Materials Chemistry A: Journal of

View Article Online
View Journal
Journal of
Materials Chemistry A
Materials for energy and sustainability
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: E. Greul, M. L.
Petrus, A. Binek, P. Docampo and T. Bein, J. Mater. Chem. A, 2017, DOI: 10.1039/C7TA06816F.
Volume 4 Number 1 7 January 2016 Pages 1–330 This is an Accepted Manuscript, which has been through the
Royal Society of Chemistry peer review process and has been
Journal of accepted for publication.
Materials Chemistry A Accepted Manuscripts are published online shortly after
Materials for energy and sustainability
acceptance, before technical editing, formatting and proof reading.

www.rsc.org/MaterialsA
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the

author guidelines.
Please note that technical editing may introduce minor changes

to the text and/or graphics, which may alter content. The journal’s
ISSN 2050-7488
standard Terms & Conditions and the ethical guidelines, outlined
in our author and reviewer resource centre, still apply. In no
PAPER
Kun Chang, Zhaorong Chang et al.
Bubble-template-assisted synthesis of hollow fullerene-like
MoS2 nanocages as a lithium ion battery anode material
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.
rsc.li/materials-a
Page 1 of 11 Please
Journal ofdo not adjustChemistry
Materials margins A
View Article Online
DOI: 10.1039/C7TA06816F
Journal Name
Journal of Materials Chemistry A Accepted Manuscript

ARTICLE
Highly stable, phase pure Cs2AgBiBr6 double perovskite thin films

Published on 05 September 2017. Downloaded by Gazi Universitesi on 06/09/2017 01:26:17.
for optoelectronic applications

Received 00th January 20xx, a a a b a
Accepted 00th January 20xx Enrico Greul, Michiel L. Petrus, Andreas Binek , Pablo Docampo and Thomas Bein *
DOI: 10.1039/x0xx00000x Hybrid lead halide perovskites have emerged as high-performing semiconductors for optoelectronic applications such as
photovoltaics. However, their toxicity and stability issues represent significant challenges. Recently, double perovskites
www.rsc.org/
have been suggested as an alternative and, especially, Cs2AgBiX6 (X = Cl, Br) has proved to be a promising material as it is
non-toxic and highly stable. However, the low solubility of precursors has so far hampered the fabrication of high quality
films. Here, we demonstrate for the first time the fabrication of Cs2AgBiBr6 films and incorporate them into working
devices. Powder X-ray diffraction measurements revealed that high annealing temperatures of at least 250 °C are required
to fully convert the precursors into Cs2AgBiBr6. After the optimization of the synthesis conditions, photovoltaic devices
comprising our Cs2AgBiBr6 films show power conversion efficiencies (PCE) close to 2.5% and a open circuit voltage (Voc)
exceeding one volt, which is currently the highest Voc reported for a bismuth halide based perovskite, showing the
potential of double perovskites as the absorber material. Furthermore, our results revealed excellent stability of the
devices upon exposure to working conditions without encapsulation. Our work opens the way to a new class of
perovskites with significant potential for optoelectronic applications.
bismuth ions cannot simply be incorporated into the hybrid

1+ 2+
Introduction three-dimensional A M X3 structure of the lead-based
3+
compounds. Generally, related hybrid Bi compounds feature
low-dimensional structures resulting in less favorable
Since the first reports of solid-state solar cells in 2012, hybrid optoelectronic properties than their lead-based
organic-inorganic lead halide perovskites emerged as some of counterparts.
15–18
the most promising absorber materials for low-cost 3+

In order to incorporate Bi ions into a three-dimensional
photovoltaic devices.1 Published PCEs of over 22% make these structure, a class of materials called elpasolites, also known as
materials potential candidates for replacing commonly double perovskites, could provide an attractive extension of
employed Si-based solar cells.2 Despite their excellent the conventional perovskite system. Elpasolites, with the
photophysical properties,3–8 lead halide perovskites still suffer 1+ 3+
general formula A2M M X6, feature a highly symmetric cubic
two serious disadvantages, namely stability issues and the double perovskite structure, with one monovalent and one
toxicity of lead.9–12 Many research groups have therefore 19
trivalent cation. Bismuth-containing elpasolites with A = Rb,
focused on the development of more stable and/or lead-free 1+
Cs, CH3NH3; M = Na, K, Tl and X = F, Cl, Br have been
perovskite materials. Substitution of the Pb2+ by the investigated for several decades.
20–24
Unfortunately,
homovalent group-14 element Sn2+ initially led to promising compounds containing alkali metals do not absorb light in the
PCEs of ∼ 6%. However, the devices showed extremely low visible range due to their large band gap energies (Eg)
stability, even under commonly used inert conditions, due to exceeding 3 eV.
20,25
Although the thallium-containing
the facile oxidation of Sn2+. So far, this makes Sn2+-based compound features a reasonable small Eg of 2.16 eV, the
materials unfeasible for typical photovoltaic applications.13,14 A severe toxicity of thallium excludes this element as a non-toxic
recent approach towards lead-free perovskite materials is the alternative for lead.
23
substitution of Pb2+ by heterovalent M3+ cations. A promising 3+

Recently, several groups have reported Bi -based double
candidate for this type of substitution is non-toxic Bi3+, which is +
perovskites with Ag as the monovalent cation featuring Eg´s of
isoelectronic with Pb2+. Due to the higher charge of Bi3+, ∼ 2 eV for the bromide-based material, photoluminescence
(PL) lifetimes of several hundred nanoseconds and calculated
charge carrier effective masses close to those calculated for
methylammonium lead iodide, making it a very interesting
19, 26–30
candidate for photovoltaic applications.
However, the preparation of high quality films of these
perovskites for optoelectronic applications is challenging,
This journal is © The Royal Society of Chemistry 20xx J. Name., 2013, 00, 1-3 | 1
Please do not adjust margins

Please
Materials margins A Page 2 of 11
View Article Online
DOI: 10.1039/C7TA06816F
ARTICLE Journal Name
26
especially for the bromide and iodide based systems. As a a vigorously stirred 0.43 mM (735 µL) solution of titanium
result of the difficulties to process this material, neither films isopropoxide (99.999%, Sigma–Aldrich) in 2-propanol (5 mL).
nor photovoltaic devices based on Ag-Bi double perovskites Afterwards, an approx. 800 nm thick, mesoporous (mp)-TiO2
have been reported so far. layer was applied by spin-coating 100 µL of a TiO2 nanoparticle

