3
This journal is © The Royal Society of Chemistry 2012

Journal of Materials Chemistry A
Page 4 of 8
View Article Online
Journal Name
COMMUNICATION
DOI: 10.1039/C4TA04969A
e3
e3
e2
e1
e0
e2
3.0
2.5
2.0
9
10.0
(a)
(c)
(e)
300 K
250 K
360 K
8
7
6
5
4
9.5
9.0
e2
e1
e0
tslow
td
e1
8.5
Forward:
0.3 V ç 0.4 V
I(Ñ) = 2.9 mA
Forward:
0.3 V ç 0.4 V
I(Ñ) = 8.9 mA
Forward:
0.3 V ç 0.4 V
I(Ñ) = 9.8 mA
8.0
7.5
tfast
e0
0
1.5
12
2
4
6
8
0
2
4
6
8
0
2
4
6
8
Time (s)
Time (s)
Time (s)
e0
e0
e-1
e-2
e-3
e-4
e-5
e0
e-1
e-2
e-3
e-4
e-5
e-6
e-7
e-8
e-9
e-10
e-11
25
20
15
10
16
15
14
13
12
11
(b)
(d)
(f)
tfast
Reverse:
0.5 V ç 0.4 V
I(Ñ) = 6.1 mA
Reverse:
0.5 V ç 0.4 V
I(Ñ) = 11.9 mA
Reverse:
0.5 V ç 0.4 V
I(Ñ) = 11.8 mA
e-1
e-2
e-3
10
8
tslow
6
300 K
250 K
360 K
e-4
0
2
4
6
8
0
2
4
6
8
0
2
4
6
8
Time (s)
Time (s)
Time (s)
Figure 3. Semi-logarithmic plot of the photocurrent response as a function of time after applying a voltage step at t = 0 (a,c,e) from 0.3 V
ç 0.4 V (forward) and (b,d,f) from 0.5 V ç 0.4 V (reverse) at (a,b) 250 K, (c,d) 300 K, and (e,f) 360 K. The spectra show that multiple
processes, designated by tfast and tslow, take place for the charging-discharging within the perovskite solar cell. The degree of hysteresis
depends on the waiting period to measure the photocurrent after applying the voltage, i.e., delay time (td).
114 The I-Vstep measurements were conducted at three different 146 is affected, and therefore enhanced photocurrent transients are
115 temperatures (250 K, 300 K, and 360 K) with Vstep sweeping from 147 produced within the perovskite film.20, 25, 35 It is also found that at
148 lower (250 K, Figs. 2a,b) and room temperatures (300 K, Figs.
149 2c,d), the photocurrent transient periods appear to be longer
150 compared to the ones at high temperature (360 K, Figs. 2e,f). This
151 may be related to the effects of capacitance at the selective-
152 contacts25, 36 and will be discussed in more details later. Additional
153 transient behaviors (generally observed above open circuit
154 voltages20) such as the ones observed in the photocurrent
155 corresponding to the applied bias held at 0.7 V at 360 K (Fig. 2e)
156 will not be discussed because it is beyond the scope of this work.
