Europium-Doped CsPbI2Br for Stable and Highly

Article

Europium-Doped CsPbI2Br for Stable and Highly Efficient Inorganic Perovskite Solar Cells Wanchun Xiang, Zaiwei Wang, Dominik J. Kubicki, ..., Giovanni Dietler, Michael Gra¨tzel, Anders Hagfeldt [email protected] (W.X.) [email protected] (A.H.)

HIGHLIGHTS We show the incorporation of europium into CsPbI2Br allinorganic perovskite A high PCE of 13.71% was achieved for europium-doped CsPbI2Br inorganic PSCs Eu-doped CsPbI2Br PSC demonstrates 370 hr thermal stability under illumination

We show the incorporation of europium into CsPbI2Br inorganic perovskite lattice. With the optimization of the doping concentration of europium, we obtained a high power-conversion efficiency of 13.71%. We found that incorporation of europium reduces non-radiative recombination to achieve a high open-circuit voltage of 1.27 V. The exceptional stability of such a device was demonstrated by retaining 93% of the initial efficiency under 100 mW cm 2 continuous illumination for 370 hr.

Xiang et al., Joule 3, 1–10 February 20, 2019 ª 2018 Elsevier Inc. https://doi.org/10.1016/j.joule.2018.10.008

Please cite this article in press as: Xiang et al., Europium-Doped CsPbI2Br for Stable and Highly Efficient Inorganic Perovskite Solar Cells, Joule (2018), https://doi.org/10.1016/j.joule.2018.10.008

Article

Europium-Doped CsPbI2Br for Stable and Highly Efficient Inorganic Perovskite Solar Cells Wanchun Xiang,1,2,7,8,* Zaiwei Wang,2,7 Dominik J. Kubicki,3,4 Wolfgang Tress,2 Jingshan Luo,3,5 Daniel Prochowicz,3 Seckin Akin,3 Lyndon Emsley,4 Jiangtao Zhou,6 Giovanni Dietler,6 Michael Gra¨tzel,3 and Anders Hagfeldt2,*

SUMMARY

Context & Scale

All-inorganic perovskite films hold promise for improving the stability of perovskite solar cells (PSCs). However, the 3D a phase of narrow-bandgap inorganic perovskites is thermodynamically unstable at room temperature, limiting the development of high-performance inorganic PSCs. Here, we show that europium doping of CsPbI2Br stabilizes the a phase of this inorganic perovskite at room temperature. We rationalize it by using solid-state nuclear magnetic resonance and high-angle annular dark-field scanning transmission electron microscopy, which show that europium is incorporated into the perovskite lattice. We demonstrate a maximum power-conversion efficiency of 13.71% for an inorganic PSC with the CsPb0.95Eu0.05I2Br perovskite and a stable power output of 13.34%. Using electroluminescence we show that incorporation of europium reduces non-radiative recombination, resulting in high open-circuit voltage of 1.27 V. The devices retain 93% of the initial efficiency after 370 hr under 100 mW cm2 continuous white light illumination under maximum-power point-tracking measurement.

The instability of the 3D a phase of narrow-bandgap inorganic perovskites such as CsPbI3 and CsPbI2Br limits the development of inorganic PSCs. We found that europium doping of the allinorganic CsPbI2Br perovskite results in stabilization of its black photoactive phase and significant improvement of its photovoltaic performance. Applying solidstate magic-angle spinning nuclear magnetic resonance, we show for the first time that europium is incorporated as B cation into the perovskite lattice on the atomic level, making it a promising modulator of the intrinsic material properties. Electroluminescence and timeresolved photoluminescence decay measurements show that incorporation of europium suppresses non-radiative chargecarrier recombination by eliminating tail states, which explains the resulting high opencircuit voltage of 1.27 V.

INTRODUCTION The past several years have witnessed rapid development of perovskites as excellent light absorbers for low-cost and solution-processable photovoltaics (PV).1–5 To date, the highest certified power-conversion efficiency (PCE) reported for a perovskite solar cell (PSC) has reached 23.3%,6 making it a viable competitor to silicon solar cells.7 However, organic components in these high-efficiency inorganic-organic hybrid PSCs, such as methylammonium (MA) and formamidinium (FA), are intrinsically unstable due to the volatility of the haloid salts with these organic cations, leading to long-term stability issues.8,9 The replacement of these volatile organic components by inorganic cations (Cs, Rb) in all-inorganic perovskite materials may fundamentally improve the stability of PSCs.10,11 However, the key problem associated with narrow-bandgap inorganic perovskites (such as CsPbI3 and CsPbI2Br) is their thermodynamic phase instability, whereby the desirable cubic a phase (‘‘black phase’’) of the narrow-bandgap material spontaneously transforms to the orthorhombic, more thermodynamically stable s phase (‘‘yellow phase’’) at room temperature.12,13 This undesired phase transition is currently one of the bottlenecks limiting the development of high-performance inorganic PSCs. The stabilization of all-inorganic perovskite phases can be realized by either tailoring the perovskite grain size (i.e., by creating quantum dots) due to the large contribution

Joule 3, 1–10, February 20, 2019 ª 2018 Elsevier Inc.

