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: Y. Cheng, Y. Han,
S. Meyer, Y. Dkhissi, K. Weber, J. Pringle, U. Bach and L. Spiccia, J. Mater. Chem. A, 2015, DOI:

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 10 Journal of Materials Chemistry A
View Article Online
DOI: 10.1039/C5TA00358J

Journal Name RSCPublishing

Journal of Materials Chemistry A Accepted Manuscript

ARTICLE

Degradation observations of encapsulated planar

Cite this: DOI: 10.1039/x0xx00000x
CH3NH3PbI3 perovskite solar cells at high
Published on 09 March 2015. Downloaded by Portland State University on 12/03/2015 13:26:25.

temperatures and humidity

Yu Han,a Steffen Meyer,b Yasmina Dkhissi,c Karl Weber,d Jennifer M. Pringle,e
Received 00th January 2012, Udo Bach,d,f Leone Spiccia,b* and Yi-Bing Chenga*
Accepted 00th January 2012

DOI: 10.1039/x0xx00000x The stability of encapsulated planar-structured CH3 NH3 PbI 3 (MAPbI 3) perovskite solar cells
(PSCs) was investigated under various simulated environmental conditions. The tests were
www.rsc.org/
performed under approximately one sun (100 mW/cm2) illumination, varying temperature (up
to 85 °C cell temperature) and humidity (up to 80%). The application of advanced sealing
techniques improved the device stability, but all devices showed significant degradation after
prolonged aging at high temperature and humidity. The degradation mechanism was studied by
post-mortem analysis of the disassembled cells using SEM and XRD. This revealed that the
degradation was mainly due to the decomposition of MAPbI 3, as a result of reaction with H2O,
and the subsequent reaction of hydroiodic acid, formed during MAPbI 3 decomposition, with
the silver back contact electrode layer.

Introduction In 2013, Burschka et al. reported a 15% efficiency mesoscopic

MAPbI3 perovskite solar cell made by the sequential deposition
CH3NH3PbI3 (MAPbI3) perovskite solar cells (PSCs) have method that showed very promising long-term stability.15 The
reached a certified efficiency of 20.1% in 2014,1,2 up from device was sealed under argon and maintained more than 80%
3.81% reported in 2009,3 making them a promising alternative of its initial photoelectric conversion efficiency (PCE) after a
to conventional thin film solar cells. These PSCs can be period of 500 h continuous operation at 45 °C at its maximum
assembled as mesoscopic or planar structured devices. power point under a light intensity of approximately 100mW
Originally, the mesoscopic structured PSCs were made using a cm−2. The efficiency decline was mainly due to a decrease in
one-step precursor deposition method,4 which led to the shunt resistance that caused a lower Voc and FF. Later,
uncontrolled morphology of perovskite films and large Leijtens et al. demonstrated that a CH3NH3PbI3-xClx PSC with a
variations in cell efficiency. A sequential deposition method mesoporous Al2O3 scaffold can overcome UV induced
was then introduced to control the morphology of the instability. The cells were encapsulated with an epoxy resin and
perovskite film. This led to a better reproducibility of device a glass coverslip in a nitrogen-filled glove box. Their devices
performance and efficiencies exceeding 15%5. In subsequent showed only a small decrease in photocurrent over 1,000 h
studies, various innovative cell fabrication techniques were under continuous full spectrum AM1.5 solar illumination at 40
developed to produce planar structured PSCs with high °C.17 However, a 50% drop in efficiency occurred primarily
photovoltaic conversion efficiencies: 15.4% by dual-source during the initial 200 h of aging. This was again mainly due to a
vapour deposition6, 12.1% by vapour-assisted solution process7, decrease of Voc and FF.17 The authors also pointed out that the
13.7% by low-temperature solution casting8, and 13.9% by a silver charge colleting electrode could corrode when in contact
solvent-induced fast crystallization method9. Recently, a gas with the perovskite layer, possibly forming silver halide.17 Noh
assisted preparation method was developed by our group and et al. found that the substitution of larger iodide ions with
the cells fabricated by this method demonstrated an average smaller bromide ions in the CH3NH3Pb(I1−xBrx)3 material with
efficiency of 15.7%10. x ≥0.2 led to a greater resistance towards high humidity (≥55%)
Despite the major improvements in device efficiency and at room temperature.18 It was postulated that bromide doping
fabrication methods made over the last few years,11–16 one of leads to a more compact and stable perovskite structure due to a
the most significant concerns for PSCs is their long-term reduction of the lattice constants and the transition to a cubic
stability.5 So far, few investigations have focused on the phase.18 Li et al. demonstrated that CH3NH3PbI3 could be
encapsulation, stability and degradation mechanisms of PSCs.