116 -1 V to +1 V and from +1 V down to -1 V in steps of ê0.1 V
117 (Figs. S3-S5). To make a convention for this work, we set the
118 voltage sweep starting from negative to the positive voltages to be
119 the forward scan direction, and the voltage starting from positive
120 to negative to be the reverse scan direction. In the I-V quadrant
121 where the solar cell outputs electrical power (Fig. 2), the
122 photocurrent response shows
a transient behavior for all
123 temperatures. Every time the voltage is increased or decreased by
124 0.1 V, a sharp initial (at t0) jump in photocurrent is followed by a
125 subsequent rise or decay, respectively. This time-dependent
157
The photocurrent signals are modelled by using exponential
158 functions. We observe that none of our raw data can be fitted with
159 a mono-exponential function. Multi-exponential functions, eqs. (1)
160 and (2), with three terms (i = 0,1,2) are needed to reproduce the
126 behavior was previously explained due to capacitive currents
23
127 present in the perovskite solar cells.18-20,
Such capacitance
128 effects were hypothesized to be originated from (i) the
129 ferroelectricity or polarization of the perovskite layer; (ii) contact 161 raw data, i.e., giving a reasonably low c2. Transient photovoltage
130 conductivity; (iii) diffusion of excess ion as interstitial defects; 162 decay technique is widely used to describe physical processes of
27
131 and/or (iv) trapping/de-trapping of charge carriers.18-20,
The 163 charge dynamics in solar cells. However, different descriptions for
132 degree of this capacitive effect is enhanced under illumination 164 the transient signals are reported in literature. For some perovskite
133 conditions (photocurrent, Fig. 2) and shows minimum influence 165 solar cells a mono-exponential function37, 38 was used while for
134 under dark conditions (dark current, Fig. 2). It has been previously 166 others a bi-exponential function23, 39 was employed. This shows
135 reported that lead halide perovskites (e.g., prepared from PbCl2 167 the complexity of the system under the analysis, because the
136 and MAI precursors) shows a giant dielectric constant (GDC) 168 charging-discharging characteristics in the cell can be strongly
137 phenomenon with a low frequency dielectric constant of the order 169 influenced by the cell architecture and/or preparation conditions.23
170 Considering the previously discussed capacitance factors27 in
171 perovskite solar cells, semi-logarithmic plots of the time-
172 dependent photocurrent signals as the response to applied voltage
173 steps shown in Fig. 2 are expected to generate a straight line after
174 subtracting the photocurrent under steady-state conditions (t ç
175 Ñ).20
138 of e ~ 1000 in dark conditions.25 More interestingly, under one sun
139 illumination, e was reported to increase by a factor of ~1000. The
140 origin of GDC and further e enhancement with light illumination is
141 still unclear at the moment,25 but it has been proposed that it may
142 originate from the ferroelectric response induced by
143 photogenerated carriers leading to structural rearrangement of
144 methylammonium ions (MA+).20, 25, 27 As the consequence of this
145 structural phase transition with light illumination, charge transport
176
This journal is © The Royal Society of Chemistry 2012
J. Name., 2012, 00, 1r3 | 3

Page 5 of 8
COMMUNICATION
Journal of Materials Chemistry A
VJieowuArrtnicalelONnlainme e
DOI: 10.1039/C4TA04969A
á
15
P
(a)
: ;
:
;
> : ;
?
+ P L + » E Í + r F +:»; ® ds F ATL lF ph
250 K
Forward
Reverse
tÜ
Ü@4
177
178
forward scan (1)
reverse scan (2)
10
5
á
P
: ;
:
;
> : ;
?
+ P L + » E Í + r F +:»; ® ATL lF
p
tÜ
Ü@4
179
180
181
0
182 where ti are the time constants of the system. Thus, as shown in
0.0
0.0
0.0
0.2
0.4
0.6
0.8
1.0
1.0
1.0
:
;
: ;
 ç, ? ¶
183 the eqs. (1) and (2), any eventual straight line in the HJ B
; C
Voltage (V)
: ;
 ç ? ¶
:
(b) 15
: ;
: ;
184 (forward scan) and HJ B ç ? ¶ C (reverse scan) versus time plots
:
;
Forward
Reverse
 ç,
300 K
185 will yield slopes with +1/t and -1/t values, respectively. The
186 representative spectra are displayed in Fig. 3 for the cases of 0.3 V
187 ç 0.4 V (forward) and 0.5 V ç 0.4 V (reverse) of applied voltage
188 steps at three different temperatures. The complete sets of
189 individual I-Vstep analysis for the different temperatures are
190 described in the supporting information (Figs. S6 to S11). The
191 semi-logarithmic plots show at least two regimes with
192 characteristic time decays. In the first time interval (t < ~1.8 s for
193 250 K and 300 K and t < ~1 s for 360 K, Fig. 3) after applying the
194 voltage step, fast charging-discharging processes, more than one
195 because of the non-linearity behavior observed in the semi-
196 logarithmic plot, are inferred and designated as tfast. Quantitative
197 description of this tfast (not a single time constant) is not possible
198 from our experimental data because of the convolution of multi-
199 exponential terms, eqs. (1) and (2). However, considering the
200 slowest process(es) in the first time interval, tfast is observed to
201 decay faster at high temperature (360 K) contributing less to
202 hysteresis effects. A more elaborated equivalent circuit with RC
203 components has been proposed by Dualeh et al.39 and our data
204 could be associated with (i) selective-contact resistance(s) and
205 capacitance(s) within the FTO/c-TiO2/perovskite/spiro-MeOTAD
10
5
0
0.2
0.4
0.6
0.8
Voltage (V)
15
(c)
360 K
Forward
Reverse
10
5
0
0.2
0.4
0.6
0.8
Voltage (V)
206 HTL/Au layers or (ii) due to conduction of ion(s) within the
20, 39
207 perovskite film18,
and/or HTM (e.g., Li+ from Li-TFSI
Figure 4. Extracted time-constants corresponding to the slowest
process (tslow) observed in the semi-logarithmic plot of the transient
part of the photocurrent response upon applied voltage.
