Preparation and characterization of methylammonium tin iodide layers as photovoltaic absorbers

Article abstract of DOI:10.1002/pssa.201532594

The preparation of absorber layers composed of methylammonium tin iodide (CH₃NH₃SnI₃) in a two-step process was investigated. This material is designed as a less toxic alternative to CH₃NH₃PbI₃ which is commonly used as active material in perovskite solar cells. Tin(II) iodide (SnI₂) layers prepared by physical vapor deposition were converted to CH₃NH₃SnI₃ by reaction with a spin-coated solution of methylammonium iodide (MAI). The perovskite particles formed in this process were over 200 nm in size and reached full surface coverage. A band gap of 1.23 eV was determined and the material, thus, absorbs over a broad part of the solar spectrum, broader even than CH₃NH₃PbI₃. The chemical composition and solid state structure of the prepared films were analyzed by X-ray photoelectron spectroscopy and X-ray diffraction, respectively. The films turned out to be remarkably stable, another key prerequisite for applications as absorber layers in perovskite solar cells. Organic metal halides provide absorber layers for solar cells, which may not only revolutionize the fundamental view on solution-processed semiconductor materials but also the feasibility of sustainable large-scale photovoltaics. Presently the chemical instability of these materials and the toxicity of predominantly used lead hinder commercialization. Replacement of lead and preparation of chemically more stable materials may represent a significant step towards a technical realization of perovskite solar cells.

Full text of DOI:10.1002/pssa.201532594

Phys. Status Solidi A 213, No. 4, 975–981 (2016) / DOI 10.1002/pssa.201532594
a
p s s
www.pss-a.com
applications and materials science
Preparation and characterization
of methylammonium tin iodide layers
as photovoltaic absorbers
Editor’s Choice
Manuel Weiss, Jonas Horn, Christoph Richter, and Derck Schlettwein*
Institute of Applied Physics and Laboratory of Materials Research (LaMa), Justus Liebig University Giessen, Heinrich-Buff-Ring 16,
5392 Giessen, Germany
3
Received 6 August 2015, revised 28 September 2015, accepted 14 October 2015
Published online 20 November 2015
3 3 3
Keywords CH NH SnI , perovskite solar cells, photovoltaics, sustainability, tin
*
Corresponding author: e-mail schlettwein@uni-giessen.de, Phone: þ49 641 99 33400, Fax: þ49 641 99 33409
The preparation of absorber layers composed of methyl- size and reached full surface coverage. A band gap of
ammonium tin iodide (CH
was investigated. This material is designed as a less toxic over a broad part of the solar spectrum, broader even than
alternative to CH NH PbI which is commonly used as CH NH PbI . The chemical composition and solid state
active material in perovskite solar cells. Tin(II) iodide structure of the prepared films were analyzed by X-ray
SnI ) layers prepared by physical vapor deposition were photoelectron spectroscopy and X-ray diffraction, respec-
converted to CH NH SnI by reaction with a spin-coated tively. The films turned out to be remarkably stable, another
3 3 3
NH SnI ) in a two-step process 1.23 eV was determined and the material, thus, absorbs
3
3
3
3
3
3
(
2
3
3
3
solution of methylammonium iodide (MAI). The perov- key prerequisite for applications as absorber layers in
skite particles formed in this process were over 200 nm in perovskite solar cells.
ß 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1
Introduction In view of fading traditional energy ABX [8] with A representing an organic cation, e.g.,
3
þ
2þ
sources like coal and natural gas, clear indications of a CH NH , B a metal cation like Pb , and X a halide anion
global climate change and increasing demand for energy in (I , Cl , Br ) [9] which results in CH NH PbI or
3
3
ꢀ
ꢀ
ꢀ
3
3
3
society, one of the most challenging tasks in the present, and CH NH PbI Cl for the best-performing devices.