Since thin films are critical for making double perovskites paste (Dyesol DSL 18NR-T) diluted in absolute ethanol (1:2
accessible for optoelectronic applications, we have developed weight ratio) onto the compact TiO2 layer at 2500 rpm for 30 s,
a synthetic route that allows for the preparation of phase followed by subsequent annealing at 500 °C for 15 min under
pure, thin films of the double perovskite Cs2AgBiBr6. Extensive ambient conditions.
powder X-ray diffraction (PXRD) investigations revealed that
high annealing temperatures of at least 250 °C are needed to Perovskite film preparation
remove side phases which form during the film synthesis. Light
absorption measurements as well as steady-state and time- The precursor solution was prepared by dissolving 268 mg
resolved PL measurements show that our films feature BiBr3 (≥ 98%, Sigma Aldrich), 112.8 mg of AgBr (99.5%, Alfa
absorption properties and lifetimes of the photoexcited Aesar) and 254 mg CsBr (99 %, Alfa Aesar) in dimethylsulfoxide
species similar to the lead-based counterparts. First steps to (1 mL, DMSO, anhydrous, 99.9%, Sigma-Aldrich). After
optimize the film synthesis conditions were undertaken to complete dissolution of the precursors, both the solution and
improve the optoelectronic properties of our Cs2AgBiBr6 films, the substrate were preheated to 75 °C before spin-coating. 100
leading to photovoltaic devices with PCEs of up to 2.43% and a µL of the hot precursor solution was spin-coated onto the
high Voc exceeding one volt. Additionally, the Cs2AgBiBr6-based TiO2-covered substrate at 2000 rpm for 30 s. The substrates
devices revealed a high stability under operating conditions. were subsequently annealed at 285 °C for 5 min under
Hence, we demonstrate the potential of highly tunable double ambient conditions to allow for a complete formation of the
perovskites as a novel class of semiconducting materials for desired double perovskite phase. The temperatures for both
optoelectronic applications. the preheating and the annealing steps were also varied to
identify the optimal conditions for phase pure films.
Experimental
Solar cell fabrication
All chemicals were used as received without any further

After film formation, the films were covered with a HTL of
purification. All synthesis steps were conducted under ambient
2,2',7,7'-tetrakis-(N,N-di-4-methoxyphenylamino)-9,9'-
conditions except the application of the hole transporting layer
spirobifluorene (spiro-OMeTAD, Borun Chemicals, 99.5%
(HTL) which was performed in a nitrogen-filled glove box.
purity). The HTL solution was prepared by dissolving 73 mg of
spiro-OMeTAD in chlorobenzene (1 mL, 99.8%, Sigma–Aldrich).
Substrate preparation The solution was filtered and mixed with 4-tert-butylpyridine
-1
(10 µL, TBP, 96% Sigma Aldrich) and a 173 mg mL
Fluorine-doped tin oxide (FTO)-coated glass sheets (7 Ωsq ,
-1 bis(trifluoromethane)sulfonamide lithium salt (LiTFSI, 99.95%,
Pilkington, USA) were patterned by etching with zinc powder Sigma–Aldrich) solution in acetonitrile (30 µL). This solution
and 3 M HCl. They were subsequently cleaned with a 2% was spin-coated dynamically at 1500 rpm for 45 s. In a second
Hellmanex solution and rinsed with deionized water and step the sample rotation was accelerated to 2000 rpm for 5 s
ethanol. Directly before applying the blocking layer, remaining to allow the solvent to dry completely. Finally, 40 nm thick
organic residues were removed by an oxygen plasma gold electrodes were thermally deposited under a high
treatment for 5 min. A compact titanium dioxide (TiO2) layer vacuum on top of the device.
was deposited by spin-coating a sol-gel precursor solution at
2000 rpm for 45 s followed by subsequent annealing at 500 °C Characterization
for 45 min. For preparing the sol-gel solution, a 27.2 mM (70
µL) solution of HCl in 2-propanol (5 mL) was added dropwise to
2 | J. Name., 2012, 00, 1-3 This journal is © The Royal Society of Chemistry 20xx

Page 3 of 11 Please
Materials margins A
View Article Online
DOI: 10.1039/C7TA06816F
Journal Name ARTICLE

Fig. 1 Schematic of the synthesis route for Cs2AgBiBr6 thin films. The film formation (3.) occurs already while the substrate is spinning.
PXRD measurements were performed in transmission mode on photo-emission. J-V curves were recorded with a Keithley 2400
a STOE Stadi MP diffractometer with a Cu Kα1 radiation source sourcemeter under simulated AM 1.5 sunlight, calibrated to
−
(λ = 1.54060 Å) operating at 40 kV and 40 mA. The 100 mW cm ² with a Fraunhofer ISE certified silicon cell. The
diffractometer was equipped with a DECTRIS MYTHEN 1K active area of the solar cells was defined with a square metal
2
solid-state strip detector. PXRD measurements on thin films aperture mask of 0.0831 cm . External quantum efficiency
were performed using a Bruker D8 Discover X-ray (EQE) measurements were performed at short circuit with a
diffractometer operating at 40 kV and 30 mA, employing Ni- bias illumination of 0.1 sun provided by an AM 1.5 solar
filtered Cu Kα radiation (λ = 1.5406 Å) and a position-sensitive simulator (Solar Light Model 16S). The measurements were
detector (LynxEye). Scanning electron microscopy (SEM) conducted under low-frequency chopped monochromatic light
images and EDX data were acquired on a FEI Helios NanoLab (12 Hz). A 150 W xenon lamp equipped with a monochromator
G3 UC microscope. The sample was fixed by silver paste. For and order-sorting filters was used as a light source. The light
the optical characterization, precursor solutions were intensity reaching the electrode was measured using a
prepared similar to those used for the devices. Similar certified Fraunhofer ISE silicon reference cell equipped with a
deposition conditions were also chosen. For UV-Vis KG5 filter at a frequency of 40 Hz. For signal detection a Signal
measurements the films were prepared on glass substrates or Recovery 7265 lock-in amplifier combined with a low-noise
on the mp-TiO2 substrates also used for solar cells. For PL current amplifier (Femto DLPCA-200) was employed.
measurements the double perovskite films were either
prepared on approximately 800 nm thick mp-TiO2 or mp-Al2O3
films deposited on a glass slide. For the mp-TiO2 on glass, Results and Discussion
similar solutions and conditions were chosen as for the solar
cell preparation. For the formation of the mp-Al2O3 film, Al2O3 So far, bismuth halide-based double perovskite crystals have
nanoparticles dispersed in isopropanol (Sigma Aldrich, <50 nm, only been synthesized by solid state reactions or solution
20%wt) were diluted with isopropanol (Sigma Aldrich, 99.8%) growth from the corresponding halide acids.19,26–29 The
in 1 : 1 ratio. 100 µL of the Al2O3 dispersion was spun on a glass solubility of the double perovskites in these acids is relatively
substrate at 2500 rpm for 30 s with subsequent annealing at low leading to double perovskite concentrations (< 0.1 M)
120 °C for 10 min. Steady-state absorption spectra were much lower than that generally used for the preparation of
acquired with a Lambda 1050 UV-Vis spectrophotometer lead or tin perovskite based films by solution based methods,4–
(Perkin Elmer) using an integrating sphere. Steady state and 6,13,14
so it is essential to find alternative solvents that can
time resolved PL measurements were conducted with a reach a higher precursor concentration in order to prepare
Fluotime 300 Spectrofluorometer (Picoquant GmbH). The Cs2AgBiBr6 films from solution. Since Cs2AgBiBr6 is hardly
excitation wavelength was fixed at 405 nm. The emission for soluble in the most commonly used solvents, we screened
time resolved measurements was monitored at 630 nm being several alternatives where dimethylsulfoxide (DMSO) showed
the wavelength of the maximum intensity of the steady state