1


Table 1. Photovoltaic Parameters of Perovskite Solar Cells Fabricated Using CsPbI2Br Doped with Europium, Measured at 100 mW,cm2 Light Intensity Perovskite Composition

VOC (mV)

JSC (mA,cm2)

FF

PCE (%)

CsPbI2Br

1.120

12.57

0.725

10.21

CsPb0.99Eu0.01I2Br

1.160

14.16

0.753

12.37

CsPb0.97Eu0.03I2Br

1.208

14.64

0.745

13.18

CsPb0.95Eu0.05I2Br

1.223

14.63

0.766

13.71

CsPb0.93Eu0.07I2Br

1.100

14.59

0.776

12.45

CsPb0.91Eu0.09I2Br

1.070

13.95

0.735

10.97

of the surface energy,14,15 or by incorporation of other components into the perovskite material.16–19 In the latter case, three possible scenarios can be invoked to describe the resulting microstructure: (1) the dopant forms a separate phase and does not affect the atomic-level structure of the perovskite phase,20,21 (2) dopants modify the surface of the perovskite crystal,22,23 and (3) the dopants are incorporated into the crystal lattice replacing one of its substituents.24 The first scenario can lead to a complex mixture of phases, as is the case for addition of rubidium and potassium into organic-inorganic lead-halide perovskites.20 The second scenario might affect the growth rate to reduce the crystal size and passivate the surface,22 while the third can substantially increase the entropy of the perovskite lattice and lead to thermodynamic stabilization, thereby solving the key problem of all-inorganic perovskites.20,24,25 Europium has been reported as an efficient sensitizer of photoluminescence in perovskites and related materials.26 Compared with lead ion, the slightly smaller europium ion is expected to be incorporated into the perovskite lattice to increase the tolerance factor and result in an enhancement of the perovskite stability.27 Here we show that stoichiometric europium doping of the CsPbI2Br all-inorganic perovskite leads to thermodynamically stable inorganic PSCs with excellent photovoltaic metrics, a PCE of 13.71%, and a stable power output of 13.34%. We use solid-state nuclear magnetic resonance (NMR) and high-angle annular dark-field scanning transmission electron microscopy (HAADF STEM) to show that europium is incorporated into the perovskite lattice on the atomic level. Electroluminescence and timeresolved photoluminescence decay measurements show that incorporation of europium suppresses non-radiative charge-carrier recombination. The cells of Eu doping maintained 93% of their initial performance under full illumination and maximumpower point tracking.

1State

Key Laboratory of Silicate Materials for Architectures, Wuhan University of Technology, 430070 Wuhan, China

2Laboratory

of Photomolecular Science, Institute of Chemical Sciences Engineering, Ecole Polytechnique Fede´rale de Lausanne (EPFL), 1015 Lausanne, Switzerland

3Laboratory

of Photonics and Interfaces, Institute of Chemical Sciences and Engineering, Ecole Polytechnique Fede´rale de Lausanne (EPFL), 1015 Lausanne, Switzerland

4Laboratory

RESULTS AND DISCUSSION We used the one-step spin-coating technique to deposit perovskite films from CsPb1xEuxI2Br (0 % x % 1) precursor solutions. Spin-coated precursor films were then annealed at 280 C for 10 min to form perovskite films. We employed the deposited films to fabricate PSCs in the following architecture: glass/fluorine-doped tin oxide (FTO)/compact TiO2/mesoporousTiO2/perovskite/2,20 ,7,70 -tetrakis(N,N-dip-methoxyphenylamine)-9,90 -spirobifluorene (Spiro-OMeTAD)/Au. Figure S1 shows cross-sectional scanning electron microscopy (SEM) images of a full device with the estimated thickness of the perovskite layer of 150 nm. Figure S2 shows the J-V characteristics measured under 100 mW cm2 light intensity for varying europium content. The corresponding PV parameters are listed in Table 1. The europium-doped devices (x = 0.01, 0.03, 0.05, 0.07, and 0.09) show outstanding performance compared with those based on pure CsPbI2Br owing to the striking improvement

2

Joule 3, 1–10, February 20, 2019

of Magnetic Resonance, Institute of Chemical Sciences and Engineering, Ecole Polytechnique Fede´rale de Lausanne (EPFL), 1015 Lausanne, Switzerland

5Institute

of Photoelectronic Thin Film Devices and Technology, College of Electronic Information and Optical Engineering, Nankai University, 300350 Tianjin, China