208 dopant39, 40) influenced by temperature. TiO2 is a semiconductor
209 material that manifests a size-dependent capacitive effect in such
210 the nanostructured form shows enhanced chemical capacitance
211 (Cm); a smaller Cm is expected on our c-TiO2, but may still provide
212 some influences on the charging-discharging processes.26, 29, 36 In
213 the second time interval (t > ~1.8 s for 250 K and 300 K and t > ~1
214 s for 360 K, Fig. 3), a linear regime prevails with the slowest
215 decay time of the system designated as tslow. Based on the previous
216 descriptions, we attribute this slow process to the intrinsic
217 ferroelectricity of the perovskite film.18-20, 23, 37 Least-square fitting
218 method was applied on the second time interval of the
219 experimental data to extract the tslow values. In the particular cases
220 of applied voltage steps shown in Fig. 3, tslow = 5.15 ê 0.02 s (Fig.
221 3a) [and 3.25 ê 0.01 s (Fig. 3b)] at 250 K, 7.49 ê 0.03 s (Fig. 3c)
222 [and 3.23 ê 0.03 s (Fig. 3d)] at 300 K, and 5.6 ê 0.2 s (Fig. 3e)
223 [and 2.01 ê 0.04 s (Fig. 3f)] at 360 K were extracted from the
224 fitted slopes in forward [and reverse bias]. A summary of the
225 extracted tslow values for the entire solar cell working bias
226 conditions (between 0 V and Voc) are shown in Fig. 4 and
227 tabulated in the supporting information (Table S1). Although
228 lower light intensity than AM1.5G was used in our experiments,
229 the extracted values are still in good agreement with previously
230 reported values ranging from seconds to hundreds of seconds.18-20
231 It has been shown that the transient time period is strongly
232 influenced by (i) the perovskite crystal size (e.g. larger crystallites
233 lead to shorter transient times) and (ii) the structural dependence of
234 the underneath TiO2 layer (e.g. the presence of mesoporous TiO2
235 leads to shorter transient times than the case with only c-TiO2.19-21,
236 Our SEM images (Fig. 1a) reveals that our perovskite films are
26
237 composed of interconnected crystallites with smaller and larger
238 grains fully covering the underneath c-TiO2 layer. The
239 determination of the grain sizes is difficult based on the SEM
240 images. No drastic variations in tslow is observed when the sample
241 is held at three different temperatures of 250 K, 300 K, and 360 K.