3
3
3ꢀx
x
in the future will be to establish renewable energy
The first perovskite solar cells were based on the
sources [1]. Solar energy is a very promising source of concept of DSSCs and utilized a liquid electrolyte which
renewable energy, because the amount of solar energy rapidly dissolved the perovskite absorber and allowed a
reaching the earth is three orders of magnitude higher than moderate efficiency of 3.8% only [6]. Replacing the liquid
the amount needed by mankind [2]. It can be converted to electrolyte by a solid hole conductor improved the efficiency
thermal energy and used as such [3], to electricity via to 10% [10, 11]. Recently, the 20-percent-barrier was passed
thermal energy in concentrating solar power systems [4] or as researchers from KRICT produced a perovskite solar cell
directly to electricity using photovoltaic systems [5]. In with a certified power conversion efficiency of 20.1% [12].
addition to well-established crystalline silicon solar cells, PSCs have, therefore, been highlighted as a great scientific
thin film photovoltaics are representing a promising breakthrough [13, 14].
alternative. Among these, technologies are sought that will
The presence of toxic lead in these cells, however, is a
allow technical application on a very large scale. The remaining problem as lead is banned from numerous
efficiency of perovskite solar cells (PSCs), which have only technical applications in many countries, and thus prevents
been introduced in 2009 [6], has quickly been exceeding the commercial use of CH NH PbX in important global
3
3
3
other emerging photovoltaics like dye-sensitized solar cells markets. Other elements from the fourth main group like tin
DSSCs), organic solar cells (OSCs), and amorphous silicon or germanium are the most obvious to replace lead [7, 15],
(
solar cells [7]. They consist of a hybrid organic–inorganic but other divalent cations provide further options [8]. There
absorber material in the perovskite crystal structure has been a report that perovskites incorporating Sn are
ß 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

a
p s s
9
76
M. Weiss et al. : Methylammonium tin iodide layers as photovoltaic absorbers
2þ
4þ
metal-like semiconductors with small band gaps which to the undesired oxidation of Sn to Sn , the reaction of
would not be suitable for photovoltaic absorber layers [16]. previously reported tin-based perovskites with air led to
2þ
Moreover, the instability of the Sn oxidation state in the chemical decomposition of the perovskite films within a few
presence of oxygen posed a major problem as formation of seconds unless the films were vigorously sealed against
4þ
the more stable Sn
further enhances the metallic air [15]. The chemical stability of CH NH SnI films has to
3 3 3
character [16–19]. Despite such problems, there are existing be improved in order to allow technical application of the
reports of functioning Sn-based hybrid perovskite solar cells materials in PSCs.
by Noel et al. and Hao et al. [15, 20]. Those cells were based
Generally, vapor-deposited hybrid perovskite films are
on solution-processed perovskite films in a mesoporous known to be more homogeneous and provide a higher
TiO2 scaffold but showed poor surface coverage and surface coverage than solution-processed films [24]. In this
inhomogeneities [15, 20] unlike the smooth and highly work we will, consequently, use vapor-deposited precursor
covering films of very successfully used CH NH PbX [21]. films in order to prepare CH NH SnI films of improved
3
3
3
3
3
3
Interfacial recombination caused by poor film quality and film quality and, hence, increased chemical stability. Thus,
low surface coverage is an important reason for the weaker materials will be presented that increase options to establish
performance of tin perovskites compared to Pb-based non-toxic PSCs with increased technical applicability.
materials [20, 22, 23]. In addition, the charge-diffusion
length in such solution-processed CH NH SnI films was
2 Experimental Methylammonium iodide (MAI)
3
3
3
determined to be only 30 nm [15], whereas CH NH PbI
was synthesized from 24 mL methylamine (33 wt.% in
3
3
3
exhibits diffusion lengths of around 100 nm and ethanol, Sigma–Aldrich) and 10 mL hydroiodic acid (57 wt.
CH NH PbI Cl of more than 1 mm [24, 25]. In addition % in water, Alfa Aesar) in a round-bottom flask which was
3
3
3ꢀx
x
cooled to 0 8C. After 2 h, the solvents were removed by
distillation at 50 8C. The product was re-dissolved in ethanol
(Roth), precipitated using diethyl ether (Roth), and filtrated.
This process was repeated multiple times for the filtrate [26].
The obtained CH NH I was sublimed at 120 8C and stored
Manuel Weiss studied advanced
materials at Justus Liebig University
Giessen. He received his BSc in 2012
while working on zinc oxide mod-
ifications for dye-sensitized solar
cells. His research on the preparation
of organometal halides for perovskite
solar cells led to an MSc at the
Institute of Applied Physics. Pres-
3
3
under dry nitrogen in a glovebox.