Please
View Article Online
DOI: 10.1039/C7TA06816F
the highest ability to dissolve the precursors, AgBr, CsBr and desired double perovskite phase. Fig. 2a & b depict PXRD
BiBr3, and the double perovskite (Table S1) patterns of Cs2AgBiBr6 powders obtained from Cs2AgBiBr6 films
As a result, the Cs2AgBiBr6 films prepared in this study were at different annealing temperatures showing the typical
26
deposited by spin-coating a DMSO-based precursor solution on reflections of the cubic elpasolite structure. The patterns

top of a substrate. Our utilized synthesis route employs two from samples annealed at temperatures below 250 °C feature
different heating steps. The first step is named “preheating additional reflections at 12.8°, 30.9° and 44.2°, which can be
step” where the substrate and precursor solution are heated assigned to Cs3Bi2Br9 and AgBr, while films annealed at 250 °C
to 75 °C prior the spin-coating. Then, the hot precursor appear to be phase-pure. This trend can also be confirmed
solution was spun on top of the hot substrate. This preheating with the corresponding PL spectra were only the 250 °C
step improves the surface coverage and film quality (Fig. S1). sample exhibits the typical PL signal of Cs2AgBiBr6 (Fig. S2). The
The second heating step named “annealing step”, is performed PXRD patterns in Fig. 2, clearly show that the vanishing of the
after the spin-coating procedure at a temperature of at least side-phases is not caused by the removal of metal-solvent
250 °C and is needed to obtain phase pure films. A scheme of intermediates, as it was shown for lead and tin based hybrid
the complete synthesis procedure is given in Fig. 1. Previous perovskites.32,33 This is also supported by the evolution of the
studies by Xiao et al. propose the facile formation of side intensity of the reflections of the side-phases with
phases due to the narrow chemical potential region for the temperature. Even at 150 °C annealing temperature, which is
formation of phase-pure Cs2AgBiBr6, which could have usually sufficient to remove the metal-solvent intermediates,
negative effects on the optoelectronic properties of the double no decrease of the reflection intensities could be observed
perovskite.31 Therefore, we performed temperature- pointing to highly stable side phases. However, McClure et al.
dependent PXRD investigations to assure the formation of demonstrated the synthesis of Cs2AgBiBr6 via solid state
phase-pure films. We found that two of the predicted side reaction at 210 °C for 10 h,19 which is close to the temperature
phases, namely Cs3Bi2Br9 and AgBr,31 actually formed during we use to obtain phase pure double perovskite films.
the film synthesis. In order to remove these side phases we Therefore, we hypothesize that micro scale solid state
performed an extensive annealing temperature study, reactions in our prepared films are responsible for the
revealing that an annealing temperature of at least 250 °C is complete conversion of the precursors and side-phases into
needed to assure complete conversion of the precursors to the the desired Cs2AgBiBr6 phase.
Fig. 2. (a, b) XRD patterns of Cs2AgBiBr6 powders obtained from our prepared films annealed at different temperatures, the asterisk (*) indicates the
position of reflections from the side phases AgBr (ICDD No. 00-006-0438) and Cs3Bi2Br9 (ICDD No. 01-070-0493), respectively. The sample labelled with RT
was kept at room temperature (ca. 25 °C) after film formation without any annealing step.

Page 5 of 11 Please
Materials margins A
View Article Online
DOI: 10.1039/C7TA06816F