242 A slight faster tslow of ~2.5 s is observed when the sample is held
243 at 360 K and voltage swept in the reverse direction (Fig. 4c). tslow
244 is proportional to the perovskite capacitance, and therefore to the
245 dielectric constant of the perovskite film. Temperature induced
246 structural phase transitions in lead halide perovskite materials
247 leading to variations in the dielectric constant, e(T), has been
248 previously reported.25, 27, 35, 41-43 For example, in the particular case
249 of CH3NH3PbI3 crystals, a steep discontinuity of e(T) at the
250 orthorhombic ç tetragonal phase transition (161.4 K) and smooth
251 e(T) transition in the tetragonal ç cubic phase (330.4 K) were
252 reported.42 Although, the description above provides an idea on the
253 phase transitions and expected relative e(T) values as a function of
254 temperature, the direct comparison with our perovskite material is
255 not valid because our study was conducted on thin film (~380 nm)
256 that differs significantly from single crystals. In addition, the
257 presence of Cl in our films might have significant influences on
258 the structural phases and transition temperatures.16, 23, 44, 45 Based
4 | J. Name., 2012, 00, 1r3
This journal is © The Royal Society of Chemistry 2012

Journal of Materials Chemistry A
Page 6 of 8
View Article Online
Journal Name
COMMUNICATION
DOI: 10.1039/C4TA04969A
(a)
(c)
(e)
Forward
Forward
Forward
250 K
15
15
10
5
6
4
2
0
I(tdçÑ
)
360 K
300 K
I( d
I( d
I( d
I( d
t
= 1 s)
t
= 200 ms)
= 40 ms)
= 0)
10
5
t
t
0
0
0.0
0.2
0.4
0.6
0.8
1.0
0.0
0.2
0.4
0.6
0.8
0.0
0.2
0.4
0.6
0.8
Voltage (V)
Voltage (V)
Voltage (V)
30
25
20
15
10
5
25
20
15
10
5
25
20
15
10
5
(b)
(d)
(f)
Reverse
Reverse
Reverse
300 K
250 K
I(td = 0)
360 K
I(td = 40 ms)
I(td = 200 ms)
I(td = 1 s)
I(tdçÑ)
0
0
0
0.0
0.2
0.4
0.6
0.8
0.0
0.2
0.4
0.6
0.8
0.0
0.2
0.4
0.6
0.8
1.0
Voltage (V)
Voltage (V)
Voltage (V)
Figure 5. Photocurrent versus voltage curves composed from the time-dependent data shown in Fig. 2 considering the different delay times
(td). The steady-state photocurrent values, I(tdçÑ), were calculated from the semi-logarithmic plots.
259 on the XRD results (Fig. 1b), our perovskite film exhibits the 289 and final values (forward) and discharging down to ~37% of the
260 orthorhombic structure3 measured at room temperature and no 290 initial value (reverse). To reach a photocurrent level that is within
261 large variations in dielectric constants at temperatures of 250 K
291 1% variation from the steady state, it takes a delay time of 5tslow
,
262 and 360 K can be inferred.43 Note that the tslow values differ when 292 which is on the order of ~30 s.18-20 This is impractically long and
263 sweeping the voltage in forward or in reverse scan direction 293 new procedures are needed. At each temperature, comparing only
264 suggesting the strong dependence on the applied voltage history
265 (remnant polarization), which is consistent with the ferroelectricity
266 nature.27
294 the I-V curves corresponding to the steady-state conditions (tdçÑ)
295 in forward and reverse scans, we notice that none of the curves are
296 observed to match each other exactly. A large discrepancy in Isc
297 values between forward and reverse I-V curves is noticed at low
298 (250 K, Figs. 5a,b) and high temperatures (360 K, Figs. 5e,f) while
299 the Isc values are closely matched at 300 K19, 20 (Figs. 5c,d and
300 Table S2). For example, at 250 K, the steady-state I-V curves in
301 the forward and reverse provide Isc (Voc and FF) values of 5.8 mA
302 (0.84 V and 0.24) and 14.9 mA (0.90 V and 0.18), respectively.