This journal is © The Royal Society of Chemistry 2013 J. Name., 2013, 00, 1-3 | 1
 Journal of Materials Chemistry A Page 2 of 10
View Article Online
ARTICLE Journal Name
DOI: 10.1039/C5TA00358J

corroded by 4-tert-butylpyridine (4-tBP) which is an important simulated light intensity. The solar cells were dismantled after
additive in the Spiro-OMeTAD hole transporting material.19 the stability tests to investigate the degradation pathways.
The corrosion was proposed to occur because 4-tBP was able to

Journal of Materials Chemistry A Accepted Manuscript

dissolve the CH3NH3PbI3 and form complexes of the type, Experimental
[PbI2(4-tBP)x]. To protect the perovskite, a montmorillonite
(MMT) layer was introduced between the perovskite and Spiro- Cell Fabrication, Sealing and Performance
OMeTAD layers, but long-term durability test results were not
reported.19
Most recently, an unsealed device with approximately 11.5%
initial PCE was reported by Mei et al. which showed excellent
Published on 09 March 2015. Downloaded by Portland State University on 12/03/2015 13:26:25.

stability under AM 1.5 simulated sunlight.20 The authors

produced a device without a hole transporting layer, using a
modified perovskite material (5-AVA)x(MA)(1-x)PbI3 and a
carbon counter electrode. They proposed that the observed
stability over 1000 h, with almost no drop in photovoltaic
conversion efficiency, was due to the perovskite being
protected by the thick carbon layer,20 but the tests were done in Figure 1. Schematic cross-section of the planar perovskite solar cell
ambient conditions.
In all previous studies, the degradation of the perovskite layer The structure of the fabricated perovskite solar cells is shown in
was attributed to a reaction with water from the atmosphere. Figure 1. The gap in the FTO layer was produced by laser
Recently, Frost et al. proposed a plausible decomposition ablation, so that the right side of the substrate acted as the
pathway for MAPbI3 in the presence of water.21 The authors photoanode and the left side of the substrate, which was
suggested that water catalyses the decomposition of MAPbI3 connected to the silver layer, acted as the photocathode. On top
through the reversible reactions (1) & (2) shown below: of the FTO coated glass substrate, after the laser cutting and
glass cleaning with surfactant and ethanol, a thin 30 nm
(1) [(CH3NH3+)PbI3]n+H2O compact TiO2 layer was deposited by spray pyrolysis of a
[(CH3NH3+)n-1(PbI3)n][H3O+] + CH3NH2 diisopropoxide bis(acetylacetonato) titanium(IV) solution at
450 °C. The perovskite layer, ca. 400 nm in thickness, was
(2) [(CH3NH3+)n-1(PbI3)n][H3O+] HI + produced by the gas-assisted deposition technique and
PbI2 + [(CH3NH3+)PbI3]n-1 + H2O subsequent annealing at 100 °C for 10 min according to the
method recently developed in our laboratory.10 The hole
Hydroiodic acid has a boiling point of -35.4 °C22 and transporting layer, ca. 300 nm in thickness, was deposited by
methylamine of around -6 °C23, thus both exist in the gas phase spin coating a solution of Spiro-OMeTAD (41.6 mg) in
at room temperature. In an open system, the continuous release chlorobenzene (0.5 mL), with 3.72 mg lithium
of these gases can drive the decomposition reaction forward bis(trifluoromethanesulfonyl)imide (LiTFSI) and 12.1 µL 4-
and, according to the proposed mechanism, even a small tBP as additives, at a speed of 3000 rpm for 30 s. Both the
amount of water is sufficient to degrade the perovskite material. deposition of perovskite layer and Spiro-OMeTAD layer was
The above studies are only the first steps on the road to meeting performed in a glove box under a dinitrogen atmosphere.
the stringent international standards for outdoor photovoltaic Finally, an 80 nm silver layer was deposited as the back contact
applications.24 As the normal operational environment for by thermal evaporation under vacuum.
photovoltaic devices is under direct sunlight, the cell To test the stability of the perovskite solar cells under simulated
temperature can be up to 45 °C higher than the environmental environmental conditions, two sealing configurations were
temperature.25 Thus, if, for example, the environmental applied, as illustrated in Figures 2 and 3. Sealing method A
temperature is 40 °C, the device temperature can be as high as used a UV-curable epoxy resin from Epotek (3035B) that was
85 °C. Moreover, daily temperature fluctuations can be as high filled between the silver counter electrode and a plain glass
as 45 °C even in mild environmental conditions.26 Inevitably, cover. Sealing method B is a more advanced encapsulation
this also increases the demand for high mechanical stability of technique adopted from the organic light emitting diode
the devices during thermal expansion and contraction. (OLED) technology. This technique was chosen as OLED
Furthermore, solar cells can become exposed to moisture from materials are also very sensitive to moisture. The encapsulated
atmospheric humidity and rain. Thus, it is critical to fully device consists of a cover glass from HanaAMT with a recess
encapsulate the devices to achieve good stability at high in the middle to accommodate a 10×15 mm piece of desiccant
temperatures in the presence of moisture. from Dynic (HG sheet of 180 µm in thickness). The purpose of
Here, we report a study on the encapsulation, stability and this absorber is to adsorb water that has permeated through the
degradation of planar structured PSCs under a number of UV epoxy sealing during long-term operation under simulated
different environmental conditions, including varying atmospheric conditions. The UV-curable epoxy adhesive used
temperature and humidity, under approximately one sun for sealing was purchased from Nagase ChemTex (XNR5516Z-