6Laboratory

of Physics of Living Matter (LPMV), Institute of Physics, Ecole Polytechnique Fede´rale de Lausanne (EPFL), 1015 Lausanne, Switzerland

7These 8Lead

authors contributed equally

Contact

*Correspondence: [email protected] (W.X.), [email protected] (A.H.) https://doi.org/10.1016/j.joule.2018.10.008


Figure 1. Photovoltaic Performance of CsPb0.95Eu0.05I2Br-Based Device (A) J-V curve of the champion CsPb 0.95 Eu 0.05 I 2 Br device. (B) Stabilized power output at maximum-power point tracking under working condition with 100 mW cm 2 irradiation. (C) IPCE and integrated J SC of the champion CsPb 0.95 Eu 0.05 I 2 Br device.

in short-circuit current density (JSC), open-circuit voltage (VOC), and fill factor (FF). The best solar cells made with CsPb0.95Eu0.05I2Br show a champion PCE of up to 13.71% at a scan rate of 10 mV s1 (Figure 1A). The stabilized output power was recorded as 13.34% at full sun illumination with a stable JSC of 13.94 mA cm2 (Figure 1B). Incident photon-to-current conversion efficiency (IPCE) measurements carried out on this cell show over 80% efficiency in the range between 420 and 650 nm, leading to an integrated current density of 14.3 mA cm2, consistent with the JSC in the J-V measurement (Figure 1C). The corresponding IPCEs of PSCs for all studied europium concentrations are also shown in Figure S3. The cells fabricated using CsPb0.95Eu0.05I2Br show moderate hysteresis with a hysteresis factor of 0.06 (Figure S4). We performed X-ray diffraction (XRD) analysis of the thin films to evaluate the effect of europium doping on the crystal structure of CsPbI2Br (Figure 2A). Pristine CsPbI2Br shows characteristic perovskite peaks at 14.6 and 29.7 . The XRD peak intensities of perovskite films are reduced with the addition of europium, indicating a decrease in the crystallite size.18 To evaluate europium incorporation into the perovskite lattice, it is necessary to apply atomic-level methods. We have recently shown that solid-state NMR allows probing the atomic-level microstructure of lead-halide perovskites, both in bulk and as thin films.20,28–30 Here, using 133 Cs solid-state magic-angle spinning (MAS) NMR we show that europium is incorporated into the parent CsPbI2Br perovskite lattice. Figures 2B and S7 show solid-state 133Cs MAS NMR of the bulk mechanoperovskites studied in this work. The paramagnetic phase synthesized by grinding and annealing of an equimolar mixture of cesium iodide and europium iodide (see Supplemental Information for further details) exhibits a single peak at 171 ppm (full width at half maximum [FWHM] 1,155 Hz) and a very short spin-lattice relaxation time, T1, of about 10 ms, characteristic of paramagnetic compounds.31 In the diamagnetic compounds, CsPbBr3, CsPbI3, and CsPbI2Br, relaxation is 3–4 orders of magnitude slower (29.5, 95.8, and 44.6 s, respectively). The reference perovskite material of this study, CsPbI2Br, exhibits a relatively broad peak at 142 ppm (FWHM 5,759 Hz) whose breadth is caused by disorder owing to the presence of cesium environments with a varying number of iodide and bromide closest neighbors. This notion is clearly illustrated by comparing it with CsPbBr3 and CsPbI3, which give relatively narrow


3


Figure 2. Characterization of Europium-Doped CsPbI2Br (A) XRD patterns of perovskite thin films of CsPb 1x Eu x I 2 Br (0 % x % 0.09). Asterisks represents the FTO peaks. (B) 133 Cs solid-state MAS NMR spectra of mechanochemical perovskite compositions measured at 11.7 T, 20 kHz MAS, and 298 K. The full spectra are given in Figure S5. Dagger denotes a transmitter artifact and asterisk a spinning sideband. (C) 133 Cs spin-lattice relaxation times, T1, of selected perovskite compositions (details in the text). The solid curves are stretched exponential fits. The dashed line is the best biexponential fit of the CsPb 0.95 Eu 0.05 I 2 Br dataset. (D) HAADF STEM image and elemental maps of CsPb0.95 Eu 0.05 I 2 Br. Scale bars, 100 nm.

lines (FWHM 219 and 367 Hz, respectively), since in these compounds cesium only has one type of neighbor anions. As for CsPb0.95Eu0.05I2Br, the 133Cs resonance shifts very slightly (from 142 to 146 ppm), which can conceivably be caused either by (1) a paramagnetic shift resulting from the hyperfine interaction of europium f electrons with cesium, (2) phase separation of cesium europium mixed halides (Br-I),