303 Meanwhile, at 300 K, the forward and reverse Isc (Voc and FF)
304 values are 15.1 mA (0.77 V and 0.35) and 14.4 mA (0.77 V and
305 0.49), respectively. Finally, at 360 K, the forward and reverse Isc
306 (Voc and FF) values are 14.2 mA (0.54 V and 0.51) and 18.3 mA
307 (0.54 V and 0.49), respectively. Excess of I- or MA+ ions are
308 proposed to be present as interstitial defects induced by photo-
309 excitation and/or during sample preparation conditions. Under
310 reverse bias with illumination, the negatively charged interstitial
311 anions (I-) migrate to the c-TiO2 layer and cations (MA+) towards
312 the spiro-MeOTAD HTL. The ionic concentrations induce energy
313 barriers for electrons and holes at their respective selective-
314 contacts. Forward bias conditions drive these interstitial ions in the
315 opposite directions. The differences in the redistribution of ions
316 within the perovskite cell when operated under forward and
317 reverse bias conditions can explain why the steady-state I-V curves
318 measured for the forward and reverse scan directions (Fig. 5) do
267
The transient photocurrent responses as the consequence of
268 capacitive effects in perovskite solar cells have strong influences
269 on the shape of I-V curves. Based on the time-dependent results
270 shown in Fig. 2, we can deduce the I-V curves (Fig. 5) with the
271 different delay times (td) as indicated in Fig. 3a. The Isc, Voc, and
272 FF extracted from the I-V curves were summarized in the
273 supporting information (Table S2). The I-V curve shapes are
274 significantly influenced by the voltage sweep direction and td.
275 Overall, low photocurrent and high photocurrent are measured in
276 the forward and reverse directions, respectively, when td is small.
277 Increasing td, a rise and decay in the photocurrent are observed in
278 forward and reverse scan directions, respectively, eventually
279 reaching the steady-state photocurrent conditions, I(tdçÑ). In the
280 particular case of the cell measured at 300 K, a hump is observed
281 to appear at ~0.5 V in the reverse direction (Fig. 5d) when the
282 voltage is swept too fast (td ~ 0-40 ms). This is an artifact of the
283 solar cell measurement conditions because the photocurrent values
284 have not reached its steady-state values.24, 46 The delay time in Fig.
285 4 provides a lower bound for the solar cell photocurrent to attain
286 its steady-state values. Considering that perovskite solar cell is
287 modelled with RC circuits,25, 26, 34, 36, 39 the time constant (t = RC)
288 characterizes the capacitor charging of ~63% between the initial
This journal is © The Royal Society of Chemistry 2012
J. Name., 2012, 00, 1r3 | 5

Page 7 of 8
COMMUNICATION
Journal of Materials Chemistry A
VJieowuArrtnicalelONnlainme e
DOI: 10.1039/C4TA04969A
319 not match exactly. Accurate photovoltaic parameters (Voc, Isc, and
320 FF) can only be analyzed, if the I-V measurements are carried out
321 under steady-state conditions.20 Interestingly, Voc decreases as the
322 cell temperature is increased (Fig. S12). In addition, the series
323 resistance is significantly enhanced when the cell is held at 250 K.
380 constants and Isc, Voc, and FF as a function of temperature. See
381 DOI: 10.1039/c000000x/
382
383 1. H. Zhou, Q. Chen, G. Li, S. Luo, T.-b. Song, H.-S. Duan, Z.
384
Hong, J. You, Y. Liu and Y. Yang, Science, 2014, 345, 542-
324 These trends may have the same origin as previously described in
385
546
40
325 solid-state dye sensitized solar cells (ssDSCs)28,
and more
386 2. O. Malinkiewicz, A. Yella, Y. H. Lee, G. M. Espallargas, M.
326 recently highlighted by the strong influences of the selective
327 contacts on the performance of perovskite solar cells.36, 47 Based
328 on a systematic study varying the combinations of TiO2 electron
329 transport layer (E), perovskite (P), and spiro-MeOTAD hole
330 transport layer (H) architectures (E/P/H, E/P, P/H, and P),
331 impedance spectroscopy measurements by Juarez-Perez et al.