2 | J. Name., 2012, 00, 1-3 This journal is © The Royal Society of Chemistry 2012
 Page 3 of 10 Journal of Materials Chemistry A
View Article Online
Journal Name ARTICLE
DOI: 10.1039/C5TA00358J

B1), and is specifically designed for OLEDs and has a low The surface and cross sectional SEM images were recorded
water vapour transmission rate of around 3 × 10-4 g/m per day with a FEI Nova dual beam, focused ion beam (FIB), SEM
according to the manufacturer. The sealing of the devices was instrument using 5kV acceleration voltage. Prior to ion beam

Journal of Materials Chemistry A Accepted Manuscript

performed in a glove box under a dinitrogen atmosphere with thinning of the cross section, two platinum protecting layers
less than 0.1 ppm oxygen and moisture. were deposited in situ using an electron beam source at 6.3 nA
Perovskite solar cell devices fabricated by the gas-assisted and an ion beam source at 0.30 nA. The milling of the cross
deposition technique have previously achieved PCEs of 15.7 ± sections was performed with a gallium ion source at a 52°
0.7%.10 In this work, it was found that the encapsulation tilting angle. An X-ray diffractometer (XRD, Philips 1140) with
processes reduced the cell efficiency significantly, possibly due Cu Kα radiation was utilized to compare the crystalline
to a lack of dioxygen doping of the Spiro-OMeTAD layer as structures of various new and aged devices, illustrated in Figure
Published on 09 March 2015. Downloaded by Portland State University on 12/03/2015 13:26:25.

both the device fabrication and encapsulation procedures were 10, with a scan rate of 1° min−1 and a step size of 0.02°.
carried out in an inert atmosphere,27 and/or a reaction of the
perovskite material with vapours outgassing from the UV Environmental Stability Testing Conditions
epoxy during the UV-curing procedure. Thus, the stability Stability testing was conducted using an ATLAS
studies were performed on devices with an initial average PCE Environmental Chamber (SC 340 MHG), which enables a
of around 12%. controlled temperature range of -20 °C to 100 °C, relative
humidity between 0% to 80% and uses a 1200 W metal halide
lamp that gives approximately one sun illumination constantly
during the entire testing period. The output of the halide lamp
was regulated to yield short circuit currents that are comparable
to the values measured using a Xe solar simulator for a standard
Si solar cell. The solar cells were kept inside the chamber at
open circuit, with a 0.16 cm2 circular aperture on top to define
the illuminated area. A K-type thermocouple was directly
attached to one solar cell of each batch to monitor the actual
cell temperature during testing. During illumination, the actual
cell temperature was found to be approximately 30 °C higher
than the environmental temperature inside the chamber. For the
ease and clarity of discussion in the following sections, the
temperatures are always presented as “environmental
temperature (actual cell temperature)”, e.g., 55 °C (85 °C). The
Figure 2. Schematic cross-section of the perovskite solar cells sealed according to
technique A (top) and B (bottom). environmental temperature, actual cell temperature and
photovoltaic performances of the devices (Voc, Jsc, FF and PCE)
were continuously monitored and recorded every 5 minutes
using an automated external I-V recording instrument (National
Instruments PXIe-1078).
The stability testing conditions were designed to closely follow
a consensus of the stability testing protocols for organic
photovoltaic materials and devices.28 Cells sealed by method A
were tested under three different environmental conditions:
i. constant temperature -20 °C (10 °C), 0% relative
humidity, one sun continuous illumination;
ii. constant temperature 55 °C (85 °C), 10% relative
humidity, one sun continuous illumination;
iii. constant temperature 55 °C (85 °C), 80% relative
humidity, one sun continuous illumination.
Cells sealed by method B were then tested and compared with
cells sealed by method A under environmental condition (iii).
Further stability testing of cells sealed by method B under
environmental condition (iv) was then conducted for over 500
hours:
iv. constant temperature 55 °C (85 °C), 50% relative
Figure 3. Photographs of solar cells sealed by method A (top) and B (bottom). humidity, one sun continuous illumination.

Scanning Electron Microscopy (SEM) images and X-ray Results and discussion
diffraction (XRD) patterns

This journal is © The Royal Society of Chemistry 2012 J. Name., 2012, 00, 1-3 | 3
 Journal of Materials Chemistry A Page 4 of 10
View Article Online
ARTICLE Journal Name
DOI: 10.1039/C5TA00358J

Device Stability
The stability of the CH3NH3PbI3 perovskite solar cells sealed by
method A was examined under three different environmental

Journal of Materials Chemistry A Accepted Manuscript

conditions (i, ii and iii). The tests were relatively short, around
120 h, but showed clear differences in device stability (Figure
4A). In the early stage of aging (< 20 h) under condition (i) and
(ii), the cells showed slightly improved PCEs with the
improvement more pronounced under condition (ii). This might
be induced by a reduction of the number of defects in the
CH3NH3PbI3 film with aging or further annealing29 as the
Published on 09 March 2015. Downloaded by Portland State University on 12/03/2015 13:26:25.