Fig. 3 (a) Optical absorption spectra of Cs2AgBiBr6 films prepared from DMSO on flat glass substrates. The films were manufactured without a preheating
step at room temperature (ca. 25 °C) or with a preheating step at 65 °C, 75 °C, 100 °C or 125 °C before spin-coating. (b) Energy level diagram of the double
perovskite and the charge extraction materials.26 (c) TCSPC decays of a Cs2AgBiBr6 film on glass and polycrystalline powder. (d) PL emission spectra of
Cs2AgBiBr6 films on mp-Al2O3, mp-TiO2 and mp-Al2O3 covered with spiro-OMeTAD. The films presented in 3d were prepared on an 800 nm thick mp scaffold.
All films were prepared according the procedure depicted in Fig. 1 with a 285 °C annealing step.
As the Cs2AgBiBr6 films were prepared for photovoltaic increased surface coverage on the substrate due to a larger
applications, sufficient light absorption is mandatory for amount of deposited double perovskite caused by the faster
efficiently working devices. In order to optimize the optical solvent evaporation at elevated temperatures. Increasing the
absorption properties of the double perovskite films, we amount of deposited Cs2AgBiBr6 by simply increasing the
added an additional preheating step to our synthesis protocol concentration of the precursors in the spin-coating solution is
where the spin-coating solution and the substrate were put on not possible since the precursor concentration of the utilized
a hotplate and heated to 75 °C before spin-coating. Fig. 3a solution is already close to the maximum of about 0.6 M.
shows absorption spectra of Cs2AgBiBr6 films on glass prepared The band gap of the Cs2AgBiBr6 films was determined from the
with and without a preheating step at different temperatures. absorption spectra given in Fig. 3a. The spectra show a steep
Without the preheating step, insufficient light absorption onset at about 550 nm corresponding to an Eg of 2.21 eV for a
would limit the photocurrent of the resulting devices to direct band gap obtained with a Tauc plot (see inset), which is
-2 19
approximately 2 mA cm , which is inadequate to prepare comparable to the value reported by McClure et al. and
26
efficient solar cells. We found that preheating the substrate consistent with other reports ranging from 1.83 eV to 2.19
19
and the solution increases the quality of the double perovskite eV . The relatively large spread of the reported Eg values most
films, by improving the surface coverage, as can be seen with likely originates from different synthesis conditions and
29
the naked eye, enhancing the optical absorption of the films measurement methods used in the different studies.
(Fig. S1). In particular, the absorption close to the onset is According to the Shockley-Queisser limit, a theoretical
significantly increased, which is also observable for films made maximum PCE of 16.4% can be obtained with an Eg of 2.2 eV,
on an mp-TiO2 scaffold as used for solar cell fabrication (Fig. thus holding promise for working double perovskite based
34
S3). We attribute the increase of the optical absorption to an photvoltaics. Although, the theoretical maximum PCE of this

Please
View Article Online
DOI: 10.1039/C7TA06816F
material does not reach the highest reported PCEs of single spiro-OMeTAD were chosen as ETM and HTM, respectively,
junction methylammonium lead iodide (MAPbI3) based while mp-Al2O3 was used as reference as no charge transfer is
photovoltaics, Cs2AgBiBr6 is a promising candidate for expected (for experimental details see the characterization
applications in tandem solar cells as it was already shown for section). Fig. 3d shows the quenching of the PL emission of

methylammonium lead bromide, which features a slightly Cs2AgBiBr6 on TiO2 by approximately 60%, while the PL was
35
larger Eg of 2.3 eV. reduced by more than 80% in contact with spiro-OMeTAD.
26
The energy level diagram in Fig. 3b shows that the These results indicate that charge transfer from the double
conduction band (CB) of the Cs2AgBiBr6 is well aligned with the perovskite to the charge-selective materials is taking place, as
CB of the electron transporting material (ETM) titania, which required for working photovoltaic devices. However, the low
was also recently shown in a computational study of Feng et PL quantum yield of Cs2AgBiBr6 indicates that the main
36 26
al. In contrast, the highest occupied molecular orbital recombination pathways in Cs2AgBiBr6 are nonradiative and
(HOMO) of the hole transporting material (HTM) 2,2',7,7'- therefore not detectable in PL quenching experiments. It has
tetrakis-(N,N-di-4-methoxyphenylamino)-9,9'-spirobifluorene to be mentioned that, the PL signal of our films is relatively
(spiro-OMeTAD) shows a strong offset compared to the broad but similar to that reported for single crystals.26 The
valence band (VB) of the double perovskite, which is likely to broad PL signal is most likely caused by defects and excitonic
limit the potential Voc of a photovoltaic device. In this work, as effects.29 Furthermore, previous studies by Slavney et al.
a first demonstration of the potential of Cs2AgBiBr6, we revealed that in addition to the direct band gap at about 2.2
focused on the preparation of devices comprising this state-of- eV, Cs2AgBiBr6 features an indirect band gap at about 1.9 eV,
the-art material which has been shown to be able to reach which was not detectable in the light absorption spectra of our
Voc’s exceeding 1.5V.37 Nevertheless, the energy level prepared films.26 Considering that the value of the indirect
alignment allows for charge extraction, and functioning
devices are expected.
Not only the band gap and energy alignment with the charge
extraction layers, but also charge collection efficiency is
important to fabricate high-efficiency solar cells. PL lifetime
measurements can provide an indication as to whether a
material can perform in this area. Here, we investigated the PL
decay times of the double perovskite films by time-correlated
single photon counting (TCSPC). For comparison, additional
TCSPC measurements were performed on polycrystalline
powder samples that were obtained by the conventional
approach from HBr solution.26 The PL decays are displayed in
Fig. 3c. Although the PL of Cs2AgBiBr6 features three differently
fast decay processes, only the value for the long-lived process
is given in Fig. 3c. We attribute the two short lifetime
processes to trap and/or surface state emission, while the long
lifetime process is suggested to be the fundamental PL decay
time of the material.26 Both the powder and the film show long
PL decay times in the range of hundreds of nanoseconds,
similar to the values observed for lead-based hybrid
perovskites,38 hinting at the potential for well-performing
devices. We note that the decay times do not differ
significantly between single crystals and polycrystalline
powders, which suggest a high defect tolerance of this
material.26 The slightly faster decay for films is likely the result
of a larger concentration of trap states in the perovskite film,
as has been previously shown for lead halide-based films.39
Here, the fast formation of the perovskite during the spin-
coating process causes a broad grain size distribution within
the film and, consequently, a large number of grain
boundaries.
PL quenching experiments are widely used to study the charge
transfer from a photoabsorber to the charge-selective
,40,41
contacts From the energy diagram (Fig. 3b) we expect
Fig. 4 (a) SEM top-view image of a Cs2AgBiBr6 film prepared with a 75 °C
favorable charge transfer behavior from Cs2AgBiBr6 to the preheating step on mp-TiO2. (b) SEM cross-section image of the film
electrodes, which was experimentally tested with PL presented in (a). (c-g) EDX elemental maps of Ti (red), Cs (turquoise), Ag
quenching experiments. For the measurements, mp-TiO2 and (purple), Bi (blue) and Br (yellow).