4



which would lead to a change in the original iodide-to-bromide ratio in CsPbI2Br, similarly to what we and Hu et al. have previously observed in rubidium- and cesium-doped mixed-cation lead-halide perovskites,20,32 or (3) bulk magnetic susceptibility effects caused by the presence of a paramagnetic phase, leading to inhomogeneity of the magnetic field inside the sample. To establish which of these two scenarios applies, it suffices to analyze the spinlattice relaxation of the europium-doped material (Figure 2C). If the material was to undergo a phase separation, one would expect to observe a biexponential T1 build-up curve with a very quickly relaxing component corresponding to the purely paramagnetic cesium europium halide phase, and a slowly relaxing component corresponding to the unchanged diamagnetic perovskite (Figure 2C, dashed line). Experimentally, we observe a stretched exponential behavior (solid lines) with stretching factors b of 0.69, 0.44, and 0.38 for 3 mol%, 5 mol%, and 8 mol% stoichiometric Pb replacement in CsPbI2Br. This result indicates the presence of a distribution of cesium relaxation times in the material, corresponding to cesium sites at different distances from the paramagnetic europium ions (the PRE effect scales approximately as inverse cube of the distance).33 We observe analogous behavior for a thin-film CsPb0.95Eu0.05I2Br sample prepared using solution processing (Figure S8), with a stretching factor b of 0.53 indicating successful incorporation of europium into the perovskite lattice. A 133Cs spectrum acquired with a recycle delay of 1 ms emphasizes cesium environments in the close vicinity of the paramagnetic europium sites, spanning over 2,000 ppm, consistent with typical hyperfine shifts of europium.34 These results taken together confirm that europium is incorporated into the CsPbI2Br perovskite lattice on the atomic level both in samples prepared by mechanosynthesis as bulk powders and by solution processing as thin films. We applied HAADF STEM imaging and energy-dispersive X-ray spectroscopy (EDX) mapping to evaluate uniformity of europium distribution inside the perovskite grains (Figures 2D and S9). The elemental maps indicate that europium is uniformly distributed within the sample, presenting negligible difference in element distribution. The uniformity is the same for all the other elements. The plane-view SEM images of the inorganic perovskite thin films are shown in Figures 3A, 3B, and S10. The CsPbI2Br film (Figure 3A) exhibits a relatively uniform surface coverage with crystal domains and clear pinholes. Upon incorporation of 1 mol% europium (CsPb0.99Eu0.01I2Br), the size of perovskite cubic crystals is decreased and new pinholes are generated. The film quality is again improved for increasing europium doping ratio, and the CsPb0.95Eu0.05I2Br (Figure 3B) film is fully uniform. Consequently, uniform coverage may be one of the reasons behind the improved PV performance of europium-doped perovskites. On the other hand, the grain size decreases for increasing europium concentration, with the average particle size less than 50 nm in the CsPb0.95Eu0.05I2Br film. Higher concentration of europium (x = 0.07 and 0.09, Figure S10) does not lead to further improvement in the film morphology. We also applied atomic force microscopy (AFM) to evaluate the film surface roughness. AFM indicates that the surface becomes smoother at a europium concentration of 5 mol% (Figure S11), with root-mean-square values of the heights of irregularities obtained by the Rayleigh-Rice theory35 of 40.7 nm and 14.0 nm for CsPbI2Br and CsPb0.95Eu0.05I2Br, respectively (Table S3). Figure 3C shows UV-vis absorption spectra and photoluminescence (PL) spectra of the perovskite thin films with CsPbI2Br and CsPb0.95Eu0.05I2Br perovskite thin films. The absorption intensity of the CsPb0.95Eu0.05I2Br films within the measuring range is


5


Figure 3. Morphology and Optical Characterization of CsPbI2Br Thin Films with and without Europium Doping (A and B) SEM images of (A) CsPbI 2 Br and (B) CsPb 0.95 Eu 0.05 I 2 Br perovskite thin films. Scale bar, 300 nm. (C) Normalized UV-vis absorption and PL emission spectra measured at 298 K. (D) Time-resolved photoluminescence decays measured at 298 K.

higher than that of CsPbI2Br, probably due to the improvement in the film morphology. A small blue shift is visible in the PL spectrum with the maxima at 649 and 644 nm for 0 mol% and 5 mol% europium, respectively. This may be caused by a change in the band structure upon europium incorporation or, more trivially, by the reduction in crystallinity and grain size, consistent with the SEM images.16 Timeresolved PL (TRPL) measurements (Figure 3D) show increased PL lifetimes upon europium incorporation. This result suggests suppression of non-radioactive recombination pathways, consistent with the improved VOC measured on europium-doped PSCs. To gain more insight into how europium incorporation suppresses charge recombination within the films, we measured the VOC as a function of light intensity and recorded the electroluminescence (EL) as a function of the injection current by measuring a CsPb0.95Eu0.05I2Br device with the highest VOC of 1.27 V (Figure S12) as well as the best CsPbI2Br device. EL measurements can be used to characterize recombination at a full solar-cell device, which is kept in the dark while a forward bias is applied to drive a current by injection of electrons at the FTO/TiO2 and holes at the spiro-OMeTAD/Au contact. They recombine, and the fraction of radiative recombination is measured in EL emission. Figure 4A shows the external EL yield of CsPbI2Br and CsPb0.95Eu0.05I2Br devices. The EL intensity recorded at a current close to JSC gives external EL yields of roughly 1 3 107 and 2 3 106, showing a considerable decrease in non-radiative recombination upon europium doping. This result predicts an increase of 80 mV in the VOC (1.19–1.27 V) according to the calculation approach used, for example, in Tress et al.36 (bandgap of 1.92 eV). Another difference is visible in the power-law coefficient (slope of the semilogarithmic plot). It decreases from 2 to approximately 0.8, implying a