process was found to be faster at higher temperature. In

comparison, the cells were the most stable and showed almost
no photodegradation at low temperature and low humidity
conditions (condition (i)), less stable at higher temperatures and
low humidity conditions (condition (ii)) and the least stable at
high temperature and high humidity conditions (condition (iii)).
At 55 °C (85 °C), 80% relative humidity and one sun
illumination, the devices sealed by method A lasted less than 20
h. The cell temperature for both tests at conditions (ii) and (iii)
were above the phase transition temperature of MAPbI3.30
However, the significant difference between the results
obtained under test conditions (ii) and (iii) indicates that
environmental humidity has a stronger impact on the device
stability than temperature (comparison between conditions (i)
and (ii)), even above the phase transition temperature. The
phase transition at around 60℃ might have an effect on stability,
but a conclusion cannot be drawn from our experimental design.
Moisture is a major factor for the device degradation and better
sealing is required to improve its environmental stability. The
more sophisticated sealing method B showed an improved Figure 4. (A) Stability testing of perovskite solar cells sealed by method A under
stability under the same condition (iii), due to better three different environmental conditions; (B) Comparison of the stability of
devices sealed by method A and B and tested under environmental condition (iii).
encapsulation and the presence of a desiccant that reduces
The efficiency of the devices before the test was set as 100% and the results
water vapour ingress and in turn reaction with MAPbI3 (Figure were all normalized and plotted against the initial efficiency. Three devices were
4B), but a significant degradation was still observed within the averaged for each test and the error bars show the standard deviation.
first 4h.
Visual inspection of the tested cells sealed by method A (Figure
5), compared to the new device in Figure 3A, showed that the
middle part of the cell tested under condition (iii) was intact
while the surrounding area was significantly degraded and had
turned yellowish, probably because of the formation of PbI2.
Moreover, the edges of the silver layer were damaged (Figure
5). This suggested that degradation primarily occurred as a
result of ingress of moisture from the environment. Glass is
impermeable to water, but the edges of the sealed device are
most vulnerable to moisture.

4 | J. Name., 2012, 00, 1-3 This journal is © The Royal Society of Chemistry 2012
 Page 5 of 10 Journal of Materials Chemistry A
View Article Online
Journal Name ARTICLE
DOI: 10.1039/C5TA00358J

would get significantly hotter than the covered area in the rest
of the cell. The severer damage observed in the exposed
circular area may be a result of combined thermal degradation

Journal of Materials Chemistry A Accepted Manuscript

and photodegradation.
Published on 09 March 2015. Downloaded by Portland State University on 12/03/2015 13:26:25.

Figure 5. Front and back photographs of a perovskite solar cell sealed by method
A after 120 h illumination at approximately one sun, 55 °C (85 °C) and 80%
relative humidity.

Figure 7. Photographs of a perovskite solar cell sealed with method B after 500 h
Additionally, a 500 h long-term test under environmental of illumination at 55 °C (85 °C) and 50 % humidity (conditions (iv)). The
condition (iv) was conducted using a new batch of three cells encapsulating cover glass was removed for a better view.
sealed by method B. All of these devices were still working
after the test, but showed significant degradation, maintaining After removal of the glass cover, cross-sectional FIB-SEM
only less than 5% of their initial efficiency. All photovoltaic images of the degraded cell were taken (Figures 8 and 9).
parameters were reduced during aging, probably because of the Compared to the new cells, a number of degradation features in
gradual decomposition of the MAPbI3 layer. The decrease in the aged cell were evident: (a) damage and fragmentation of the
FF could also be arising from the gradual increase of the total silver layer; (b) voids formed in the Spiro-OMeTAD layer; (c)
series resistance of the cell, as shown in Figure 6, which was voids formed in the perovskite layer and detachment of the
also measured in-situ during the aging test. A detailed post- perovskite layer from the TiO2 layer; (d) formation of new
mortem analysis of the degraded cells was conducted and the particles containing elements of high atomic numbers (brighter
results are discussed in the following sections. contrast in the SEM image) in the perovskite layer. Further
analyses of the degraded devices are presented in the following
sections.

Figure 6. Gradual increase of the total series resistance of the cells tested at
constant temperature (55 °C (85 °C)) and humidity (50%) under continuous one
sun illumination (environmental condition (iv)). Resistance values from three
devices were averaged and the error bars show the standard deviation.

Investigation of the Degradation Pathways of Perovskite Solar

Cells
For a post-mortem examination of the devices tested for 500 h
at 55 °C (85 °C), the encapsulating cover glass was carefully
removed without damaging the active photovoltaic area. From
the back side it was visually evident that the most severely
degraded area was the middle circular area of the silver layer
that was directly exposed to illumination. The remaining area
that was covered by a circular aperture (Figure 7) showed some
damage but less severe. Under illumination, the uncovered area

This journal is © The Royal Society of Chemistry 2012 J. Name., 2012, 00, 1-3 | 5
 Journal of Materials Chemistry A Page 6 of 10
View Article Online
ARTICLE Journal Name
DOI: 10.1039/C5TA00358J

at 55 °C (85 °C) and 50% humidity under 100 mW cm−2

continuous illumination (condition (iv)) for 500 h in the
environmental chamber. Afterwards, the samples were analysed

Journal of Materials Chemistry A Accepted Manuscript

and compared with the new devices using SEM and powder X-
ray diffraction (XRD).
Published on 09 March 2015. Downloaded by Portland State University on 12/03/2015 13:26:25.