Tin(II) iodide was obtained by refluxing 2 g tin powder
(Merck) and 2.8 g iodine (Scharlau) in 20 mL 2 M hydro-
chloric acid (Roth) at 115 8C in a round-bottom flask. After
the solution had turned yellow, it was filtrated through glass
wool into a heated second round-bottom flask. While the hot
solution slowly cooled to room temperature, tin(II) iodide
crystallized. It was filtrated and rinsed with a mixture of
water and hydrochloric acid [27]. After drying at 150 8C in
ently, he is working as a PhD student in the field of solid
ion conductors at the Institute of Physical Chemistry of
Justus Liebig University Giessen.
vacuum, the orange–red SnI was stored under dry nitrogen.
2
Derck Schlettwein studied chemistry
at the Universities of M €u nster and
Bremen (Germany) with a focus on
physical and macromolecular chem-
istry. He received a PhD (1991)
following research in photoelectro-
chemistry under the guidance of N.
Jaeger and D. W €o hrle (Bremen).
Following PostDoc years in the US,
FTO-coated glass substrates (TEC 7, Sigma–Aldrich)
were subsequently cleaned with RBS 25 solution (Roth),
acetone (Roth), and 2-propanol (Roth) in an ultrasonic bath
for 15 min each and stored in 2-propanol. About 100 nm
(determined by a calibrated quartz crystal microbalance) of
tin(II) iodide were deposited on these substrates by thermal
evaporation at a rate of 2 nm min in a vacuum <10 Pa.
The films were transferred to dry nitrogen without
intermittent contact to air and a solution of MAI in
ꢀ
1
ꢀ3
Japan, and Germany he received a degree for teaching at a
University (habilitation) in Physical Chemistry from
Bremen University in 2000 and was appointed lecturer.
From 2000–2003, he served as Senior Scientist and
Temporary Professor of Physical Chemistry at the
University of Oldenburg before he was appointed
Professor of Applied Physics at Justus Liebig University
Giessen in 2004 where he works until today. His research
is dedicated to the preparation, characterization, and
optimization of organic and organic/inorganic hybrid
materials as new electrodes for photovoltaic cells, organic
field-effect transistors, or electrochromic applications.
2-propanol was spin-coated on top of the SnI films as
2
specified in the text. The perovskite formed during spin
coating. Some films were annealed at 60–100 8C as also
indicated in the text.
Absorption spectra of the films were measured at air
using a diode array spectrometer LOE-USB evaluation line
MMS1 by tec5 with a spectrum of FTO-coated glass serving
as a reference. For X-ray photoelectron spectroscopy,
samples were transferred under dry nitrogen from the
glovebox to a PHI 5000 VersaProbe (base pressure below
ꢀ
6
10 Pa) by use of a transfer module without any
intermittent contact to air. Analysis was performed by
use of monochromatic Al K at a pass energy of 187.85 eV
a
ß 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.pss-a.com

Editor’s
Choice
Phys. Status Solidi A 213, No. 4 (2016)
977
(overview scan) and 23.5 eV (scan of individual peaks).
This is in good agreement with previous reports [15, 28,
X-ray diffraction measurements of thin films were 29]. The band gap was determined to be 1.23 eV using the
performed at air with a Siemens D5000 whose Cu K tube Tauc plot characteristic for direct allowed transitions shown
a
was operated at 30 kV and 40 mA. The slits were set to 2 mm in Fig. 1b which is close to the ideal band gap for
for the pre- and post-sample slit and to 1 mm for the detector photovoltaic applications [30] and in agreement with the
slit. Film morphology was analyzed after transfer through value determined earlier for CH NH SnI [15].
3
3
3
air using a Philips SEM 525 scanning electron microscope
at 20 kV acceleration voltage and 140 mA emission current scanning electron microscopy. The obtained images for
with a secondary electron detector. CH NH SnI perovskite films prepared with different MAI
Morphological characterization was carried out by
3
3
3
concentrations are shown in Fig. 2. It can be seen that the
3
Results and discussion Tin(II) iodide was syn- films in Fig. 2b–f consisted of cube-shaped crystallites
thesized, dried, and deposited on FTO-coated glass sheets covering the whole surface. Bare FTO is not visible. Such
by thermal evaporation. FTO was used as a standardized preferred morphology could be obtained because the
reference substrate to provide conductive samples for evaporated SnI (Fig. 2a) already covers the substrate to
2
various investigation methods. Also the rough FTO surface a significantly higher extent than that reached by solution-
is closer to the actual morphology in a solar cell in based processing [15, 20]. CH NH SnI is known to exist in
3
3
3
comparison with the surface of glass or silicon. SnI was a tetragonal crystal structure with the space group P4mm at
2
also prepared on compact or mesoporous TiO and films of room temperature [17]. As the lattice constant c is longer
2
similar appearance albeit slightly more scattering were than a by just 0.02% the structure conforms to the observed
obtained.
cube-shaped crystallites and is sometimes referred to as
After evaporation of SnI the samples were transferred pseudocubic. This morphology differs clearly from that of
2
to a nitrogen-filled glovebox without any intermediate SnI , which is known to exist in a monoclinic structure [31].