6



Figure 4. Optoelectronic Characterization (A) External quantum yield of electroluminescence (EQE EL ) measured while performing a currentvoltage sweep (10 mV/s) and plotted as a function of the injection current. The curves follow a power law, resulting in the given ideality factors n ID . (B) V OC as a function of the light intensity extracted from the transient experiment shown in Figure S10. The resulting n ID are slightly dependent on the scan direction. (C) Ideality factor differentially deduced from (B) and corrected for transients. Arrows indicate the scan direction. (D) Normalized PCE of unencapsulated CsPbI 2 Br and CsPb0.95 Eu 0.05 I 2 Br devices monitored under continuous white light exposure as a function of time.

reduction of the ideality factor from 3 to 1.8.37 To verify whether this trend is present also under light illumination, we analyzed the VOC as a function of the light intensity. Figure 4B shows the VOC measured for two consecutive stepwise intensity sweeps from 0 to 1 sun (forward and backward) (the transient data are shown in Figure S13). The common slow transient response known from the J-V hysteresis38 makes it difficult to accurately determine the ideality factor. Therefore, the values deduced from this plot contain a large error and depend on how fast the intensity sweep is performed. To avoid the direct influence of the transients, we propose to determine the ideality factor for each intensity step at the voltage response within 1 s (Figure 4C). This reproduces well the trend of the ideality factor seen in EL, although with slightly lower average values. The high ideality factor (>2) for the reference CsPbI2Br device cannot be explained in the framework of the established recombination theory, where values less than or equal to 2 are expected. Larger values have been explained by the presence of large defect distributions39 or coupled multi-level defect recombination. The latter has been observed in particular at certain positions in the device such as local shunts, which is consistent with the pinholes observed in the SEM images (cf. Figure 3A). Upon europium incorporation this effect disappears, and the ideality factor %1.5 is indicative of Shockley-Read-Hall recombination at regions with high background charge-carrier densities, such as interfaces.37 Also, this stepwise determination of the ideality factor yields differences between forward and backward intensity sweeps, indicating


7


that recombination varies dynamically. Taken together, these results suggest that incorporation of europium into the CsPbI2Br lattice significantly reduces non-radiative recombination and leads to a more stable VOC. To examine the long-term stability of the materials, we prepared thin films of CsPbI2Br and CsPb0.95Eu0.05I2Br and exposed them to ambient air at room temperature and under ca. 40% relative humidity. Figure S14 shows photographs of the two films monitored over the course of 6 months. We found that CsPb0.95Eu0.05I2Br remained dark during the whole monitoring period, while the CsPbI2Br film became yellow-white in less than 50 hr. After 1,200 hr we examined the two perovskite films by XRD (Figure S15B). Sharp perovskite peaks at 14.2 and 29.6 are visible for CsPb0.95Eu0.05I2Br and none of them are present in the undoped CsPbI2Br film, indicating that the perovskite phase has completely disappeared in the latter case. To study the phase of degraded CsPb0.95Eu0.05I2Br, we stored the film in humid conditions for 100 hr (humidity under ca. 50%–70%, 298 G 5 K) and we eventually achieved the completely degraded d perovskite phase (Figure S15C).We also monitored the evolution of UV-vis spectra of CsPbI2Br and CsPb0.95Eu0.05I2Br perovskite films under relative humidity of 50%–70% (Figure S16). Although both films degraded into yellow phase eventually with no photoactivity in the visible range, the europium-doped sample demonstrated better stability. At the same time, there was no shift of absorption onset during the measurement. We subjected one of the best performing, unencapsulated CsPb0.95Eu0.05I2Br-based devices to a long-term stability test under continuous white light LED illumination at 100 mW cm2 with N2 gas flow during 370 hr (a detailed description of the setup can be found in Supplemental Information). An undoped CsPbI2Br device was monitored under identical conditions as a reference. The CsPb0.95Eu0.05I2Br device showed no signs of degradation during the first 300 hr and retained 93% of its initial efficiency at the end of the test. Conversely, the efficiency of the CsPbI2Br device dropped to half of its initial value in less than 30 hr and retained only about 20% of its initial value at the end of the test, probably due to thermodynamic instability of this perovskite phase. The exceptional stability of the CsPb0.95Eu0.05I2Br device can be explained by the excellent thermodynamic phase stability of the europiumdoped perovskite. This may be due to the increase of tolerance factor after europium doping, which could induce less distortion of the perovskite lattice.27 In addition, the increase of surface-to-volume ratio leads to the increase of surface energy. Therefore, more energy is required to conquer the barrier for the transformation from a perovskite phase to d non-perovskite phase.40 Conclusion In conclusion, we have reported an all-inorganic europium-doped CsPbI2Br perovskite for photovoltaic applications. It is used as an absorber layer in mesoscopic solar cells that obtained power-conversion efficiencies of up to 13.71% and a stable power output of 13.34%. Using solid-state NMR and HAADF STEM we have shown that europium is incorporated into the perovskite lattice on the atomic level. Meanwhile, europium doping leads to exceptionally high thermodynamic stability of the CsPbI2Br perovskite, which translates to excellent operational stability of the resulting solar cells.