Figure 10. Design of samples with three different structures used for the aging
tests; (A) Sample DA without a perovskite layer; (B) Sample DB without a Spiro-
OMeTAD layer; (C) Sample DC with only the perovskite layer, all on a TiO2
blocking layer coated FTO glass. All samples were encapsulated using the sealing
method B illustrated in Figure 2.

Figure 8. Cross-sectional FIB-SEM images of a new cell (A) and different areas of In the perovskite free sample (DA) the microstructures of the
the aged cell (B-D) at constant temperature (55 °C (85 °C)) and humidity (50%) silver, Spiro-OMeTAD, TiO2 and FTO layers in both the new
for 500 h. Degradation features are marked as follows: voids in the Spiro-
and aged cells showed almost no difference in the cross
OMeTAD layer (□); voids in the perovskite layer and detachment of the
perovskite layer from the TiO2 layer (Δ); damaged and discontinuous silver layer sectional images after the aging test (Figure 11 B and C). In
(○) and formation of particles with higher atomic numbers, likely PbI2 (◊). addition, there was no evidence of degradation of the top Ag
layer, from comparison of Figure 11A and Figure 9A. This
indicated that Spiro-OMeTAD was stable in contact with silver
and TiO2 under condition (iv) (55 °C (85 °C), 50% humidity
and 500 h). Thus, it can be concluded that the damages to the
silver layer and the formation of voids in the Spiro-OMeTAD
layer in the aged solar cell (Figures 8) were mainly associated
with the degradation of the perovskite material at high
temperature and humidity.

Figure 9. SEM images of the surface of the silver layer in a new (A) and an aged
(B) perovskite solar cell that was tested at constant temperature (55 °C (85 °C))
and humidity (50%) for 500 h.

To analyse the causes of the microstructural changes observed

in the SEM images of Figures 8 and 9, and to better understand
the degradation pathways, additional samples with three
Figure 11. SEM images of the silver surface (A), and cross sections of new (B) and
different structures were prepared (Figure 10). Each new aged (C) devices for device DA, silver/ Spiro-OMeTAD/TiO2/FTO/glass structure
sample was designed to omit one of the functional layers in a after testing at 55 °C (85 °C), 50% humidity for 500 h.
normal solar cell device; sample DA was without the perovskite
layer, DB without the Spiro-OMeTAD layer and DC without In the sample without the Spiro-OMeTAD layer (sample DB),
the Spiro-OMeTAD and silver layers. The samples were sealed severe morphological changes were evident as a result of
using the sealing method B (Figure 2) and subsequently tested degradation of both the perovskite and silver layers after the

6 | J. Name., 2012, 00, 1-3 This journal is © The Royal Society of Chemistry 2012
 Page 7 of 10 Journal of Materials Chemistry A
View Article Online
Journal Name ARTICLE
DOI: 10.1039/C5TA00358J

aging test (Figures 12 B and C). More precisely, the FIB cross-
sectional SEM images in Figures 12 D and E show that the
perovskite layer has changed dramatically. That is, a significant

Journal of Materials Chemistry A Accepted Manuscript

amount of it has shrunk to form lumps of residue, leaving voids
in some areas after aging.
Published on 09 March 2015. Downloaded by Portland State University on 12/03/2015 13:26:25.

Figure 13. Powder XRD patterns of (A) new Ag coated on TiO2/FTO/glass; (B)
new, (C) and (D) aged sample DB (Ag/perovskite/TiO2/FTO/glass) at 55 °C (85 °C),
50% humidity; (Δ) peaks from the CH3NH3PbI3 perovskite phase.