2
contact to air. A solution of MAI (CH NH I) in 2-propanol
The films formed upon treatment with solutions of
3
3
was spin coated onto the SnI . During spin coating, the dark differentMAI concentrationsshowed significant variationsin
2
ꢀ
1
brown to almost black perovskite formed. It was not crystallite size. For 6 mgmL , some big crystallites ranging
necessary to anneal the resulting CH NH SnI films. from 3 to 4 mm were present whereas most of crystallites
3
3
3
Figure 1a shows the absorption spectrum. The films absorb showed sizes between 100 and 200 nm. This agrees to the
over a broad spectral range with the absorption edge at about experiments by Im et al. on CH NH PbI which resulted in
3
3
3
1000 nm.
some large perovskite cuboids on top of a large number of
smaller particles when 6 mg mL MAI solutions were spin-
ꢀ
1
coated on top of PbI and the soaking time was low [32]. At
2
increasedMAIconcentrationsthedistributionofthe observed
CH NH SnI crystals was more uniform with the mean
3
3
3
crystallite size also increasing (Fig. 2c and d). For the films
ꢀ
1
prepared by use of a 40 mg mL solution, however, the
typical grain size decreased again. Smaller crystallite size for
suchhigherMAIconcentrationcanbeexplainedbythehigher
density of seed crystals limiting the final crystal size. At
smaller concentrations, however, the nucleation centers were
far apart from each other and growth of larger crystallites was
ꢀ
1
possible [32]. Moreover, at 40 mg mL
large domains
appeared for which no distinct crystalline shape could be
detected.
By annealing, even for short periods of time (10 min) at
reasonably low temperatures (808C), the crystallites sintered
together (Fig. 2f). Grain boundaries were thereby reduced
which typically should improve the photovoltaic perfor-
mance. Therefore, itwouldbepossibletousesuchmoderately
heated absorber layers even on plastic substrates in order to
produce flexible solar cells when combined with an electron-
accepting layer like ZnO [33] or TiO in rutile structure [34]
2
which both can be prepared at low temperatures.
The crystallite size of CH NH SnI was determined
3
3
3
from the SEM images for different rotation rates and
different MAI concentrations. The results are shown in
0 mg mL MAI solution at 4000 min without any subsequent Fig. 3. As qualitatively discussed above, higher concen-
Figure 1 Absorption spectrum of a CH
3 3 3
NH SnI film prepared by
vapor-deposition of 100 nm SnI and subsequent spin coating of a
2
ꢀ
1
ꢀ1
2
annealing (a) and a Tauc plot (b) to determine the band gap.
trations of MAI in the solution generally led to smaller
www.pss-a.com
ß 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

a
p s s
9
78
M. Weiss et al. : Methylammonium tin iodide layers as photovoltaic absorbers
Figure 2 SEM images of different films: 100 nm vapour-deposited SnI
2
(a), non-annealed CH
3
NH
3
SnI
3
perovskite films prepared by
ꢀ
1
ꢀ1
ꢀ1
ꢀ1
spin-coating of 6 mg mL (b), 10 mg mL (c), 20 mg mL (d), 40 mg mL (e) MAI solutions, and a CH NH SnI perovskite film
3
3
3
ꢀ
1
prepared by use of 20 mg mL MAI and subsequent annealing at 80 8C for 10 min (f).
crystallites. Lower rotation rates resulted in larger crystals efficiencies because of increased internal light
because of more time for the crystallites to grow before scattering [32].