Page 7 of 11 Please
Materials margins A
View Article Online
DOI: 10.1039/C7TA06816F
band gap matches quite closely the position of the observed PL cross-section of the film presented in Fig. 4a and the elemental
maximum, we believe that the relatively large red shift of the maps of Ti, Cs, Ag, Bi and Br, respectively. The cross-section
maximum of the PL signal with respect to the absorption onset image and EDX maps clearly establish complete infiltration of
can be assigned to contributions of this indirect band gap at Cs2AgBiBr6 into the mp-TiO2 scaffold, which is important to

about 1.9 eV. Nevertheless, the PL results indicate that obtain efficient devices, while some additional Cs2AgBiBr6
Cs2AgBiBr6 films can be incorporated in working photovoltaic agglomerates are formed on top of the scaffold.
devices. Based on the promising results of the above investigations, we
The morphology of the double perovskite films prepared incorporated the newly developed Cs2AgBiBr6 films into
according the procedure shown in Fig. 1 was investigated with photovoltaic devices featuring a layer assembly of fluorine
scanning electron microscopy (SEM). Fig. 4a shows a SEM top- doped tin oxide (FTO)/dense TiO2/mp-TiO2/Cs2AgBiBr6/spiro-
view image of a Cs2AgBiBr6 film, on an 800 nm thick mp-TiO2 OMeTAD/Au to investigate their photovoltaic characteristics.
scaffold as used for the preparation of photovoltaic devices. As mentioned before, the thickness of the utilized mp-TiO2
Here, no dense and homogeneous capping layer on top of the layer was approx. 800 nm as depicted in Fig. 4b.
mp-TiO2 is observable, instead many agglomerates at the scale Fig. 5a displays the statistical parameter distribution of 48
of about one micron have formed. This is most likely caused by solar cells (per treatment temperature) that were assembled
the fast crystallization process during the spin-coating with Cs2AgBiBr6 films prepared without a preheating step and
procedure induced by fast solvent evaporation due to the with a preheating step at different temperatures. The studies
performed preheating step, as shown in Fig. 1. We believe that revealed a PCE maximum at 75 °C, we attribute to the
this fast crystallization process impedes homogeneous crystal improved optical properties of those films shown in Fig. 3a and
growth, leading to the observed agglomerates. Energy S3. Furthermore, the J-V curves given in Fig. S4 show that the
dispersive X-ray spectroscopy (EDX) measurements were increased device performance at 75 °C preheating is not only
performed to evaluate the penetration depth of the double caused by an increased short-circuit current (Jsc) due to the
perovskite into the mp-TiO2 scaffold. Fig. 4b-g show a SEM enhanced optical absorption properties because of the
Fig. 5 (a) Device performance as a function of preheating temperature, values obtained from 48 individual devices per temperature. All films were annealed
at 250 °C after spin-coating to achieve phase pure Cs2AgBiBr6 films. (b) Device performance of solar cells assembled with Cs2AgBiBr6 films preheated at 75 °C
as a function of annealing temperature, values obtained from 48 individual devices per temperature. (c) J-V curve of the best performing device (d) EQE
spectrum (black) and integrated predicted current (red) of a device showing a Jsc of 3.7 mA cm-².

Please
View Article Online
DOI: 10.1039/C7TA06816F
preheating, but also by the enhanced fill factor (FF) and by the J-V curve of the best performing device exhibiting a PCE of
Voc increase observed with increasing preheating temperature. 2.43%. In particular, the Voc of 0.98 V is significantly higher
We hypothesize that the preheating also improves the mp- than that of other bismuth-based compounds featuring
43-45
TiO2/double perovskite interface, for e.g. by reducing trap comparable Eg´s. Moreover, Voc´s exceeding 1 V were
42