Sample DC only had a perovskite layer coated on the TiO2

blocking layer. There was almost no change in the dark colour
of the perovskite layer after aging but an increase in grain size
after aging was observed in the SEM images (Figure 14). From
the powder XRD patterns, the (001) peak of PbI2 can also be
Figure 12. Photographs of the new and aged sample DB found in the aged DC sample (Figure 15). This showed that the
(Silver/Perovskite/TiO2/FTO/glass) before (A) and after (B) the aging test at 55 °C
(85 °C), 50% humidity for 500 h; SEM images of the silver surface of the aged DB
decomposition of perovskite with ingression of water will occur
(C), cross section of the aged DB (D and E). The area where the perovskite layer when it is exposed to high temperature (85 °C in this work) but
totally disappeared is marked by (Δ) in (D); lump shaped degradation by- the degradation was not as severe as that observed in sample
products and voids are marked by (○) in (C and E). Platinum layer deposited for
Focused Ion Beam (FIB) cutting to reveal the cross section of the sample (□) was
DB (Figure 12E). Therefore, it can be seen that the presence of
marked in (E). silver is not beneficial to the stability of the perovskite solar
cells.
To obtain more compositional information, powder XRD was
used to analyse the surface of the new and aged DB samples
(Figure 13). The XRD patterns showed that there was a clear
reduction in the intensity of the Ag and MAPbI3 perovskite
diffraction peaks with the progression of aging, compared to the
new sample. Furthermore, the diffraction peak (111) from γ-
AgI31 and the diffraction peak (001) from PbI2 15can be clearly
observed in the aged samples and their intensities are shown to
increase with time. Thus, it can be concluded that the Ag layer
could be corroded through either the chemical reaction in direct
contact with the perovskite layer or gaseous by-products, most Figure 14. SEM images of the perovskite layer in Device DC, (A) new sample and
(B) aged sample at 55 °C (85 °C), 50% humidity for 500 h.
likely HI, from the perovskite decomposition (see also reaction
2 in the introduction).21 Hence, the lumps observed in the SEM
image in Figures 12 C to E were likely corroded silver and
PbI2. In a closed system like the solar cell encapsulated by
method B, the reaction between silver and HI gas would
continue until all of the Ag was consumed by HI or all the
perovskite material was decomposed. This would result in a
gradual increase of resistivity of the silver layer, as was
observed in Figure 6.

This journal is © The Royal Society of Chemistry 2012 J. Name., 2012, 00, 1-3 | 7
 Journal of Materials Chemistry A Page 8 of 10
View Article Online
ARTICLE Journal Name
DOI: 10.1039/C5TA00358J

gases are trapped within the Spiro-OMeTAD

layer, forming voids, and some reach the silver
layer, causing corrosion of silver by HI, which

Journal of Materials Chemistry A Accepted Manuscript

eventually leads to the gradual increase of series
resistance of the cell observed in Figure 7. Such
reactions are severe when silver is directly
deposited on top of the perovskite and are
slowed down by the insertion of a Spiro-
OMeTAD layer between the CH3NH3PbI3 and
silver in a fully functioning perovskite solar cell.
Published on 09 March 2015. Downloaded by Portland State University on 12/03/2015 13:26:25.

iv. The presence of silver as a back contact will

accelerate the degradation process as the silver
will consume HI and shift the equilibrium of the
decomposition reaction (2) to the right. The
Figure 15. Powder XRD patterns for sample DC, (A) TiO2 on FTO glass; (B) aged corrosive reaction will not stop until the entire
sample of perovskite/TiO2/FTO at 55 °C (85 °C), 50% humidity for 500 h; (C) a CH3NH3PbI3 layer is decomposed or the entire
new sample of perovskite/TiO2/FTO; (Δ) CH3NH3PbI3 perovskite peaks; (□) (001)
peak from PbI2. silver layer is corroded. This also infers that the
use of other materials that can react with acid,
Thus, from the partially degraded perovskite solar cells in such as a ZnO blocking layer and/or ITO
Figure 8 and the investigations of the three different conductive layer, may also accelerate the
configurations (DA, DB, DC), the following observations can perovskite degradation process. Thus, they
be made: would not be compatible with the CH3NH3PbI3
i. In an open system or not properly sealed layer in PSCs.
environment (e.g., devices sealed by method A), Based on the investigations above, a scheme for the degradation
CH3NH3PbI3 can decompose in the presence of processes of encapsulated CH3NH3PbI3 perovskite solar cells
water as a catalyst, forming PbI2, and releasing has been proposed and is depicted in Figure 16.
gases such as HI and CH3NH2 according to
reactions (1) and (2). This process is accelerated
by heat in the case of sealed solar cells as the
sealing becomes more permeable. The release of
the vapours will accelerate the decomposition
reaction (according to reaction 2), so that given
enough time, CH3NH3PbI3 will fully decompose
into PbI2.
ii. In a closed system or properly sealed
environment (e.g., devices sealed by method B),
CH3NH3PbI3 decomposes at a slower rate
(sample DC) due to less ingress of moisture
through the sealing material. Although
degradation is less severe and much slower in a
properly sealed environment, the protecting
material used is a desiccant and cannot consume
the decomposition products of perovskite in this
structure, namely HI and CH3NH2. Therefore,
the vapour pressure of HI and/or CH3NH2
increases and saturates the sealed environment.
In such a situation, the perovskite decomposition
Figure 16. Scheme of the degradation processes of encapsulated
reaction (2) reaches a dynamic equilibrium and methylammonium lead iodide perovskite solar cells illuminated at elevated
thus the decomposition slows down significantly. temperatures and high humidity.
iii. The voids observed in the Spiro-OMeTAD layer
and perovskite layer are likely due to the release
of gases from the decomposition of the
Conclusions
perovskite material. As Spiro-OMeTAD is a The stability of planar methylammonium lead iodide perovskite
softer material compared to the TiO2/FTO/glass solar cells at elevated temperature (up to 85 °C) and humidity
substrate, especially at high temperature, some (up to 80%) has been investigated. The encapsulated devices