solvent evaporation was completed. The films prepared at
Upon heating at 80–100 8C, Ostwald ripening of the
ꢀ
1
4000 min , while having by far the largest crystallites, particles was observed. The surface energy was decreased
were a lot less homogeneous than the other ones. For these when the particles merged, and thus enlarged their volume-
films, SnI might have been partially dissolved because of to-surface ratio while the total grain boundary energy is also
2
the longer contact time of solvent and film causing holes in minimized [26]. Upon heating at temperatures lower than
ꢀ
1
the perovskite layer. Samples prepared at 6000 min and 60 8C, there was no increase in size noticeable and the
ꢀ
1
8000 min
were of greater homogeneity. The films observed morphology did not change. Films heated at
ꢀ
1
prepared at 6000 min represented the optimum of good 60 8C should, therefore, have identical properties as non-
homogeneity and large crystallites, and thus were used for annealed films. At higher temperatures the thermal energy
the following experiments since films with higher crystallite washighenough to causemass flowwhichled tothe observed
size have been reported to deliver higher photo-conversion increase in average crystal size. The results shown in Fig. 4
3 3 3 3 3 3
Figure 3 CH NH SnI crystallite size for a given rotation rate Figure 4 CH NH SnI crystallite size depending on heating
ꢀ
1
depending on the concentration of the MAI solution used for spin temperature subsequent to spin coating at 6000 min of MAI
coating as determined from the SEM images. The films were not solutions with a concentration indicated in the plot. Each film was
heated after spin coating.
heated for 10 min at the given temperature.
ß 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.pss-a.com

Editor’s
Choice
Phys. Status Solidi A 213, No. 4 (2016)
979
were obtained at films prepared by use of solutions with a would also form during decomposition of CH NH SnI . The
3
3
3
ꢀ
1
concentration of 10 or 20 mgmL
MAI which led to signal intensity of SnO relative to CH NH SnI also did not
2 3 3 3
homogeneous films of rather large crystallites (Fig. 3). For increase upon annealing or storage, all confirming the
ꢀ
1
films prepared by use of solutions with 20 mgmL MAI the material’s superior stability.
crystallites tendedtobea little larger already before annealing
XPS measurements served to analyze the elemental
and the increase upon thermally induced ripening also composition of the prepared CH NH SnI perovskite films
3
3
3
appeared to be slightly higher. This trend was confirmed by and the oxidation state of Sn in the films. Following film
peak analysis in XRD (Fig. 5) based on the Scherrer formula preparation in a nitrogen-filled glovebox the samples were
ꢀ
1
leading to sizes of crystalline domains of 50 nm (10mg mL
transferred directly to the XPS chamber without intermittent
ꢀ
1
MAI) and 65 nm (20 mgmL ) which have, however, to be contact to air. The resulting spectrum of a film prepared from
ꢀ
1
considered as lower limits because of non-compensated 20 mgmL MAI solution is shown in Fig. 6. In the overview
instrumental peak broadening and close spacing of peaks. scan (Fig. 6a), a slight contamination by sodium and oxygen
The crystal structure of the CH NH SnI films was from glass fragments present on the surface after cracking of
3
3
3
analyzed by XRD. XRD patterns of a film before and after the sample was detected. Apart from that, however, the
annealing are shown in Fig. 5. The most robust preparation spectrum corresponds well to the elements expected in
ꢀ
1
conditions of a concentration of 10 or 20 mgmL and a CH NH SnI . The atomic distribution C:N:Sn:I in stoichio-
3
3
3
ꢀ
1
rotation of 6000 min were chosen. The results match well metric CH NH SnI should equal 1:1:1:3. The distribution
3
3
3
with literature data on single crystals and confirm the 1.2:0.8:1.0:2.9 determined experimentally for the present
tetragonal structure of a-CH NH SnI whichwasexpectedto film corresponds quite well to this, confirming the high purity
3
3
3
be present at room temperature [17]. SnO peaks were of the material already concluded from the XRD analysis.