332 showed that the photovoltaic parameters (Isc, Voc, and FF) are
333 strongly affected by the hole selective contact.36 Thus, the
334 observed decrease in Isc in our perovskite cell at low temperature is
335 possibly due to the decrease in hole mobility in the organic spiro-
336 MeOTAD HTM.36, 47 On the other hand, the observed increase in
337 FF at high temperature (360 K) is consistent with the reduced
338 series resistance due to high diffusion current.36, 47 In parallel, the
339 charge-recombination rate is expected to increase with temperature
340 inducing an overall decrease in charge density within the cell,
341 which reduces Voc.36, 47 Based on our quantitative analysis shown
342 in Fig. 5, the major influences of temperature are on tfast processes
343 (rather than on the tslow process) that perhaps are associated with
344 the selective-contacts. Further investigation is needed to determine
345 unambiguously the origin of the tfast processes.
387
388
Graetzel, M. K. Nazeeruddin and H. J. Bolink, Nat.
Photonics, 2014, 8, 128-132
389 3. M. Z. Liu, M. B. Johnston and H. J. Snaith, Nature, 2013,
390 501, 395-398
391 4. N. J. Jeon, J. H. Noh, Y. C. Kim, W. S. Yang, S. Ryu and S.
392 I. Seok, Nature Mater., 2014, 13, 897-903
393 5. K. Wojciechowski, M. Saliba, T. Leijtens, A. Abate and H. J.
394 Snaith, Energy Environ. Sci., 2014, 7, 1142-1147
395 6. M. M. Lee, J. Teuscher, T. Miyasaka, T. N. Murakami and H.
396 J. Snaith, Science, 2012, 338, 643-647
397 7. National Renewable Energy
398 http://www.nrel.gov/ncpv/images/efficiency_chart.jpg.,
399 8. L. K. Ono, S. Wang, Y. Kato, S. R. Raga and Y. Qi, Energy
400 Environ. Sci., 2014, DOI: 10.1039/C4EE02539C
401 9. A. T. Barrows, A. J. Pearson, C. K. Kwak, A. D. F. Dunbar,
Laboratory
(NREL),
402
A. R. Buckley and D. G. Lidzey, Energy Env. Sci., 2014, 7,
2944-2950
403
404 10. A. Mei, X. Li, L. Liu, Z. Ku, T. Liu, Y. Rong, M. Xu, M. Hu,
405
406
J. Chen, Y. Yang, M. Grätzel and H. Han, Science, 2014, 345,
295-298
346
407 11. N. K. Noel, S. D. Stranks, A. Abate, C. Wehrenfennig, S.
347 Conclusions
408
409
410
Guarnera, A.-A. Haghighirad, A. Sadhanala, G. E. Eperon, S.
K. Pathak, M. B. Johnston, A. Petrozza, L. M. Herz and H. J.
Snaith, Energy Environ. Sci., 2014, 7, 3061-3068
348
In summary, staircase voltage sweep measurements at 250 K,
349 300 K, and 360 K conducted on perovskite solar cell reveal a
350 complex time-dependent photocurrent transient signal. Our
351 photocurrent data suggest multiple charging-discharging processes
352 take place within perovskite cell. Semi-logarithmic plots of the
353 photocurrent responses reveal a linear regime showing the slowest
354 transient process (well-defined mono-exponential trend) with a
411 12. F. Hao, C. C. Stoumpos, D. H. Cao, R. P. H. Chang and M.
412 G. Kanatzidis, Nature Photon., 2014, 8, 489-494
413 13. M. H. Kumar, S. Dharani, W. L. Leong, P. P. Boix, R. R.
414
Prabhakar, T. Baikie, C. Shi, H. Ding, R. Ramesh, M. Asta,
M. Graetzel, S. G. Mhaisalkar and N. Mathews, Adv. Mater.,
2014, DOI: 10.1002/adma.201401991
415
355 time constant (tslow) in the order of seconds. This process was 416
356 interpreted to be originated from the polarization response of the 417 14. P.-Y. Chen, J. Qi, M. T. Klug, X. Dang, P. T. Hammond and
357 perovskite layer. Additional studies are needed to describe the 418
A. M. Belcher, Energy Environ. Sci., 2014, 7, 3659-3665
419 15. M. Gratzel, Nature Mat., 2014, 13, 838-842
420 16. H. Yu, F. Wang, F. Xie, W. Li, J. Chen and N. Zhao, Adv.