8 | J. Name., 2012, 00, 1-3 This journal is © The Royal Society of Chemistry 2012
 Page 9 of 10 Journal of Materials Chemistry A
View Article Online
Journal Name ARTICLE
DOI: 10.1039/C5TA00358J

*
showed good performance at room temperature and zero Corresponding authors. Email: yibing.cheng@monash.edu,
humidity, but degraded at high temperature and humidity. leone.spiccia@moansh.edu.
Improved stability was demonstrated by adopting a sealing

Journal of Materials Chemistry A Accepted Manuscript

technique from the OLED technology that effectively reduced 1. M. a. Green, A. Ho-Baillie, and H. J. Snaith, Nat. Photonics, 2014, 8,
the ingress of moisture into the cell at high temperature. From 506–514.
three purposely designed configurations, the degradation 2. N. J. Jeon, J. H. Noh, W. S. Yang, Y. C. Kim, S. Ryu, J. Seo, and S.
mechanism was analyzed using cross-sectional FIB-SEM and Il Seok, Nature, 2015, 517, 476–480.
powder XRD analyses, and it was found that decomposition of 3. A. Kojima, K. Teshima, Y. Shirai, and T. Miyasaka, J. Am. Chem.
the CH3NH3PbI3 phase became significant at 85°C in the Soc., 2009, 131, 6051.
presence of even a very small amount of water. A number of 4. A. Kojima, K. Teshima, Y. Shirai, and T. Miyasaka, J. Am. Chem.
Published on 09 March 2015. Downloaded by Portland State University on 12/03/2015 13:26:25.

processes took place in conjunction with the perovskite Soc., 2009, 131, 6050–6051.
decomposition: (i) corrosion of the thin silver layer and the 5. M. Grätzel, Nat. Mater., 2014, 13, 838–842.
formation of voids in the Spiro-OMeTAD layer by the gaseous 6. M. Liu, M. B. Johnston, and H. J. Snaith, Nature, 2013, 501, 395–
products from the decomposition of the perovskite material; (ii) 398.
the formation of voids in the perovskite layer, plus detachment 7. Q. Chen, H. Zhou, Z. Hong, S. Luo, H.-S. Duan, H.-H. Wang, Y. Liu,
of the perovskite layer from the TiO2 layer and the formation of G. Li, and Y. Yang, J. Am. Chem. Soc., 2014, 136, 622–625.
PbI2. In a well encapsulated environment, the generated HI gas 8. A. Yella, L.-P. Heiniger, P. Gao, M. K. Nazeeruddin, and M. Grätzel,
is consumed by reaction with the Ag, which promotes further Nano Lett., 2014, 14, 2591–2596.
decomposition of the perovskite material until the cell is fully 9. M. Xiao, F. Huang, W. Huang, Y. Dkhissi, Y. Zhu, J. Etheridge, A.
degraded. Thus, replacement of the silver back contact by a Gray-Weale, U. Bach, Y.-B. Cheng, and L. Spiccia, Angew. Chemie,
chemically more stable conducting material, such as gold or 2014, 126, 10056–10061.
carbonaceous materials, and selection of sealing materials that 10. F. Huang, Y. Dkhissi, W. Huang, M. Xiao, I. Benesperi, S. Rubanov,
have better high temperature performance to further enhance Y. Zhu, X. Lin, L. Jiang, Y. Zhou, A. Gray-Weale, J. Etheridge, C. R.
the sealing/encapsulation of the devices are useful strategies to McNeill, R. A. Caruso, U. Bach, L. Spiccia, and Y.-B. Cheng, Nano
improve the stability of the perovskite solar cells at high Energy, 2014, 10, 10–18.
temperatures and/or in the presence of water. 11. J.-H. Im, C.-R. Lee, J.-W. Lee, S.-W. Park, and N.-G. Park,
Nanoscale, 2011, 3, 4088–4093.
12. I. Chung, B. Lee, J. He, R. P. H. Chang, and M. G. Kanatzidis,
Acknowledgement Nature, 2012, 485, 486–489.
13. H.-S. Kim, C.-R. Lee, J.-H. Im, K.-B. Lee, T. Moehl, A. Marchioro,
S.-J. Moon, R. Humphry-Baker, J.-H. Yum, J. E. Moser, M. Gratzel,
The study was financially supported by the Australian
and N.-G. Park, Sci. Rep., 2012, 2, 591.
Renewable Energy Agency, the Australian Centre for
14. M. M. Lee, J. Teuscher, T. Miyasaka, T. N. Murakami, and H. J.
Advanced Photovoltaics and the Victorian State Government.
Snaith, Science, 2012, 338, 643–647.
YH is grateful for a Monash Graduate Scholarship. SM
15. J. Burschka, N. Pellet, S.-J. Moon, R. Humphry-Baker, P. Gao, M. K.
gratefully acknowledges the Alexander von Humboldt
Nazeeruddin, and M. Grätzel, Nature, 2013, 499, 316–319.
Foundation for a Feodor Lynen Fellowship. Electron
16. M. A. Green, K. Emery, Y. Hishikawa, W. Warta, and E. D. Dunlop,
microscopy access provided by the Monash University Centre
Prog. Photovoltaics Res. Appl., 2013, 21, 827–837.
for Electron Microscopy and the Advanced Microscopy Facility
17. T. Leijtens, G. E. Eperon, S. Pathak, A. Abate, M. M. Lee, and H. J.
at The University of Melbourne are greatly acknowledged.
Snaith, Nat. Commun., 2013, 4, 2885.
18. J. H. Noh, S. H. Im, J. H. Heo, T. N. Mandal, and S. Il Seok, Nano
Notes and references Lett., 2013, 13, 1764–1769.
a
Australian Centre for Advanced Photovoltaics (ACAP), Department of
19. W. Li, H. Dong, L. Wang, N. Li, X. Guo, J. Li, and Y. Qiu, J. Mater.
Materials Engineering, Monash University, Victoria 3800, Australia
b
Chem. A, 2014, 2, 13587–13592.
Australian Centre for Advanced Photovoltaics (ACAP), School of
20. a. Mei, X. Li, L. Liu, Z. Ku, T. Liu, Y. Rong, M. Xu, M. Hu, J. Chen,
Chemistry, Monash University, Victoria 3800, Australia
Y. Yang, M. Gratzel, and H. Han, Science, 2014, 345, 295–298.
c PFPC, School of Chemistry, The University of Melbourne, Victoria
21. J. M. Frost, K. T. Butler, F. Brivio, C. H. Hendon, M. van
3010, Australia
d
Schilfgaarde, and A. Walsh, Nano Lett., 2014, 14, 2584–2590.
CSIRO Manufacturing Flagship, Bayview Ave, Clayton, Victoria 3168,
22. W. F. Giauque and R. Wiebe, J. Am. Chem. Soc., 1929, 51, 1441–
Australia
e
1449.
Australian Research Council Centre of Excellence for Electromaterials
23. J. G. Aston, C. W. Siller, and G. H. Messerly, J. Am. Chem. Soc.,
Science, Deakin University, Melbourne Burwood Campus, 221 Burwood
1937, 59, 1743–1751.
Highway, Burwood, Victoria 3125, Australia
f
24. D. C. Jordan and S. R. Kurtz, Prog. Photovoltaics Res. Appl., 2013,
Tech Fellow, The Melbourne Centre for Nanofabrication, 151
21, 12–29.
Wellington Road, Clayton 3168 Victoria, Australia
25. E. Cuddihy, C. Coulbert, A. Gupta, and R. Liang, JPL Publ., 1986,
VII, 86–31.