EXPERIMENTAL PROCEDURES Full experimental procedures are provided in the Supplemental Information.

8



SUPPLEMENTAL INFORMATION Supplemental Information includes Supplemental Experimental Procedures, 16 figures, and 3 tables and can be found with this article online at https://doi.org/10. 1016/j.joule.2018.10.008.

ACKNOWLEDGMENTS This work was funded by National Science Foundation of China (no. 51502224), with financial support from the Chinese Scholarship Council. S.A. would like to thank TUBITAK – 2214-A International Doctoral Research Fellowship Program.

AUTHOR CONTRIBUTIONS W.X. conceived the idea of the work. W.X. and Z.W. designed the project. W.X. fabricated and characterized devices and performed photoluminescence and UV-vis. D.J.K. performed ss-NMR and its analysis. D.P. performed p-XRD. W.T. performed EL and VOC versus light intensity measurement and their analyses. W.X. and J.L. performed SEM and (HAADF) STEM and conducted the XRD measurements. W.X. and Z.W. analyzed and discussed the XRD, SEM, and (HAADF) STEM data. W.X. and J.Z. performed AFM. W.X. and Z.W. wrote the manuscript. D.J.K. and A.H. revised the manuscript. All authors contributed toward finalizing the draft.

DECLARATION OF INTERESTS The authors declare no competing interests. Received: June 10, 2018 Revised: August 31, 2018 Accepted: October 2, 2018 Published: October 26, 2018

REFERENCES 1. Kojima, A., Teshima, K., Shirai, Y., and Miyasaka, T. (2009). Organometal halide perovskites as visible-light sensitizers for photovoltaic cells. J. Am. Chem. Soc. 131, 6050–6051. 2. Kim, H.S., Lee, C.R., Im, J.H., Lee, K.B., Moehl, T., Marchioro, A., Moon, S.J., Humphry-Baker, R., Yum, J.H., Moser, J.E., et al. (2012). Lead iodide perovskite sensitized all-solid-state submicron thin film mesoscopic solar cell with efficiency exceeding 9%. Sci. Rep. 2, 591. 3. Lee, M.M., Teuscher, J., Miyasaka, T., Murakami, T.N., and Snaith, H.J. (2012). Efficient hybrid solar cells based on mesosuperstructured organometal halide perovskites. Science 338, 643–647. 4. Jeon, N.J., Noh, J.H., Yang, W.S., Kim, Y.C., Ryu, S., Seo, J., and Seok, S.I. (2015). Compositional engineering of perovskite materials for high-performance solar cells. Nature 517, 476–480. 5. Burschka, J., Pellet, N., Moon, S.J., HumphryBaker, R., Gao, P., Nazeeruddin, M.K., and Gratzel, M. (2013). Sequential deposition as a route to high-performance perovskitesensitized solar cells. Nature 499, 316–319. 6. The National Renewable Energy Laboratory (NREL). (2018). https://www.nrel.gov/pv/ assets/pdfs/pv-efficiencies-07-17-2018.pdf.

7. Polman, A., Knight, M., Garnett, E.C., Ehrler, B., and Sinke, W.C. (2016). Photovoltaic materials: present efficiencies and future challenges. Science 352, 307. 8. Berhe, T.A., Su, W.N., Chen, C.H., Pan, C.J., Cheng, J.H., Chen, H.M., Tsai, M.C., Chen, L.Y., Dubale, A.A., and Hwang, B.J. (2016). Organometal halide perovskite solar cells: degradation and stability. Energy Environ. Sci. 9, 323–356. 9. Robert, F. (2016). Cesium fortifies nextgeneration solar cells. Science 351, 113–114.