2
detected caused by the FTO substrates the CH NH SnI
The slightly increased amount of carbon found in the film
3
3
3
perovskite films were prepared on. The absence of SnI2 could be caused by typical contaminations even under UHV
reflexes shows that the evaporated tin(II) iodide was conditions. It is important to note that no excess MAI was
completely converted to CH NH SnI during spin coating found in the film which would have led to an increased
3
3
3
the solution of MAI. Peaks arising from unreacted MAI [35] nitrogen content. The Sn3d5 peak (Fig. 6b) showed almost
were observed mainly on the film before annealing, were exclusively one type of tin with only a small shoulder at
found significantly smaller after annealing and were higher binding energy. The main band at 487 eV can be
2þ
completely absent on a film stored for 20 days. This latter assigned to Sn whereas a small additional signal around
4þ
finding isofhighrelevanceasdecomposition ofCH NH SnI
488 eV might be caused by the presence of Sn at a content,
3
3
3
which could also lead to MAI was not observed in this study. however, of less than 1%. Since, considerably smaller
Further, there is no indication of the formation of SnI which
2
Figure 5 Diffractograms of a perovskite film from spin coating a
ꢀ
1
ꢀ1
solution of 20 mg mL MAI in 2-propanol at 6000 min before
and after annealing compared to a diffractogram of a non-annealed
ꢀ
1
ꢀ1
sample 20 days after preparation (10 mg mL ; 6000 min ). The
peaks for CH NH SnI are assigned as (hkl) [17] and marked by Figure 6 X-ray photoelectron spectrum of a perovskite film
3
3
3
ꢀ
1
*
, signals of MAI [35] are marked by 4, and substrate signals prepared by spin coating 20 mg mL
MAI in 2-propanol at
ꢁ
ꢀ1
from SnO [JCPDS #00-041-1445] are marked by . Signals of 6000 min onto evaporated SnI and subsequent annealing for
2
2
SnI
2
[JCPDS #00-025-0975] were not detected.
30 min at 60 8C (a) and details of Sn3d5 (b).
www.pss-a.com
ß 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

a
p s s
9
80
M. Weiss et al. : Methylammonium tin iodide layers as photovoltaic absorbers
concentrations of Sn(IV) already led to a metallic conduction nevertheless was still high. Significantly, higher stability of
in CH NH SnI which is destructive in a photovoltaic the present films compared to earlier reports was clearly
3
3
3
absorber layer [16, 17, 20], additional resistance measure- confirmed, but still lacks a mechanistic explanation.
ments were performed. Rather high resistivity was found for The stability of the present films was also investigated
the samples speaking in favor of a sufficiently low content of under illumination. It was found earlier that CH NH SnI
3
3
3
Sn(IV) to open options of using such material as absorber can degrade rapidly when exposed to light where the films
layer (Supporting Information, Fig. S2, online at: www.pss-a. can lose their color within a few minutes [15]. The present
com).
samples, however, were able to withstand degradation to a
The CH NH SnI perovskite films prepared according higher degree. After 90 min, illumination of the sample was
3
3
3
to the method proposed in this study turned out to be started which led to a certain decrease in absorbance. This
remarkably stable. The films used for the depicted XRD decrease, however, was still low compared to earlier reports.
patterns (Fig. 5) have been exposed to air for about an hour The increased absorbance below 500 nm might be due to
and exhibited only minor color changes compared to the as- increased light scattering as a consequence of film ripening
prepared films. In the preceding studies, it was reported that which is currently still under investigation. After 40 min of
decoloration of air-exposed CH NH SnI films occurred illumination and an overall exposure to air of more than 2 h
3
3
3
within seconds and that the resulting diffractograms were the film retained remarkable 80% of its initial absorbance at
equivalent to sheer MAI [15], very much contrasting to the 550 nm and a color change was hardly noticeable.
situation in the present study (Fig. 5). The measurement of a
2
0 days old sample showed no presence of MAI nor SnI2.
This sample had been exposed to air for about half an hour deposition and solution-based techniques was proposed for
on the day of its preparation and was then stored for 20 days the preparation of CH NH SnI perovskite films that could
in the nitrogen-filled glovebox) until another measurement serve as an alternative to toxic CH NH PbX as absorber
was performed. Even after this long period of time following layers in perovskite solar cells. Complete surface coverage
initial contact to air there was no degradation detected.