358 convoluted tfast processes (multi-exponential terms), which had
359 stronger influence of temperature. I-V curves under steady-state
360 conditions were composed from the transient photocurrent data.
361 The hysteresis effects were smaller at 360 K and higher at 300 K
362 and 250 K. On the basis of our study, in order to compare the
363 results from different laboratories, it is essential to establish a
364 protocol for extracting hysteresis-free I-V curves on perovskite
365 solar cells corresponding to the steady-state conditions. The
366 extrapolation method used in this work to extract the steady-state
367 photocurrents is suggested as a possible method.
421
422 17. C. Wehrenfennig, M. Z. Liu, H. J. Snaith, M. B. Johnston and
423 L. M. Herz, Energy Environ. Sci., 2014, 7, 2269-2275
424 18. H. J. Snaith, A. Abate, J. M. Ball, G. E. Eperon, T. Leijtens,
Func. Mater., 2014, DOI: 10.1002/adfm.201401872
425
N. K. Noel, S. D. Stranks, J. T.-W. Wang, K. Wojciechowski
426
and W. Zhang, J. Phys. Chem. Lett., 2014, 5, 1511-1515
427 19. H.-S. Kim and N.-G. Park, J. Phys. Chem. Lett., 2014, 5,
428 2927-2934
429 20. E. L. Unger, E. T. Hoke, C. D. Bailie, W. H. Nguyen, A. R.
368
430
431
Bowring, T. Heumuller, M. G. Christoforo and M. D.
McGehee, Energy Environ. Sci., 2014, 7, 3690-3698
369 Acknowledgements
370 This work was financially supported by Okinawa Institute of
371 Science and Technology Graduate University in Japan.
372
432 21. Z. Xiao, Q. Dong, C. Bi, Y. Shao, Y. Yuan and J. Huang,
433 Adv. Mater., 2014, 26, 6503-6509
434 22. C.-W. Chen, H.-W. Kang, S.-Y. Hsiao, P.-F. Yang, K.-M.
435 Chiang and H.-W. Lin, Adv. Mater., 2014, 26, 6647–6652
436 23. R. S. Sanchez, V. Gonzalez-Pedro, J.-W. Lee, N.-G. Park, Y.
373 Notes and references
374 Energy Materials and Surface Sciences Unit, Okinawa Institute of
375 Science and Technology Graduate University, 1919-1 Tancha, Onna-
376 son, Okinawa, 904-0495, Japan. E-mail: yabing.qi@oist.jp; Fax: +81-
377 098-966-1062.
437
S. Kang, I. Mora-Sero and J. Bisquert, J. Phys. Chem. Lett.,
2014, 5, 2357-2363
438
439 24. H. M. Tian, J. Y. Zhang, X. Y. Wang, T. Yu and Z. G. Zou,
440 Measurement, 2011, 44, 1551-1555
441 25. E. J. Juarez-Perez, R. S. Sanchez, L. Badia, G. Garcia-
442
443
378 † Electronic Supplementary Information (ESI) available: Experimental
379 section, semi-logarithmic plots of photocurrents, and tabulated time
Belmonte, Y. S. Kang, I. Mora-Sero and J. Bisquert, J. Phys.
Chem. Lett., 2014, 5, 2390-2394
6 | J. Name., 2012, 00, 1r3
This journal is © The Royal Society of Chemistry 2012

Journal of Materials Chemistry A
Page 8 of 8
View Article Online
Journal Name
COMMUNICATION
DOI: 10.1039/C4TA04969A
444 26. H. S. Kim, I. Mora-Sero, V. Gonzalez-Pedro, F. Fabregat-
445
Santiago, E. J. Juarez-Perez, N. G. Park and J. Bisquert, Nat.
Commun., 2013, 4, 2242
446
447 27. C. C. Stoumpos, C. D. Malliakas and M. G. Kanatzidis,
448 Inorg. Chem., 2013, 52, 9019-9038
449 28. H. J. Snaith, L. Schmidt-Mende, M. Gratzel and M. Chiesa,
450 Phys. Rev. B, 2006, 74, 045306
451 29. S. R. Raga and F. Fabregat-Santiago, Phys. Chem. Chem.