13. Eperon, G.E., Paterno`, G.M., Sutton, R.J., Zampetti, A., Haghighirad, A.A., Cacialli, F., and Snaith, H.J. (2015). Inorganic caesium lead iodide perovskite solar cells. J. Mater. Chem. A 3, 19688–19695. 14. Swarnkar, A., Marshall, A.R., Sanehira, E.M., Chernomordik, B.D., Moore, D.T., Christians, J.A., Chakrabarti, T., and Luther, J.M. (2016). Quantum dot-induced phase stabilization of alpha-CsPbI3 perovskite for highefficiency photovoltaics. Science 354, 92–95.

10. Liang, J., Liu, J., and Jin, Z. (2017). All-inorganic halide perovskites for optoelectronics: progress and prospects. Solar RRL 1, https:// doi.org/10.1002/solr.201700086.

15. Bian, H., Bai, D., Jin, Z., Wang, K., Liang, L., Wang, H., Zhang, J., Wang, Q., and Liu, S. (2018). Graded bandgap CsPbI2+xBr1x perovskite solar cells with a stabilized efficiency of 14.4%. Joule 2, 1500–1510.

11. Beal, R.E., Slotcavage, D.J., Leijtens, T., Bowring, A.R., Belisle, R.A., Nguyen, W.H., Burkhard, G.F., Hoke, E.T., and McGehee, M.D. (2016). Cesium lead halide perovskites with improved stability for tandem solar cells. J. Phys. Chem. Lett. 7, 746–751.

16. Taiyang, Z., Dar, M.I., Ge, L., Feng, X., Guo, N., Michael, G., and Zhao, Y. (2017). Bication lead iodide 2D perovskite component to stabilize inorganic a-CsPbI3 perovskite phase for high-efficiency solar cells. Sci. Adv. 3, e1700841.

12. Stoumpos, C.C., Malliakas, C.D., and Kanatzidis, M.G. (2013). Semiconducting tin and lead iodide perovskites with organic cations: phase transitions, high mobilities, and near-infrared photoluminescent properties. Inorg. Chem. 52, 9019–9038.

17. Lau, C.F.J., Zhang, M., Deng, X., Zheng, J., Bing, J., Ma, Q., Kim, J., Hu, L., Green, M.A., Huang, S., and Ho-Baillie, A. (2017). Strontiumdoped low-temperature-processed CsPbI2Br perovskite solar cells. ACS Energy Lett. 2, 2319–2325.


9


18. Hu, Y., Bai, F., Liu, X., Ji, Q., Miao, X., Qiu, T., and Zhang, S. (2017). Bismuth incorporation stabilized a-CsPbI3 for fully inorganic perovskite solar cells. ACS Energy Lett. 2, 2219–2227. 19. Jiang, Y., Yuan, J., Ni, Y., Yang, J., Wang, Y., Jiu, T., Yuan, M., and Chen, J. (2018). Reduceddimensional a-CsPbX3 perovskites for efficient and stable photovoltaics. Joule 2, 1356–1368. 20. Kubicki, D.J., Prochowicz, D., Hofstetter, A., Zakeeruddin, S.M., Gratzel, M., and Emsley, L. (2017). Phase segregation in Cs-, Rb- and Kdoped mixed-cation (MA)x(FA)1-xPbI3 hybrid perovskites from solid-state NMR. J. Am. Chem. Soc. 139, 14173–14180. 21. Rosales, B.A., Men, L., Cady, S.D., Hanrahan, M.P., Rossini, A.J., and Vela, J. (2016). Persistent dopants and phase segregation in organolead mixed-halide perovskites. Chem. Mater. 28, 6848–6859. 22. Bai, D., Zhang, J., Jin, Z., Bian, H., Wang, K., Wang, H., Liang, L., Wang, Q., and Liu, S.F. (2018). Interstitial Mn2+-driven high-aspectratio grain growth for low-trap-density microcrystalline films for record efficiency CsPbI2Br solar cells. ACS Energy Lett. 3, 970–978. 23. Pan, J., Shang, Y., Yin, J., De Bastiani, M., Peng, W., Dursun, I., Sinatra, L., El-Zohry, A.M., Hedhili, M.N., Emwas, A.H., et al. (2018). Bidentate ligand-passivated CsPbI3 perovskite nanocrystals for stable near-unity photoluminescence quantum yield and efficient red light-emitting diodes. J. Am. Chem. Soc. 140, 562–565. 24. Liang, J., Zhao, P., Wang, C., Wang, Y., Hu, Y., Zhu, G., Ma, L., Liu, J., and Jin, Z. (2017). CsPb0.9Sn0.1IBr2 based all-inorganic perovskite solar cells with exceptional efficiency and stability. J. Am. Chem. Soc. 139, 14009–14012. 25. Yi, C., Luo, J., Meloni, S., Boziki, A., AshariAstani, N., Gra¨tzel, C., Zakeeruddin, S.M., Ro¨thlisberger, U., and Gra¨tzel, M. (2016). Entropic stabilization of mixed A-cation ABX3 metal halide perovskites for high performance