states, resulting in an enhanced performance. The large measured, which are, with the best of our knowledge, the
spread of the device performance observable for films highest Voc values reported for bismuth halide based materials
preheated at 25 °C and 65 °C indicates that the improvement (see Fig. S7), confirming the good alignment of the conduction
of the TiO2/double perovskite interface is not as reliable as for band of the perovskite with TiO2 as predicted from the energy
36
preheating at 75 °C, which results in the smallest performance level diagram in Fig. 3b and in computational studies.
spread of all investigated preheating temperatures. The PCE These PCEs are viewed as an excellent starting point for double
decrease and the broadening of the PCE distribution at higher perovskite absorbers in solar cells if one considers that the
preheating temperatures (>100 °C) is tentatively attributed to spectroscopically predicted maximum efficiency of Cs2AgBiBr6
the formation of a top layer consisting of large crystallites on is reaching 7.92% determined from density functional theory
46
the mp-TiO2 scaffold, caused by a strongly accelerated calculations by Savory et al.. As the band gap of Cs2AgBiBr6 is
precipitation of the double perovskite due to the relatively 2.2 eV, the current of solar cells comprising this absorber is
-2
high deposition temperature, as shown in Fig. S5, which is limited to ∼6.8 mA cm under 1 sun illumination. External
apparently too thick for efficient charge transport. We note quantum efficiency (EQE) measurements show an integrated
-2
that Cs2AgBiBr6 films prepared without a preheating step current of 3.6 mA cm (Fig. 5d), which is in good agreement
-2
feature the most homogeneous morphology, with the lowest with the Jsc of 3.7 mA cm obtained from the J-V curve (Fig.
number of agglomerates on top the mp-TiO2 scaffold (see Fig S8). The EQE spectrum (Fig. 5d) shows an onset around 550
S6), without resulting in the best performing devices. nm, which is in agreement with the absorbance data, and is
Accordingly, it is likely that preheating to 75 °C results in above 50% from 375 to 470 nm with a maximum of 60% at 415
double perovskite films featuring strongly improved nm. According to Fig. 5c the J-V curves show hysteresis,
optoelectronic properties, which overcome the possible comparable to behavior observed in devices employing lead-
47–50
detrimental effects of a less homogeneous morphology due to based perovskites. Computational studies by Eames et al.
the agglomerate formation on top of the mp-TiO2 induced by reveal that the migration of the halide anions contributes to
the preheating. the hysteresis in hybrid lead halide perovskite based
51
Although a full conversion of the precursors into the double photovoltaics due to their very low activation barrier.
perovskite phase is achieved by 250 °C annealing after film Therefore, we believe that similar effects, including
47, 52
formation, the obtained PCEs showed a large spread. trapping/de-trapping of charge carriers, are likely to be
Therefore, an extensive study of the influence of the annealing responsible for the hysteresis in our Cs2AgBiBr6 based devices
temperature on device performance was conducted. Fig. 5b since the crystal structure of the double perovskite is very
shows the relation between annealing temperature and the similar.
statistical distribution of PCEs of 48 individual devices. The Besides performance, stability is also an important issue for
highest and most reproducible PCEs were obtained for an widespread applications as devices need to work for many
annealing temperature of 285 °C. The increase in PCE with years at high intensity illumination. Accordingly, we conducted
increasing annealing temperature is dominated by an increase the first stability studies on Cs2AgBiBr6-based solar cells. Fig. 6a
of Jsc and FF at annealing temperatures below 250 °C, whereas displays the stabilized power output under ambient
only the Jsc rises significantly at higher annealing temperatures conditions. The device shows a rapid response after
(Fig. S7). Up to 250 °C, this trend is most likely caused by a illumination, resulting in a PCE of around 2.0%. During the next
decrease of the side phases observed in the PXRD pattern (Fig. 30 s, only a very small decay of around 5% in the PCE was
2a & b), until the phase-pure Cs2AgBiBr6 has formed at 250 °C. observed, after which the PCE stabilizes just above 1.9% and
This temperature is the lower end of the range of the no further decay was observed over a period of five minutes,
annealing temperature required to fully convert the precursors indicating a similar performance stability like lead halide based
into the double perovskite phase. Annealing at slightly higher solar cells. Another device was illuminated using an AM1.5
temperatures (285 °C) make this process more robust, leading solar simulator under ambient conditions for a period of 100
to slightly higher PCEs and more importantly, better min while at regular intervals J-V scans were taken (Fig 6b &
reproducibility. The decrease in PCE at even higher S9). During the first couple of minutes an increase in the PCE
temperature (300 °C) is associated with a reduction in Jsc, was observed as a result of an increase in the Voc, which we
53
which we attribute to the degradation of the Cs2AgBiBr6 films assign to light soaking effects. This indicates the presence of
as the double perovskite slowly starts to come off the (interfacial) defects that can be neutralized with
53–59
substrate at annealing temperatures at about 300 °C (Fig. S7). photogenerated charge carriers. After this period the Voc
The application of our optimized synthesis conditions resulted and Jsc turn out to be very stable. Some fluctuation in the FF
in devices showing PCEs of more than two percent. (Fig. S9) are observed, which are also reflected in the PCE,
Remarkable is the relatively narrow distribution in the nevertheless after 100 min the PCE did not significantly decay
performance, taking into consideration the surface of the film and at 2.06%, it was even higher than the initial value
that contains the perovskite agglomerates. Fig. 5c displays the measured before the light soaking. The results presented in

Page 9 of 11 Please
Materials margins A
View Article Online
DOI: 10.1039/C7TA06816F
Figure 6b & S9 indicate a high device stability under powders, and solar cells comprising these new films exhibit
permanent illumination, higher than that of reported MAPbI3 very promising PCEs close to 2.5% with Voc´s exceeding one
based devices showing a significant drop of Jsc within the first volt, demonstrating the suitability of double perovskites for
60
two hours. Furthermore, initial longer-term stability studies optoelectronic devices. Furthermore, our Cs2AgBiBr6-based

on the new Cs2AgBiBr6-based devices revealed an excellent devices feature a high stability under constant illumination at
stability under ambient conditions for at least 25 days (Fig. ambient conditions, higher than that of reported MAPbI3
S10). The devices were stored in the dark to avoid any possible based solar cells. Since the elpasolite structure allows for
light driven degradation processes of the double perovskite as numerous different combinations of elements, this work
26
indicated by Slavney et al.. shows the potential for developing a wide range of double
perovskites for diverse environmentally friendly optoelectronic
applications.
Conflict of interests
There are no conflicts to declare.
Acknowledgements
We thank Julian Rotter from the Chemistry Department of the
University of Munich (LMU) for the scanning electron
microscopy and energy dispersive x-ray scattering
measurements. We acknowledge funding by the Bavarian
State Ministry of the Environment and Consumer Protection,
the Bavarian network “Solar Technologies Go Hybrid”, the DFG
Excellence Cluster Nanosystems Initiative Munich (NIM), the
German Federal Ministry of Education and Research (BMBF)
under the Project ID 03SF0516B and the EPSRC
(EP/P03148X/1). We gratefully acknowledge support from the
European Union through the award of a Marie Curie Intra-
European Fellowship.
Notes and references

1 M. M. Lee, J. Teuscher, T. Miyasaka, T. N. Murakami and H. J.
Snaith, Science, 2012, 338, 643–647.
2 NREL chart,
http://www.nrel.gov/ncpv/images/efficiency_chart.jpg,
(accessed July 2017).
3 T. M. Brenner, D. A. Egger, L. Kronik, G. Hodes and D. Cahen,
Fig. 6 (a) Stabilized power output and current density measured under Nat. Rev. Mater., 2016, 1, 15007.
ambient conditions without encapsulation. (b) Photovoltaic performance as 4 S. De Wolf, J. Holovsky, S.-J. Moon, P. SLoiper, B. Niesen, M.
a function of time under continuous illumination under ambient conditions. Ledinsky, F.-J. Haug, Yum and C. Ballif J. Phys. Chem. Lett.,
All devices were manufactured according the procedure described in Fig. 1 2014, 5, 1035−1039.
with a 285 °C annealing step. 5 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, 2013, 342, 341−344.
6 C. Wehrenfennig, G. E. Eperon, M. B. Johnston, H. J. Snaith
Conclusions and L- M. Herz, Adv. Mater., 2014, 26, 1584−1589.
7 R. E. Brandt, V. Stevanović, D. S. Ginley and T. Buonassisi,
In conclusion, we have demonstrated the solution-based MRS Commun., 2015, 5, 265−275.
formation of double perovskite films via a fast and efficient 8 A. Walsh, D. O. Scanlon, S. Chen, X. G. Gong and S.-H. Wei,
spin-coating method. Extensive studies of the synthesis Angew. Chem., Int. Ed., 2015, 54, 1791−1794.
9 Y.-Y Zhang, S. Chen, P. Xu, H. Xiang, X.-G. Gong, A. Walsh and
conditions revealed the necessity of a high-temperature S.-H. Wei, arXiv:1506.01301, 2015.
annealing step to fully convert the precursors into the desired 10 G. P. Nagabhushana, R. Shivaramaiah and A. Navrotsky, Proc.
double perovskite phase, and the requirement of an additional Natl. Acad. Sci. U. S. A., 2016, 113, 7717−7721.
preheating step to improve the optical properties of the 11 N. Aristidou, I. Sanchez-Molina, T. Chotchuangchutchaval, M.
Cs2AgBiBr6 films. The resulting Cs2AgBiBr6 films feature an Eg Brown, L. Martinez, T. Rath and S. A. Haque, Angew. Chem.,
Int. Ed., 2015, 54, 8208−8212.
comparable to that of single crystals and polycrystalline