452 Phys., 2013, 15, 2328-2336
453 30. U. B. Cappel, T. Daeneke and U. Bach, Nano Lett., 2012, 12,
454 4925-4931
455 31. L. K. Ono, P. Schulz, J. J. Endres, G. O. Nikiforov, Y. Kato,
456 A. Kahn and Y. Qi, J. Phys. Chem. Lett., 2014, 5, 1374-1379
457 32. G. E. Eperon, V. M. Burlakov, P. Docampo, A. Goriely and
458 H. J. Snaith, Adv. Funct. Mater., 2014, 24, 151-157
459 33. E. Edri, S. Kirmayer, A. Henning, S. Mukhopadhyay, K.
460
461
Gartsman, Y. Rosenwaks, G. Hodes and D. Cahen, Nano
Lett., 2014, 14, 1000-1004
462 34. V. Gonzalez-Pedro, E. J. Juarez-Perez, W.-S. Arsyad, E. M.
463
Barea, F. Fabregat-Santiago, I. Mora-Sero and J. Bisquert,
464
Nano Lett., 2014, 14, 888-893
465 35. J. M. Frost, K. T. Butler, F. Brivio, C. H. Hendon, M. van
466 Schilfgaarde and A. Walsh, Nano Lett., 2014, 14, 2584-2590
467 36. E. J. Juarez-Perez, M. Wuꢀler, F. Fabregat-Santiago, K.
468
469
Lakus-Wollny, E. Mankel, T. Mayer, W. Jaegermann and I.
Mora-Sero, J. Chem. Phys. Lett., 2014, 5, 680-685
470 37. D. Q. Bi, L. Yang, G. Boschloo, A. Hagfeldt and E. M. J.
471 Johansson, J Phys Chem Lett, 2013, 4, 1532-1536
472 38. Y. Zhao, A. M. Nardes and K. Zhu, J. Phys. Chem. Lett.,
473 2014, 5, 490-494
474 39. A. Dualeh, T. Moehl, N. Tétreault, J. Teuscher, P. Gao, M. K.
475 Nazeeruddin and M. Grätzel, Acs Nano, 2014, 8, 362-373
476 40. H. J. Snaith, S. M. Zakeeruddin, L. Schmidt-Mende, C. Klein
477
and M. Gratzel, Angew. Chem. Int. Edit., 2005, 44, 6413-
478
6417
479 41. J. M. Frost, K. T. Butler and A. Walsh, APL Mater., 2014, 2,
480 081506
481 42. N. Onoda-Yamamuro, T. Matsuo and H. Suga, J. Phys.
482 Chem. Solids, 1992, 53, 935-939
483 43. T. Baikie, Y. N. Fang, J. M. Kadro, M. Schreyer, F. X. Wei,
484
S. G. Mhaisalkar, M. Graetzel and T. J. White, J. Mater.
Chem. A, 2013, 1, 5628-5641
485
486 44. S. Colella, E. Mosconi, P. Fedeli, A. Listorti, F. Gazza, F.
487
488
489
Orlandi, P. Ferro, T. Besagni, A. Rizzo, G. Calestani, G.
Gigli, F. De Angelis and R. Mosca, Chem. Mater., 2013, 25,
4613-4618
490 45. B. Suarez, V. Gonzalez-Pedro, T. S. Ripolles, R. S. Sanchez,
491
492
L. Otero and I. Mora-Sero, J. Phys. Chem. Lett., 2014, 5,
1628–1635
493 46. H. Tian, L. Liu, B. Liu, S. Yuan, X. Wang, Y. Wang, T. Yu
494 and Z. Zou, J. Phys. D Appl. Phys., 2009, 42
495 47. F. Fabregat-Santiago, G. Garcia-Belmonte, I. Mora-Sero and
496
J. Bisquert, Phys. Chem. Chem. Phys., 2011, 13, 9083-9118
497
This journal is © The Royal Society of Chemistry 2012
J. Name., 2012, 00, 1r3 | 7