10


perovskite solar cells. Energy Environ. Sci. 9, 656–662. 26. Lefevre, G., Herfurth, A., Kohlmann, H., Sayede, A., Wylezich, T., Welinski, S., Duarte Vaz, P., Parker, S.F., Blach, J.-F., Goldner, P., and Kunkel, N. (2018). Electron-phonon coupling in luminescent europium doped hydride perovskites studied by luminescence spectroscopy, inelastic neutron scattering, and first-principle calculations. J. Phys. Chem. C 122, 10501–10509. 27. Swarnkar, A., Mir, W.J., and Nag, A. (2018). Can B-site doping or alloying improve thermal- and phase-stability of all-inorganic CsPbX3 (X = Cl, Br, I) perovskites? ACS Energy Lett. 3, 286–289. 28. Kubicki, D.J., Prochowicz, D., Hofstetter, A., Pechy, P., Zakeeruddin, S.M., Gratzel, M., and Emsley, L. (2017). Cation dynamics in mixedcation (MA)x(FA)1-xPbI3 hybrid perovskites from solid-state NMR. J. Am. Chem. Soc. 139, 10055–10061. 29. Kubicki, D.J., Prochowicz, D., Hofstetter, A., Saski, M., Yadav, P., Bi, D., Pellet, N., Lewinski, J., Zakeeruddin, S.M., Gratzel, M., and Emsley, L. (2018). Formation of stable mixed guanidinium-methylammonium phases with exceptionally long carrier lifetimes for highefficiency lead iodide-based perovskite photovoltaics. J. Am. Chem. Soc. 140, 3345– 3351. 30. Kubicki, D.J., Prochowicz, D., Hofstetter, A., Zakeeruddin, S.M., Gratzel, M., and Emsley, L. (2018). Phase segregation in potassium-doped lead halide perovskites from (39)K solid-state NMR at 21.1 T. J. Am. Chem. Soc. 140, 7232– 7238. 31. Knight, M.J., Felli, I.C., Pierattelli, R., Emsley, L., and Pintacuda, G. (2013). Magic angle spinning NMR of paramagnetic proteins. Acc. Chem. Res. 46, 2108–2116. 32. Hu, Y., Aygu¨ler, M.F., Petrus, M.L., Bein, T., and Docampo, P. (2017). Impact of rubidium and cesium cations on the moisture stability of multiple-cation mixed-halide perovskites. ACS Energy Lett. 2, 2212–2218.

33. Solomon, I. (1955). Relaxation processes in a system of two spins. Phys. Rev. 99, 559. 34. Clayton, A.N., Dobson, C.M., and Grey, C.P. (1990). High resolution 13C MAS NMR spectra of paramagnetic lanthanide complexes. J. Chem. Soc. Chem. Commun. 72–74. 35. Franta, D., Ohlı´dal, I., and Necas, D. (2008). Optical quantities of rough films calculated by Rayleigh-Rice theory. Phys. Status Solidi C 5, 1395–1398. 36. Tress, W., Marinova, N., Ingana¨s, O., Nazeeruddin, M.K., Zakeeruddin, S.M., and Graetzel, M. (2015). Predicting the open-circuit voltage of CH3NH3PbI3 perovskite solar cells using electroluminescence and photovoltaic quantum efficiency spectra: the role of radiative and non-radiative recombination. Adv. Energy Mater. 3, 6. 37. Tress, W., Yavari, M., Domanski, K., Yadav, P., Niesen, B., Correa Baena, J.P., Hagfeldt, A., and Graetzel, M. (2018). Interpretation and evolution of open-circuit voltage, recombination, ideality factor and subgap defect states during reversible light-soaking and irreversible degradation of perovskite solar cells. Energy Environ. Sci. 11, 151–165. 38. Tress, W., Marinova, N., Moehl, T., Zakeeruddin, S.M., Nazeeruddin, M.K., and Gra¨tzel, M. (2015). Understanding the ratedependent J-V hysteresis, slow time component, and aging in CH3NH3PbI3 perovskite solar cells: the role of a compensated electric field. Energy Environ. Sci. 8, 995–1004. 39. Giebink, N.C., Wiederrecht, G.P., Wasielewski, M.R., and Forrest, S.R. (2010). Ideal diode equation for organic heterojunctions. I. Derivation and application. Phys. Rev. B 82, 155305. 40. Ahmad, W., Khan, J., Niu, G., and Tang, J. (2017). Inorganic CsPbI3 perovskite-based solar cells: a choice for a tandem device. Solar RRL 1, 1700048.