was reached which was not reached until now by one-
The decoloration of air-exposed films was, therefore, step-approaches utilizing a solution of all precursors. In this
explicitly investigated by UV/vis spectroscopy. Figure 7 study, two-step processing of a preformed SnI film turned
shows absorption spectra at different times after the sample out to be particularly successful when using vapor-deposited
was removed from the glovebox in which it was prepared. SnI rather than spin-coated SnI (Supporting Information,
The first spectrum, (as-prepared), was measured 7 min after Fig. S3), presumably because of a more homogeneous
the first contact to air. Within the first subsequent 30 min coating by SnI and, hence, good conversion and coverage
corresponding to typical measurements of XRD (Fig. 5) the by CH NH SnI even at the short contact times needed to
change in absorbance was still negligible. After 60 and avoid dissolution of SnI . Complete surface coverage,
0 min, the absorbance decreased more noticeably but however, is of high relevance to suppress leakage at the
4 Conclusions A two-step process utilizing vapor-
3
3
3
(
3
3
3
2
2
2
2
3
3
3
2
9
contact between electron and hole conductor. We found that
an increase of the MAI concentration in the solution used for
conversion of SnI to CH NH SnI during spin-coating led
2
3
3
3
to smaller particles and also increased the homogeneity of
the films. The optimum balance between a particle size of
about 250 nm and good film homogeneity was reached at a
ꢀ
1
concentration of 10–20 mg mL MAI and a spin-coating
ꢀ
1
frequency of 6000 min . XPS and XRD showed that the
films consisted of pure CH NH SnI in the tetragonal
3
3
3
a-structure and that no contaminations by precursors nor
degradation products could be observed in the films. Tin was
almost completely present as Sn(II), another important
aspect for photovoltaic applicability of the material (see
Supporting Information). It was shown that the SnI2
precursor was completely converted to the perovskite
already during spin-coating even without subsequent
annealing which, however, served to increase film quality.
Finally, and perhaps most important, the films were of
improved stability against decomposition at air and even
under illumination compared to CH NH SnI materials
3
3
3
3 3 3
Figure 7 Absorption spectra of a CH NH SnI film during
stability tests in air. Measurements started in the dark and after
studied earlier [15, 20]. Thus, these films consist of a
material that has good prospects to be successfully used as
absorber layers in lead-free perovskite solar cells, which
might become a key step in their commercialization.
9
0 min a continuous illumination of the film started. The film had
ꢀ
1
been prepared by spin coating a solution of 20 mg mL MAI at
ꢀ
1
6
000 min and subsequent annealing for 30 min at 60 8C.
ß 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.pss-a.com

Editor’s
Choice
Phys. Status Solidi A 213, No. 4 (2016)
981
Supporting Information
Additional supporting information may be found in the
online version of this article at the publisher’s web-site.
[18] D. B. Mitzi, in: Progress in Inorganic Chemistry, edited by
K. D. Karlin (Wiley-Blackwell, New York 2007).
[
[
[
19] D. B. Mitzi, C. D. Dimitrakopoulos, and L. L. Kosbar, Chem.
Mater. 13, 3728 (2001).
20] F. Hao, C. C. Stoumpos, D. H. Cao, R. P. H. Chang, and
M. G. Kanatzidis, Nature Photon. 8, 489 (2014).
21] Y.-J. Jeon, S. Lee, R. Kang, J.-E. Kim, J.-S. Yeo, S.-H. Lee,
S.-S. Kim, J.-M. Yun, and D.-Y. Kim, Sci. Rep. 4, 6953
Acknowledgement The authors are grateful for support
within LaMa by J. Sann in XPS and by M. Kracht and M. Serafin in
XRD.
(
2014).
References
[
22] J. Burschka, N. Pellet, S.-J. Moon, R. Humphry-Baker,
P. Gao, M. K. Nazeeruddin, and M. Gr €a tzel, Nature 499, 316
(2013).
[
1] S. Kazim, M. K. Nazeeruddin, M. Gr €a tzel, and S. Ahmad,
Angew. Chem. Int. Ed. 53, 2812 (2014).
[
2] N. S. Lewis and D. G. Nocera, Proc. Natl. Acad. Sci. USA [23] J. M. Ball, M. M. Lee, A. Hey, and H. J. Snaith, Energy
1
03, 15729 (2006).
Environ. Sci. 6, 1739 (2013).
[
[
3] S. A. Kalogirou, Prog. Energy Combust. 30, 231 (2004).
4] H. L. Zhang, J. Baeyens, J. Degr ꢀe ve, and G. Cac ꢀe res, Renew.
Sustain. Energy Rev. 22, 466 (2013).
[24] M. Liu, M. B. Johnston, and H. J. Snaith, Nature 501, 395
(2013).
[25] S. D. Stranks, G. E. Eperon, G. Grancini, C. Menelaou,
M. J. P. Alcocer, T. Leijtens, L. M. Herz, A. Petrozza, and
H. J. Snaith, Science 342, 341 (2013).
[
[
[
5] C. Breyer, O. Koskinen, and P. Blechinger, Renew. Sustain.