This journal is © The Royal Society of Chemistry 2012 J. Name., 2012, 00, 1-3 | 9
 Journal of Materials Chemistry A Page 10 of 10
View Article Online
ARTICLE Journal Name
DOI: 10.1039/C5TA00358J

26. M. C. Alonso García and J. L. Balenzategui, Renew. Energy, 2004,

29, 1997–2010.
27. J. Burschka, A. Dualeh, F. Kessler, E. Baranoff, N.-L. Cevey-Ha, C.

Journal of Materials Chemistry A Accepted Manuscript

Yi, M. K. Nazeeruddin, and M. Grätzel, J. Am. Chem. Soc., 2011,
133, 18042–18045.
28. M. O. Reese, S. a. Gevorgyan, M. Jørgensen, E. Bundgaard, S. R.
Kurtz, D. S. Ginley, D. C. Olson, M. T. Lloyd, P. Morvillo, E. a.
Katz, A. Elschner, O. Haillant, T. R. Currier, V. Shrotriya, M.
Hermenau, M. Riede, K. R. Kirov, G. Trimmel, T. Rath, O. Inganäs,
F. Zhang, M. Andersson, K. Tvingstedt, M. Lira-Cantu, D. Laird, C.
Published on 09 March 2015. Downloaded by Portland State University on 12/03/2015 13:26:25.

McGuiness, S. (Jimmy) Gowrisanker, M. Pannone, M. Xiao, J.

Hauch, R. Steim, D. M. DeLongchamp, R. Rösch, H. Hoppe, N.
Espinosa, A. Urbina, G. Yaman-Uzunoglu, J.-B. Bonekamp, A. J. J.
M. van Breemen, C. Girotto, E. Voroshazi, and F. C. Krebs, Sol.
Energy Mater. Sol. Cells, 2011, 95, 1253–1267.
29. Y. Yamada, M. Endo, A. Wakamiya, and Y. Kanemitsu, J. Phys.
Chem. Lett., 2015, 482–486.
30. Y. Kawamura, H. Mashiyama, and K. Hasebe, J. Phys. Soc. Japan,
2002, 71, 1694–1697.
31. M. Vukic, D. Veselinovic, and V. Markovic, J. Serbian Chem. Soc.,
2007, 72, 857–868.

10 | J. Name., 2012, 00, 1-3 This journal is © The Royal Society of Chemistry 2012