Please
View Article Online
DOI: 10.1039/C7TA06816F
12 Babayigit, A.; Ethirajan, A.; Muller, M.; Conings, B. Nat. 40 D. S. Ginger and N. C. Greenham, Phys. Rev. B, 1999, 59,
Mater., 2016, 15, 247−251. 10622–10629.
13 F. Hao, C. C. Stoumpos, D. H. Cao, R. P. H. Chang and M. G. 41 P. Docampo, J. M. Ball, M. Darwich, G. E. Eperon and H. J.
Kanatzidis, Nat. Photonics, 2014, 8, 489–494. Snaith, Nat. commun., 2013, 4, 2761.
14 N. K. Noel, S. D. Stranks, A. Abate, C. Wehrenfennig, S. 42 M. Abdi-Jalebi, M. I. Dar, A. Sadhanala, S. P. Senanayak, F.

Guarnera, A. A. Haghighirad, A. Sadhanala, G. E. Eperon, S. K. Giordano, S. M. Zakeeruddin, M. Grätzel and R. H. Friend, J.
Pathak, M. B. Johnston, A. Petrozza, L. M. Herz and H. J. Phys. Chem. Lett., 2016, 7, 3264–3269.
Snaith, Energy Environ. Sci., 2014, 7, 3061–3068. 43 Y. Kim, Z. Yang, A. Jain, O. Voznyy, G.-H. Kim, M. Liu, L. N.
15 R. Jakubas, J. Zaleski and L. Sobczyk, Ferroelectrics, 1990, Quan, F. p. García de Arquer, R. Comin, J. Z. Fan and E. H.
108, 109−114. Sargent, Angew. Chem., Int. Ed., 2016, 55, 9586–9590.
16 S. Sun, S. Tominaka, J-H. Lee, F. Xie, P. D. Bristowe, A. K. 44 B.-W. Park, B. Philippe, X. Zhang, H. Rensmo, G. Boschloo and
Cheetham, APL Mater., 2016, 4, 031101. E. M. Johansson, J. Adv. Mater., 2015, 27, 6806–6813.
17 B. Chabot and E. Parthé, Acta Cryst., 1978, B34, 645–648. 45 I. Turkevych, S. Kazaoui, E. Ito, T. Urano, K. Yamada, H.
18 A. J. Lehner, D. H. Fabini, H. A. Evans, C. A. Hébert, S. R. Tomiyasu, H. Yamagishi, M. Kondo and S. Aramaki,

Smock, J. Hu, H. Wang, J. W. Zwanziger, M. I. Chabinyc and R. ChemSusChem, 2017, 10, 1–7.
Seshadri, R. Chem. Mater., 2015, 27, 7137–7148. 46 C. N. Savory, A. Walsh and D. O. Scanlon, ACS Energy Lett.,
19 E. T. McClure, M. R. Ball, W. Windl and P. M. Woodward, 2016, 1, 949–955.
Chem. Mater., 2016, 28, 1348−1354. 47 H.J. Snaith, A. Abate, J. M. Ball, G. E. Eperon, T. Leijtens, N. K.
20 L. R. Morrs and W. R. Robinson, Acta Crystallogr., Sect. B: Noel, S. D. Stranks, J. T. W. Wang, K. Wojciechowski and W.
Struct. Crystallogr. Cryst. Chem., 1972, 28, 653−654. Zhang, J. Phys. Chem. Lett., 2014, 5, 1511–1515.
21 F. Pelle, B. Jacquier, J. Denis and B. J. Blanzat, Lumin., 1978, 48 E. L. Unger, E. T. Hoke, C. D. Bailie, W. H. Nguyen, A. R.
17, 61−72. Bowring, T. Heumüller, M. G. Christoforo and M. D.
22 W. Smit, G. Dirksen and D. Stufkens, J. Phys. Chem. Solids., McGehee, Energy Environ. Sci., 2014, 7, 3690–3698.
1990, 51, 189−196. 49 H. S. Kim and N. G. Park, J. Phys. Chem. Lett., 2014, 5, 2927–
23 Z. Deng, F. Wei, S. Sun, G. Kieslich, A. K. Cheetham and P. D. 2937.
Bristowe, J. Mater. Chem., A 2016, 4, 12025−12029. 50 R. S. Sanchez, V. Gonzalez-Pedro, J. W. Lee, N. G. Park, Y. S.
24 A. Tressaud, S. Khaïroun, J. P. Chaminade and M. Couzi, Phys. Kang, I. Mora-Sero and J. Bisquert, J. Phys. Chem. Lett., 2014,
Status Solidi A 1986, 98, 417–421. 5, 2357–2363.
25 F. Wei, Z. Deng, S. Sun, F. Xie, G. Kieslich, D.M. Evans, M. A. 51 C. Eames, J. M. Frost, P. R. F. Barnes, B. C. O´Regan, A. Walsh
Carpenter, P. D. Bristowe and A. K. Cheetham, Mater. Horiz., and M. S. Islam, Nat. Commun., 2015, 7497.
2016, 3, 328−332. 52 A. Dualeh, T. Moehl, N. Tétreault, J. Teuscher, P. Gao, M. K.
26 A. H. Slavney, T. Hu, A. M. Lindenberg and H. I. Karunadasa, J. Nazeeruddin and M. Grätzel, ACS Nano, 2013, 8, 362–373.
Am. Chem. Soc., 2016, 138, 2138−2141. 53 C. Zhao, B. Chen, X. Qiao, L. Luan, K. Lu and B. Hu, Adv.
27 G. Volonakis, M. R. Filip, A. A. Haghighirad, N. Sakai, B. Energy Mater., 2015, 5, 1500279.
Wenger, H. J. Snaith, and F. Giustino, J. Phys. Chem. Lett., 54 P. Tiwana, P. Docampo, M. B. Johnston, L. M. Herz and H. J.
2016, 7, 1254−1259. Snaith, Energy Environ. Sci., 2012, 5, 9566–9573.
28 F. Wei, Z. Deng, S. Sun, F. Zhang, D. M. Evans, K. Kieslich, S. 55 L. Cabau, L. Pellejà, J. N. Clifford, C. V. Kumar and E.
Tominaka, M. A. Carpenter, J. Zhang, P. D. Bristowe and A. K. Palomares, J. Mater. Chem. A, 2013, 1(31), 8994–9000.
Cheetham, Chem. Mat., 2017, 29, 1089–1094. 56 T. Kobayashi, H. Yamaguchi and T. Nakada, Prog. Photovolt:
29 M. R. Filip, S. Hillman, A. A. Haghighirad, H. J. Snaith and F. Res. Appl., 2014, 22, 115–121.
Giustino, J.Phys.Chem.Lett., 2016, 7, 2579−2585. 57 L. Yang, B. Xu, D. Bi, H. Tian, G. Boschloo, L. Sun and E. M.
30 M. R. Filip, C. Verdi and F. Giustino, J. Phys. Chem. C, 2015, Johansson, J. Am. Chem. Soc., 2013, 135, 7378–7385.
119, 25209–25219. 58 J. Kim, G. Kim, Y. Choi, J. Lee, S. Heum Park and K. Lee, Appl.
31 Z. Xiao, W. Meng, J. Wang and Y. Yan, ChemSusChem, 2016, Phys., 2012, 111, 114511.
9, 2628–2633. 59 S. Trost, K. Zilberberg, A. Behrendt, A. Polywka, P. Görrn, P.
32 N. J. Jeon, J. H. Noh, Y. C. Kim, W. S. Yang, S. Ryuand and S. I. Reckers, J. Maibach, T. Mayer and T. Riedl, Adv. Energy
Seok, Nat. Mater., 2014, 13, 897–903. Mater., 2013, 3, 1437–1444.
33 F. Hao, C. C. Stoumpos, P. Guo, N. Zhou, T. J. Marks, R. P. H. 60 T. Leijtens, G. E. Eperon, S. Pathak, A. Abate, M. M. Lee and
Chang and M. G. Kanatzidis, J. Am. Chem. Soc., 2015, 137, H. J. Snaith Nat. Commun., 2013, 4, 2885.
11445−11452
34 S. Rühle, Phys. Status Solidi A, 2017, 1600955.
35 R. Sheng, A. W. Y. Ho-Baillie, S. Huang, M. Keevers, X. Hao, L.
Jiang, Y.-B Cheng, and M. A. Green, J. Phys. Chem. Lett.,
2015, 6, 3931–3934.
36 H.-J. Feng, W. Deng, K. Yang, J. Huang and X. C. Zeng, J. Phys.
Chem. C, 2017, 121, 4471–4480.
37 S. Gholipour, A. M. Ali, J.-P. Correa-Baena, S.-H. Turren-Cruz,
F. Tajabadi, W. Tress, N. Taghavinia, M. Grätzel, A. Abate, F.
De Angelis, C. A. Gaggioli, E. Mosconi, A. Hagfeldt and M.
Saliba, Adv. Mater., 2017, 1702005.
38 D. Shi, V. Adinolfi, R. Comin, M. Yuan, E. Alarousu, A. Buin, Y.
Chen, S. Hoogland, A. Rothenberger, K. Katsiev, Y. Losovyj, X.
Zhang, P. A. Dowben O. F. Mohammed, E. H. Sargent and O.
M. Bakr, Science, 2015, 347, 519–522.
39 D. W. deQuilettes, S. M. Vorpahl, S. D. Stranks, H. Nagaoka,
G. E. Eperon, M. E. Ziffer, H. J. Snaith, D. S. Ginger, Science,
2015, aaa5333.

Page 11 of 11 Journal of Materials Chemistry A
View Article Online
DOI: 10.1039/C7TA06816F
Cs2AgBiBr6 double perovskite thin films were prepared and incorporated into photovoltaic devices
featuring power conversion efficiencies close to 2.5%.


Materials Chemistry A: Journal of

Uploaded by

Copyright:

Available Formats

Materials Chemistry A: Journal of

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Materials Chemistry A: Journal of

Uploaded by

Copyright:

Available Formats

View Article Online

acceptance, before technical editing, formatting and proof reading.

You can find more information about Accepted Manuscripts in the

Please note that technical editing may introduce minor changes

Journal of Materials Chemistry A Accepted Manuscript

Highly stable, phase pure Cs2AgBiBr6 double perovskite thin films

for optoelectronic applications

bismuth ions cannot simply be incorporated into the hybrid

the most promising absorber materials for low-cost 3+

substitution of Pb2+ by heterovalent M3+ cations. A promising 3+

Please do not adjust margins

Journal of Materials Chemistry A Accepted Manuscript

All chemicals were used as received without any further

Please do not adjust margins

Journal of Materials Chemistry A Accepted Manuscript

Please do not adjust margins

Journal of Materials Chemistry A Accepted Manuscript

Please do not adjust margins

Journal of Materials Chemistry A Accepted Manuscript

Please do not adjust margins

Journal of Materials Chemistry A Accepted Manuscript

Please do not adjust margins

Journal of Materials Chemistry A Accepted Manuscript

Please do not adjust margins

Journal of Materials Chemistry A Accepted Manuscript

Please do not adjust margins

Journal of Materials Chemistry A Accepted Manuscript

Notes and references

Please do not adjust margins

Journal of Materials Chemistry A Accepted Manuscript

18 A. J. Lehner, D. H. Fabini, H. A. Evans, C. A. Hébert, S. R. Tomiyasu, H. Yamagishi, M. Kondo and S. Aramaki,

Please do not adjust margins

Journal of Materials Chemistry A Accepted Manuscript

You might also like