Energy Rev. 49, 610 (2015).
6] A. Kojima, K. Teshima, Y. Shirai, and T. Miyasaka, J. Am. [26] Q. Chen, H. Zhou, Z. Hong, S. Luo, H.-S. Duan, H.-H. Wang,
Chem. Soc. 131, 6050 (2009).
Y. Liu, G. Li, andY. Yang, J. Am. Chem. Soc. 136, 622 (2014).
7] P. P. Boix, K. Nonomura, N. Mathews, and S. G. Mhaisalkar, [27] A. M. Stolzenberg, T. P. Gray, S. Chingin, G. S. Girolami, and
Mater. Today 17, 16 (2014).
8] H. J. Snaith, J. Phys. Chem. Lett. 4, 3623 (2013).
A. P. Sattelberger, Inorganic Syntheses (Wiley, Hoboken, NJ,
2014).
[
[
9] M. A. Loi and J. C. Hummelen, Nature Mater. 12, 1087 [28] F. Hao, C. C. Stoumpos, Z. Liu, R. P. H. Chang, and M. G.
(
2013).
Kanatzidis, J. Am. Chem. Soc. 136, 16411 (2014).
[
10] H.-S. Kim, C.-R. Lee, J.-H. Im, K.-B. Lee, T. Moehl, [29] M. H. Kumar, S. Dharani, W. L. Leong, P. P. Boix, R. R.
A. Marchioro, S.-J. Moon, R. Humphry-Baker, J.-H. Yum,
J.E.Moser, M.Gr €a tzel, andN.-G. Park, Sci. Rep.2, 591(2012).
11] M. M. Lee, J. Teuscher, T. Miyasaka, T. N. Murakami, and
H. J. Snaith, Science 338, 643 (2012).
Prabhakar, T. Baikie, C. Shi, H. Ding, R. Ramesh, M. Asta,
M. Gr €a tzel, S. G. Mhaisalkar, and N. Mathews, Adv. Mater.
26, 7122 (2014).
[
[
[
[
[
[30] W. Shockley and H. J. Queisser, J. Appl. Phys. 32, 510 (1961).
12] M. A. Green, K. Emery, Y. Hishikawa, W. Warta, and [31] R. A. Howie, W. Moser, and I. C. Trevena, Acta Cryst. B 28,
E. D. Dunlop, Prog. Photovolt. 23, 1 (2015).
2965 (1972).
13] P. Gao, M. Gr €a tzel, and M. K. Nazeeruddin, Energy Environ. [32] J.-H. Im, I.-H. Jang, N. Pellet, M. Gr €a tzel, and N.-G. Park,
Sci. 7, 2448 (2014).
Nature Nanotechnol. 9, 927 (2014).
14] H.-S. Kim, S. H. Im, and N.-G. Park, J. Phys. Chem. C 118, [33] T. Yoshida, J. Zhang, D. Komatsu, S. Sawatani, H. Minoura,
5
615 (2014).
T. Pauport ꢁe , D. Lincot, T. Oekermann, D. Schlettwein,
H. Tada, D. W €o hrle, K. Funabiki, M. Matsui, and H. Yanagi,
Adv. Funct. Mater. 19, 17 (2009).
15] N. K. Noel, S. D. Stranks, A. Abate, C. Wehrenfennig,
S. Guarnera, A. A. Haghighirad, A. Sadhanala, G. E. Eperon,
S. K. Pathak, M. B. Johnston, A. Petrozza, L. M. Herz, and [34] A. Yella, L.-P. Heiniger, P. Gao, M. K. Nazeeruddin,
H. J. Snaith, Energy Environ. Sci. 7, 3061 (2014).
16] D. B. Mitzi, S. Wang, C. A. Feild, C. A. Chess, and A. M.
Guloy, Science 267, 1473 (1995).
M. Gr €a tzel, and M. K. Nazeeruddin, Nano Lett. 14, 2591
(2014).
[35] O. Yamamuro, T. Matsuo, H. Suga, W. I. F. David, R. M.
Ibberson, and A. J. Leadbetter, Acta Cryst. B 48, 329
(1992).
[
[
17] C. C. Stoumpos, C. D. Malliakas, and M. G. Kanatzidis,
Inorg. Chem. 52, 9019 (2013).
www.pss-a.com
ß 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim